<p>In this dissertation, the principles of anodic sectioning have been discussed and satisfactory anodic sectioning methods have been developed for V, Mo, Nb and Ta. In addition, W has peen reconsidered. It will emerge that the use of suitable non-aqueous electrolytes permits the formation of impurity-containing (doped) anodic films which have enhanced solubility, thus permitting anodic sectioning.</p> <p>Thicknesses of film-formed and metal-removed have been determined separately using a conventional gravimetric method for anodizations at ≥ 30 votes. In the thin-film region (≤ 100 Å), new methods of determining small thicknesses, based on ion-range-profiles, have been developed leading to a sensitivity which is suggested to be as good as, if not better than, ellipsometry.</p> <p>The stoichiometry of the anodic films has been inferred from the thickness calibrations of metal-removed and film-formed, and the matter has been further investigated for V, Mo and W using reflection electron diffraction. The results were that the films are V₂O₅, MoO₃ and WO₃ respectively. Examination of as-formed films showed that thin anodic films on V, Mo, Ta and W are crystalline with likely stoichiometries H₆V₄O₁₂, MoO₃, Ta₂O₅ and WO₃, whereas all thick films (and also the thin films on Nb) are amorphous in nature.</p> <p>The sectioning methods developed here have been used in determining the range profiles of 5-40 keV Kr⁸⁵ in polycrystalline metals and evidence is presented to show that the stopping process involved tho usual mixture of random and channeled ion trajectories.</p> <p>Similar work on range profiles in amorphous anodic films on V has also been pursued with the main result being that the film properties were found to vary as the bombardment dose increased. The properties include solubility (decreased with dose), sputtering coefficient (decreased with dose from ~44 to ~12 atoms/ion) and stoichiometry (progressed from V₂O₅ to V₂O₃ to VO). These results suggest that all ion bombardment work with oxides must be accompanied by structural investigation, and in particular makes suspect some of the previous work on sputtering coefficients, especially that with Nb₂O₅, WO₃ and TiO₂.</p> / Doctor of Philosophy (PhD)

<p>In this thesis the results of an experimental investigation and theoretical analysis of high temperature oxidation properties of cobalt-iron alloys are presented. Cobalt-iron alloys containing up to 70% iron were subjected to pure oxygen atmospheres with pressures ranging from 10ˉ⁴ to 1 atm at 1200°C. The reaction kinetics, oxide scale morphologies and spatial distributions of the reacting species in the oxidized specimens were determined. It has been established that the reaction is controlled by metal diffusion through the oxide scales separating the reactants. A ternary diffusion model was invoked to describe the high temperature oxidation behaviour of the investigated alloys. The consistency of the model is tested by evaluating the concentration profiles from metal diffusion coefficients and nonstoichiometry of ternary oxides using the Wagner phenomenological equations. The comparison of the results from theoretical calculation suggests that the theoretical analysis adapted in this work is consistent with the actual physical and chemical processes governing the oxidation reaction.</p> / Doctor of Philosophy (PhD)

<p>This thesis concerned with the experimental determinations of the thermodynamic parameters describing the Fe-Ni-O and Ni-Cr-O ternaries between 900°C and 1100°C.</p> <p>Equilibrium oxygen pressures for Fe-Ni and Ni-Cr allows equilibrated with their oxides have been measured by an electrochemical cell with a calcia stabilised zirconia electrolyte. These pressures were related to compositional determinations of the solid phases, using an electron probe microanalyser, and correlated with the ternary isotherms. Both the Fe-Ni-O and Ni-Cr-O systems have been described by equilibrium oxygen pressure diagrams which relate oxyren activity to the compositions of the equilibrated phases.</p> <p>The variation of metal activities with alloy composition for Fe-Ni and Ni-Cr alloys has been determined between 900°C and 1100°C. Spinel-alloy equilibria in the Fe-Ni-O system have been described by a Gibba-Duhem calculation and by a structural model for the spinel based on a lattice point defect model.</p> <p>Experimental observations of the scales formed on Fe-Ni-Cr alloys are qualitatively discussed with respect to thermodynamic considerations.</p> / Doctor of Philosophy (PhD)

<p>An improved theory of vacancy formation energy and volume, based on local pseudopotential theory, is presented and applied to vacancies in the alkali metals and in aluminum. Good agreement with the existing experimental results is obtained.</p> <p>The solubility of hydrogen in magnesium has been measured in the temperature interval 400°C - 600°C and for pressures 50 - 760 mmHg, from which the heat of solution of hydrogen has been determined.</p> <p>A theory of the heat of solution of hydrogen is developed. The host lattice is treated with the local pseudopotential theory.</p> <p>Nonlinear theory is used for the screening of the proton and a method for an approximate self-consistent solution proposed. The theoretical results for the heats of solution in aluminum and magnesium are in good agreement with the experimental values, particularly when one considers the magnitude of the numbers involved.</p> / Doctor of Philosophy (PhD)

