<p>Solar photovoltaics (PV), the technology that converts sunlight into electricity, has immense potential to become a significant electricity source. Nevertheless, the laws of economics dictate that to grow from the current 2% of U.S. electricity generation and to achieve large scale adoption of solar PV, the cost needs to be reduced to the point where it achieves grid parity. For silicon solar cells, which form 90% of the PV market, a significant and slowly declining component of the cost is due to the high-temperature (> 900 &deg;C) processing required to form p-n junctions. In this thesis, the replacement of the high-temperature p-n junction with a low-temperature amorphous titanium dioxide (TiO₂)/silicon heterojunction is investigated. The TiO₂/Si heterojunction forms an electron-selective, hole-blocking contact. A chemical vapor deposition method using only one precursor is utilized, leading to a maximum deposition condition of 100 &deg;C. High-quality passivation of the TiO₂/Si interface is achieved, with a minimum surface recombination velocity of 28 cm/s. This passivated TiO₂ is used in a double-sided PEDOT/n-Si/TiO₂ solar cell, demonstrating an open-circuit voltage increase of 45 mV. Further, a heterojunction bipolar transistor (HBT) method is developed to investigate the current mechanisms across the TiO₂/p-Si heterojunction, leading to the determination that 4nm of TiO₂ provides the optimal thickness. And finally, an analytical model is developed to explain the current mechanisms observed across the TiO₂/Si interface. From this model, it is determined that once &#916;E_V (TiO₂/Si) is large enough (400 meV), the two key parameters are the Schottky barrier height (resulting in band-bending in silicon) and the recombination velocity at the TiO₂/Si interface. Data corroborates this, indicating the hole-blocking mechanism is due to band-bending induced by the unpinning of the Al/Si interface and TiO₂ charge, as opposed to due to the TiO₂ valence band edge.

<p> The operating temperature of energy conversion and electronic devices affects their efficiency and efficacy. In many devices, however, the reference values of the thermal properties of the materials used are no longer applicable due to processing techniques performed. This leads to challenges in thermal management and thermal engineering that demand accurate predictive tools and high fidelity measurements. The thermal conductivity of strained, nanostructured, and ultra-thin dielectrics are predicted computationally using solutions to the Boltzmann transport equation. Experimental measurements of thermal diffusivity are performed using transient grating spectroscopy.</p><p> The thermal conductivities of argon, modeled using the Lennard-Jones potential, and silicon, modeled using density functional theory, are predicted under compressive and tensile strain from lattice dynamics calculations. The thermal conductivity of silicon is found to be invariant with compression, a result that is in disagreement with previous computational efforts. This difference is attributed to the more accurate force constants calculated from density functional theory. The invariance is found to be a result of competing effects of increased phonon group velocities and decreased phonon lifetimes, demonstrating how the anharmonic contribution of the atomic potential can scale differently than the harmonic contribution.</p><p> Using three Monte Carlo techniques, the phonon-boundary scattering and the subsequent thermal conductivity reduction are predicted for nanoporous silicon thin films. The Monte Carlo techniques used are free path sampling, isotropic ray-tracing, and a new technique, modal ray-tracing. The thermal conductivity predictions from all three techniques are observed to be comparable to previous experimental measurements on nanoporous silicon films. The phonon mean free paths predicted from isotropic ray-tracing, however, are unphysical as compared to those predicted by free path sampling. Removing the isotropic assumption, leading to the formulation of modal ray-tracing, corrects the mean free path distribution. The effect of phonon line-of-sight is investigated in nanoporous silicon films using free path sampling. When the line-of-sight is cut off there is a distinct change in thermal conductivity versus porosity. By analyzing the free paths of an obstructed phonon mode, it is concluded that the trend change is due to a hard upper limit on the free paths that can exist due to the nanopore geometry in the material.</p><p> The transient grating technique is an optical contact-less laser based experiment for measuring the in-plane thermal diffusivity of thin films and membranes. The theory of operation and physical setup of a transient grating experiment is detailed. The procedure for extracting the thermal diffusivity from the raw experimental signal is improved upon by removing arbitrary user choice in the fitting parameters used and constructing a parameterless error minimizing procedure.</p><p> The thermal conductivity of ultra-thin argon films modeled with the Lennard-Jones potential is calculated from both the Monte Carlo free path sampling technique and from explicit reduced dimensionality lattice dynamics calculations. In these ultra-thin films, the phonon properties are altered in more than a perturbative manner, referred to as the confinement regime. The free path sampling technique, which is a perturbative method, is compared to a reduced dimensionality lattice dynamics calculation where the entire film thickness is taken as the unit cell. Divergence in thermal conductivity magnitude and trend is found at few unit cell thick argon films. Although the phonon group velocities and lifetimes are affected, it is found that alterations to the phonon density of states are the primary cause of the deviation in thermal conductivity in the confinement regime.</p><p>

