Optics has the inherent advantages of parallelism and wide bandwidths in processing information. However, the need to interface with electronics creates a bottleneck that eliminates many of these advantages. The proposed research explores novel optical devices and techniques to overcome some of these bottlenecks. To address parallelism issues we take a specific example of a content-addressable memory that can recognize images. Image recognition is an important task that in principle can be done rapidly using the natural parallelism of optics. However in practice, when presented with incomplete or erroneous information, image recognition often fails to give the correct answer. To address this problem we examine a scheme based on free-space interconnects implemented with diffractive optics. For bandwidth issues, we study possible ways to eliminate the electronic conversion bottleneck by exploring all-optical buffer memories and all-optical processing elements. For buffer memories we examine the specific example of slow light delay lines. Although this is currently a popular research topic, there are fundamental issues of the delay-time-bandwidth product that must be solved before slow light delay lines can find practical applications. For all-optical processing we examine the feasibility of constructing circuit elements that operate directly at optical frequencies to perform simple processing tasks. Here we concentrate on the simplest element, a sub-wavelength optical wire, along with a grating coupler to interface with conventional optical elements such as lenses and fibers. Even such a simple element as a wire has numerous potential applications. In conclusion, information processing by all-optical devices are demonstrated with an associative memory using diffractive optics, an all-optical delay line using room temperature slow light in photorefractive crystals, and a subwavelength optical circuit by surface plasmon effects.

High-speed optical networks will require all-optical signalprocessing to avoid bottleneck due to optical-to-electrical (O/E)and electrical-to-optical (E/O) conversion. Enabling of opticalprocessing tasks such as optical buffering and data synchronizationwill require large tunable delay. Recently, slow-light physics gotwide attention to generate tunable delay. However, for a slow-lightsystem large delay comes at the expense of increased distortion.This dissertation presents a study of the slow-light systems andquantifies the limitations imposed on delay due to distortion andsystem resource constraints. Optimal designs for two- and three-lineBrillouin slow-light systems showed fractional pulse delay of up to1.7 compared to a single-line gain system. Optimal designs forbroadband Brillouin gain system showed upto 100\% delay improvementcompared to the Gaussian pump. Wavelength conversion and dispersionbased tunable delay systems showed bit delay of 15 bits. An opticalbuffer based on photorefractive medium for real-time data storagewas demonstrated by storing and retrieving a 7-bit data sequence.

<p>All-optical devices have attracted many research interests due to their ultimately low heat dissipation compared to conventional devices based on electric-optical conversion. With recent advances in nonlinear optics, it is now possible to design the optical properties of a medium via all-optical nonlinear effects in a table-top device or even on a chip.</p><p>In this thesis, I realize all-optical control of the optical group velocity using the nonlinear process of stimulated Brillouin scattering (SBS) in optical fibers. The SBS-based techniques generally require very low pump power and offer a wide transparent window and a large tunable range. Moreover, my invention of the arbitrary SBS resonance tailoring technique enables engineering of the optical properties to optimize desired function performance,</p><p>which has made the SBS techniques particularly widely adapted for</p><p>various applications.</p><p>I demonstrate theoretically and experimentally how the all-optical</p><p>control of group velocity is achieved using SBS in optical fibers.</p><p>Particularly, I demonstrate that the frequency dependence of the</p><p>wavevector experienced by the signal beam can be tailored using</p><p>multi-line and broadband pump beams in the SBS process. Based on the theoretical framework, I engineer the spectral profile</p><p> to achieve two different application goals: a uniform low group velocity (slow light) within a broadband spectrum, and a group velocity with a linear dependence on the frequency detuning (group velocity dispersion or GVD).</p><p>In the broadband SBS slow light experiment, I develop a novel noise current modulation method that arbitrarily tailors the spectrum of a diode laser. Applying this method, I obtain a 5-GHz broadband SBS gain with optimized flat-topped profile, in comparison to the ~40 MHz natural linewidth of the SBS resonance. Based on the broadband SBS resonance, I build a 5-GHz optical buffer and use this optical buffer to delay a return-to-zero data sequence of rate 2.5 GHz (pulse width 200 ps). The fast noise modulation method significantly stabilizes the SBS gain and improves the signal fidelity. I obtain a tunable delay up to one pulse-width with a peak signal-to-noise ratio of 7. I also find that SBS slow light performance can be improved by avoiding competing nonlinear effects. A gain-bandwidth product of 344 dB.GHz is obtained in our system with a highly-nonlinear optical fiber.</p><p>Besides the slow light applications, I realize that group velocity dispersion is also optically controlled via the SBS process. In the very recent GVD experiment, I use a dual-line SBS resonance and obtain a tunable GVD parameter of 7.5 ns$^2$/m, which is 10$^9$ times larger than the value found in a single-mode fiber. The large GVD system is used to disperse an optical pulse with a pulse width of 28 ns, which is beyond the capability for current dispersion techniques working in the picosecond and sub picosecond region. The SBS-based all-optical control of GVD is also widely tunable and can</p><p>be applied to any wavelength within the transparent window of the</p><p>optical fiber. I expect many future extensions following this work</p><p>on the SBS-based all-optical GVD control using the readily developed SBS tailoring techniques.</p><p>Finally, I extend the basic theory of backwards SBS to describe the forward SBS observed in a highly nonlinear fiber, where asymmetric forward SBS resonances are observed at the gigahertz range. An especially large gain coefficient of 34.7 W$^{-1}$ is observed at the resonance frequency of 933.8 MHz. This is due to good overlap between the optical wave and the high order guided radial acoustic wave. The interplay from the competing process known as the Kerr effect is also accounted for in the theory.</p> / Dissertation