<p>The temperature and time dependence of the thermoelastic martensitic transformation in Cu-Zn-Al shape memory alloys have been studied by means of internal friction, DSC (Differential Scanning Calorimetry), X-ray diffraction and metallography. A physical model and mathematical treatment have been developed to derive an expression for internal friction associated with the thermoelastic martensitic transformation. This expression has been able to explain the amplitude dependence of the internal friction in various ranges of stress amplitude, where other theories have failed. In the special case where the applied stress amplitude is much higher than the critical stresses for the stress-induced martensitic transformation, the internal friction is found to be linearly proportional to the reciprocal of the stress amplitude, in good agreement with the experimental results. The time evolution of internal friction and shape change has been measured simultaneously in different microstructural states of the shape memory alloys. In the single-martensite phase state, the monotonic decrease of internal friction with time, without any accompanying shape change, has been attributed to the progressive pinning of the martensite/martensite interfaces. When the parent phase is present, the appearance of an internal friction peak, accompanied by an evident shape change, is found to be associated with the stress-induced martensitic transformation and the subsequent pinning of the martensite/parent and martensite/martensite interfaces.</p> <p>The effects of aging in both martensite and parent phases have been studied. Aging in parent phase caused the precipitation of $\alpha$-phase which hindered the subsequent martensitic transformation and resulted in a suppressed internal friction peak and a deteriorated DSC peak. Aging in the martensite phase resulted in the pinning of the martensite/martensite interfaces by vacancies, which significantly obstructed the subsequent transformations upon heating and cooling. This has been evidenced by both internal friction and DSC measurements.</p> / Doctor of Philosophy (PhD)

<p>SnO₂ films have been prepared by reactive sputtering in an oxygen-containing glow discharge, by ion-beam sputtering in an accelerator, and by anodizing with an ethylen-glycol-based electrolyte. Preparation of SnO₂ films by bombarding metallic tin with oxygen ions has also been attempted. The sputter-deposited films were amorphous if deposited on KCl or Ta at temperatures from -100° to +200°C, though they were crystalline if deposited on SnO₂ at 200°C. The anodic films were variously crystalline or microcrystalline, though clearly not amorphous. Films formed by bombarding metallic tin with oxygen ions turned out to be crystalline α-SnO at all doses from 1 x 10¹⁵ to 3 x 10¹⁷ ions/cm². The crystallization temperatures ranged from <200°C for crystalline SnO₂ substrates, to 250°-300°C for KCl substrates, to 400°-450°C for Ta substrates, to 475°-550°C for unsupported films. The crystallization product was consistently cassiterite, i.e., normal SnO₂. Once crystallinity was attained in unsupported sputtered films, the grain size remained in the vicinity of 400 Å to a temperature of about 1000°C. This is in good agreement with the behaviour of anodic films; which retained their microcrystalline structure in heat treatment again until a temperature of about 1000°C was reached. The thickness of the reactively sputtered films was estimated from the observation of their interference colours when deposited on a Ta substrate, while the thickness of the anodic films was estimated by sputtering the films with 20-keV Kr ions until metal was exposed and noting the weight change. The anodization of Sn has a rather low efficiency (4-21%), and this could be shown to be due to electronic conduction rather than dissolution. Kr-ion bombardment of SnO₂ thin films has shown that this oxide presents a very high value of the sputtering coefficient (21.5 ± 1.5 atoms/ion for 10-keV Kr impact), a result which can be attributed to thermal sputtering (i.e., bombardment induced vaporization). As far as the structure of the target material after bombardment is concerned, the experimental evidence was that SnO₂ undergoes amorphization. The effects of ion bombardment of bulk samples (either SnO₂ sintered powder or natural cassaterite) were investigated using the following techniques: (I) reflection electron diffraction, (II) marker-release spectrometry, (III) dissolution measurements, and (IV) resistivity measurements. Through the use of techniques I and II we have established that the annealing (to the original singlecrystalline phase) of bombardment induced amorphousness occurs in two stages: (i) a homogeneous transformation to a polycrystalline phase showing the normal cassiterite structure; (ii) epitaxial recrystallization to single-crystal cassiterite. Dissolution measurements have established the depth of amorphization due to ion bombardment, while resistivity measurements have indicated that bombardment does not cause SnO₂ to change its stoichiometry. Finally the various results allowed us to expand the formalism on which marker-release spectrometry is based, namely we have identified a previously overlooked release process (Stage IC) due to epitaxial crystallization.</p> / Doctor of Philosophy (PhD)

