In this work I report on the development and optimization of fabrication process of three dimensional optical nano devices directly on the tip of optical fiber.

The system we use in order to realize these optical elements is the commercial Nanoscribe™ system which is based on polymerization of negative tone photoresist, and using two-photon absorption.

Many efforts were invested in this research in two main directions. One is the optimization of the element’s optical quality, by reducing its surface’s roughness and second is the system’s adaptation to print elements directly on an optical fiber tip, in accurate form and exact alignment to the optical axis of the fiber.

We utilize the ability to print arbitrary real 3D volumetric structures in photoresist at the nanoscale with our Nanoscribe tool directly onto a fiber tip in convenient and accurate way, and this ability gives us wide leeway for realizing sophisticated optical elements that are directly interact with the beam delivered by the fiber.

Spectrally dispersed light from a fine resolution waveguide grating router (WGR) of 25 GHz free spectral range (FSR) that radiates to free-space is spatially filtered at ~1 GHz resolution using a liquid crystal on Silicon (LCoS) spatial light modulator (SLM). Fabrication imperfections leading to phase errors on the 32 waveguide arms of the WGR are measured by the pair-wise far-field interference of adjacent waveguide pairs. The phase errors are then corrected using a UV pulsed laser to inscribe permanent optical path changes to the waveguides. WGR phase errors are permanently trimmed waveguide-by-waveguide with an excimer laser by inducing stress in the glass cladding above the waveguide for coarse setting and using the photosensitivity effect for fine setting. The WGR was then mated with an LCoS SLM located at the Fourier plane to form a photonic spectral processor (PSP), for arbitrary spectral amplitude and phase manipulations.

Ultrashort optical pulses are widely and increasingly used in many diverse fields of science and technology. By providing high temporal resolution they enable investigation and measurement of fundamental physical, chemical and biological phenomena that occur on picosecond time scales or shorter. In addition ultrashort pulses are an essential enabling tool for high speed optical communications and data processing technologies, as well as in advanced manufacturing and photomedicine applications. In all of these areas precise measurement and control of ultrashort optical pulses is vital - advances in ever shorter pulse generation must be accompanied by new methods to characterise and manipulate them.

This thesis presents work on the ongoing development of an ultrashort pulse measurement and manipulation technique known as time-to-space conversion. Time-to-space conversion uses sum-frequency generation between spectrally resolved ultrashort pulses to transfer information from the time domain to the space domain; in other words to create the real-time spatial image of an ultrashort pulse. Mapping the pulse temporal intensity envelope and phase onto a quasi-static spatial image allows high resolution measurement of these quantities, overcoming the difficulty of optoelectronic detection of ultrashort pulses directly in the time domain. Furthermore, the spectrally resolved nature of time-to-space conversion results in a large time window of operation. This enables a series of ultrashort pulses to be simultaneously transferred to spatially separated locations via interaction with a single reference pulse, thereby performing an all-optical demultiplexing operation.

The two main developments introduced in this thesis are: a) greater feasibility of time-to-space conversion for all-optical demultiplexing of a high speed optical communications channel by demonstrating the technique in a planar nonlinear waveguide and b) the demonstration of full-field characterisation of ultrashort pulses by using interferometric detection after the time-to-space conversion. The practicality of time-to-space conversion for all-optical demultiplexing depends on minimising its optical power consumption. This can be achieved by implementation of the conversion process in the guided-wave regime, as opposed to the free-space regime in which it has previously been demonstrated. The first three papers presented in this thesis describe the preliminary steps towards this goal, namely the demonstration of non wavelength-degenerate and background-free collinearly phase-matched time-tospace conversion and the demonstration of time-to-space conversion in a planar nonlinear waveguide. Full-field characterisation of ultrashort pulses by time-to-space conversion is enabled by the quasi-monochromaticity of the output sum-frequency signal, a feature which follows from the unique geometry of the oppositely dispersed waves of the pulse to be measured and the reference pulse. The quasi-monochromatic converted signal can be mixed with a narrow linewidth local oscillator for interferometric measurement of the ultrashort pulse field amplitude and phase. The final two papers included here describe the first time demonstration of full-field measurement of bandwidth-limited and chirped pulses by time-to-space conversion and of single-shot coherent detection of a phase modulated ultrashort pulse train.

Taken together, the work presented in this thesis has achieved an increase in the utility of time-to-space conversion as an ultrashort optical pulse measurement and manipulation technique, with potential applications in optical communications and data processing and in the field of ultrashort pulse measurement.

A novel approach for a multi-port Wavelength Selective Switch (WSS) is shown in this work. The switching is performed from a series of 8 input fibers to a series of 24 output fibers. This device can be useful in reducing the complexity of optical communication nodes based on conventional switch with only one input fiber

The multi-port switching is based on a spatial separation, of light beams, according to input port and wavelength channel on a dynamic steering device – LCoS (Liquid Crystal on Silicon) SLM (Spatial light modulator).