La photonique sur silicium est envisagée comme une solution technologique très prometteuse pour le remplacement des interconnexions électriques par des interconnexions optiques devant se produire dans les prochaines années. Des dispositifs optoélectroniques comme des sources lasers, des modulateurs et des détecteurs, ont été développés pour la réalisation de circuits intégrant des émetteurs/récepteurs. Parmi les défis devant être relevés pour faire avancé la photonique sur silicium, la réduction de la consommation électrique du modulateur est un point crucial. L’intégration des composants passifs et actifs en utilisant une seule et même technologie est également un enjeu majeur pour les futurs systèmes de communication optique. Grâce au développement de l'intégration hybride de semi-conducteurs III-V sur silicium pour la réalisation de sources laser sur silicium, de nouvelles voies peuvent être envisagée pour réaliser des modulateurs optiques et des photodétecteurs efficaces et compacts. De plus, les cristaux photoniques 2D (PhC) et spécifiquement les structures à ondes lentes, qui sont connues pour renforcer les interactions entre la lumière et la matière peuvent apporter des solutions intéressantes pour diminuer de manière ultime la puissance consommée.Dans ce contexte, les travaux menés durant ma thèse ont porté plus spécifiquement sur la conception, la fabrication et la caractérisation de modulateurs à électro-absorption à onde lente en semiconducteur III-V sur silicium. Dans une première partie consacrée à la modélisation, une attention particulière est portée à la conception du cristal photonique et au couplage de la lumière du guide silicium vers l’onde lente. Les performances de la structure optimisée sont aussi analysées, donnant un modulateur de seulement 18.75 µm de longueur fonctionnant à 15 GHz avec un taux d’extinction supérieure à 5 dB sur une gamme spectrale supérieure à 10 nm. Par la suite, l’ensemble des procédés de nanotechnologies durant la thèse pour la fabrication des dispositifs sont présentés. Enfin, les résultats expérimentaux obtenus au cours de cette thèse démontrent l’effet Stark Confiné Quantiquement et l’effet de photodétection obtenu sur les structures intégrées.Les perspectives de ce travail de thèse concernent la réalisation de circuits intégrés photoniques complets, incluant sources lasers, modulateurs à électroabsorption et photodétecteurs en utilisant une seule et même technologie. / Silicon photonics is considered as a promising solution to replace electrical interconnections in the next years. Among the remaining challenges, the driving power of the active devices has to be minimized. Furthermore the use of a common technological platform for the realization of Silicon (Si) photonics passive and active devices would present a great interest in term of fabrication complexity and cost. III-V on Si is a good candidate for such a common technological platform as the physical properties of III-V semiconductors allow for active functionalities such as III-V on Si laser which have already been successfully demonstrated. In this perspective, 2D photonic crystals (PhCs) and slow light structures, which are known to intrinsically reinforce light/matter interactioncan alsobring interesting opportunities.In this context, the work is focused on the design, fabrication and characterization of slow-light III-V- on-silicon electroabsorption modulators. In a first part, the photonic crystal structure and light coupling from silicon waveguide to slowlight III-V waveguide are designed and modeled. The performance of the optimized structure is analyzed, showing a modulator operating at 15 GHz and exhibiting an extinction ratio of more than 5 dB over a spectral range of more than 10 nm, using a 18.75 µ;m-long modulator. Subsequently, the masks and fabrication steps for a hybrid III-V photonic crystalon Si modulators are presented. Finally, the experimental results obtained during this thesis are presented, showing Quantum Confined Stark Effect and photodetection in the waveguide integrated structures.The reported works open perspective towards the integrating of optical modulators with III-V on silicon nanolasers and photodetectors using a single technology.

