FePt alloys with the CuAuI L10-ordered structure are important magnetic materials since their large uniaxial magnetocrystalline anisotropy can overcome the thermal stability problem of high-density magnetic recording media. In this study, transmission electron microscopy (TEM) and energy-filtered TEM (EFTEM) are employed to characterize the microstructure of a series of FePt/MgO multilayered thin films with different FePt layer thicknesses. The as-deposited FePt layers transform into the L10-ordered phase after annealing at 700oC for 30 minutes. The laminated FePt/MgO structure effectively preserves a strong L10 [001] texture normal to the thin film plane during annealing. The annealed films exhibit large coercivities from 6.5 kOe to 11.5 kOe due to the formation of the L10-ordered phase. A reduction in coercivity with increasing FePt layer thickness is attributed to the differences in the microstructure after annealing. For the films with thinner FePt layers, annealing produces discontinuous FePt layers that reduce exchange coupling and increase coercivity. As the FePt layer thickness increases, the magnetic layers become more continuous after annealing and thereby the exchange coupled layers have lower coercivity.

Semi-classical transport models based on Boltzmann and Fermi-Dirac statistics have been very effective in identifying the pertinent physical parameters responsible for thermoelectric performance in bulk materials. Reliance on Boltzmann-based models has produced a culture of smaller is better research, where the reduction in size is expected to produce limitless increase in performance. Experimental observations especially in the case of thermoelectric performance of nanoscale devices have not exhibited this behavior. The semi-classical Boltzmann models are based on the relaxation-time approximation and cannot model strong non-equilibrium transport. In addition, wave effects in these models are included through correction terms that cannot suitably capture their influence on transport. <p> A coupled quantum-scattering model to study thermoelectric performance of nanoscale structures is proposed through the nonequilibrium Greens function method. The model includes all the pertinent physics of the wave nature of electrons while coupling electron-phonon scattering effects. The NEGF method is used to study the performance of silicon nano-films and nanowires as well as strained quantum well Si/Ge/Si superlattices as a function of doping, effective mass and in the case of superlattices, substrate strain and superlattice geometry. Results suggest that the power factor of nanostructured materials is dominated by the electrical conductivity which in turn is strongly influenced by quantum confinement effects and electron-phonon scattering effects. No significant improvement in the Seebeck coefficient is observed due to the decrease in dimensionality of the structure. <p> The NEGF method can be used as a tool to design structures with optimized values of doping, effective mass, substrate strain and superlattice geometry taking into consideration the effects of electron confinement and scattering. The model developed in this research can be used as a framework to guide further studies on performance of highly scaled thermoelectric devices in order to obtain optimal value of ZT. This effort represents the first reported use of the nonequilibrium Greens function method to predict thermoelectric performance.

This dissertation presents the radiation-induced charge trapping studies of upcoming material systems of Si and SiC for future low power and high power technologies. HfO2/Si with metal gates has already been announced as the material system that will power the future technology scaling for low power devices. SiO2/SiC based devices are possible candidates for the upcoming high power device technologies. This dissertation provides significant insights into the charge trapping characteristics in these devices exposed to high-energy ionizing radiation. Charge trapping is studied as a function of dose, processing and gate oxide fields with extensive materials characterization performed before irradiations. The results provide additional information for establishing reliable design rules for future MOS devices intended for both, high and low operating voltage when exposed to a radiation environment.

