In this thesis, a novel silica nanosphere (SNS) lithography technique has been developed to offer a fast, cost-effective, and large area applicable nano-lithography approach. The SNS can be easily deposited with a simple spin-coating process after introducing a N,N-dimethyl-formamide (DMF) solvent which can produce a highly close packed SNS monolayer over large silicon (Si) surface area, since DMF offers greatly improved wetting, capillary and convective forces in addition to slow solvent evaporation rate. Since the period and dimension of the surface pattern can be conveniently changed and controlled by introducing a desired size of SNS, and additional SNS size reduction with dry etching process, using SNS for lithography provides a highly effective nano-lithography approach for periodically arrayed nano-/micro-scale surface patterns with a desired dimension and period. Various Si nanostructures (i.e., nanopillar, nanotip, inverted pyramid, nanohole) are successfully fabricated with the SNS nano-lithography technique by using different etching technique like anisotropic alkaline solution (i.e., KOH) etching, reactive-ion etching (RIE), and metal-assisted chemical etching (MaCE).

In this research, computational optical modeling is also introduced to design the Si nanostructure, specifically nanopillars (NPs) with a desired period and dimension. The optical properties of Si NP are calculated with two different optical modeling techniques, which are the rigorous coupled wave analysis (RCWA) and finite-difference time-domain (FDTD) methods. By using these two different optical modeling techniques, the optical properties of Si NPs with different periods and dimensions have been investigated to design ideal Si NP which can be potentially used for thin c-Si solar cell applications. From the results of the computational and experimental work, it was observed that low aspect ratio Si NPs fabricated in a periodic hexagonal array can provide highly enhanced light absorption for the target spectral range (600 ~ 1100nm), which is attributed to (1) the effective confinement of resonant scattering within the Si NP and (2) increased high order diffraction of transmitted light providing an extended absorption length. From the research, therefore, it is successfully demonstrated that the nano-fabrication process with SNS lithography can offer enhanced lithographical accuracy to fabricate desired Si nanostructures which can realize enhanced light absorption for thin Si solar cell.

From 2D planar MOSFET to 3D FinFET, the geometry of semiconductor devices is getting more and more complex. Correspondingly, the number of mesh grid points increases largely to maintain the accuracy of carrier transport and heat transfer simulations. By substituting the conventional uniform mesh with non-uniform mesh, one can reduce the number of grid points. However, the problem of how to solve governing equations on non-uniform mesh is then imposed to the numerical solver. Moreover, if a device simulator is integrated into a multi-scale simulator, the problem size will be further increased. Consequently, there exist two challenges for the current numerical solver. One is to increase the functionality to accommodate non-uniform mesh. The other is to solve governing physical equations fast and accurately on a large number of mesh grid points.

This research rst discusses a 2D planar MOSFET simulator and its numerical solver, pointing out its performance limit. By analyzing the algorithm complexity, Multigrid method is proposed to replace conventional Successive-Over-Relaxation method in a numerical solver. A variety of Multigrid methods (standard Multigrid, Algebraic Multigrid, Full Approximation Scheme, and Full Multigrid) are discussed and implemented. Their properties are examined through a set of numerical experiments. Finally, Algebraic Multigrid, Full Approximation Scheme and Full Multigrid are integrated into one advanced numerical solver based on the exact requirements of a semiconductor device simulator. A 2D MOSFET device is used to benchmark the performance, showing that the advanced Multigrid method has higher speed, accuracy and robustness.

Crystalline silicon has a relatively low absorption coefficient, and therefore, in thin silicon solar cells surface texturization plays a vital role in enhancing light absorption. Texturization is needed to increase the path length of light through the active absorbing layer. The most popular choice for surface texturization of crystalline silicon is the anisotropic wet-etching that yields pyramid-like structures. These structures have shown to be both simple to fabricate and efficient in increasing the path length; they outperform most competing surface texture. Recent studies have also shown these pyramid-like structures are not truly square-based 54.7 degree pyramids but have variable base angles and shapes. In addition, their distribution is not regular -- as is often assumed in optical models -- but random. For accurate prediction of performance of silicon solar cells, it is important to investigate the true nature of the surface texture that is achieved using anisotropic wet-etching, and its impact on light trapping. We have used atomic force microscopy (AFM) to characterize the surface topology by obtaining actual height maps that serve as input to ray tracing software. The height map also yields the base angle distribution, which is compared to the base angle distribution obtained by analyzing the angular reflectance distribution measured by spectrophotometer to validate the shape of the structures. Further validation of the measured AFM maps is done by performing pyramid density comparison with SEM micrograph of the texture. Last method employed for validation is Focused Ion Beam (FIB) that is used to mill the long section of pyramids to reveal their profile and so from that the base angle distribution is measured. After that the measured map is modified and the maps are generated keeping the positional randomness (the positions of pyramids) and height of the pyramids the same, but changing their base angles. In the end a ray tracing software is used to compare the actual measured AFM map and also the modified maps using their reflectance, transmittance, angular scattering and most importantly path length enhancement, absorbance and short circuit current with lambertian scatterer.