The objective of this dissertation is to provide, for a fiber reinforced MMC, an understanding of the evolution of strains as a result of thermo-mechanical loadings and the relationship of the deformation evolution to microstructural factors. Both analytical and finite element approaches were used to predict global and local properties. The analytical approach involved rigorous modifications to the shear-lag model to account for fiber end effects. It was demonstrated that the modification not only results in a correct prediction of the modulus increases in the small aspect ratio regime, but is also able to correctly predict the values of local stress variations in the matrix and whisker. The elasto-plastic behavior of the composite including fiber/fiber interactions was investigated in detail. Cyclic stress-strain hysteresis and Bauschinger effect due to the presence of fibers was also analyzed. It was found that plastic constraint generates a triaxiality in the matrix and so gives a substantial increase of fiber axial stress and an elevation of composite flow stress. Detailed deformation evolution was evaluated both under monotonic and fatigue conditions. Role of thermal residual stresses due to thermal mismatch in MMCs were investigated using the thermo-elasto-plastic FEA with temperature dependent matrix properties. It was found that the spatial distribution of the residual stresses during cooling is sensitive to constraint effects. The evolution of the "caging" plasticity surrounding fibers during cooling including the shape and size of the plastic zones were determined. While FEA solutions give good results, the application of FEA to composites requires careful attention to the geometry of the optimum mesh used in the analysis. The optimization strategy was based on an error energy norm calculation for global convergence and traction differential approach for local convergence at the fiber/matrix interface. It was shown that this optimization approach provides the optimum mesh with a much more rapid convergence than conventional automatic codes. The mesh patterns generated are also shown to be significantly different from using this approach. A converged local property values can be obtained using a significantly lower degree of freedom than by conventional methods.

In recent years, supercritical fluids (SCF), specifically carbon dioxide (CO2), have been tested and applied as alternative solvents for polymer processing and modification. The principal utility of CO2 in heterogeneous polymer systems lies in the sorption of significant mass fractions of CO2, which influence properties that are driven by free volume. The effects include depressed glass transition temperatures, enhanced transport within the dilated polymer and decreased viscosity. The exploitation of these effects in multicomponent systems requires an understanding of the influence of compressible fluid sorption on polymer-polymer compatibility that to date has been unexplored. In this dissertation, it is demonstrated for the first time, that sorption of SCF's can induce phase segregation in polymer systems exhibiting Lower Critical Solution Temperature-type (LCST) behavior at temperatures hundreds of degrees below the ambient pressure transition. For LCST systems, the relative compressibilities of the components play a dominant role, which can be exacerbated by sorption of SCF's. For example, fluorescence quenching experiments indicate that sorbed gas (CO2) depresses the LCST's of blends of polystyrene/poly (vinyl methyl ether) (PS/PVME) by over 100°C at modest pressures (around 20 bar) of the gas with negligible dependence on temperature and molecular weights of the polymer components. Absorbed CO2 has similar effects on blends of deutrated-polybutadiene/polyisoprene as studied by Small Angle Neutron Scattering. The phase behavior of PS/PVME in the presence of CO2 has been modeled using the Sanchez-Lacombe equation of state, which indicates that the polymer blend phase separation is driven primarily by the selective dilation of PVME by CO2 relative to PS. Ethane, with a weaker selectivity also induces phase separation in PS/PVME system, but is significantly different from the effect of CO2 with respect to temperature and polymer molecular weights, indicating the role of selectivity of poor solvents that is superimposed on compressibility effects. Finally, the design, development and application of neutron reflectivity to high-pressure systems for in situ measurements are discussed including results on swelling of thin homopolymer films and investigations of the phase behavior of a diblock copolymers of polystyrene and poly (n-butyl methacrylate) that exhibit Lower Disorder-Order Transition (LDOT).

