La détection de gaz suscite depuis une dizaine d'années, un intérêt grandissant voir galopant, cela pour diverses raisons : environnementales, de santé publique, de sécurité (Hommes et infrastructures), etc... Ce dynamisme a pour conséquence un besoin de renouvellement et de perfectionnement des moyens de détection, qui croît à mesure que les normes existantes et les secteurs d'application respectivement, se complexifient et se diversifient. Une telle frénésie entraîne naturellement l'augmentation du nombre de projets de recherche sur cette thématique : le projet ANR PEPS, débuté en 2010, est l'un d'entre-eux. Acronyme de "Pellet Photonique Sensor", il vise à démontrer la faisabilité d'une plateforme de détection photonique multi-gaz exploitant l'effet thermo-optique, grâce à la combinaison de deux éléments centraux : des nanopoudres catalytiques sélectives aux gaz visés (dihydrogène ou monoxyde de carbone) et un composant photonique planaire très sensible aux variations d'indice de réfraction. Ce manuscrit est principalement consacré à la conception (sélection et optimisation) du composant photonique. / For about a decade, gas detection has known a tremendous interest, due to several reasons: environmental issues, public health, people and building safety, etc... This trend, triggered a spectacular and sustainable need of gas detection means improvement and development, which has grown with both standards sophistication and scope extension. As a matter of fact, the number of research projects related to this particular topic increased: PEPS ANR project, begun in 2010, is one of them. This research project, which is the acronym of Pellet Photonique Sensor, aim at developing a photonic multigas detection system based on thermo-optical effect, thanks to the combination of on one hand, catalitic nanopowders which react selectively with the target gas (hydrogen or carbon monoxide) and on the other hand, an high sensitive to refractive index variations planar photonic component. This thesis manuscript mostly treats of the photonic component design (choice and optimization).

Dans le moyen-infrarouge, les barrettes de lasers à cascade quantique sont d’un grand intérêt pour la réalisation de sources large bande intégrables dans les systèmes de spectroscopie laser. Une excellente finesse spectrale, la présence d’un seul mode spatial et une gamme d’accordabilité large sont ainsi rassemblées sur une seule puce, compacte et intrinsèquement stable. Afin de bénéficier de l’ensemble des longueurs d’onde sur une sortie unique, les défis majeurs résident dans l’association de technologies pour rassembler les différentes sorties en une seule via l’utilisation de circuits photoniques intégrés (CPI). Ce CPI peut être séparé en trois briques élémentaires : une filière de guidage passif, un combineur de longueurs d’onde et un coupleur actif/passif. Pour la mise-en-forme du faisceau, nous reportons la fabrication et la caractérisation de guides d’onde en InP/InGaAs/InP gravés profondément, avec des performances proches de l’état de l’art. Nous fabriquons et caractérisons des multiplexeurs basés sur des réseaux de diffraction intégrés, sur filière InP et SiGe. Un multiplexeur de 60-vers-1 voies couvrant la gamme de 7-8,5 µm est réalisé. Une méthode innovante mettant en œuvre des multiplexeurs inter-digités et fonctionnant sur trois ordres de diffraction est démontrée. Finalement, nous réalisons des barrettes de laser à cascade quantique sur InP et sur silicium. Un coupleur adiabatique est dimensionné, fabriqué et caractérisé pour associer efficacement les guides actifs et passifs. Des intégrations de types hétérogène et hybride sont envisagées, avec la première démonstration d’une source accordable utilisant une barrette de lasers et un multiplexeur InP. / In the mid-infrared, arrays of distributed feedback quantum cascade lasers have been developed as a serious alternative to obtain extended wavelength operation range of laser-based gas sensing systems. Narrow-linewidth, single mode operation and wide tunability are then gathered together on a single chip with high compactness and intrinsic stability. In order to benefit from this extended wavelength range in a single output beam, the key challenge resides in the combination of different technologies to merge the output of different sources via the use of mid-IR photonic integrated circuits (PIC). The PIC can be split into three main blocks: the passive waveguide platform, the beam combiner and the active/passive coupler. For beam handling and guiding, we report fabrication and characterization of deeply etched InP/InGaAs/InP waveguides with state of the art performances. We fabricate and characterize multiplexers based on echelle and arrayed waveguide gratings on InP and SiGe platforms. A 60-to-1 spectral multiplexer operating in the 7-to-8.5 µm range is demonstrated. An advanced multiplexing scheme using interleaved and cross-order operations is also exposed. Finally, we realize quantum cascade laser arrays on InP and silicon. We design, fabricate and characterize an adiabatic coupler to efficiently and monolithically integrate active and passive waveguides. Heterogenous and hybrid integration are also considered with the demonstration of a tunable source using laser array and InP-based multiplexer.