<p>The mechanical behaviour and the microstructural changes preceeding fatigue fracture have been investigated in ferrous systems subjected to controlled cyclic straining. Attention has been focussed on three basic aspects of the fatigue process:</p> <p>(1) The Bauschinger Effect and the Polarisation of Plastic Flow in Single- and Two-Phase Iron-Based Alloys: the investigation involved the study of the influence of dislocation distribution on the magnitude of the Bauschinger effect, and the study of the Bauschinger effect in two-phase materials as a function of both forward strain and volume fraction of particles. The results are interpreted in terms of some of the currently available theories for the deformation in plastically inhomogeneous materials.</p> <p>(2) Fatigue Softening in Single-Phase Low-Carbon Iron: the investigation involved the study of the microstructural changes occurring in prestrained specimens during cyclic softening, and the influence of initial dislocation arrangement on the rate and extent of fatigue softening. Simple models were developed to rationalize the kinetics of fatigue softening.</p> <p>(3) Instability of Plastic Flow in Quenched-Aged Armco Iron: the investigation involved the study of the influence of various particle dispersions on the stability of plastic flow in fatigue, and the detailed microstructural features developing in fatigue which lead to the localization of plastic flow and onset of fatigue cracks. The problem of instability has been rationalized, macroscopically, in terms of the mechanical equation of state, and microscopically, in terms of the detailed structure of the localized regions.</p> <p>Structural changes were monitored by Transmission Electron Microscopy and ancillary techniques in order to provide a detailed characterisation of the fatigue state.</p> / Doctor of Philosophy (PhD)

<p>The present work has examined the role of shear bands, distribution of second-phase particles, and changes in strain path on deformation and recrystallization processes in single phase and two-phase Al alloys.</p> <p>The development of shear bands and their influence on the overall recrystallization kinetics in an Al-4.8% Mg alloy was studied at both the macroscopic and microscopic level. Shear bands which contained high angle boundaries acted as preferential nucleation sites for recrystallization nucleation, and influenced the early states of recrystallization kinetics.</p> <p>The evolution of the substructure around non-deformable silicon particles was studied in Al-Si-Cu single crystals which were oriented for single and mutiple slips. The relationship of the local substructure around particles was related to the subsequent recrystallization behaviour by using electron diffraction. The influence of the spatial distribution of second-phase particles was correlated to the occurrence of recrystallization nuclei in particle-stimulated nucleation (PSN) by using the Dirichlet cell technique. PSN appears to involve only a small fraction of particles in the distribution, indicating that recrystallization nucleation depends on particle clustering as well as particle size and the level of imposed strain.</p> <p>The influence of change in strain path on mechanical response (deformation and strain hardening) and microstructural evolution during deformation and recrystallization processes were examined. The rate of dynamic recovery appears to be enhanced by the change in strain path. The change in strain path is reflected in changes in recrystallization behaviour and can be rationalized in terms of the energy stored by various strain paths. An attempt has been made to correlate the experimental findings with phenomenological models of dynamic recovery available in the literature.</p> / Doctor of Philosophy (PhD)

<p>Measurements of the internal, friction of several aluminum alloys were made over the temperature range from 80 to 300 K using the two component resonator technique. The technique was improved, by using a metallic bond between the two components, so that measurements in the Qˉ¹ ≃ 10ˉ⁶ range could be made at three frequencies. Preliminary experiments using quartz crystals as specimens were performed in order to check the properties of the metallic joint. A method of checking the bond, by measuring the resonator properties at about half the fundamental frequency of the quartz transducer alone, was developed.</p> <p>Comparisons of the internal friction spectra of hydrogen charged samples with spectra of vacuum annealed samples were used, in conjunction with a simple model of the distribution of hydrogen, to determine upper limits on the binding energy between hydrogen and substitutional solutes in the aluminum matrix.</p> / Doctor of Philosophy (PhD)

<p>Diffusion of nitrogen in zirconium nitride was studied using the l5N gas-solid exchange technique.</p> <p>Two kinds of samples were used: ZrN 0.71-0.79 spheres, 60u and 90u diameter and ZrN 0.93 slabs 254u thick. The spheres were prepared by dropping 99% powdered zirconium nitride through an electric arc while the slabs were obtained by direct nitridation of 99.995% zirconium sheets at 1150°C in very pure nitrogen at 1 atm. These samples were carefully characterized by chemical analysis, density and lattice parameter measurements, electron microanalysis, X-ray diffraction and scanning electron microscopy techniques.</p> <p>The diffusion annealings were performed at a pressure of 200 Torr and in the temperature range 1000-1200°C. The rate of depletion of 15N in the isotopically enriched gas phase was measured by a mass spectrometer and used to calculate the self-diffusion coefficient.</p> <p>The diffusion coefficients for spheres and slabs could be respectively represented by</p> <p>Dspheres(cm^2/sec) = (2.97+18.76-2.56)x 10^-10 exp (-(23,000+/-5,400)/RT)</p> <p>and</p> <p>Dslabs(cm^2/sec) = (4.08+88.23-3.90) x 10^-6 exp(-(36,000+/-4,500)/RT)</p> <p>These results are in good agreement with the chemical diffusion coefficients available in the literature. It is shown that the rather low activation energies, in particular in the case of spheres, do not preclude a vacancy mechanism for nitrogen diffusion.</p> / Master of Science (MS)