Increasing the conversion efficiencies of photovoltaic (PV) cells beyond the single junction theoretical limit is the driving force behind much of third generation solar cell research. Over the last half century, the experimental conversion efficiency of both single junction and tandem solar cells has plateaued as manufacturers and researchers have optimized various materials and structures. While existing materials and technologies have remarkably good conversion efficiencies, they are approaching their own limits. For example, tandem solar cells are currently well developed commercially but further improvements through increasing the number of junctions struggle with various issues related to material interfacial defects. Thus, there is a need for novel theoretical and experimental approaches leading to new third generation cell structures. Multiple exciton generation (MEG) and intermediate band (IB) solar cells have been proposed as third generation alternatives and theoretical modeling suggests they can surpass the detailed balance efficiency limits of single junction and tandem solar cells. MEG or IB solar cell has a variety of advantages enabling the use of low bandgap materials. Integrating MEG and IB with other cell types to make novel solar cells (such as MEG with tandem, IB with tandem or MEG with IB) potentially offers improvements by employing multi-physics effects in one device. This hybrid solar cell should improve the properties of conventional solar cells with a reduced number of junction, increased light-generated current and extended material selections. These multi-physics effects in hybrid solar cells can be achieved through the use of nanostructures taking advantage of the carrier confinement while using existing solar cell materials with excellent characteristics. This reduces the additional cost to develop novel materials and structures. In this dissertation, the author develops thermodynamic models for several novel types of solar cells and uses these models to optimize and compare their properties to those of existing PV cells. The results demonstrate multiple advantages from combining MEG and IB technology with existing solar cell structures.

GaAs-based solar cells have attracted much interest because of their high conversion efficiencies of ~28% under one sun illumination. The main carrier recombination mechanisms in the GaAs-based solar cells are surface recombination, radiative recombination and non-radiative recombination. Photon recycling reduces the effect of radiative recombination and is an approach to obtain the device performance described by detailed balance theory. The photon recycling model has been developed and was applied to investigate the loss mechanisms in the state-of-the-art GaAs-based solar cell structures using PC1D software. A standard fabrication process of the GaAs-based solar cells is as follows: wafer preparation, individual cell isolation by mesa, n- and p-type metallization, rapid thermal annealing (RTA), cap layer etching, and anti-reflection coating (ARC). The growth rate for GaAs-based materials is one of critical factors to determine the cost for the growth of GaAs-based solar cells. The cost for fabricating GaAs-based solar cells can be reduced if the growth rate is increased without degrading the crystalline quality. The solar cell wafers grown at different growth rates of 14 μm/hour and 55 μm/hour were discussed in this work. The structural properties of the wafers were characterized by X-ray diffraction (XRD) to identify the crystalline quality, and then the as-grown wafers were fabricated into solar cell devices under the same process conditions. The optical and electrical properties such as surface reflection, external quantum efficiency (EQE), dark I-V, Suns-Voc, and illuminated I-V under one sun using a solar simulator were measured to compare the performances of the solar cells with different growth rates. Some simulations in PC1D have been demonstrated to investigate the reasons of the different device performances between fast growth and slow growth structures. A further analysis of the minority carrier lifetime is needed to investigate into the difference in device performances.

New technologies enable the exploration of space, high-fidelity defense systems, lighting fast intercontinental communication systems as well as medical technologies that extend and improve patient lives. The basis for these technologies is high reliability electronics devised to meet stringent design goals and to operate consistently for many years deployed in the field. An on-going concern for engineers is the consequences of ionizing radiation exposure, specifically total dose effects. For many of the different applications, there is a likelihood of exposure to radiation, which can result in device degradation and potentially failure. While the total dose effects and the resulting degradation are a well-studied field and methodologies to help mitigate degradation have been developed, there is still a need for simulation techniques to help designers understand total dose effects within their design. To that end, the work presented here details simulation techniques to analyze as well as predict the total dose response of a circuit. In this dissertation the total dose effects are broken into two sub-categories, intra-device and inter-device effects in CMOS technology. Intra-device effects degrade the performance of both n-channel and p-channel transistors, while inter-device effects result in loss of device isolation. In this work, multiple case studies are presented for which total dose degradation is of concern. Through the simulation techniques, the individual device and circuit responses are modeled post-irradiation. The use of these simulation techniques by circuit designers allow predictive simulation of total dose effects, allowing focused design changes to be implemented to increase radiation tolerance of high reliability electronics.

