This thesis investigates the properties of exciton states in semiconductor nanocrystals using effective mass models. We focus on type-I core-shell nanocrystals in which the staggered band alignments of the core and shell material mean the lowest energy states for electrons and holes lie in different spatial regions, giving rise to spatially indirect excitons. The technological potential of type-I! nanocrystals provides motivation for understanding single exciton states which determine many important optical properties. In the first research chapter we study CdTe/CdSe and CdSe/CdTe nanocrystals using a single-band and (2,6)-band effective mass model. The (2,6)-band model is based on a multiband k·p theory previously developed for spherical quantum dot heterostructures. We calculate exciton energies as a function of the core radius a and shell width as of the heterostructure, and assign six exciton transitions in the experimental absorption spectra of CdTe/CdSe nanocrystals. The second research chapter is concerned with strained ZnTe/ZnSe nanocrystals. The (2,6)-band model is modified to incorporate strain using a continuum elasticity model. Exciton energies from absorption spectra are compared with the predictions of the strained and unstrained nanocrystal models, showing that they only describe the lowest exciton energy of one of the three size series. Improved agreement is found for the change in exciton energy due to a particular shell width, with the strained nanocrystal model giving much better fits to the as-dependence. The as-dependence of nanocrystals with alloyed heterointerfaces is better described by the unstrained nanocrystal model, indicating alloying relaxes the strain in this system. In the final chapter we model spatial correlations between the electron and hole in CdTe/CdSe and CdSe/CdTe nanocrystals using a configuration interaction approach developed in the framework of the (2,6)-band k·p theory. We find that the single-particle basis can, be restricted without changing t~e resulting exciton energies significantly; using this decoupled configuration ihter~ction approach we calculate exciton energy shifts due to correlation as a function of aand as. Dielectric confinement increases correlation for many heterostructure designs by shifting carrier wavefunctions away from the surface, and the interparticle Coulomb interaction leads to large changes in radial probability density for the uniform dielectric constant case. Dielectric confinement affects the correlated hole more than the electron so that excitons in the CdSe/CdTe heterostructure are more affected by the dielectric environment than those in the CdTe/CdSe heterostructure. The overall behaviour of the correlated charge density is due to the net effect of the type-I! spatial confinement, interparticle Coulomb attraction, dielectric confinement and single-particle electronic structure.

New approaches for the synthesis of highly luminescent InP and InP/ZnS nanocrystals were developed by stepwise systematic investigation of the parameters during the reaction stages. The parameters, including solvents, precursors, ligands, capping agents, protic agents, Lewis acids and bases and temperature were discussed in detail in different chapters. The investigation processes helped increase understanding of understand the reaction and surface passivation mechanisms and to develop convenient synthesis approaches. Highly luminescent InP NCs were prepared with an in-situ indium chloride complex in the presence of zinc carboxylates or zinc dithiocarbamates - convenient nucleation initiators and stabilisers. The nanocrystals prepared covered a wide photoluminescence emission range from blue to the near infra-red. This synthesis method also allowed the in-situ growth of highly luminescent InP/ZnS core-shell nanocrystals as well. The principles of selection of reagents are also applicable in the synthesis of other nanocrystals. The as-prepared high-quality InP and InP/ZnS nanocrystals have been exploited in applications in the fields of light emitting diodes (LEOs) and bioimaging.

