Diamond is a material with many exceptional properties. In this thesis methods for fabrication of microstructures as well as several applications of such structures in optics, microfluidics and electrochemistry are presented. A method for etching deep and highly precise gratings is described. This method was used to fabricate circularly symmetric half wave plates for use in vector vortex coronagraphs. Such coronagraphs are a very promising approach to the direct imaging of extrasolar planets. By varying the lateral etch rate of the aluminum mask during diamond etching in an inductively coupled plasma, the sidewall angle of the etched structures could be controlled. This method was used to make smooth sloped sides on a waveguide for coupling light into it. Antireflective structures that drastically reduced the surface reflection in a wavelength band between 10 and 50 µm were also fabricated. An array of boron doped diamond microelectrodes for electrochemical measurements in a microchannel was fabricated and tested, showing very good stability and reusability. Several hundred hours of use did not adversely affect their performance and no damage to them could be detected by atomic force microscopy or scanning electron microscopy. Superhydrophobic surfaces in diamond were demonstrated, using both hydrogen and fluorine termination. Hydrogen termination on a flat surface gives contact angles below 90°. To achieve a superhydrophobic surface with this low intrinsic hydrophobicity, structures looking like microscopic nail heads were fabricated. The effect of water pressure on immersed superhydrophobic surfaces was also studied and it was found that the collapse of the superhydrophobic state due to pressure was sometimes reversible as the pressure was lowered. Finally, a method was tested for functionalizing diamond surfaces using block copolymers of polyethylene oxide and polypropylene oxide to both passivate the surface and to attach synthetic binder molecules. This method was found to give very high signal to noise ratios when detecting C-reactive protein.

Propagation of radiation in a waveguides is theoretically described by Maxwell&amp / #8217 / s equations. The gradient of refractive index and an influence on the waveguide by a superstrate requires a numerical solution of the differential equation. Iterative methods such as the Runge-Kutta approaches are used to calculate the effective refractive index in the waveguide depending on the superstrate&amp / #8217 / s and the waveguide&amp / #8217 / s local refractive indices. In this study,the refractive indices, and the model fields of the TE00 modes are calculated. The calculated fields of the 00 TE modes give information about the propagation of the light in the waveguide. Also, the precision of the Runge-Kutta aproaches has been tested. The advantages and disadvantages of the Runge-Kutta aproaches are discussed.

No description available.

Microwave communication devices have become ubiquitous in the past decade. As an increasing number of systems compete for spectrum, guard bands have shrunk to increase bandwidth efficiency. The frequency behaviour of microwave devices is affected by thermal expansion. In order to avoid interference with adjacent bands, microwave components must exhibit high temperature-stability in most communications applications. Thermally stable materials can be used to construct temperature-stable components. However, this approach requires an expensive mass and cost trade-off. Temperature compensated aluminum resonators and filters provide major advantages in cost and mass. This work proposes that a compensating tuning screw with a temperature-dependent effective length be constructed by mounting a bimetallic compensator at the end of a mounting screw. This so-called bimetal tuning-screw can be used to produce temperature-compensated resonators and filters. There are several advantages to this approach. Compensation can be tuned by adjusting the depth of the bimetal, simply by adjusting the mounting screw. Since there are no moving parts inside the cavity or filter, and the bimetal can be plated, there are no additional sources of passive intermodulation. Also, this design is simple to implement for waveguide designs in general. In order to compensate for temperature drift, it is useful to quantify uncompensated drift. Temperature drift for a lossless linearly expanding RF component is derived from Maxwell's equations. For the lossy case, it is demonstrated that the resulting formula is approximately true, and that the quality of this approximation is excellent for practical levels of temperature range and thermal expansion. Experimental results are provided that demonstrate bimetal compensation under uniform-temperature conditions for a single aluminum resonator. Measured drift of the compensated resonator is -0.38 ppm/°C, compared to -23 ppm/°C for an uncompensated resonator. Measured drift for a bimetal-compensated 4-pole filter prototype is 2.35 ppm/°C. A method for adjusting compensation for a filter is also provided. Multiphysics simulations are used to examine power handling for bimetal-compensated filters. It is demonstrated that power-handling can be improved by reducing the effective length of the compensator to improve heat conduction to the cavity or filter.