The goal of this research work is to develop a particle-based device simulator for modeling strained silicon devices. Two separate modules had to be developed for that purpose: A generic bulk Monte Carlo simulation code which in the long-time limit solves the Boltzmann transport equation for electrons; and an extension to this code that solves for the bulk properties of strained silicon. One scattering table is needed for conventional silicon, whereas, because of the strain breaking the symmetry of the system, three scattering tables are needed for modeling strained silicon material. Simulation results for the average drift velocity and the average electron energy are in close agreement with published data. A Monte Carlo device simulation tool has also been employed to integrate the effects of self-heating into device simulation for Silicon on Insulator devices. The effects of different types of materials for buried oxide layers have been studied. Sapphire, Aluminum Nitride (AlN), Silicon dioxide (SiO2) and Diamond have been used as target materials of interest in the analysis and the effects of varying insulator layer thickness have also been investigated. It was observed that although AlN exhibits the best isothermal behavior, diamond is the best choice when thermal effects are accounted for.

GaN high electron mobility transistors (HEMTs) based on the III-V nitride material system have been under extensive investigation because of their superb performance as high power RF devices. Two dimensional electron gas(2-DEG) with charge density ten times higher than that of GaAs-based HEMT and mobility much higher than Si enables a low on-resistance required for RF devices. Self-heating issues with GaN HEMT and lack of understanding of various phenomena are hindering their widespread commercial development. There is a need to understand device operation by developing a model which could be used to optimize electrical and thermal characteristics of GaN HEMT design for high power and high frequency operation. In this thesis work a physical simulation model of AlGaN/GaN HEMT is developed using commercially available software ATLAS from SILVACO Int. based on the energy balance/hydrodynamic carrier transport equations. The model is calibrated against experimental data. Transfer and output characteristics are the key focus in the analysis along with saturation drain current. The resultant IV curves showed a close correspondence with experimental results. Various combinations of electron mobility, velocity saturation, momentum and energy relaxation times and gate work functions were attempted to improve IV curve correlation. Thermal effects were also investigated to get a better understanding on the role of self-heating effects on the electrical characteristics of GaN HEMTs. The temperature profiles across the device were observed. Hot spots were found along the channel in the gate-drain spacing. These preliminary results indicate that the thermal effects do have an impact on the electrical device characteristics at large biases even though the amount of self-heating is underestimated with respect to thermal particle-based simulations that solve the energy balance equations for acoustic and optical phonons as well (thus take proper account of the formation of the hot-spot). The decrease in drain current is due to decrease in saturation carrier velocity. The necessity of including hydrodynamic/energy balance transport models for accurate simulations is demonstrated. Possible ways for improving model accuracy are discussed in conjunction with future research.

A proposed visible spectrum nanoscale imaging method requires material with permittivity values much larger than those available in real world materials to shrink the visible wavelength to attain the desired resolution. It has been proposed that the extraordinarily slow propagation experienced by light guided along plasmon resonant structures is a viable approach to obtaining these short wavelengths. To assess the feasibility of such a system, an effective medium model of a chain of Noble metal plasmonic nanospheres is developed, leading to a straightforward calculation of the waveguiding properties. Evaluation of other models for such structures that have appeared in the literature, including an eigenvalue problem nearest neighbor approximation, a multi- neighbor approximation with retardation, and a method-of-moments method for a finite chain, show conflicting expectations of such a structure. In particular, recent publications suggest the possibility of regions of invalidity for eigenvalue problem solutions that are considered far below the onset of guidance, and for solutions that assume the loss is low enough to justify perturbation approximations. Even the published method-of-moments approach suffers from an unjustified assumption in the original interpretation, leading to overly optimistic estimations of the attenuation of the plasmon guided wave. In this work it is shown that the method of moments approach solution was dominated by the radiation from the source dipole, and not the waveguiding behavior claimed. If this dipolar radiation is removed the remaining fields ought to contain the desired guided wave information. Using a Prony's-method-based algorithm the dispersion properties of the chain of spheres are assessed at two frequencies, and shown to be dramatically different from the optimistic expectations in much of the literature. A reliable alternative to these models is to replace the chain of spheres with an effective medium model, thus mapping the chain problem into the well-known problem of the dielectric rod. The solution of the Green function problem for excitation of the symmetric longitudinal mode (TM01) is performed by numerical integration. Using this method the frequency ranges over which the rod guides and the associated attenuation are clearly seen. The effective medium model readily allows for variation of the sphere size and separation, and can be taken to the limit where instead of a chain of spheres we have a solid Noble metal rod. This latter case turns out to be the optimal for minimizing the attenuation of the guided wave. Future work is proposed to simulate the chain of photonic nanospheres and the nanowire using finite-difference time-domain to verify observed guided behavior in the Green's function method devised in this thesis and to simulate the proposed nanosensing devices.