A transportation method was used to evaluate the activity coefficients of the minor elements, As, Bi, Pb, and Sb, in both Cu - Fe and Ni - Cu mattes, white metal, and molten Ni₃S₂ as a function of their concentration. Values for the activity coefficients are compared to values obtained by other investigators. With the Cu - Fe mattes and white metal, the analysis was conducted at Cu/Fe molar ratio between 1 to ∞, at sulfur deficiencies from -0.02 to +0.02, and at temperatures between 1493 and 1573 K. Activities of As and Sb in the melts are both concentration dependent at weight percents less than 0.3. No such dependency is observed for Bi or Pb. First and second order self interaction parameters are also reported for As and Sb in Cu - Fe matte. With the Ni-Cu mattes only the sulfur deficiency and the Ni/Cu molar ratio were varied. The experiments were conducted at 1473 K at Ni/Cu molar ratios of ∞, 2, and 1, and at SD values from 0 to -0.1. As and Sb are substantially more stable in Ni - Cu mattes than in Cu - Fe mattes, while for Pb and Bi the presence of Ni does not change the activity of Pb and Bi significantly. The Henrian activity coefficients of these minor elements in both Cu - Fe and Ni - Cu mattes are reported. (Abstract shortened by UMI.)

The adsorption of selected surfactants and inorganic ions of interest to semiconductor processing onto metallic, semiconductor, and dielectric surfaces has been investigated using a quartz crystal microbalance (QCM) technique. The effects of variables such as solution pH and concentration on the extent of adsorption have been characterized. Improvement of the sensitivity of the technique using an electrical bias to the crystal electrode has been explored. The etch rate of sputter coated silicon, sputter coated silica, and thermally grown silica of interest to semiconductor processing has been investigated using a thickness shear mode (TSM) quartz crystal microbalance (QCM) technique. In this research, silicon and sputtered silica and low temperature thermal oxide were investigated in ammonia peroxide solutions using a quartz crystal microbalance. The results obtained have been compared with the literature results to show that a QCM is a valuable in situ measurement technique to follow low levels of etch rate. In an application of the QCM technique to the chemical mechanical polishing (CMP) process, the static etch rate and chemical mechanical polishing rate of an Al-1%Si-0.5%Cu alloy were investigated in abrasive-free solutions containing a proprietary amine at alkaline pH values. The effect of lixiviant, oxidizing agents, complexing agents, temperature, and applied pressure on polishing behavior were investigated. The results have shown that it is possible to polish aluminum alloy films with a high degree of selectivity over SiO2 using abrasive-free amine based solutions containing a glycine based complexing agent.

This study was undertaken with the goal of comparing the calculated nucleation rate for phase separation with experimental measurements for a simple glass system. The magnitude and the temperature dependence of the nucleation rate for a sodium silicate glass composition in the binodal regime was calculated. These calculations used a minimum of assumptions in order to determine the limits on certain thermodynamic variables, chiefly surface energy. Many of the values used in these calculations were determined from growth and coarsening measurements made on this system. Nucleation rates, as well as growth and coarsening rates, were then measured in this system for this comparison to theory. It was found that the free energy of mixing models are unable to predict nucleation behavior at temperatures near the immiscibility boundary. In addition, these models predict that the nucleating composition lies outside of the binodal. Although the values measured for the activation energy correspond well to those in the literature, their incorporation into the nucleation expression does not correct for the temperature behavior of the free energy of mixing. It is also unlikely that a temperature dependent surface energy term could account for the poor predictive nature of classical nucleation theory at small undercoolings.

Jackson, Gilmer and Temkin used a Spin-1 kinetic Ising model to simulate non-equilibrium binary alloy solidification. In this dissertation the detailed relationship of this model to the solidification of binary alloys is reported. The phase transformation kinetics of the model is investigated as a function of growth rate, surface roughness, liquid diffusivity, equilibrium segregation coefficient, entropy of fusion, and composition of the liquid. Simulations for pure silicon predict a growth rate dependence on orientation and undercooling in accord with experimental results. Simulation results for the binary material show an increase in the non-equilibrium segregation coefficient (k(neq)) with surface smoothness, growth velocity and decreasing liquid diffusivity. Simulations for the orientation and growth velocity dependence of the segregation coefficient are in accord for experimental results for the solidification of bismuth doped silicon due to Aziz et al. Simulation results on the dependence of k(neq) on the equilibrium segregation coefficient, k(eq) are also consistent with experiment. The non-equilibrium segregation coefficient was found to increase with concentration of the liquid, but the effect is small at low concentrations.

