Thesis (Ph. D. Philosophy (Faculty of Education))--University of the Western Cape, 2007. / Includes bibliographical references (leaves 292-314).

The numerical treatment of nonlocal problems, which taking into account material microstructures, by means of meshless approaches is promising due to its efficiency in addressing integropartial differential equations. This thesis focuses on the investigation of meshless methods to nonlocal elasticity. Firstly, mathematical constructions of meshless shape functions are introduced and their properties are discussed. Shape functions based upon different radial basis function (RBF) approximations are implemented and solutions are compared. Interpolation errors of different meshless shape functions are examined. Secondly, the Point Collocation Method (PCM), which is a strong-form meshless method, and the Local Integral Equation Method (LIEM) that bases on the weak-form, are presented. RBF approximations are employed both in PCM and LIEM. The influences of support domains, different kinds of RBFs and free parameters are studied in PCM. While in LIEM, analytical forms of integrals, which is new in meshless method, is addressed. And, the number of straight lines that enclose the local integral domain as well as the integral radius are analyzed. Several examples are conducted to demonstrate the accuracy of PCM and LIEM. Besides, comparisons are made with Abaqus solutions. Then, PCM and LIEM are applied to nonlocal elastostatics based on the Eringen's model. Formulations of both methods are reported in the nonlocal frame. Numerical examples are presented and comparisons between solutions obtained from both methods are made, validating the accuracy and effectiveness of meshless methods for solving static nonlocal problems. Simultaneously, the influence of characteristic length and portion factors are investigated. Finally, LIEM is employed to solve nonlocal elastodynamic problems. The Laplace transform method and the time-domain technique are implemented in LIEM respectively as the time marching schemes. Numerical solutions of both approaches are compared, showing reasonable agreements. The influence of characteristic length and portion factors are investigated in nonlocal dynamic cases as well.

<p>CASTRO, SUSANA PATRICIA. Characterization of the Boron Doping Process UsingBoron Nitride Solid Source DiffusionThe purpose of this research has been to develop an optimum process for the borondoping of implants and polysilicon gates of metal-oxide-semiconductor (MOS) devices.An experimental design was constructed to determine the effects of diffusiontemperature, time, and ambient on characteristics of the doping process. A temperaturerange of 800 to 1000 degrees Celsius was studied with a diffusion time between 10 and60 minutes. Two diffusion ambients were used for doping processes, a pure nitrogenambient and a nitrogen-oxygen gaseous mixture. Device wafers were fabricated, and thetesting of MOS capacitors and van der Pauw test structures was performed to determinethe effect of diffusion conditions on flatband voltage and poly gate doping. Materialscharacterization techniques were used on monitor wafers for each diffusion process todetermine the wafer structure formed for each process and evaluate the effectiveness ofthe deglaze etch. The processes that resulted in the best device characteristics withoutsuffering from significant poly depletion effects and flatband voltage shifts were wafersdoped at 800 degrees Celsius in a pure nitrogen atmosphere for 20 minutes and 45minutes. The presence of oxygen in the atmosphere caused the depletion of boron fromthe Si wafer surface. The formation of the Si-B phase only occurred on devices processedat 1000 degrees Celsius. The deglaze process used in this experiment did not fullyremove this layer, and thus all devices doped at this temperature were seriously degraded.<P>

<p>The characteristics of clean n-type GaN surfaces and the interface between this surface and Pt, Au and Ag, have been investigated. Gallium-terminated (0001) surfaces of GaN, free of carbon and oxygen within the detection limits of XPS have been achieved by annealing in ammonia at 860°C for 15 minutes. Additional, in-situ surface analysis indicated a flat, stoichometric, and unreconstructed surface free of other contaminants. The electron affinity of this surface was 3.1 ± 0.2 eV. The valence band maximum was located 3.0 ± 0.1 eV below the Fermi level, indicating the presence of a surface state near the valence band maximum. Individual layers of Pt, Au or Ag were deposited in-situ on the cleaned surface and the interfaces characterized using XPS, UPS, LEED and TEM. All as-deposited metal/GaN interfaces were abrupt and unreacted; the Pt and Au were deposited epitaxially. The Schottky barrier heights obtained from photoemission measurements were 1.2, 0.9 and 0.5 ± 0.2 eV for Pt, Au and Ag, respectively. Values of the metal work function from UPS results were 5.7, 5.3 and 4.4 ± 0.2 eV for Pt, Au and Ag, respectively. Schottky barrier heights determined via ex-situ current-voltage measurements were 1.15, 0.88 and 0.56 ± 0.05 eV for Pt, Au and Ag, respectively. Capacitance-voltage measurements yielded barrier heights of 1.25 and 0.96 ± 0.05 eV, for Pt and Au, respectively. These results indicate that the Fermi level of the cleaned surface is not pinned. Upon annealing the aforementioned contacts from 400 to 800°C for 3 minutes each. The rectifying behavior of the Pt and Au contacts degraded as a function of temperature during annealing at 400, 600 and 800°C for 3 minutes each until they became ohmic. This was correlated with TEM of the annealed interfaces, which displayed increased chemical reaction and roughening as a function of temperature. <P>