Over the past years, the synthesis of zeolite films has attracted considerable attention due to their potential applications as membranes, sensors, and insulators with low dielectric constant. Therefore, the preparation of high quality defect free zeolite films with controlled film orientation and thickness is highly desired. Silicalite-l nanocrystals were functionalised via ultrasonic treatment in alcohol (methanol, ethanol, 2-propanol and I-butanol) suspensions. The formation of methoxy groups on the external surface of silicalite-l was confirmed by nitrogen adsorption porosimetry analysis. Pressure-area Langmuir isotherms and BAM analysis confirmed the enhanced hydrophobicity of alcohol modified silicalite-l nanocrystals and film formation at the air-water interface, respectively. Silicalite-l films were successfully fabricated onto silicon substrates for the first time using the Langmuir-Blodgett (LB) method. SEM and AFM characterisations confmned that the monolayer nature of the LB films extended over an area of several mm2 • Secondary growth of the monolayers LB seeds produced smooth and continuous polycrystalline zeolite film with preferred b-orientation evidenced by XRD analysis. In addition, we have also demonstrated the synthesis of continuous and homogeneous silicalite-l films by steaming of spin-coated precursor coatings according to the SEM images. This simplified method benefits from the limited amount of structure directing template used, facile recovery of the supported films and negligible amounts of waste products. We have also shown the first successful use of synchrotron IR spectroscopy in grazing incidence mode for the characterisation of zeolite precursor LB and spin-coated films on gold-coated glass substrates. The spectra confirmed the existence of incipient zeolite nanoparticles with MFI type structure in clear silicalite-l synthesis solution. The supported LB zeolite precursor particles were used as seed layers to prepare highly crystalline and uniformly oriented silicalite-l films by secondary growth. HR-TEM images confirmed that the steamed LB precursor films were ultra-thin, < 10 nm thick but structural characterisation is required. The results confirmed the nm dimensions of the silicalite-l precursor species.

The motivation for the work- presented in this thesis originates in the semiconductor industry's drive to create increasingly scaled transistors. In view of current device dimensions approaching fundamental atomic scales, the industry is looking to alternative structures to provide continued scaling capabilities. The use of metal, usually silicide. source and drain regions to create Schottky barrier (SB-)MOSFETs is one such approach. Previous work on static and RF electrical characterisation as well as simulations has shown this device to provide a number of scaling benefits to the planar MOSFET structure. In addition, it provides simpler and more cost effective fabrication. In this work, the electrical properties under DC bias conditions of p-channel SBMOSFETs with PtSi sources and drains are explored at room temperature and down to 80 K. High room temperature ON currents up to 545 mA/mm and transconductances up to 640 mS/mm for 85 nm gate length devices are measured and performance factors are found to satisfy ITRS recommendations. Increasing silicide anneal temperatures lead to increases in Schottky barrier height and corresponding decreases in drain current are observed. The radio-frequency performance at frequencies up to 110 GHz is studied using a novel measurement deembedding technique. High unity-gain cutoff frequencies up to 71 GHz are extracted. Finally, two-dimensional simulations using the drift-diffusion simulator MEDICI are performed and fitted to the measured electrical characteristics.

SiC is seen as the material of choice for the next generation of power semiconductor devices because of its potential to deliver high power densities whilst maintaining high efficiency levels. The novel contact formation examined in this· thesis is aimed at improving the robustness and ohmicity of contacts to 4H-SiC. Innovative triple and multiple metallisations based on a combination of Al/Ti and Al/Ni layers have been fabricated, which are aimed at reducing the contact resistance and improving the contact morphology to p-type 4H-SiC. The lowest specific contact resistivity obtained was based on a triple layer AI/Ti/Al contact, measured at 5.0 x 10-6 Ocm2 • This value approaches the state of the art. The multiple layered contacts all dis- . played Schottky-like characteristics even after high temperature annealing. X-ray diffraction scans have indicated enhanced formation of metal silicides with respect to the ohmic triple layered contacts based on Al/Ti (Ti3SiC2) and Al/Ni (NiSix ). Low levels of silicide formation have been found in multiple layered contacts. Scanning electron microscope examination and wavelength dispersive analysis point to the fact that multiple layered contacts display a superior morphology, reducing the amount of Al spreading, which is known to be problematic during a high temperature anneal. A physical and electrical investigation of silicon contact layers grown by molecular beam epitaxy on n-type 4H-SiC (Si/SiC) has been conducted. The effect of the silicon doping concentration, growth temperature and SiC surface cleaning prior to growth have been evaluated. X-ray diffraction and scanning electron microscope analysis have been utilised for the purpose of evaluating the morphology and crystal structure of grown silicon layers. I-V and e-V measurements have shown the rectifying properties of the Si/SiC heterojunction along with their corresponding built-in potentials and energy band offsets. Modelling of Si/SiC heterojunction diodes revealed that the current characteristics can be explained by an isojunction drift-diffusion and thermionic emission model. The modelling results point to the fact that I-V properties of Si/SiC n-N heterojunction diodes are relatively independent of the doping concentration of the grown silicon layer. Preliminary . results obtained from p-type silicon grown by molecular beam epitaxy on n-type 4H-SiC have been presented. A possible application for Si/SiC layers is a vertical heterojunction MOSFET. The establishment of 4H-SiC PiN diode technology is discussed. Scanning electron microscope examination has been utilised for the purpose of evaluating PiN diode fabrication based on shadow masking and liftoff techniques. Reverse bias I-V measurements highlight the fact that mesa isolated PiN diodes display around 5 orders of magnitude lower leakage currents with respect to their un-isolated counterparts. Forward bias I-V measurements have compared performances of diodes with different anode metallisations based on AI, Ni and Ti layers. PiN diodes based on Ti show lower turn-on voltages while mesa isolated PiN diodes displayed higher current densities and more ideal forward characteristics when compared to their un-isolated counterparts. __.