This thesis investigates the macroscopic yield behavior and microscopic energy absorption mechanisms in single and multiphase polymers. One unique aspect is the evaluation of polymers under multiaxial loading conditions. This is important because in many applications polymers are subjected to complex loading conditions and hence optimal design requires experimental evaluation and modeling of behavior under multiaxial stress states. This work has resulted in a more quantitative understanding of yield and energy absorption in the different polymers considered. Multiaxial stress states are achieved using thin-walled hollow cylinder specimens. The hollow tubes are simultaneously subjected to internal pressure and axial load, leading to biaxial stress states. Stress states ranging from uniaxial compression to equibiaxial tension are interrogated using the same specimen geometry, a procedure uncovering true material behavior. In the first part of this study, a generalized model for the yield behavior of single-phase polymers is evaluated for a polycarbonate system. The generalized model accounts not only accounts for viscoelasticity (i.e., rate and temperature dependence) but also the effect of pressure on yield behavior. The effects of physical aging on the behavior of amorphous polycarbonate are also highlighted. For rubber-modified polymers, existing models for both macroscopic yield behavior and the onset of microscopic damage (e.g., cavitation) are evaluated under multiaxial conditions (chapter 3). Serious discrepancies are found for both cases, prompting an investigation into the nature of energy absorption mechanisms in the materials. Apart from the chosen rubber-modified systems, a toughening mechanism in the form of overlapping parallel cracks is identified to be generic to a range of polymers (chapter 4). The last part of the thesis (chapter 5) involves a quantitative investigation of interactions in overlapping crack patterns. This effort is vital, because for better design of materials, the interaction has to be related to intrinsic material properties. The interactions in an infinite 2D array of parallel and overlapping cracks are analyzed using a complex stress function method. The size and number density of cracks in the array are related to intrinsic material properties and conditions for damage stability are identified.

The enhanced functionality and tunability of electronic nanomaterials enables the development of next-generation photovoltaic, optoelectronic, and electronic devices, as well as biomolecular tags. Design and efficient synthesis of such semiconductor nanomaterials require a fundamental understanding of the underlying process-structure/composition-property-function relationships. To this end, this thesis focuses on a systematic, comprehensive analysis of the physical and chemical phenomena that determine the composition and properties of semiconductor nanomaterials. Through synergistic combination of computational modeling and experimental studies, the thesis addresses the thermodynamics and kinetics that are relevant during synthesis and processing and their resulting impact on the properties of silicon thin films and ternary quantum dots (TQDs) of compound semiconductors. The thesis presents a computational study of the growth mechanisms of plasma deposited a-Si:H thin films based on kinetic Monte Carlo (KMC) simulations according to a transition probability database constructed by first-principles density functional theory (DFT) calculations. Based on the results, a comprehensive model is proposed for a-Si:H thin-film growth by plasma deposition under conditions that make the silyl (SiH3) radical the dominant deposition precursor. It is found that the relative roles of surface coordination defects are crucial in determining the surface composition of plasma deposited a-Si:H films and should be properly accounted for. The KMC predictions for the temperature dependence (over the range from 300 K to 700 K) of the surface concentration of SiHx(s) (x = 1,2,3) surface hydride species, the surface hydrogen content, and the surface dangling-bond coverage are in agreement with experimental measurements. In addition, the thesis details a systematic analysis of equilibrium compositional distribution in TQDs and their effects on the electronic and optoelectronic properties. Formation of hetero-nanostructures, such as core/shell-like structures, through atomic-scale assembly driven by equilibrium surface segregation is studied as a function of nanocrystal size, composition, and temperatures for TQD morphologies that include faceted equilibrium nanocrystal shapes for ZnSe1-xTex and InxGa1-xAs TQDs; the results are based on coupled compositional, structural, and volume relaxation of the nanocrystals according to Monte Carlo and conjugate-gradient methods employing a DFT-parameterized description of interatomic interactions. A phenomenological species transport theory also is developed that explains the concentration profiles due to surface-segregation-induced ordering of constituent and dopant atoms in the dilute limit. The nm-scale diffusion lengths in nanocrystals introduce an interesting interplay between the kinetic and thermodynamic stability of interfaces. The thermodynamic stability of such interfaces in ZnSe 1-xSx TQDs are investigated based on DFT calculations combined with X-ray photoelectron spectroscopy (XPS) and photoluminescence (PL) spectra of TQDs that are synthesized and annealed using colloidal methods. The results demonstrate the possibility of compositional redistribution that causes degradation over time of core/shell TQD electronic properties, with far reaching implications for the use of such nanostructures in devices. Electronic structure calculations of ZnSe1-xSx (type-I) and ZnSe1-xTe x (type-II) TQDs elucidate the impact of composition and compositional distribution on the electron density distribution, density of states, and band gap of the TQDs. The resulting relationships with respect to the distributions in the TQDs of constituent/dopant/impurity atoms (core/shell vs. alloyed TQDs) provide an interpretation for the key features observed in the PL spectra, as well as useful guidelines for improving the design and device performance of TQDs.