abstract: Photonic integrated circuit (PIC) in the visible spectrum opens up new opportunities for frequency metrology, neurophotonics, and quantum technologies. Group III nitride (III-N) compound semiconductor is a new emerging material platform for PIC in visible spectrum. The ultra-wide bandgap of aluminum nitride (AlN) allows broadband transparency. The high quantum efficiency of indium gallium nitride (InGaN) quantum well is the major enabler for solid-state lighting and provides the opportunities for active photonic integration. Additionally, the two-dimensional electron gas induced by spontaneous and polarization charges within III-N materials exhibit large electron mobility, which is promising for the development of high frequency transistors. Moreover, the noncentrosymmetric crystalline structure gives nonzero second order susceptibility, beneficial for the application of second harmonic generation and entangled photon generation in nonlinear and quantum optical technologies. Despite the promising features of III-N materials, the investigations on the III-N based PICs are still primitive, mainly due to the difficulties in material growth and the lack of knowledge on fundamental material parameters. In this work, firstly, the fundamental nonlinear optical properties of III-N materials will be characterized. Then, the fabrication process flow of III-N materials will be established. Thirdly, the waveguide performance will be theoretically and experimentally evaluated. At last, the supercontinuum generation from visible to infrared will be demonstrated by utilizing soliton dynamics in high order guided modes. The outcome from this work paves the way towards fully integrated optical comb in UV and visible spectrum. / Dissertation/Thesis / Doctoral Dissertation Electrical Engineering 2020

Suite au développement remarquable d'Internet, il est attendu que le trafic numérique augmente de manière exponentielle, ainsi que la nécessité d'utiliser des liens de communication très hauts débits. Pour adresser ces problématiques, il est donc essentiel de proposer des systèmes performants avec une consommation énergétique réduite. La photonique sur silicium est une solution prometteuse qui répond à ce besoin en intégrant des fonctionnalités optiques dans un circuit intégré utilisant les procédés de fabrication de pointe de la microélectronique. Dans ce contexte le sujet de thèse porte sur le modulateur optique. Celui-ci doit supporter des hauts débits de transmission, avoir de faibles pertes optiques, et être peu énergivore. Pour respecter ces objectifs, plusieurs paramètres doivent être optimisés en tenant compte des contraintes de fabrications, afin de trouver le meilleur compromis entre ces différents facteurs de mérite. Durant cette thèse, la recherche de l'obtention de meilleures performances du composant a été faite de trois manières. La première approche a été d'améliorer les simulations de la technologie existante de modulateurs à jonction PN. En intégrant les étapes de fabrication dans le processus d'optimisation des performances du composant, les résultats numériques sont plus réalistes. Le point clé de cette étude est la comparaison entre les caractérisations et les simulations de deux architectures différentes de modulateur obtenues dans les mêmes conditions de fabrication. Une partie importante de la thèse a aussi été consacrée au développement de nouveaux modulateurs basés sur l'utilisation d'une capacité verticale intégrée au milieu d'un guide d'onde. Des outils numériques dédiés ont permis de dimensionner deux nouvelles architectures de modulateurs possédant une grande efficacité. Un nouveau procédé de fabrication a été mis en place, et les premiers lots d'étude nous ont permis d'extraire des informations utiles pour la fabrication de ces composants. Enfin, une étude comparative générale entre trois des modulateurs étudiés au cours de cette thèse a été faite. Les résultats permettent de déterminer la configuration optimale pour chaque type de modulateur en fonction de l'application visée. De plus, l'ensemble des données nous a permis de générer un modèle compact pour optimiser rapidement le composant en un temps de simulation réduit. / With the outstanding development of the internet, it is expected that global network traffic will grow exponentially, as well as the concern about the need for high-speed links and interconnections. To address these issues, it is then essential to propose performant systems that will support high speed transmission with low power consumption. Silicon photonics is a promising solution and integrate complex optical functions in a silicon chip, by using standard fabrication process used in microelectronic. In this context, the subject of my PhD is focused on the optical modulator which should support high speed transmission, have low optical losses, and have low power consumption. To obtain these constraints, several parameters need to be optimized while taking account fabrication constraints in order to find the best compromise between the different figures of merit. During this PhD, the improvement of the performances of the component was made by three different ways. The first optimization is related to the simulations for the current technology of modulators based on PN junctions. By integrating the fabrication process in the optimization process, more reliable numerical results are obtain. The key point of this study is the comparison of experimental characterizations and numerical simulations of two architectures of modulator. A substantial part of the PhD was also focused on the development of new modulators based on vertical capacitive junctions. The use of dedicated numerical tools reveals several key aspects of these components, and allow us to optimize two different architectures in order to obtain high efficient modulator. A new fabrication process has been established, and several information were extracted from the first run of fabrication. Then, a comparative study between most of modulators reviewed during this PhD was performed. The results allow us to determine which configuration has the best performances depending of the targeted application. In addition, a compact model was generated to optimize the component in a reduced simulation time.