Stereodeposition is a freeform fabrication technique which accomplishes the computer-controlled, layerwise buildup of an object through direct placement of a fluid which rapidly solidifies. Current materials compatible with stereodeposition include functional ceramics and metals, engineering polymers, and composites. The key to this flexibility is stereodeposition's ability to operate under a wide range of liquid-to-solid transformation rates. Understanding and controlling the material parameters involved in the liquid-to-solid transition is critical, as solidification ultimately impacts the precision and quality of the final object. A model of the liquid-to-solid transition has been developed in which a bead spreading on a curved surface is followed as a series of state "snapshots", whereby an applied force produces an incremental bead motion in an increment of time. This approach differs from most liquid spreading models, but allows flexibility in the time and geometry dependence of forces associated with a solidifying stereodeposition liquid. The model predicts bead contact angle as a function of time based on initial liquid properties (surface tension, viscosity, yield strength) and the solidification strategy employed, namely rheology control, mass transfer, or thermal transfer. Three parameter groupings are identified: alpha, which controls final contact angle, beta, which controls spreading rate, and delta, which controls amount of liquid transformation occurring during spreading. To validate the model, dynamic measurements are performed on the spreading of slurries of silica particles in various liquids. The model is found to predict the early stage of spreading (5 s) under one of the two proposed boundary conditions, and is shown to be equivalent in the limit to the more traditional dynamic wetting models employing an energy rate balance. An explanation is presented for the failure of the model to accurately predict the final contact angle for highly shear-thinning slurries.

Solid freeform fabrication (SFF) methods enable the creation of new structures with multiple materials. While the ability to put any material in any location during the building process allows the freedom to create most any combination, it does not readily suggest the best combination for a given task. The performance demands of a structural member differ from those of a sensor and hence would have different criteria for optimal structures. The methods used herein begin to show how the interaction of the different materials impacts their performance experimentally and can be modeled to determine preferred structures. In this work structures were formed by SFF comprising multiple materials with two areas examined: metal-ceramic monoliths and embedded polymer sensors within composite structures. The Metal-ceramic monoliths consisted of a SFF ceramic preform which was subsequently infiltrated with metal resulting in a graded structure, pure ceramic on one side and nearly pure metal on the other. These structures showed improved toughness over pure ceramic structures when tested in bending. Different metal-ceramic interface gradings were modeled based on the experimental samples, including variations of ceramic content. The model showed that the optimal structure was dependent on the orientation during mechanical testing, or application, as well as the ceramic content of the monolith. Embedded poly vinylidene fluoride (PVF2) sensors were used to monitor internal stresses in composite systems. The PVF2 sensors were shown to be capable of detecting damage over the range light tapping to severe impact. More importantly the sensors were able to detect barely visible impact damage (BVID), which can lead to deterioration of mechanical performance without visible evidence. Additionally the PVF2 sensors were used to monitor cure of epoxy systems by sensing the modulus of the matrix. It was shown that for a fixed impact level the sensor response varied as the relative modulus of matrix to sensor changed. Modeling confirmed that when the sensor modulus is much higher than the matrix the stress level in the sensor is higher. The model also showed that the stress level in the sensor is dependent on the geometry and loading, with smaller sensors performing better.

As the MOS devices continue to scale down in feature size, the gate oxide thickness is approaching the nanometer node. High leakage current densities caused by tunneling is becoming a serious problem. Replacing silicon oxide with a high kappa material as the gate dielectrics is becoming very critical. In recent years, research has been focused on a few promising candidates, such as ZrO₂, HfO₂, Al₂O₃, Ta₂O₅, and some silicates. However, unary metal oxides tend to crystallize at relatively low temperatures (less than 700°C). Crystallized films usually have a very small grain size and high leakage current due to the grain boundaries. The alternatives are high κ oxides which are single crystal or amorphous. Silicates remain amorphous at high temperatures, but have some problems such as phase separation, interface reaction, and lower κ value. In this work, we addressed the crystallization problems of zirconium oxide and hafnium oxide thin films. Both of these two thin films were deposited by DC reactive magnetron sputtering so that very dense films were deposited with little damage. A specially designed system was set up in order to have good control of the deposition process. The crystallization behavior of as-deposited amorphous ZrO₂ and HfO₂ films was studied. It was found that the films tended to have higher crystallization temperature when the films were thinner than a critical thickness of approximately 5 nm. However, it was still well below 900°C. The crystallization temperature was significantly increased by sandwiching the high kappa oxide layer between two silica layers. Ultra thin HfO₂ films of 5nm thickness remained amorphous up to 900°C. This is the highest crystallization temperature which has been reported. The mechanisms for this effect are proposed. Electrical properties of these high kappa dielectric films were also studied. It was found that ultra thin amorphous HfO₂ and ZrO₂ films had superior electrical properties to crystalline films. The leakage current density of ultra thin amorphous films was at least two orders of magnitude lower than that of crystallized films. Amorphous films also showed much less hysteresis in the capacitance-voltage curve than uncapped crystallized films. The mechanisms for the electrical property differences between ultra thin crystalline and amorphous films were studied. Due to successful control of the low dielectric interfacial layer thickness, an effective oxide thickness of 1.2 and 1.4 nm was obtained for HfO₂ and ZrO₂ films, respectively.