Semiconductor nanocrystals, or quantum dots (QDs), are interesting nanomaterials whose size-dependent, tunable optical and electronic properties make them ideal for applications in biological sensing and imaging, light-emitting devices, displays, and solar cells. The commercial exploitation of these materials requires the development of synthesis techniques that are scalable, economical, and environmentally friendly, while enabling precise control of the size, shape and size distribution of the nanocrystals. The most common synthesis technique for these nanocrystals employs small batch reactors in which nanocrystals grow as a function of time following a rapid injection of organometallic precursors into a hot coordinating solvent. The limitations of this process for large-scale commercial exploitation stem from the incomplete mixing of the precursors in large batches that can lead to non-uniform nucleation and broad particle size distributions. Limitations also include the high cost, flammability, and toxicity of the organometallic precursors and its operator-intensive nature. Templated synthesis techniques for nanocrystals have distinct advantages over other methods, including more precise control of particle size, shape, and size distribution and easier scalability for commercial applications. This thesis presents the templated synthesis of semiconducting nanocrystals in stable microemulsions and liquid crystals, formed by the self-assembly of an amphiphilic block copolymer in the presence of a polar and non-polar solvent. The work of this thesis investigates microemulsion templates for the scalable synthesis of semiconductor nanocrystals including: materials composition and particle size control, continuous production of nanocrystals, improvement of optical properties, and alternative non-toxic reactants. The nanocrystals were formed by reacting a group-II salt dissolved in the dispersed phase of the template with a group-VI hydride gas inside the nanodomains. The versatility of nanomaterials and precision of size control of this synthesis method were demonstrated by adjusting the metal salt composition and concentration. The scalability of this technique was displayed by developing a counter-current flow, packed-bed reactor for continuous synthesis of nanocrystals in templating microemulsions. Limitations of the optical properties of nanoparticles synthesized with microemulsion template were addressed by post-processing techniques including extraction and functionalization of the nanocrystals, annealing, and overcoating the quantum dots with an inorganic shell to optimize fluorescence emission and quantum yield. This post-process annealing allowed for the investigation of Mn-dopant incorporation and expulsion from the ZnSe nanocrystal. To eliminate the toxic and flammable group-VI hydride gases, a microwave-assisted templated synthesis route was developed. This employed bursts of microwaves to selectively heat the aqueous, dispersed droplets of water-in-oil microemulsions that contain the water-soluble precursors of the group-II and VI elements, thus leading to nucleation and formation of a single nanocrystal inside each nanodomain.

Magnetic mesoscopic and nano-sized structures have promising applications such as high-density data storage, magnetic field sensors, and microwave devices. Patterned magnetic structures are especially interesting because their constitutive material, sizes and geometry are easily adjustable in fabrication. This makes manipulation of electromagnetic properties possible and creates many novel features never discovered in conventional bulk materials. The artificial magnetic structures that can be engineered to meet specific application purposes are called magnetic metamaterials. This thesis aims to investigate magnetic materials nanostructured to produce high permeability and low loss performance at gigahertz (GHz) frequency region. Such property is highly desired for communication devices with miniaturized size, reduced energy consumption and enhanced signal detection sensitivity. Antennas, microwave field sensors are the examples of applications. We first analyze the single domain model for ac magnetization to get theoretical understanding and prediction. Then we evaluate all free energy terms for a magnetic dipole to know which energies (or fields) are contributing to the effective magnetic field in our real experiments. Secondly experiment work including fabrication, dc characterization and ac characterization of Permalloy and cobalt nanoscale magnetic structures, as well as FePt nanoparticles are covered. Different microwave techniques regarding sensitive magnetic permeability measurements are discussed in detail for comparison. In the last chapter, micromagnetic simulations are performed to obtain broadband ac magnetization response spectrum for a single Permalloy nanowire and two interacting Permalloy nanowires.