Les centres de données subissent une augmentation exponentielle du trafic de données, à laquelle la photonique sur silicium apporte une solution grâce à de forts volumes de production, à faibles coûts et à haut rendements. Aujourd'hui la consommation énergétique est un défi supplémentaire à relever face à la densification du réseau. Le modulateur capacitif permet en particulier de réduire la part d'énergie consommée par l'émetteur du circuit photonique qui réalise la modulation d'intensité optique. Les travaux d'intégration et de conception réalisés au cours de cette thèse consistent à introduire un isolant diélectrique au centre du guide d'onde, pour former une capacité silicium/ oxyde/ poly-silicium en régime d'accumulation. Une première version consiste à empiler l'isolant horizontal entre les deux semiconducteurs. Une seconde version complémentaire emploie un isolant vertical au centre du guide d'onde ruban à fente. La première génération de ces composants permet de démontrer leur faisabilité dans la plateforme industrielle de STMicroelectronics. Les mesures conduites au C2N, au CEA-LETI et à STMicroelectronics évaluent les compromis entre l'efficacité, atout du modulateur capacitif, les pertes d'insertion et la bande passante du composant à isolant horizontal. Des pertes de 3 dB/mm sont extraites, dont seulement 0.5 dB/mm dues au poly-silicium. Un taux d'extinction de 2 dB est mesuré à travers 700 µm sur un diagramme de l'œil à 10 Gb/s, grâce à un produit VpLp=5.5 V.mm à 15 nm d'épaisseur d'oxyde (1.2 pF/mm). La consommation électrique du modulateur capacitif est finalement optimisée pour atteindre 1 pJ/bit à 0.9 Vpp. / Global datacenter data exchange is exponentially growing and silicon photonics is the key answer, thanks to high production volume, at low cost and high yield. Today, energy consumption is a new challenge highlighted by network densification. Capacitive modulators address a specific reduction of the power dedicated to the photonic circuit emitter for light intensity modulation. Design and integration of capacitive modulators are carried out during this thesis. It consists of inserting a dielectric insulator within the optical waveguide center, in order to shape a capacitance with a silicon/ oxide/poly-silicon stack in accumulation regime. A first device is made up of an horizontal insulator stacked between the semiconductors. A second device type comprises a vertical insulator in the center of a slot rib waveguide. The first fabrication release demonstrates device feasibility within STMicroelectronics industrial platform. Characterization results of the first device type from C2N, CEA-LETI and STMicroelectronics evaluate the trade-offs between efficiency - featured by capacitive modulators - insertion losses and bandwidth. 3 dB/mm insertion losses are measured, including 0.5 dB/mm poly-silicon absorption only. 2 dB extinction ratio through 700 µm is evaluated on a 10 Gb/s eye diagram, thanks to a VpLp =5.5 V.mm at 15 nm oxide thickness (1.2 pF/mm). Capacitive modulator power consumption is eventually optimized for 1 pJ/bit at 0.9 Vpp.