No description available.

No description available.

Light-matter interactions are fundamentally based on the quantum mechanical principles that govern photons, electrons and other fundamental particles. One very interesting phenomenon within all of light-matter interactions is Electromagnetically Induced Transparency(EIT). This phenomenon causes an otherwise absorbing atomic transition to stop absorbing through quantum mechanical interference of probability wave functions. Corresponding to that change in absorption, will be a sudden, large change in the index of refraction. This change in the index of refraction leads to another phenomenon in which the group velocity of light can be slowed down dramatically. In the past, many researchers have been able to achieve both EIT and slow light in bulk atomic vapor cells. In an attempt to miniaturize this process and we have been using a platform of Anti Resonant Reflecting Optical Waveguides (ARROW) devices to both guide light and contain the interacting matter. However, the platform creates a whole new set of challenges when integrating rubidium vapor into the hollow waveguides as rubidium is highly reactive and it is difficult to maintain an inert atmosphere for the rubidium vapor. A variety of sealing methods were attempted and their appropriateness and effectiveness was analyzed. Among these sealing methods were PMMA, Crystal Wax, Active Solder, Epoxy, and Indium Solder. PMMA, Crystal Wax and Active Solder each had major faults in one or more of the sealing requirements. We have used a high temperature epoxy with relative success to contain the rubidium vapor. However, the epoxy degrades very quickly at the high temperatures required for EIT testing. Indium solder is the most recent application method. It has high potential although we have yet to fully test its effectiveness. We were able to successfully demonstrate the first EIT and slow light on a chip with our ARROW atomic vapor cell system. In the slow light experiment, we were able to slow light down to 2.5x105m/s. The group velocity of light decreased from the standard 3x108m/s by a factor of 1200. We believe we can achieve even lower group velocities using this same platform through further experimentation.

Los moduladores son elementos claves para la transmisión de la señal y el procesamiento de la información. Las técnicas de fabricación avanzadas "complementary metal-oxide semiconductor" (CMOS) permiten reducir drásticamente las dimensiones de estos dispositivos de interés para la implementación a gran escala en un chip de silicito a bajo coste. El trabajo realizado en esta tesis se centra en el diseño, la fabricación y la caracterización de estructuras de onda lenta con el objetivo de realizar moduladores compactos y eficientes integrados en un chip de silicio. El trabajo se divide en cuatro capítulos y un capítulo de conclusión y perspectivas. El capítulo uno introduce los fundamentos de física del estado sólido y de los mecanismos básicos de propagación guiada de la luz por reflexión total interna. El capítulo dos presenta los parámetros importantes de los moduladroes electro-ópticos así como un trabajo de recopilación de todos los mecanismos físicos que pueden ser empleados para modular la luz en silicio. Además, se presenta el estado del arte de los moduladores basados en silicio. El capítulo tres presenta el diseño , fabricación y caracterización de un modulador electro-óptico en silicio compacto y eficiente basado en el efecto de onda lenta en una estructura periódica unidimensional integrada, cuya geometría, similar a la de una red de Bragg, permite reducir la velocidad de grupo de un paquetes de ondas. Dicho efecto, se emplea para incrementar la interacción luz-materia y por lo tanto la eficiencia del modulador electro-óptico. El capítulo cuatro demuestra experimentalmente que dicha guía unidimensional periódica puede ser mejorada a fin de conseguir que el efecto de baja velocidad de grupo suceda en un rango mayor de longitudes de onda para posibles aplicaciones como la multiplexación por división de longitudinal de onda. En el capítulo cinco, se proporcionan conclusiones y perspectivas sobre el trabajo realizado. / Brimont ., ACJ. (2011). Towards Compact and High Speed Silicon Modulators [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/14345