Liquid crystalline polymers (LCPs) are a new class of high strength materials. Their rheological behavior is different from that of ordinary isotropic polymers because they are inherently anisotropic in the melt state. Crystallization in rigid-rod mesogenic systems in the nematic melt and super-cooled states has been studied by small amplitude oscillatory shear, which has been found to be in some ways more sensitive than conventional techniques like DSC or X-ray scattering (in that changes are measured by orders of magnitude instead of merely on a linear scale). A scheme of preheating is suggested for unsubstituted, rigid-rod polymers whereby a metastable nematic melt can be achieved, effectively suppressing crystallization and enabling a thorough rheological characterization. For example, an excellent agreement has been obtained for the stress relaxation modulus between experimental values and those calculated from dynamic oscillatory measurements, thereby confirming the existence of a linear viscoelastic range for the LCP. Non-linear creep studies illustrate how rheological properties are strongly affected by structural changes upon deformation. Structure in unsubstituted rigid-rod nematic systems is hypothesized to exist at two levels--Non-Periodic Layer (NPL) Crystallites and Domains/Disclination Network. Another aspect of research concerns a comparative study of the phase behavior of liquid crystalline components in closely related blends and copolymers.

Porous materials have long been a research interest due to their practical importance in traditional chemical industries such as catalysis and separation processes. The successful synthesis of porous materials requires further understanding of the fundamental physics that govern the formation of these materials. In this thesis, we apply molecular modeling methods and develop novel models to study the formation mechanism of ordered porous materials. The improved understanding provides an opportunity to rational control pore size, pore shape, surface reactivity and may lead to new design of tailor-made materials. To attain detailed structural evolution of silicate materials, an atomistic model with explicitly representation of silicon and oxygen atoms is developed. Our model is based on rigid tetrahedra (representing SiO4) occupying the sites of a body centered cubic (bcc) lattice. The model serves as the base model to study the formation of silica materials. We first carried out Monte Carlo simulations to describe the polymerization process of silica without template molecules starting from a solution of silicic acid in water at pH 2. We predicted Qn evolutions during silica polymerization and good agreement was found compared with experimental data, where Qn is the fraction of Si atoms with n bridging oxygens. The model captures the basic kinetics of silica polymerization and provides structural evolution information. Next we generalize the application of this atomic lattice model to materials with tetrahedral (T) and bridging (B) atoms and apply parallel tempering Monte Carlo methods to search for ground states. We show that the atomic lattice model can be applied to silica and related materials with a rich variety of structures including known chalcogenides, zeolite analogs, and layered materials. We find that whereas canonical Monte Carlo simulations of the model consistently produce the amorphous solids studied in our previous work, parallel tempering Monte Carlo gives rise to ordered nanoporous solids. The utility of parallel tempering highlights the existence of barriers between amorphous and crystalline phases of our model. The role of template molecules during synthesis of ordered mesoporous materials was investigated. Implemented surfactant lattice model of Larson, together with atomic tetrahedral model for silica, we successfully modeled the formation of surfactant-templated mesoporous silica (MCM-41), with explicit representation of silicic acid condensation and surfactant self-assembly. Lamellar and hexagonal mesophases form spontaneously at different synthesis conditions, consistent with published experimental observations. Under conditions where silica polymerization is negligible, reversible transformation between hexagonal and lamellar phases were observed by changing synthesis temperatures. Upon long-time simulation that allows condensation of silanol groups, the inorganic phases of mesoporous structures were found with thicker walls that are amorphous and lack of crystallinity. Compared with bulk amorphous silica, the wall-domain of mesoporous silicas are less ordered withlarger fractions of three- and four-membered rings and wider ring-size distributions. It is the first molecular simulation study of explicit representations of both silicic acid condensation and surfactant self-assembly.