The accurate calculation of the total energy, ET of an electronic system, using as our starting point simply the properties of the interactions between the particles that constitute such a system, is a source of immense power. With it we can understand and control, properties of materials, chemical and biological processes. The problem is complex; approximations must be introduced. There exist two main families of methods. One tries to follow the theory that describes these interactions:. Many Body Perturbation Theory (MBPT). The other uses· a different formulation, based on the electronic density, Density Functional Theory, (DFT) and a new technique: Kohn-Sham scheme. The first is numerically expensive and can't be used for most systems. The second is simple, but lacks accuracy. We introduce a third way, based on a generalised Kohn-Sham (GKS) scheme, which retains some of the simplicity of the standard Kohn-Sham scheme, but being more flexible, allows the use of non-local potentials and orbital-dependent quantities. We realised this meant the Kohn-Sham scheme was now able to process concepts from MBPT; this last theory is based on non-local quantities like the self-energy. operator, ~, and the Green's function, G, which, if we ignore frequency dependence in the self-energy operator, can be represented by the one-particle orbitals. We produced models for screened interaction, W, and ~ and calculated ET for different systems. Results showed the strong points and weaknesses of these models. With this experience we introduce a new method, based on a frequency-dependent W, and theGW approximation to obtain a model ~. This final model, gave excellent results for homogeneous systems, and promising results for inhomogeneous systems; the way to proceed and the remaining tests that need to be done, are also mentioned.

This thesis describes the characterisation of microplasma configurations which include hollow cathode (MHC) and parallel plate (PP) structures. A custom built microplasma source was used to investigate its DC microdischarge characteristics at high pressure (p > 10 Torr). High resolution (± 12 μm) control of the gap geometry allowed an investigation of pressure-geometric scaling by varying the discharge gap down to < 10 μm, as well as pressure, allowing for the first time a detailed comparison. The MHC configurations described in this thesis vary in external and internal gap geometry and anode dimensions. Voltage-ampere (V-I) measurements obtained for PP sources and the external anode MHC displayed distinctive low (Mode 1), and high (Mode 2), current modes separated by an oscillation regime. These V - I characteristics were compared to traditional scaling relationships to distinguish mode behaviour. Analysis showed j/p2 scaling holds for the Mode 1 gap plasma and j2p scaling for the Mode 2 plasma within the hollow cavity. Electron temperatures were estimated to be between 2 eV and 3 eV using spectroscopy and a collisional radiative model. The hollow cathode sheath models of both ionising and non-ionising sheath conditions were examined analytically. Sheath thicknesses yielded from the models indicated the only realistic assumption is that that these MHC discharges operate with an ionising sheath. The diffusion equation was solved numerically to show a domed plasma density profile (nplpeak > 1013 cm-3). When the anode was positioned within the hollow cathode, a stable plasma was generated and sustained inside the cathode. V-I characteristics from this configuration showed a new high current mode, Mode 3. Analysis of these characteristics showed that the Mode 2 plasma varies with length depending on current. The placing of the anode further into the cathode has been shown to extend the plasma along its axis, and hence offers a way of creating plasma surface interactions in long tubes. When the Mode 2 plasma approaches the end of the cathode, the plasma expands outside, radially across the cylinder wall, leading to Mode 3 operation.