Los efectos de la luz lenta y luz rápida (SFL) han mostrado unas capacidades excepcionales sobre el control dinámico de la velocidad de la luz en diferentes medios. Una de las motivaciones más estimulantes redica en la potente aplicación de estos sistemas en el marco del procesado fotónico de señales de radio frecuencia (RF). En esta tesis doctoral, se evalúan las prestaciones de las plataformas de SFL actuales para desarrollar múltiples tareas que se requieren en el campo de la fotónica de microondas (MWP) con el valor añadido de sintonizabilidad y operación en banda ancha. En esta contexto, el scattering de Brillouin estimulado (SBS) tanto en fibra estándar como en fibra mantenedora de polarización (PMF), de redes Bragg (FBG), amplificadores ópticos de semiconductor (SOA) y cristales fotónicos (PhC) han sido las tecnologías bajo estudio. Desde escalas del orden de km hata mm, estas plataformas de SFL representan la evolución hacia la consolidación de componentes y subsistemas de MWP en circuitos fotónicos integrados (PIC). Diversos modelos analíticos y numéricos se han desarrollado con el objetivo de entender los procesos físicos que goboernan la porpagación a través de las diferentes plataformas de SFL, así como para describir los enfoques de MWP propuestos. Además, a través de las plataformas presentadas se ha llevado a cabo el análisis de las prestaciones de dos de las funcionalidades clave que se requieren para el procesado fotónico de señales de microondas, desfasadores sintonizables y retardos verdaderos (TTD). Se ha propuesto un sistema de TTD basado en la llamada técnica de sintonización separada de la portadora (SCT) a través de los efectos de SBS en fibrras estándr. Se ha evaluado la interacción del SBS en PMF con el propósito de desarrollar redes de Brillouin dinámicas (DBG), cuya fase generada ha sido fruto de estudio. Por otro lado, también se ha demostrado un sistema de densado distribuido basado en la reflexión continua de un pulso estrecho a lo la / Sancho Durá, J. (2012). Photonic-assisted RF Signal Processing based on Slow and Fast Light Technological Platforms [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/16471

This thesis set out to investigate light storage based on dynamic electromagnetically induced transparency (EIT) in a room-temperature atomic ensemble of rubidium as a means to provide a quantum memory for single-photons created by a single rubidium atom coupled to a high-finesse optical resonator. Setting up the light storage medium presented a new addition to the research group's portfolio of experimental techniques and led to investigations of EIT, slow light and stored light in warm rubidium-87 vapour. Lambda level schemes connecting Zeeman or hyperfine substates on the D₁ and D₂ lines were addressed in rubidium vapour cells containing different buffer gases and different isotopic fractions of rubidium-87 and rubidium-85. Single beam spectroscopy with a weak probe was used to characterise the vapour cells. A numerical method to fit the D line spectrum to a theoretical model to include isotopic fractions and collisional broadening of a buffer gas has been implemented. Temperature and isotopic fractions could be reliably extracted from the fit parameters. For an offset-stabilisation of two lasers to address a lambda level scheme connecting the two different hyperfine groundstates in rubidium a phase locked loop including a frequency divider has been designed and implemented. Light storage and retrieval has been demonstrated using a Zeeman scheme on the D1 line. Two microsecond long classical light pulses containing one million photons on average were stored and retrieved with an efficiency of 15&percnt; after a delay of one microsecond. Several methods of attenuating the strong co-propagating control laser beam to allow for lowering the signal pulse intensity in future experiments are discussed.