The LCoS SLM was extensively characterized in order to understand the capabilities and limitations for better system design.

The system design is extensively discussed in this paper and a proof of concept experiment demonstrates that indeed this concept can be realized.

A record performance metric arrayed waveguide grating (AWG) design with a 200 GHz free spectral range (FSR) capable of resolving sub-one GHz resolution spectral features is developed for a fine resolution photonic spectral processor (PSP). The AWGʼs FSR was designed to support sub-channel add/drop from a super-channel of 1Tb/s capacity. Due to fabrication imperfections we introduce phase corrections to the light beams emerging from the 250 waveguides of the AWG output using a liquid crystal on Silicon (LCoS) phase spatial light modulator (SLM) placed in an imaging configuration. A second LCoS SLM is located at the Fourier plane, for arbitrary spectral amplitude and phase manipulations. The PSP is utilized in different experiments, such as flexible spectral shaping and sub-carrier drop demultiplexer with sub-GHz spectral resolution.

In recent decades, the use of fast optical signals has become increasingly dominant, both in scientific research and in engineering applications. High speed photonics serves as the core of modern worldwide communication networks, as well as in many optical signal processing applications. Such applications rely on the ability to control, filter and manipulate large bandwidth signals. Traditionally, such control can be realized using fast electronics. However, continuous growth in data rates makes this option impractical, since the signals become too fast to control even for cutting edge electric circuit technology. The alternative is to use an all-optical system, where signal control is done in the frequency (spectral) domain. Such a system must be capable of manipulating large bandwidth signals with high spectral resolution. Such optical systems are essential in optical communication networks, for performing signal conditioning, impairment mitigation and WDM channel power equalization.

In this work I explore a family of optical sub-systems combining guided-wave and free-space optics for spectrally resolving optical signals at unprecedented resolution, and actively manipulating the spectral components with spatial light modulator (SLM) technology. The ability to combine the employed cutting edge technologies, including a high resolution planar lightwave circuit (PLC) arrayed waveguide grating (AWG), together with the state-of-the-art phase SLM, which was adapted from the light projection industry, enables the design and demonstration of high resolution photonic spectral processors (PSP). This system is capable of applying arbitrary spectral phase and amplitude at high spectral resolution to an optical signal and of controlling its properties in the time domain. A PSP can be configured for addressing the entire conventional optical communication band, at a price of poor resolution due to the finite space-bandwidth trade-off. Alternatively, the PSP can be designed as a colorless adaptive device, operating with a free spectral range (FSR) matching the channel plan, e.g. with a 100-GHz FSR, for in-band high-resolution wavelength division multiplexing (WDM) filtering applications. By using two-dimensional free-space optics achieved by crossing the PLC AWG with a bulk grating, a new broadband processor was introduced. This PSP is capable of controlling independent WDM channels on the 100 GHz grid at the high resolution of the colourless solution, thereby shattering the space-bandwidth limitation.

Based on these concepts, a family of novel systems and implementations were developed and investigated. In this thesis I introduce six papers which demonstrate the design and implementation of three PSP systems, based on hybrid waveguide/free space optics arrangements. The papers are divided into two groups: in the first group, three papers present the evolution of the spectral processing device, from the simplest version of colorless PSP up to two dimensional PSP arrangement with full spectral, control along the c-band. The second group contains three papers describing several implementations of these technologies, including amplitude filtering applications (Nyquist-WDM generation), phase filtering applications (tunable chromatic dispersion compensation and group delay stairs generation) and a demonstration of a new fiber laser which was built using the PSP platform. These high spectral resolution devices and systems can serve as an important element in controlling dispersion, enhancing signal quality and optimally filtering a distorted signal, and their development is essential for the progress in the optical fiber communication world.

In this work I report on the development of a platform of a polymeric waveguide composed of Cytop as the cladding and PFCB as the core. These two polymers were chosen due to their low loss in the optical communication regime (0.26 dB/cm for PFCB core and 0.022dB/cm for Cytop cladding). PFCB and Cytop have refractive indexes of 1.48 and 1.34 respectively and therefore offer high index contrast in comparison to glass waveguides. PFCB was chosen as the waveguide core since it has been proven a good host for nanocrystals (NCs).
In this work a lot of effort was invested in making the fabrication process compatible with semiconductor NCs that will in the long term be mixed in the PFCB core. Doping the core with nanocrystals is of interest, since the NCs properties are diverse, flexible and controllable. Choosing NCs with high third order susceptibility will allow us to fabricate nonlinear waveguides. Furthermore,specifying the NCs shape and size will allow us to align them by applying external electric voltage and by that enhance the macroscopic nonlinear properties of the composite.
Two fabrication configuration are proposed. Both are aimed at fabricating a square waveguide. The first configuration is the ridge-method where the PFCB core undergoes reactive ion etching (RIE).This method carries on with previously proposed methodology at the Photonic Devices Laboratory of Dr. Marom [1], however several key improvements were made.
The second configuration is the trench-method where only the Cytop undergoes RIE. By that method we wish to prevent roughness that might occur in the alternative method due to etching a composite made of PFCB and the NCs at the same time. In addition not all NC materials we would like to use are allowed into the RIE chamber since they may cause contamination to the RIE machine. Replacing the ridge method with the trench one will obviously overcome this obstacle. However both methods have their own challenges. In this work I tried to overcome some of the challenges and to produce reliable and reproducible method for fabricating a square polymeric waveguide compatible with NCs.

