A general technique for solving Maxwell's equations exactly, based on expansion of the solution in a complete set of vector basis functions has been developed. These vector eigenfunctions are derived from the complete set of separable solutions to the scalar Helmholtz equation in a particular coordinate system and are shown to form a complete set. The method is applicable to a variety of problems including the study of near and far field electromagnetic scattering from particles with arbitrary shapes, plasmon resonances in spherical nanoparticles with spherically concentric 'shells' and the calculation of plasmon resonances in the sphere-plane geometry. An exact method for solving the inhomogenous Maxwell's equation (i.e., in the presence of charges and currents) is also outlined.

Low frequency noise in giant magnetoresistive spin valves has been studied as a means of optimizing signal to noise ratios and characterizing device performance. The devices studied were sputter deposited NiFe/Cu/NiFe/FeMn spin valves with D R/R ∼ 4%. Static measurements demonstrated a strong dependence of the magnetic coupling and giant magnetoresistance (GMR) ratio on the thickness and quality of the Cu spacer layer and the bottom NiFe layer (free layer). These parameters were varied to determine how the noise in spin valve sensors would be affected. Noise power spectra were measured in patterned spin valves. The noise was observed to have a 1/f slope at low frequencies. The fluctuation-dissipation relation relating thermal fluctuations in magnetization to the resistance fluctuations was used to explain the origin of the 1/f noise. The noise was found to be sensitive to the anisotropy and defect density of the free layer. The noise was minimized for spin valves operating with parallel anisotropy axes and an applied field aligned along the hard axis of magnetization. Dynamic fields were used to measure the Barkhausen noise in the sense layer of the spin valve. The low frequency noise in the presence of dynamic fields was much greater than the 1/f noise background. Clustering of Barkhausen jumps was used to explain the observed dependence of the noise power on the magnitude and frequency of the applied field. Higher frequency signals resulted in lower Barkhausen noise. The noise was reduced when the applied field was aligned along the hard axis of magnetization.

Electromagnetic levitation, which can provide rapid heating and melting, homogeneity of melt and minimal specimen contamination, is an important branch of containerless processing. The longitudinal electromagnetic levitator is a new type of levitation device, which was invented recently and has a potential to become a containerless manufacturing processing tool. It has some unique advantages, such as good visual access to the sample, capability to support multiple samples, large loads and cylindrical shape sample availability. In this thesis, a brief review of the history and application of electromagnetic levitation is presented. Then the detailed theoretical analysis coupled with experimental work validating the theoretical models of the longitudinal electromagnetic levitator are presented. First, a new electric current model is introduced, which is more appropriate for the computation of the electromagnetic force field in the levitated specimen. Based on this new model, the essential equations for the electromagnetic field and the lifting force field for a cylindrical sample are derived, the current density distribution and the averaged power in the sample are analyzed. Additionally, both lifting force and lifting capacity for the longitudinal levitator are investigated analytically, and compared with experimental data with good agreement. These theoretical predictions can be used to design longitudinal levitators, to select suitable material for levitation, and to provide the framework for further investigation of materials processing using the longitudinal levitator. In addition, temperature distribution simulation for the sample levitated in the longitudinal electromagnetic levitator is implemented by analytical and numerical ways. Isothermal case, steady state and lumped system are discussed respectively as some special cases. The exact solution and numerical simulation of the temperature distribution for the levitated sample are compared with good agreement. The flow motion within the levitated sample and the numerical simulation of the temperature distribution with flow convection has also been investigated. The results provide important information of the levitation phenomena which are very useful for scientific and engineering applications, especially for materials processing.

The plasmon resonance in metallic nanoshells can be used to efficiently harvest solar energy and convert it into thermal or electronic form. Possible applications include improved optical coupling into silicon photodiodes, solar water heaters, and photocatalysis. We use standard optimization algorithms to theoretically determine the best mixture of different nanoshell species ([core, shell] sizes) for two practical scenarios. We show that a mixture of nanoshell species [r1, r2] = [47, 58] nm and [r1, r2] = [28, 42] run in a 6:5 volume ratio is optimal for absorbing AM 1.5 sunlight when deposited on a silicon surface. Surprisingly, we find that a single particle species is very good for scattering AM 1.5 light on a glass surface, and that very little benefit is gained by mixing different shells. Assumptions and approximations made in the analysis are discussed.