An algorithm for the synthesis of microwave cross-coupled filters based on Control Theory has been developed. This has enabled a visualisation of the filter network using Signal Flow Graphs and made it possible to synthesise a wide range of filter topologies meeting arbitrary transfer functions. The model was also employed to establish a methodology to account for stray couplings in quartet filter sections. In parallel, an algorithm for extracting coupling parameters of microwave filters based on their parameters was also developed. The extraction method makes direct use of poles and zeros and hence does not require curve fitting or optimization routines. Distorting effects that cannot be accounted for in the coupling matrix are alleviated using vector rotations. These two analytical devices, for synthesis and tuning, are then used for two different applications; an X-band double passband filter realized using cylindrical wave guides and an L-band quadruplet microstrip filter in which stray couplings have been accounted for.

Low frequency distortion components of the output voltage of a cycloconverter are largely responsible for the restriction on its practical range: of frequencies, and the object of this thesis is to show that these components can be attenuated by the application of sampling techniques to the control system. After a general description of the operation and control of the cycloconverter, the distortion of the output waveform due to low frequency components is discussed. Under these circumstances, the fundamental repetition frequency of the waveform is less than the wanted output frequency, and two methods of determining it for given input and output frequencies are developed. The characteristics and properties of the low frequency distortion components, and the requirements for attenuating them are analysed. The particular effects on the magnitudes of these components due to operation of the cycloconverter in the inhibited mode, rather than the circulating-current mode, are examined and the requirements for attenuating them are identified. It is shown that the communications engineering processes of pulse width modulation and of natural sampling can be identified in the control of the cycloconverter. Regular sampling is more widely used in communications engineering, and its effect on the low frequency distortion components in the cycloconverter output is compared with natural sampling. A modified control method for the inhibited cycloconverter is then developed to attenuate these components. Digital computer programs were written to test the effect of introducing modifications to the control of the cycloconverter, and the more significant results are given in graphical and tabulated form. An experimental cycloconverter, with an inhibition control circuit designed for this project, was constructed to check the validity of the computer programs. The design details are described, and the experimental results are discussed.

The purpose of this work was to produce novel Phase Locked Loops (PLL) architectures for antenna array applications. The operational characteristics of these PLLs architectures are presented and are analysed theoretically. The first type ofPLL ar.c. hitecture presented is called the Phase Conjugating Locked Loops (PCLL). Three different topologies for this type are designed to provide the phase conjugation operation necessary for retrodirective antenna array action. PCLL .Type 1 operates under the circumstance that input radio frequency roRF is different . from the VCO output frequency rovco while Type 2 and 3 operate when both ofthese frequencies are equal to each other. I I The second type of PLL architecture, which is called the Variable Phase Locked Loop (VPLL), is designed to operate in transmit only operation as a spatial phase shifter. The beamforming operation of a VPLL based antenna array is obtained by baseband tuning ofthe VCO voltage. The last type of PLL architecture studied, the Phase Buffering Locked Loop (PBLL), is suitable for self-tracking receive only operation. The embodiment of this PLL is architecturally similar to that of PCLL Type 1 with the exception that a down' converter mixer is included after the VCO and an inverter is inserted before the VCO. For each of the PLL architectures comprehensive derivations of the output phase noise jitter noise equations including the effects of the non-linearity and non-idealistic mixer characteristics is presented. These RMS phase noise jitter equations are utilised in order to evaluate the steady state error of each PLL topology. Following the investigation of noise performance, the stability of PLL architectures is examined. Bode Plot and Polar Plot have been deployed to facilitate the investigation. The final investigation concentrates on the impact of multiple signals on PBLL self-tracking receiver antenna array.