This research aimed to measure the operational impact of various strategies for utility loading behavior for commercial spent nuclear fuel (CSNF), as well as thermal design parameters involving waste transportation and emplacement of CSNF during the pre-closure period at Yucca Mountain, NV (YM). Specific objectives included: (1) generation of probable scenarios for waste stream quantities and thermal properties, (2) identification of performance trade-offs in surface facility design parameters and operating decisions, and (3) evaluation of safety and throughput implications of pre-closure strategies. These objectives were accomplished by modeling the pre-closure phase of YM operations as an integrated waste management system. Through application of this model, a better understanding was gained of how various waste management strategies influence facility design assumptions. Advantages and disadvantages of various operating strategies were identified in terms of safety considerations and throughput of surface facilities. This led to an opportunity for making recommendations that have potential to result in improved performance. Strategies that required fewer canisters to be aged and those that contained fewer dual-purpose canisters (DPCs) resulted in lower frequencies of dropped canisters and spent nuclear fuel assemblies. Results indicate that DOE estimates are within the likely range of frequencies for a drop event that is generated by this analysis. The consequences of these events are low enough that their frequency will not likely impact the ability to meet regulatory requirements. Still, DOE could strengthen its preliminary hazards analyses by adopting a probabilistic approach to estimating the drop frequency in evaluating pre-closure safety. This would help to communicate the risk posed by the drop events, and allow for an analysis that is less reliant on deterministic estimates of waste stream quantities and crane drop rates. Pending future agreements between waste generators and the Department of Energy, the capacity of the Wet Handling Facility may need to be increased in order to efficiently handle assemblies arriving in DPCs and bare-fuel transportation casks. Furthermore, because design goals for throughput have been defined in terms of surface buildings as opposed to functional needs, there is a redundancy in functional capability of the Canister Receipt and Closure Facility and the Receipt Facility which should be taken into account in order to design a more efficient system. An alternative design for throughput requirements, based on functional demands of the system, is proposed.

Silicon carbide has been preferred over other wide band-gap semiconductors for high power applications because of its unique ability to grow a thermal oxide, challenges lie in the quality of the dielectric and of the SiO2/SiC interface. This thesis focuses on the electrical properties and the reliability of the oxide and its interface with silicon carbide. In particular, the effects of processing parameters, such as implant activation, oxidation conditions (partial pressure), and post-oxidation anneal (nitridation), are considered. Tests are performed on metal-oxide-semiconductor (MOS) capacitors probed by capacitance-voltage measurements (CV), carrier injection (tunneling, photo-emission, irradiation), and time-dependent dielectric breakdown (TDDB). The most important new finding is that nitrogen, required for passivating the complex SiO2/SiC interface, can be detrimental to its reliability. Indeed, nitridation leads to the formation of hole traps, yielding large voltage instabilities.

INTERDISCIPLINARY STUDIES: NEUROGENETICS ASSOCIATION STUDY OF TWO SLC6A4 POLYMORPHISMS WITH AUTISM JACQUELYN RECKTENWALD Thesis under the direction of Professor James Sutcliffe Autism is a heritable neurodevelopmental disorder characterized by impairments in language development and use, social interaction and repetitive behaviors, restricted interests and resistance to change. Serotonergic system dysregulation is implicated in autism since ~25% of individuals with autism have blood platelet hyperserotonemia, selective serotonin reuptake inhibitors are efficacious in treating repetitive behaviors, anger and anxiety in autism, and PET studies indicate children with autism have less capacity for CNS serotonin synthesis than children without autism. These findings indicate the serotonin transporter (SERT) as a strong candidate susceptibility gene and studies point toward linkage in a region containing the SERT gene (SLC6A4) locus with autism. Two SLC6A4 repeat polymorphisms, HTTLPR and VNTR, are located in the promoter region upstream of SLC6A4 and in the second intron, respectively, and regulate SERT expression. Association studies of HTTLPR or VNTR polymorphisms with autism are inconclusive. We hypothesized that there would be preferential transmission of an HTTLPR polymorphism, VNTR polymorphism and a haplotype containing these polymorphisms to affected individuals in our sample. We found no association of any HTTLPR polymorphisms or VNTR polymorphisms with autism in our sample and further we did not find any association of a haplotype containing these polymorphisms with autism in our sample. Paternal inheritance of VNTR alleles was associated with the social intent and milestones (p =.02 and p = .05, respectively) components of the ADI-R, however, these two association results may be due to chance.

To support the world's insatiable desire for energy in the coming century without risking environmental catastrophe, paradigm-shifting research into next-generation photovoltaics and solid-state white lighting is necessary. The Rosenthal group has been pursuing a fully solid-state, inorganic nanostructured photovoltaic featuring pyrolytically synthesized semiconductor nanocrystals deposited within a mesoporous nanocrystalline anatase (meso-nc-TiO2) framework. The synthesis and characterization of meso-nc-TiO2 pores and nanotubes with deposited nanocrystals represents a significant milestone in the fabrication of our next-generation photovoltaic devices. Additionally, a variety of nanocrystal systems were imaged using aberration-corrected atomic number contrast scanning transmission electron microscopes (Z-STEM). Most significantly, the first ever Z-STEM images of ultra-small white-light emitting CdSe nanocrystals were collected, which will aid in the development of a solid-state white-light source.