The explosive growth in the information technology industry requires high-performance optical sources. In recent years, wavelength-tunable optical pulse sources are of interest for applications in optical instrumentation, communications, and sensing. This thesis demonstrates and analyzes the generation of wavelength tunable, picosecond pulses from mode-locked semiconductor fiber ring lasers. One structure using an intracavity electro-optic modulator and the other an injected optical control signal, are investigated and experimentally characterized. A single or superimposed linearly chirped fiber Bragg gratings are used to provide wavelength selectivity, tunability, and multi-wavelength operation. The semiconductor optical amplifier as the gain media makes it possible to obtain stable simultaneous oscillation of several wavelengths at any wavelength band with very small channel spacing. We have successfully generated picosecond pulses at one or two wavelengths over the reflection bandwidth(s) of the grating(s) by simply changing the modulation frequency.

A thorough investigation of a pulsed plasma x-ray source is presented with the intent of improving its design and maximizing its x-ray emission efficiency. In this approach a hollow gas column is puffed in the z-pinch diode by magnetically ionizing and compressing an inert gas in an annular plenum. This gas column is preionized by the radiation coming out from the plasma in the plenum and pinched by a fast electrical discharge. / It is shown that the system can be improved considerably by modifying the gas puff design. Three gas puffs developed in this work are optimized for x-ray emission from argon, krypton, and neon gases. In the optimized conditions the output x-ray energies of 0.5 J from Ar-K shell, 2 J from Kr-L shell, and more than 2 J from Ne-K shell are obtained. / The implosion dynamics is studied with different gases under varying conditions. The average implosion velocity, the final pinch diameter, the current waveform, and the emitted x-ray energy are measured. The pinched plasma parameters such as temperature, density, and the average ionic state are estimated using the corona model calculations, and the pinched current waveforms. The spectrum of the neon radiation clearly shows the characteristic H-like and He-like lines. The neon spectrum is used to estimate the plasma temperature. / The dynamic performance of the magnetically induced compression gas puff is studied carefully. A special ion probe was developed for studying the dynamic parameters of the gas puff. The ion measurements with the probe have led to the characterization of the gas puff performance under varying operating conditions. It is shown that ions are generated through photoionization of the injected gas by the UV light emitted from the inside of the gas puff plenum through the nozzle. It is found that the jet velocity and ion density can be in excess of $3 times10 sp3$ m/s and $2 times10 sp{14}$ cm$ sp{-3}$, respectively. / A theoretical model is developed to simulate the plasma evolution in the gas puff. This model uses the magneto hydrodynamic (MHD) equations solved by the finite difference method. The magnetic field in the vacuum is calculated using the Laplace equation and self consistent boundary conditions. The model predicts the evolution of plasma variables such as density, temperature, velocity, and magnetic field. It also calculates the variation of the total mass flow rate, optical output, and the ionic signal. The simulation results are shown to compare favourably with the experimental measurements.

As the demand for wireless communications increases, high speed and low cost electronics are desired. Traditionally, RF circuits are implemented using high performance technologies such as GaAs or SiGe in order to minimize noise and achieve high gain. However, those high performance processes are incompatible with mainstream digital circuitry, which are usually implemented in CMOS technologies. / In this thesis, an RF receiver frontend which consists of a differential low noise amplifier, active mixers, passive mixers, and a quadrature voltage-controlled oscillator, for 5 GHz applications are designed and manufactured in a digital CMOS process, in order to demonstrate the RF potential of CMOS processes. We explore the use of simple circuit topologies and common packaging to build CMOS receivers that can operate from IV supplies and lower, while providing reasonable image rejection without the use of any special image rejection filters. In addition, a high image rejection receiver is explored after designing it based on good and simple frequency planning. / Apart from inductors and capacitors, transformers are very useful passive components in RF applications. For example, transformers can act as on-chip single-ended to differential converters. In this thesis, a comprehensive study of transformer modeling is presented and discussed. A modeling program has been developed, and its accuracy verified through measurements of several transformer prototypes fabricated in a variety of state-of-the-art CMOS technologies. The program allows the generation of compact lumped transformer models to be used in circuit simulation.

The speed of the Finite Element Method (FEM) is an obstacle to the fast calculation of magnetic fields. A fast Local Refinement Method (LRM) using the first-order FEM is presented for quickly tracking the magnetic field changes while electromagnetic models have small changes made to their shape. This method resolves the potentials in the local mesh or submesh extracted from the whole mesh, with a boundary condition that is calculated by the initial solution based on the whole mesh. Instead of being re-meshed in the local area, the extracted submesh is coarsened and reshaped by the LRM to speed up the calculation time by sharply decreasing the time used for building the S matrix and solving the matrix equation Ax = b. The new potentials in the submesh are, with an acceptable error, embedded back into the whole problem to update the magnetic fields which provide designers or users with a fast visual feedback to their adjustment.