The scaling trend of microelectronics over the past 50 years, quantified by Moore’s Law, has faced insurmountable bottlenecks, necessitating the use of optical communication with its high bandwidth and energy efficiency to further improve computing performance. Silicon photonics, compatible with CMOS platform manufacturing, presents a promising means to achieve on-chip optical links, employing highly sensitive microring resonator devices that demand electronic feedback and control due to fabrication variations. Achieving the full potential of both technologies requires tight integration to realize the ultimate benefits of both realms of technology, leading to the convergence of microelectronics and photonics. A promising approach for achieving this convergence is the monolithic integration of electronics and photonics on CMOS platforms. A critical milestone was reached in 2015 with the demonstration of the first microprocessor featuring photonic I/O (Chen et al, Nature 2015), accomplished by integrating transistors and photonic devices on a single chip using a monolithic CMOS silicon-on-insulator (SOI) platform (GlobalFoundries 45RFSOI, 45 nm SOI process) without process modifications, thus known as the "zero-change" approach. This dissertation focuses on leveraging the fabrication capabilities of advanced monolithic electronic-photonic 45 nm CMOS platforms, specifically high-resolution lithography and small feature size doping implants, to realize photonic devices with subwavelength features that could potentially provide the next leap in integrated optical links performance, beyond microring resonator based links. Photonic crystal (PhC) nanobeam cavities can support high-quality resonance modes while confining light in a small volume, enhancing light-matter interactions and potentially enabling ultimate efficiencies in active devices such as modulators and photodetectors. However, PhC cavities have been overshadowed by microring resonators due to two challenges. First, their fabrication demands high lithography resolution, which excludes most standard SOI photonic platforms as viable options for creating these devices. Secondly, the standing-wave nature of PhC nanobeam cavities complicates their integration into wavelength-division multiplexing (WDM) optical links, causing unwanted reflections when coupled evanescently to a bus waveguide. In this work, we present PhC nanobeam cavities with the smallest footprint, largest intrinsic quality factor, and smallest mode volume to be demonstrated to date in a monolithic CMOS platform. The devices were fabricated in a 45 nm monolithic electronics–photonics CMOS platform optimized for silicon photonics, GlobalFoundries 45CLO, exhibiting a quality factor in excess of 100,000 the highest among fully cladded PhC nanobeam cavities in any SOI platform. Furthermore to eliminate reflections, we demonstrate an approach using pairs of PhC nanobeam cavities with opposite spatial mode symmetries to mimic traveling-wave-like ring behavior, enabling efficient and seamless WDM link integration. This concept was extended to realize a reflectionless microring resonator unit with two microrings operating as standing-wave cavities. Using this scheme with standing-wave microring resonators could lead to an optimum geometry for microring modulators with interdigitated p-n junctions in terms of modulation efficiency in a manner that allows for straightforward WDM cascading. This work also presents the first demonstration of resonant-structure-based modulators in the GlobalFoundries 45CLO platform. We report the first-ever demonstration of a PhC modulator in a CMOS platform, featuring a novel design with sub-wavelength contacts on one side allowing it to benefit from the "reflection-less"' architecture. Additionally, we also report the first demonstration of microring modulators. The most efficient devices exhibited electro-optical bandwidths up to 30 GHz, and 25 Gbps non-return-to-zero (NRZ) on-off-keyed (OOK) modulation with 1 dB insertion loss and 3.1 dB extinction ratio. Finally, as the complexity of silicon photonic systems-on-a-chip (SoC) increases to enable new applications such as low-energy data links, quantum optics, and neuromorphic computing, the need for in-situ characterization of individual components becomes increasingly important. By combining Near-field scanning optical microscopy (NSOM) with a flip-chip post-processing technique, this dissertation demonstrates a method to non-invasively perform NSOM scans of a photonic device within a large-scale CMOS-photonic circuit, without interfering with the performance and packaging of the photonics and electronics, making it a valuable tool for future development of high performance photonic circuits and systems.

Silicon (Si) and silicon nitride (Si3N4) have become the dominant photonic integrated circuit (PIC) material platforms, due to their low-cost, wafer-scale production of high-performance circuits. However, novel materials can offer additional functionalities that cannot be easily accessed in Si and Si3N4, such as light emission. Tellurium oxide (TeO2) is a novel material of interest because of its large linear and non-linear refractive indices, low material losses and large rare-earth dopant solubility, with applications including compact low-loss waveguides and on-chip light sources and amplifiers. This thesis investigates the post-processing integration of TeO2 devices onto standardized Si and Si3N4 chips to incorporate TeO2 material advantages into high-performance PICs. Chapter 1 introduces the state-of-the-art functionality for various integrated photonic materials as well as methods for integrating multiple materials onto single chips. Chapter 2 presents the development of a high-quality TeO2 thin film fabrication process by reactive RF sputtering, with material refractive indices of 2.07 and optical propagation losses of <0.1 dB/cm at 1550 nm. Chapter 3 investigates a conformally coated TeO2-Si3N4 waveguide platform capable of large TeO2 optical confinement and tight bending radii, characterizing fiber-chip edge couplers down to ~5 dB/facet, waveguide propagation losses of <0.5 dB/cm, directional couplers with 100% cross-over ratio, and microresonators with internal Q factors of 7.3 × 105. In Chapter 4 a spectroscopic study of TeO2:Er3+-coated Si3N4 waveguide amplifiers was undertaken, with internal net gains of up to 1.4 dB/cm in a 2.2-cm-long waveguide and 5 dB total in a 6.7-cm-long sample demonstrated, predicted to reach >10 dB could 150 mW of pump power be launched based on a developed rate-equation model. Chapter 5 demonstrates TeO2-coated microtrench resonators coupled to silicon waveguides, with internal Q factors of up to 2.1×105 and investigates environmental sensing metrics of devices. Chapter 6 summarizes the thesis and provides avenues for future work. / Thesis / Doctor of Philosophy (PhD)