A new deposition technique for high performance polymer films, resonant infrared laser ablation (RIR-LA) is presented. Ultraviolet laser deposition techniques have been shown to cause decomposition and depolymerization of the deposited polymer films. We hypothesized that the infrared radiation would be a gentler technique compared to ultraviolet radiation and should leave the polymer structure intact. We proposed a technique where a solution-based polymeric precursor is frozen in liquid nitrogen, placed in vacuum chamber, and ablated by a rastered infrared laser beam. Then the ejected material is collected on a substrate forming a thin polymeric film. First we tested the technique on a 15 weight % pyromellitic dianhydride-co-4,4-oxidianiline (PMDA-ODA) in N-methylpyrrolidinone (NMP), the polymeric precursor to polyimide. PMDA-ODA is converted to polyimide by a thermal cure near 250 °C. Fourier transform infrared spectroscopy results confirmed that the PMDA-ODA was transferred intact and without curing by RIR-LA. Molecular weight studies show that only a small portion of the original molecular weight is lost, allowing for the preservation of strength and structural properties. The technique was then tested with other polymers including polyamide imide and polyether imide. Both polymers were successfully transferred intact with no signs of curing. Polyamide imide boasts an even lower cure temperature than polyimide at only 150°C, illustrating how effective RIR-LA is at avoiding thermal transformations.

<p>This research discovers if there are early indicators of the conduct of clinical trials that related to predicting the success of a clinical trial. Successful trials were those achieving at least 100% of the minimum projected accrual goal necessary to statistically support the scientific objective. Two primary aspects of a clinical trial are investigated that influence success are identified as 1) factors related to the design and development stage of a trial and 2) factors that are observed once a clinical trial is implemented, or available for patient enrollment. The research is conducted using National Cancer Institute (NCI) Cancer Therapy Evaluation Program (CTEP) sponsored trials for all therapeutic Phase I, I/II, II and III trials between January 1, 2000 and December 31, 2007. </p> <p>Overall observations of the trials in the sample uncover that approximately two out of every five therapeutic studies conducted through NCI-CTEP will fail to achieve the minimum accrual goal. Furthermore, it is observed that phase III clinical trials are unique from non-phase III clinical trials; Phase III trials have longer development times, greater trial size with respect to accrual goals, and a higher rate failure in terms of inability to achieve the accrual goal. </p> <p>The findings of the research suggest that studies with shorter development time have a higher likelihood of achieving accrual success at trial closure; inversely, studies with longer development time have a decreased likelihood of achieving accrual success. Both the time-to-first accrual and the accrual performance at the expected period are both identified as potential accrual milestone indicators of predicting eventual accrual success. Finally, clinical trial characteristics, particularly the number of scientific reviews, can be used to explain the variance in the development time. </p>

The development phase of systems engineering is one of the most critical phases in the entire life of a project because it is at that point the majority of the projects technology and committed costs are determined (Buede 2000). Yet, the decision tools available to assist systems engineers in choosing between technology alternatives at this phase are extremely limited, consisting mainly of the financial models of net present value and return on investment. Conceptually, if we view the alternatives available as a portfolio of cost, performance, and schedule options, then we can extend the financial portfolio model of Real Options Analysis (ROA) into system engineering. Such an extension is the first element of my research. However, because there is a need to extend beyond strictly the financial realm, I then introduce an advanced performance model, Technology Options Analysis (TOA), which is employed to address a systems performance options. Technology options, while similar to financial options, use performance dimensions such as availability of current and future technologies, technology risks, and industry clock speeds (Fine 1998). The purpose of Technology Options Analysis is to provide a tool that decision makers can use to understand the value of waiting to make a technology selection until some or all of a technologys performance uncertainty has been resolved. Finally, to create a more complete picture of the decision trade space, I combined Real Options and Technology Options into a unified algorithm that incorporates both performance and cost uncertainty into the decision making process. The aspect of schedule is beyond the scope of this dissertation and is left for future research. Applicability of the models are illustrated in the context of one case study.