<p>The surface morphology and atomic structure of nitrogen doped, n-type 6H-SiC(0001)Si wafers before and after various surface preparation techniques were investigated. As-received wafers were exposed to in-situ cleaning with or without excess silicon to obtain either a (rt3 x rt3)R30&#176 or a (3 x 3) reconstructed surface. The resulting surfaces were characterized using reflection high-energy electron diffraction, photo-electron emission microscopy, and atomic force microscopy. An atomically clean, reconstructed surface was obtained via thermal annealing at 950ºC. Cleaning with excess silicon resulted in the formation of silicon islands on the surface. The surface morphology of hydrogen etched wafers depended upon their doping concentrations. Wafers with doping concentrations of greater than or equal 2.5 x 10E18 and less than 7 x 10E17 (ND-NA)/cm3 were investigated with the former exhibiting more surface features. The microstructure of all the samples showed regions with full and half unit cell high steps. An atomically clean, ordered, stepped surface was achieved via annealing at 1030 degrees Celcius. Chemical vapor cleaning resulted in the formation of silicon islands. The initial growth of AlN and GaN thin films on the cleaned, hydrogen etched 6H-SiC(0001) substrates were investigated using PEEM and AFM. The AlN films nucleated immediately and coalesced, except in the areas of the substrate surface which contained half unit cell height steps where pits were observed. The GaN films grown at 800ºC for 2.5 minutes exhibited nucleation and three-dimensional growth along the steps. The GaN films deposited at 700&#176 C for 2 minutes grew three-dimensionally with coalescence of the film dependent upon the step structure. Almost complete coalescence occurred in regions with unit cell high steps and incomplete coalesce occurred in regions with half unit cell height steps. Films of AlN grown for 30 minutes via GSMBE on hydrogen etched surfaces exhibited two-dimensional growth and had an RMS roughness value of 4 &Aring. Films grown at 1000 &#176 C exhibited an SK growth mode and had rocking curve FWHM of 150-200 arcsecs. MOCVD grown films on hydrogen etched wafers had an RMS roughness value of 4 &Aring and a XRD rocking curve FWHM of approximately 260 acrsecs. <P>

<p>The purpose of this research has been to use off-line characterization techniques to establish material-specific properties of gate-stack constituents (i.e., high-k dielectric stacks and electrodes) and complete gate-stack structures. Hence, the characterization methodologies were established to evaluate high-k dielectrics at various processing levels, which, in part, determine the final characteristics of an advanced gate-stack device. Material systems that were investigated include: Al-O, Hf-Si-O, Zr-Si-O, Ti-O, Ta-O and Sr-Ti-O. Various physical and electrical characterization techniques were used to establish fundamental understandings of the materials selected, thin-film growth/deposition processes, and gate-stack structures. General conclusions for stable and unstable gate-dielectric materials have been establishedregarding the presence of a problematic interfacial layer at the Si/dielectric interface, graded dielectric layers, and the stability of gate electrodes on high-k dielectrics.The nanometer-scale chemistry of a gate-stack capacitor whose expected structure is Si/SiOxNy/Ta2O5/TiN/Al was studied by high-resolution electron-energy-loss spectroscopy in a scanning transmission electron microscope. Elemental profiles with near-atomic-level resolution for Si, Ti, N, Al, and O demonstrate that the device structure deviates drastically from the expectation and is chemically complex.It is concluded that the graded distribution of certain elements across the gate-stack capacitor completely precludes a band-structure model that assumes abrupt interfaces and chemically discrete layers. This study impacted on subsequent interpretations of flatband voltage extractions and electrical degradation following backside metallization/postmetallization annealing for capacitors whose dielectric-stack was based on Ta-O.Detailed and extensive electrical characterizations of Pt/SiOx/Sr-Ti-O/Si MOS capacitors were carried out to investigate reliability issues in a bi-layer gate dielectric. Based on these studies, models are proposed to describe the carrier transport and dielectric degradation for a Sr-Ti-O capacitor. It is concluded that conduction is dominated by Frenkel-Poole emission from mid-gap trap levels. The trap barrier height is estimated to be 1.51eV. A model based on the atomic and electronic structure of oxygen vacancies can account for the reported leakage-current characteristics. In addition, it is tentatively proposed that anode-hole injection and hole trapping control the dielectric degradation under gate injection.<P>