Nobel metal nanoparticles that exhibit plasmon resonances in the visible and near infrared have been of great interest in recent years. Strong light-matter interactions on the nanoscale have a range of interesting properties that may be useful in applications in medicine, sensing, solar energy harvesting and information processing. Depending on the application, particle materials and geometries can be optimized for performance. A novel method of quantifying individual nanoparticle scattering cross-sections by comparing experiments with analytical theory for gold nanospheres is proposed and utilized. Results show that elongated particles scatter very brightly for their volumes. This brightness is due to a strong longitudinal plasmon resonance that occurs in the near infrared – where gold has minimal loss. Elongated particles, such as nanorods, are therefore, ideal for applications that rely on particles scattering brightly in small spaces, such as biological imaging. Next, gold nanobelts are discussed and characterized. These novel structures are akin to nanowires, but with a small, rectangular cross-sectional geometry. Gold nanobelts are shown to exhibit a strong transverse resonance that has never been reported previously in nanowires. The transverse resonance is shown to shift linearly with crosssectional aspect ratio. Other interesting products from the nanobelt synthesis, tapered and split nanobelts, are discussed. Gold nanobelts also support longitudinal propagating plasmons, and have the smallest cross-sectional area of any elongated plasmonic structure that has been reported to do so. By analyzing the output tip signal of propagating plasmons for nanobelts of different lengths, the decay length is measured. Finite Difference Time Domain simulations and polarization measurements show the fundamental, azimuthally symmetric mode is very strong for thin structures such as these, but decays much more quickly than a higher-order mode, which begins to dominate at longer lengths. The cross-sectional mode area is given, illustrating the high confinement of plasmons in these structures. A figure of merit that takes into account both confinement and propagation length is calculated to be 1300 for the higher-order mode, the highest reported for nanoscale plasmonic waveguides. The high figure of merit makes gold nanobelts excellent candidates for studying strong coupling between plasmonic structures and objects that exhibit quantum behavior.

Microwave communication devices have become ubiquitous in the past decade. As an increasing number of systems compete for spectrum, guard bands have shrunk to increase bandwidth efficiency. The frequency behaviour of microwave devices is affected by thermal expansion. In order to avoid interference with adjacent bands, microwave components must exhibit high temperature-stability in most communications applications. Thermally stable materials can be used to construct temperature-stable components. However, this approach requires an expensive mass and cost trade-off. Temperature compensated aluminum resonators and filters provide major advantages in cost and mass. This work proposes that a compensating tuning screw with a temperature-dependent effective length be constructed by mounting a bimetallic compensator at the end of a mounting screw. This so-called bimetal tuning-screw can be used to produce temperature-compensated resonators and filters. There are several advantages to this approach. Compensation can be tuned by adjusting the depth of the bimetal, simply by adjusting the mounting screw. Since there are no moving parts inside the cavity or filter, and the bimetal can be plated, there are no additional sources of passive intermodulation. Also, this design is simple to implement for waveguide designs in general. In order to compensate for temperature drift, it is useful to quantify uncompensated drift. Temperature drift for a lossless linearly expanding RF component is derived from Maxwell's equations. For the lossy case, it is demonstrated that the resulting formula is approximately true, and that the quality of this approximation is excellent for practical levels of temperature range and thermal expansion. Experimental results are provided that demonstrate bimetal compensation under uniform-temperature conditions for a single aluminum resonator. Measured drift of the compensated resonator is -0.38 ppm/°C, compared to -23 ppm/°C for an uncompensated resonator. Measured drift for a bimetal-compensated 4-pole filter prototype is 2.35 ppm/°C. A method for adjusting compensation for a filter is also provided. Multiphysics simulations are used to examine power handling for bimetal-compensated filters. It is demonstrated that power-handling can be improved by reducing the effective length of the compensator to improve heat conduction to the cavity or filter.

This thesis reports on a new design of photonic crystal waveguides (PCWs) to achieve large guiding bandwidth, linear dispersion, single-mode behavior, good coupling efficiency to dielectric waveguides, and small loss. The design is based on using the linear dispersion region of one PCW in the photonic bandgap (PBG) of another PCW. While perturbing the period can result in a PCW with linear dispersion and large guiding bandwidth, it introduces an odd mode at those frequencies, as well. By using another perturbation scheme, it is shown that single-mode behavior can also be achieved. The linear dispersion of these waveguides and their operation at lower frequencies of the PBG, where the density of states of radiation modes is smaller, gives rise to very small loss coefficients as verified experimentally. Full characterization of a waveguide requires the measurement of not only the transmission coefficient, but also the dispersion and spectral phase. We have developed a real-time characterization technique based on spectral interferometry with femtosecond laser pulses at optical communication wavelengths to measure the spectral phase of waveguides. This haracterization technique can be used to study fast dynamics in timevarying structures and makes the alignment easy.