A modification of Scherer's Self-Consistent model and a new model, the Interlocking Cell model, have been developed to characterize the sintering behavior of mixtures of two glass powders. The theoretical sintering curves predicted by both these models are compared to each other and to experimental densification behaviors. Viscosities extrapolated from homogeneous sintering curves of sol-gel derived powders are transient and cannot be predicted based on composition alone. These transient viscosities have a significant effect on the sintering kinetics. The Self-Consistent and Interlocking Cell models assume very different microstructural changes during sintering. However, differences between the two models can just be distinguished using experimental densification curves of sol-gel cordierite based glass mixtures.

The formation of small particles is important in many areas of science, including astronomy, material science, and the study of pollution. In an effort to elucidate the details of this process, this research explored the production and characterization of carbon particles. The particles were produced by the resistive heating of carbon rods in a reaction chamber at various pressures of helium. Samples were collected on substrates at distances of 15, 25, and 45 mm from the carbon rods. The average particle size increased with the sample collection distance from the carbon rods and with increasing chamber pressure. Crystals of disordered graphite and fullerenes were detected by electron diffraction in the samples collected at a distance of 45 mm and in a pressure of 100 torr. The ultraviolet/visible spectrum results indicated that fullerene concentration increased with sample collection distance from the arc.

The two-phase Al-Si alloy is ideal for wear applications due to the high hardness of the Si phase, which imparts excellent wear resistance to the alloy. The shape, size and distribution of the Si phase have significant effects on the wear resistance. Microstructural modification is the logical approach to improving wear resistance, and further improvements may be achieved through refinement of the Al and Si phases to the nanometric regime (< 100 nm). However, bulk nanostructured materials lack ductility and are often associated with high manufacturing costs. A cladding approach, exploits the advantages of the nanostructured properties on the surface while maintaining the bulk properties of the substrate. Electrospark Deposition (ESD) and Spark Plasma Sintering (SPS) will be evaluated as potential methods to generate nanostructured Al-Si claddings. The solidification behaviour of the Al-12Si system using the ESD process was studied. The production and sintering of a nanostructured Al-Si powder was also investigated followed by a cladding operation in the SPS. The final claddings were then subjected to dry sliding wear testing. ESD was proven as a viable method of depositing nanostructured Al-Si. It was determined that the optimal pulse energy input required for the deposition of the binary Al-Si was 0.55 J. This resulted in a consistent layer ~30um thick after a single deposition. The minimum grain size of the aluminum and silicon phases was determined to range from ~52 to 76 nm and 5–11 nm respectively. It was determined that while the silicon content of the starting electrode had an effect on the microstructure of the deposition, the starting microstructure of the electrode did not. A binary Al-12Si powder was cryomilled and after 4 hours of cryomilling the grain size inside the powder reached 46 nm with a lattice strain of 0.09 %. Extension of the solubility limit of silicon in aluminum was achieved during cryomilling up to 1.72 at% of Si in solid solution. The extended solubility was removed after 1 h of annealing at 100 C and grain growth was observed when the temperature is further increased. The effects of different sintering times and temperatures using SPS on the densification of the powder, the growth of the Si phase and the morphology change of the Si phase were investigated. The Al and Si phase growth were investigated and it was observed that the Si coarsening rate is increased due to the increased volume of grain boundaries. As the Si coarsens, any pinning effect on the Al grains would be lost, resulting in a highly unstable microstructure that coarsens rapidly. The cryomilling was effective in breaking up the eutectic phase in the powder to modify the morphology of the Si phase from plate-like to spherical. A cladding of Al-Si created from nanostructured powder was successfully bonded to a wrought AA7075 substrate using Spark Plasma Sintering. The final state of the substrate regardless of the starting microstructure had the η (MgZn2) phase present indicating an over aged microstructure. The peak hardness of the substrate was not achievable via a