Current major demands in SPR sensor development are system miniaturization and throughput improvement. Structuring an array of integrated optical SPR sensor heads on a semiconductor based optoelectronic platform could be a promising solution for those issues, since integrated optical waveguides have highly miniaturized dimension and the optoelectronic platform enables on-chip optical-to-electrical signal conversion. Utilizing a semiconductor based platform to achieve optoelectronic functionality poses requirements to the senor head; the sensor head needs to have reasonably small size while it should have reasonable sensitivity and fabrication tolerance. This research proposes a novel type of SPR sensor head and demonstrates a fabricated device with an array of integrated optical SPR sensor heads endowed with optoelectronic functionality. The novel integrated optical SPR sensor head relies on mode conversion efficiency for its operational principle. The beauty of this type of sensor head is it can produce clear contrast in SPR spectrum with a highly miniaturized and simple structure, in contrast to several-millimeter-scale conventional absorption type or interferometer type sensor heads. The integrated optical SPR sensor with optoelectronic functionality has been realized by structuring a dielectric waveguide based SPR sensor head on a photodetector-integrated semiconductor substrate. A large number of unit sensors have been fabricated on a substrate with a batch fabrication process, which promises a high throughput SPR sensor system or low-priced disposable sensors.

Interest in nanophotonics continues to grow as integrated optics provides an affordable platform for areas like telecommunications, quantum information processing, and biosensing. Designing and characterizing integrated photonics components and circuits, however, remains a major bottleneck. This is especially true when complex circuits or devices are required to study a particular phenomenon.To address this challenge, this work develops and experimentally validates a novel machine learning design framework for nanophotonic devices that is both practical and intuitive. As case studies, artificial neural networks are trained to model strip waveguides, integrated chirped Bragg gratings, and microring resonators using a small number of simple input and output parameters relevant to designers. Once trained, the models significantly decrease the computational cost relative to traditional design methodologies. To illustrate the power of the new design paradigm, both forward and inverse design tools enabled by the new design paradigm are demonstrated. These tools are directly used to design and fabricate several integrated Bragg grating devices and ring resonator filters. The method's predictions match the experimental measurements well and do not require any post-fabrication training adjustments.

Computing as a service is rapidly becoming the new normal for many sectors of the economy. The widespread availability of broadband internet has allowed an extensive range of services to be delivered on-demand from centralized computing systems known as ‘data centers’. These systems have evolved to be enormously complex. Optical-based communication is desired to increase data center capability and efficiency, however traditional optical technologies are not feasible due to cost and size. Silicon photonics aims to deliver optical communications on an integrated and affordable platform for use in data centers by leveraging the existing capabilities of complementary metal-oxide semiconductor manufacturing. This thesis contains a description of the development of monolithic silicon photodiodes for use in photonic integrated circuits in, and beyond, the current telecommunications wavelength windows. The focus is on methods which are compatible with standard silicon processing techniques. This is in contrast to the current approaches which rely on hybrid material systems that increase fabrication complexity. Chapter 1 and 2 provide background information to place this work into context. Chapter 3 presents an experimental study of resonant devices with lattice defects which determines the refractive index change in silicon-on-insulator waveguides. High-speed operation of resonant photodiodes is demonstrated and is found to be limited by resonance instability. Chapter 4 demonstrates high responsivity avalanche photodetectors using lattice defects. The detectors are shown to operate error-free at 10 Gbit/s, thus confirming their capability for optical interconnects. Chapter 5 presents photodiodes operating with absorption through surface-state defects. These detectors show fast operation (10 Gbit/s) and have an extremely simple fabrication process. Chapter 6 demonstrates photodiodes operating beyond the traditional telecommunications window. Operation at 20 Gbit/s, at a wavelength of 1.96 µm is demonstrated, offering potential for their use in the next generation of optical communication systems which will exploit the thulium doped fiber amplifier. / Thesis / Doctor of Philosophy (PhD) / This thesis describes photodiodes constructed on silicon optical waveguides. The photodiodes are notable for their high-speed performance and simple fabrication methods. Such devices may find use within chip-integrated optical transceivers, which are desired for optical interconnects within large-scale computing systems such as data centers.