Selectively erbium (Er) doped titanium (Ti) in-diffused optical waveguide amplifiers on lithium niobate (LiNbO3) substrate have been fabricated and characterized in the wavelength regime around λ = 1.53μm using counter-directional pumping at λP = 1.48μm. LiNbO3 waveguide amplifiers are desirable for providing gain in optical circuit chips through integration with other optical elements on a single substrate. A prerequisite for achieving useful gain rests on the optimization of overlap between the incident guided optical signal mode distribution and the evolving emission from excited Er ions. The extent of overlap can be controlled by adjusting fabrication parameters. Fabrication parameters for Er-doped Ti in-diffused waveguide amplifiers of useful optical gain have been optimized by diffusing selective patterns of vacuum-deposited 17nm-thick erbium film at 1100˚C for 100 hours into LiNbO3, and integrating with 7μm-wide single mode straight channel waveguides formed by diffusing 1070Å thick titanium film into the LiNbO3. Small-signal gain characterization was carried out with a -30 dBm of transmitted input signal power at λS=1531nm with counter-directionally launched pump power ranging between 0 to 119mW at λP=1488nm, using TM polarization for both the signal and pump beams. At a maximum launched pump power of 119mW, a signal enhancement of 8.8dBm for 25mm-long erbium doped region, and 11.6dBm for 35mm-long erbium doped region were obtained. The corresponding calculated net gain values are 1.8dB and 2.8dB, for the 25mm-long and 35mm-long Er-doped regions, respectively.

The dielectric ridge waveguide is an important passive component used in the optical communication system. Compared to its cousins¡Xthe buried ridge waveguides, it is less expensive to process but harder to design due to its inherent complex field structure (it has a larger index contrast between the core and the cladding). Therefore, it is crucial to develop an efficient and accurate method to analyze the modal characteristic of ridge waveguides. We began with the expansion of one-dimensional three-layer dielectric slab waveguide using simple orthogonal basis functions. We examined both the step-index profile and the graded-index profile waveguides to confirm the feasibility of this method and to understand the level of accuracy our technique can reach. We then proceeded to derive the vector formulation of two-dimension dielectric waveguides and compared our results against the exact numerical solutions of optical fiber modes using Bessel functions. Our 2-D Cartesian mode solver gave up to 6 significant digits of the fiber propagation constants. Finally, for rectangular dielectric waveguides, we use the tensor product of 1-D guiding-mode bases as an improvement over the simple orthogonal bases to speed-up the numerical convergence and cut the storage requirement by a factor of ten without loss of accuracy which is around 4 to 5 digits over a wide range of parameters and mode types. We will use these bases to solve for the mode field distribution of ridge-waveguides and other complex structures.

Hybrid waveguides based on antiresonant reflecting optical waveguide (ARROW) structure on Si substrates is investigated. The core layer of the waveguide is separated from the Si substrate by interference cladding which consists of a high index first cladding layer and a low index second cladding layer. The Ta2O5 first cladding layer was grown by rf magnetron sputtering system. The SiON second cladding layer was deposited by plasma-enhanced chemical vapor deposition (PECVD) based on the reaction of SiH4/N2O mixtures. Typical propagation losses of the waveguides using this SiON material system are less than 0.15 dB/cm. The high quality cladding layers are prepared to form the Fabry-Perot cavities to ensure low loss operation of the ARROW device at antiresonant conditions. Two methods were proposed to demonstrate the enormous applications of hybrid ARROW waveguides. First, polyimide/Ta2O5/SiON ARROWs were fabricated to achieve high extinction ratio waveguide polarizers. This is accomplished by tuning the operating point of the TM0 mode from antiresonant condition toward high-order antiresonance of the second cladding layer utilizing birefringence of the waveguide core. The measured extinction ratio and insertion loss of a 2-cm-long ARROW polarizer are 40 dB and 2.2 dB, respectively. In addition, the temperature dependence of refractive index of organic-inorganic sol-gel glasses was measured by Mach-Zehnder interferometry using the ARROW waveguides. The Sol-gel/Ta2O5/SiO2 ARROWs were fabricated to characterize index-to-temperature coefficients (dn/dT) of the sol-gel glasses because the optical confinement factor of the ARROW is very close to 1. The measured index-to-temperature coefficients of the sol-gel glasses with different compositions are negative and are on the order of 10-4.