secondary aging treatment after sintering, but was recovered after full solutionization and aging. The bond between the cladding and the substrate was tested by three-point bending and in all tests, fracture occurred in the Al-Si cladding.Sliding wear testing was performed and the final wear rates measured were 8.0x10^-3 mm^3N^-1m^-1 (as-cast), 1.3x10^-2 mm^3N^-1m^-1 (SPS), and 2.2x10^-3 mm^3N^-1m^-1 (ESD) after 2000 cycles of sliding. The ESD sample exhibited lower wear rates than the as-cast and SPS samples, which showed similar wear rates, in spite of having different initial hardnesses. The SPS sample appeared to suffer from subsurface cracking. The extreme refinement of the Al and Si phases achieved for sample ESD was much higher than what resulted during the wear of the As-cast and SPS surfaces and as a result showed lower wear rates throughout testing. / L'alliage à deux phases Al-Si est idéal pour des applications de résistance dû à la dureté de la phase Si qui donne une excellente résistance à l'usage. La géométrie, grandeur et distribution de la phase Si ont des effets significatifs. Modifier la microstructure est l'approche logique pour améliorer la résistance à l'usage. D'autres améliorations peuvent être atteints par le raffinement des phases Al et Si. Par contre, les matériaux nanostructurés en vrac manquent de ductilité et s'associent souvent à des coûts de production élevés. Un approach utilisant le placage exploite les avantages des propriétés de nanostructures sur la surface tout en gardant la majorité des propriétés du substrat. Electrospark Deposition (ESD) et Spark Plasma Sintering (SPS) seront évalués comme méthodes potentielles pour générer des placages d'Al-Si nanostructuré. Le comportement de solidification du system Al-12Si en utilisant le ESD était étudié. La production et le frittage d'une poudre d'Al-Si nanostructurée étaient examinés suivi par un placage dans l'SPS. Les placages ont subi un test de résistance par glissage sec. L'ESD est une méthod éfficace pour la déposition de l'Al-Si nanostructuré. Ce fut déterminé que l'énergie optimale pour la déposition de l'Al-12Si était 0.55 J. Le résultat était une simple déposition consistante et dense d'une épaisseur de ~ 30 um. La grosseur minimale des grains des phases d'Al et de Si était entre 52-76 nm et 5-11 nm respectivement. Ce fut déterminé que pendent que le contenu de silicon de l'électrode premier avait une effet sur la microstructure du déposition, la microstructure débutante n'avait pas d'effet. La poudre d'Al-12Si a subit un broyage cryogénique et après 4 h la grosseur des grains à l'intérieur de la poudre a atteint 46 nm avec une tension du réseau de 0.09 %. Une extension de la limite de solubilité du Si dans l'Al fut atteint après le moulage, jusqu'à 1.72 %at de Si en solution solide. Cette extension de solubilité fut enlevé après un traitment de chauffage d'1 h à 100C et une croissance des grains fut observé une fois que la température s'est accrut. Les effets de temps et températures variés de frittage en utilisant l'SPS sur la densification de la poudre, l'accroissement de la phase Si et la morphologie dans le changment du phase Si furent examinés. La croissance des phases Si et Al fut examinée et ce fut observé que la vitesse d'alourdissement du Si est accru dû à l'accroissement des limites des grains. Pendant que le Si s'alourdisse, il aurait une perte d'effet d'épinglement résultant d'une microstructure instable. Le broyage cryogénique fut efficace pour défaire la phase eutectique de la poudre et modifier la morphology de la phase Si de sphérique à plat. Une placage de l'Al-Si de la poudre nanostructurée fut joint à un substrat AA7075 par SPS. L'état final du substrac avait la phase η (MgZn2) present qui indique une microstructure suragée. La dureté maximale du substrat ne fut pas achevée par un traitement de vieillissement secondaire suivant frittage. Ce fut recouvert après une mise en solution complète et un traitement de vieillissement. Le joint entre les deux materiaux fut testé par un plissage à trois points et tous les tests de fracture se sont produits dans le placage du Al-Si. Des tests d'usage par glissage fut performés et les vitesses d'usage finales étaient 8.0x10^-3 mm^3N^-1m^-1 (AC), 1.3x10^-2 mm^3N^-1m^-1 (SPS) et 2.2x10^-3 mm^3N^-1m^-1 (ESD) après 2000 cycles. L'échantillon du ESD a démontré des vitesses d'usage plus bases que ceux du AC et d'SPS, qui démontraient des vitesses d'usage pareils, même s'ils avaient des duretés initiales différentes. L'échantillon du SPS semblait souffrir de fractures sous-surface. Le raffinement extrême des phases d'Al et Si atteint pour l'échantillon d'ESD était beaucoup plus élevée que ce qui était résultant durant l'usage des surfaces as-cast et SPS et comme résultat démontrait des vitesses d'usage plus bases pendent l'examination.