Crystals of Cu(In1-xGax)3Se 5 were grown by the horizontal and vertical Bridgman methods. A non-contact carbon coating was used to avoid the adhesion between Cu(In1-x Gax)3Se5 ingots and the inner ampoule walls. The composition along and across the as grown ingots with different starting Ga contents was analyzed and the results were interpreted by the established pseudobinary phase diagrams. Results of XRD confirmed that the lattice constants of the Cu(In1-xGax)3Se 5 crystals varied linearly with the Ga content. Results of X-ray Laue back-reflection showed that the Cu(In1-xGax) 3Se5 ingots contained large single crystal regions. Hall effect measurements carried out on the grown samples revealed that the Cu(In 1-xGax)3Se5 crystals were highly resistive with rather low carrier concentrations. The morphology of as-grown or cleaved sample surfaces of the Cu(In1-xGax) 3Se5 ingots were also studied under optical microscope and SEM. / Metallic Na was, for the first time, introduced into Cu(In1-x Gax)3Se5 compounds to observe the doping effects. The introduction of Na increased the electron concentration significantly for CuIn3Se5 samples (x = 0) but did not show a significant effect on Cu(In1-xGax) 3Se5 samples with x > 0. The increase in electron concentration in the CuIn3Se5 samples after the Na diffusion could be explained by defect generation related to Se and In sites. / Crystals of CuInSe2 were also grown by the horizontal Bridgman method for the first time with the addition of metallic sodium. Degradation in crystalline quality and a change of conductivity type from p- to n-type were observed in ingots grown from melts containing more than 0.25 at. % Na. Experiments of Na diffusion were also carried out on CuInSe 2 crystals in a sealed glass ampoule to observe the doping effect. Hot probe measurements indicated that the sodium-treated CuInSe2 samples changed from p-type to n-type. / MIS devices were fabricated on the Na-treated CuIn3Se 5 material for electrical characterization. Dark current density-voltage characteristics and differential capacitance-voltage characteristics of the MIS devices were measured at room temperature. An energy band diagram of the MIS devices has been constructed based on the band lineup data reported in the literature. The current transport mechanism was examined and a dominant multi-step tunneling process was proposed. / Samples of Cu(In1-xGax)3Se 5 with x ≤ 0.5 were found to be strongly photoconductive over the wavelength range from 700 to 1100 nm even at room temperature. It was observed that the sensitivity of photoconductivity was greatly influenced by surface preparation conditions. Chemically etched samples showed the highest photoconductivity, believed to be due to the reduced surface recombination velocity. / Capacitance measurements were carried out to investigate the interface and bulk properties of ZnO/CdS/Cu(In1-xGax)Se 2 solar cells. Results from the steady state C-V measurements showed evidence of interface or surface states, especially for the samples without annealing. DLTS technique was used to determine the deep levels in Cu(In 1-xGax)Se2 crystals with Ga content varying from 0 to 1. Different deep levels for holes with different DLTS spectra were found in the Cu(In1-xGax)Se2 crystals, with different Ga contents. The present results showed that the Ga content has an important effect on the formation of deep levels in Cu(In1-x Gax)Se2 crystals.

A robust method has been developed for the test and characterization of analog and mixed-signal integrated circuits. The method relies on a compact, robust, and easily synthesized integrated test core capable of emulating the function of external automatic test equipment. The core consists of a 2 x N memory whose contents are periodically circulated, a coarse analog filter, and a voltage comparator. One half of the circular memory is used to generate analog signals without the need for multi-bit digital-to-analog converters. The second half is used to generate extremely accurate DC levels, the latter being programmed using a clever software encoding technique that relies on some form of sigma-delta modulation. The DC levels, in combination with the comparator, enable multi-bit digitization using a progressive multiple conversion pass procedure. In order to accommodate broadband circuit phenomena, a delayed-clock sub-sampling mechanism is also employed, in which the digitizer sample clock is consistently delayed over multiple runs of the periodic test signal. One method of delaying the clock is to use a voltage-controlled delay line tuned by a delay-locked loop. The timing resolution of this approach is determined by the value of the consistent clock delay and not its period. / A divide-and-conquer approach to the test of deeply embedded analog integrated circuits using the proposed test core is described. Multiple test configurations are presented that can span a wide range of phenomena to be tested both internally to the integrated circuit and externally through I/O interfaces. The applicability of these configurations to increasing test parallelism both at the core and die levels is investigated. Performance limits of the proposed test core are also derived by drawing a comparison to conventional circuits used for data-conversion applications. The same fundamental limitations on integrated circuit performance are shown to affect the test core electronics, although test-specific requirements, such as forcing periodicity and the reliance on software signal processing, help further enhance on-chip measurement accuracy and repeatability. Finally, several successful experimental prototypes that demonstrate the viability of the proposed approach are presented. The prototypes range from concept proving test core integrated circuits to ones containing multiple simultaneously operated test cores and completely embedded circuits under test. In total, several hundred different test cores have been demonstrated, which is further testimony to the practicality of the proposed techniques.