Photonic analog to digital converters (ADC) have been the focus of much research interest in recent years, because of their potential for very high bandwidth and sampling rates. Using photonic techniques may help to surpass the limitations of traditional electronic analog to digital converters, providing unprecedented performance. A key parameter of any ADC is its conversion resolution. This works explores the technique of spatial oversampling as a means to increase resolution in photonic ADCs. Spatial oversampling is shown to be equivalent to temporal oversampling, a commonly used technique in the field of digital signal processing. The properties, benefits and requirements of spatial oversampling are derived, and the concept is demonstrated theoretically and experimentally. A photonic ADC design based on this technique is described, and an implementation as a photonic integrated circuit is presented. The design is based on electro-optic phase modulation, interferometric detection and spatial oversampling. The abilities and performance of this photonic ADC concept are demonstrated experimentally by digitizing analog signals with frequencies of up to 13GHz.

The simulation and analysis of a temporal soliton perturbation (interaction) with a dispersive truncated Airy pulse traveling in a nonlinear fiber at the same center wavelength (or frequency). True Airy pulses remain self-similar while propagating along a ballistic trajectory. However, they are infinite in energy due to the infinite tail that prevents the energy integral from converging. In order to be realized, Airy pulses must therefore, be truncated. The truncation is carried out by apodizing the infinite Airy tail. Despite the truncation Airy pulses remain self-similar over extended ranges while the ballistic trajectory is completely preserved. This allows them to interact with a nearby soliton on account of the accelerating wavefront property.
The interactions are governed by the Nonlinear Schrödinger equation for which no analytical solution currently exists for these initial conditions. Therefore, numerical simulations are required. The numerical method chosen is the split step Fourier method which is a mathematical algorithm for propagation of the pulses. By providing the simulation program with the initial launch conditions we are able to follow the interactions as they progress.
Analysis of the simulation is carried out by tracking the fundamental parameters of the emergent soliton during propagation—time position, amplitude, phase and frequency—that alter due to the primary collision with the Airy main lobe and the continuous co-propagation with the dispersed Airy background. Following the collision, the soliton intensity oscillates as it relaxes in the dispersed Airy background, trying to settle in to a new soliton state. Further, by varying the initial parameters of the Airy pulse such as initial phase, amplitude and time position, different outcomes are witnessed which allows for a broader understanding of the interaction.
Due to the spectral repositioning of the Airy spectrum by dispersion, the interaction is found to resemble coherent interactions at times and incoherent at others. The results indicate that in certain cases permanent change in frequency and intensity occurs, depending on the configuration of the initial parameters chosen. These changes are made apparent through changes in time position and in the accumulated phase of the soliton. Furthermore, according to the perturbation theory local changes in time position and phase can also occur independently from the frequency change and intensity change, respectively.

The expected permittivity and third order nonlinear susceptibility, of a low filling fraction composite consisting of semiconductor nanorods dispersed in a polymer host are derived, using the Maxwell-Garnett model for anisotropic nonlinear inclusions. The semiconductor nanorods are modeled both as prolate spheroids and more realistic capsule shapes. A new generalized model is presented for various nanorod axis orientation statistics, achieved by an aligning electric field. The angular distribution function of the nanorods is calculated for nanorods with a permanent electrical dipole moment, which assists the alignment of the nanorods. Using the angular distribution function, the composite macroscopic characteristics are found for a composite with random orientation, partially aligned and nematic array nanorods. As the alignment field strength increases, the composite optical properties
asymptotically converge towards the nematic case. Different parameters relate to the nanorods geometry are examined, concluding that the main parameter influencing the alignment is the single NR volume, while for the nematic array the single nanorod axes aspect-ration is the major parameter. Due to the symmetry of the nanorods, the composite characteristics depend on the polarization of the optical electrical field, with a symmetry that resembles a uniaxial crystal.
A nonlinear waveguide with a core made of such a composite is simulated, in order to find the nonlinear parameter of the waveguide. The model takes into account two electrodes for the alignment process, far enough from the waveguide core, in order to avoid losses to the optical mode. Significant optical response can be achieved even for randomly oriented nanorods composite, with a nonlinear parameter of 68(W×m)-1. The alignment process increases the nonlinear parameter significantly even at elevated temperature that are needed for polymerization of the polymer host, typically 150^oC. Aligning field strength of 107 V/m results with very high value for the nonlinear parameter – 120(W×m)^-1, much higher than ordinary glass based nonlinear optical fibers, that result with nonlinear parameter up to 50(W×m)^-1.