<p>Characterization of nitride films on 6H-SiC (0001) wafers via x-ray, TEM, and AFM was accomplished on standard GaN thin films with AlN or AlGaN buffer layers. TEM sample thinning capability was improved through the use of Nomarski in an optical microscope to gauge the thickness of the sample during preparation. TEM analysis was then completed of Au and Pt films deposited on chemical vapor cleaned GaN with annealed up to 800°C. Chemical reactions were detected in x-ray measurements of the 800°C Pt samples and GaN/metal interface roughening were confirmed by TEM images in both metals. Interface roughening is attributed to the chemical reactions and interfacial stresses greater than the yield stress of the metal created during heat treatments by the difference in the thermal expansion coefficients of the GaN and the metals. The GaN rocking curves were found to track very closely to the values of the underlying substrate and changes in buffer layer growth temperatures were found to change the screw and edge dislocation populations of subsequent GaN layers. GaN grown on 1030°C AlN buffer layers showed the lowest edge dislocation populations when compared against buffers grown in the range of 1010-1220°C, even though the 1220°C AlN was much smoother. AlGaN buffer layers provided more edge dislocation reduction, with a 1090°C Al0.2Ga0.8N layer yielding the best GaN rocking curve values found in this work. GaN films with AlN buffer layers grown on hydrogen etched SiC substrates did not show rocking curve improvement when compared against samples with unetched substrates. The AlN layers showed extremely narrow, substrate limited, on-axis rocking curve values, but it is not clear as to whether additional defects are present that may broaden the off-axis rocking curves, causing the poorer results seen in the GaN films. Reciprocal space maps of uncoalesced, maskless pendeo epitaxy samples revealed that the wing regions are shielded from poor substrate material when compared against the seed material. The wing regions also have lower strain and rocking curve widths than the corresponding seed material.<P>

<p>LOESING, RAINER, Development of High Resolution Depth Profiling Of Ultra Shallow Dopant Implants with SIMS (under the direction of Phillip E. Russell. Secondary Ion Mass Spectrometry (SIMS) is considered a reliable technique for precise and accurate dopant depth profiling in Si with respect to junction depth and implanted dose. The junction depths of source drain extension structures are predicted to be between 19-33nm for the 0.1µm MOSFET generation. Accurate high depth resolution analysis of these ultra-shallow junctions by SIMS can only be provided if atomic mixing caused by energetic primary ion bombardment is minimized and extensive beam induced crater bottom roughening is avoided. For quantitative measurements, the influence of primary ion implantation, sputter rate changes and beam induced crater bottom roughness on secondary ion intensities has to be known. In this work SIMS was used to develop techniques for the accurate analysis of ultra shallow B, P and As implants in Si.Low energy O2+ primary ion bombardment was found to give the highest depth resolution for the analysis of B and P in Si, while low energy Cs+ and CsC6- primary ion bombardment resulted in the highest depth resolution for the analysis of As in Si. To obtain a more accurate profile shape and depth scale it was found to be essential to limit beam induced crater bottom roughness by means of sample rotation, variations of primary ion angle of incidence or change in sample chamber vacuum conditions. Beam induced crater bottom roughness was investigated for low energy O2+, Cs+ and CsC6- ion bombardment using atomic force microscopy (AFM) and optical profilometer (OP) measurements. OP was found to be a valuable tool for investigating small changes in sputter rate in the initial stages of a SIMS depth profile. It was shown that dose measurements of ultra shallow implants can be improved by using a correction procedure based on bulk doped standards. SIMS was proven to be a valuable tool for the characterization of ultra shallow implants in Si, but careful consideration of analysis conditions and SIMS artifacts is required for accurate analysis.<P>

<p>Free-standing single crystals of bulk GaN were grown via unseeded vapor phase transport at 1130C on hexagonal BN surfaces via direct reaction of Ga with ammonia. The temperature and stability of the Ga source were critical in terms of uniform nucleation and growth. The source temperature was maintained at 1260C to minimize a rapid reaction leading to the formation of GaN and the subsequent decomposition beneath the surface and consequent spattering of Ga. A maximum crystal growth temperature of 1130C was determined in which the GaN growth kinetics were much greater than decomposition. The number of nucleation events was reduced and the crystal size increased by a novel nucleation technique wherein ammonia was introduced at high temperatures. The resulting crystals were either needles or platelets depending on the process variables employed. Low V/III ratios achieved via ammonia flow rates 75sccm and/or ammonia total pressures 430Torr favored lateral growth. The average lateral growth rate for the platelets was ~50micron/hr; the average vertical growth rate for the needles was ~500micron/hr. Growth rates in all other directions for each of these two morphologies were very low. Seeded growth of both needle and platelet crystals was also achieved; however, the growth rate decreased at longer times and higher pressures due to reaction with hydrogen from the increased decomposition of ammonia. Nitrogen dilution of ammonia reduced the amount of hydrogen generated as a result of ammonia decomposition and increased the kinetic barrier to desorption of reactants from the GaN surface and then alleviated the enhanced decomposition of GaN crystals. A 2mm x 1.5mm needle and a 2.3mm x 1.8mm x 0.3mm platelet of GaN were grown with minimal decomposition in a 66.7% ammonia + 33.3% nitrogen gas mixture. Excellent crystalline quality was confirmed by Raman spectroscopy and Photoluminescence.Crystal growth using a Ga- 5at%Al source was conducted in an attempt to increase growth rate and inhibit decomposition. No notable change in growth rate was observed and hollow crystals were formed, indicating that Al promotes vertical growth under otherwise similar conditions for GaN growth. In addition, fine-grained AlN was formed within the binary Ga-Al source, thus, the supply of Al was progressively reduced. Nucleation control via addition of Si resulted in a slightly reduced number of larger crystals. X-ray Photoelectron Spectroscopy suggested that amorphous silicon nitride was formed on the BN substrate and nucleation rate was slightly reduced. Silicon was not detected within the sensitivity of Energy Dispersive Spectroscopy. Raman spectroscopy revealed insignificant amount of Si present in the crystal. Smoother surface morphology of the crystals grown in the presence of Si was observed by Scanning Electron Microscopy.<P>

<p>The purpose of the research presented herein has been to determine the underlyingmechanisms of and to optimize the growth parameters for the growth of smooth surfaceson InGaN and GaN thin films via metalorganic vapor phase epitaxy. Relationshipsamong dislocation density, film thickness, flow rates of the reactants, kinetic growthregime, and thermodynamic growth mode with the surface morphology and surfaceroughness were determined. The two chief parameters affecting template surface roughness in both growth ofGaN above 1000ºC were determined to be temperature and layer thickness. An optimumtemperature of 1020ºC was found for the former process, below which the islands formedin the growth on AlN buffer layers did not coalesce properly, and above which a hillockgrowth instability was pervasive on the surface. Increasing the GaN film depositiontemperature to 1100°C for GaN film deposition via PE enhanced sidewall growth;however, surface roughness was increased on the (0001) growth plane through theformation of hillocks. Template thickness above 2.5 microns had the lowest root mean squaresurface roughness of 0.48nm over 100 square microns. This was attributed to reductions indislocation density, as measured by corresponding 50% reductions in symmetric andasymmetric full width half maximum values of X-ray rocking curves. GaN films were grown at 780ºC to remove the influence of indium incorporationon the surface roughness. V-defects covered the surface at a density of 2E9 per square centimeter andwere linked with a boundary dragging effect. Growth parameters that affect Inincorporation into the InGaN films were investigated and measured using roomtemperature photoluminescence, x-ray diffraction, and x-ray photoelectron spectroscopy.Temperature and growth rate had the greatest effect on incorporation over the range of760 to 820ºC and 25 and 180nm/hr, respectively, through kinetically limiting InNdecomposition. Additions of In into the GaN film produced hillock islands that wereattributed to a strain relief mechanism. The V-defects were also observed in InGaN films;however, their formation was suppressed below a nominal thickness of 25nm<P>