In this thesis, we study the surface flattening kinetics at a roughened rutile (1 x 1)(110) surface in an ultrahigh vacuum chamber by low energy electron diffraction spot profile analysis (SPA-LEED). Surface roughness was introduced by argon ion bombardment at a low sputtering dose, instead of deposition such as molecular beam epitaxy (MBE) used elsewhere. In addition, we further investigated the effect of annealing on the terrace height distribution by measuring spot profiles versus incident electron energy at room temperature after annealing for different times. The terrace height distribution was determined by analyzing the relative weight of the Bragg peak in the spot profiles versus the incident electron energy. By means of the results of the terrace height distribution the interface width versus annealing time was determined. (Abstract shortened by UMI.)

Since the dawn of the information age, the high technology industry has been driven to display more data and to create large area high-resolution displays. Organic light emitting diodes, introduced in the late 1980s, are emerging as an attractive alternative to other display technologies. These devices offer cheap materials, low operating voltages, high brightness and a wide viewing angle. The ability to spin coat molecular materials on large rigid or flexible substrates is unparalleled in existing technologies. Several different device geometries are possible, ranging from the simple homostructure, where there is only one organic material present, to the complex microcavity, where dielectric stacks are used to selectively filter out specific wavelengths of light. No matter which geometry is chosen, all of these devices have several common features: a high work function anode (usually transparent) and a low work function cathode. Due to the low work function requirement of the cathode, we are limited to the very reactive metals in the periodic table. In this thesis cathodes based on pure lanthanum, lanthanum-silver bilayers, and lanthanum-yttrium alloys were investigated. (Abstract shortened by UMI.)

New electronic structure calculations of the electric field gradient (EFG) at octahedral Fe2+ sites in layer silicates are discussed. These calculations were done with the aim of providing a link between quadrupole splitting distributions from Mossbauer spectroscopy and the physical distributions of local environments within the material. Various distortions were applied to FeO610- and Fe(OH)6 4- clusters to simulate different local environments and the corresponding EFG's calculated. The electronic structure calculations were performed with the General Atomic and Molecular Electronic Structure System (GAMESS) and a self-consistent-charge-Xalpha method. An analytic model of the EFG using a classical electrostatic point charge model and crystal field theory is used to complement the electronic structure calculations. There is good qualitative agreement between the electronic structure calculations, the analytic model, and with experimental quadrupole splittings in micas. A geometric model of the octahedral sheet in a layer silicate is described, based on isometric flattening and counter-rotation as the main distortions, which can have one, two, or three unique sites. EFG distributions are then calculated using a variety of cases based on the geometric model and the calculated EFG vs. distortion curves. The most realistic distribution results from a case that assumes two unique site-types in a ratio of 2:1, with the height of each site and the inter-cation distance held constant throughout the sheet and the Fe--O bond length of one site-type allowed to vary with a Gaussian distribution.

The cation distributions in europium iron garnet Eu3Fe5 5-xAlxO12 series have not been determined yet since they were synthesized and studied by Yamaguchi and Sakuraba twenty years ago. In this thesis, the Fe-Al distributions over the octahedral ( a) and tetrahedral (d) sites in the europium iron garnet Eu3Fe5-xAlxO12 series have been determined by powder X-ray diffraction (XRD) and 57Fe Mossbauer spectroscopy (57Fe MS). The 57Fe MS spectra are collected at temperatures above the Neel temperatures of the studied samples and the cation distributions are derived from both the raw and thickness-corrected MS spectra. It is found that Al atoms occupy both the a and d sites. Higher Al concentration in the d sites is interpreted as the evidence that Al atoms enter preferentially the d sites due to the smaller size of Al ions compared with Fe ions. MS gives more reliable results of the Fe-Al distributions than XRD, which could be ascribed to the sensitivity of MS to the local environment around the iron nucleus at the specific site.

We have studied the magnetic and structural properties of synthetic and meteoritic face centred cubic (FCC) Fe-Ni alloys in an effort to understand the underlying mechanisms responsible for the Invar and anti-Invar effects. We have performed a detailed 57Fe Mossbauer spectroscopic and electronic imaging study of the Santa Catharina meteorite, an FCC Fe-Ni meteorite with a bulk composition close to that of Invar (Fe65Ni 35). Its microstructure consists of islands of tetrataenite in a matrix of Fe-rich antitaenite. Using Mossbauer spectroscopy (MS) and electron probe microanalysis (EPMA), we have determined that antitaenite is a low moment alloy with an Fe content of 88 +/- 2 at. %. Antitaenite appears to be stabilized in the FCC phase by its epitaxial relation with tetrataenite. Its Neel temperature is ∼60 K and its isomer shift (IS) is close to that of gamma-Fe. The effect of the magnetic interaction with tetrataenite has been studied by Monte Carlo (MC) simulations and have been shown to be consistent with the anomalous temperature dependence of the width of the antitaenite's Mossbauer spectrum. Using electronic structure calculations (ESCs), we have studied the T = 0 K properties of chemically ordered and disordered FCC Fe-Ni alloys in the ferromagnetic (FM), non-magnetic (NM) and disordered local moment (DLM) states. A significant drop in the IS and in the atomic volume are shown to occur at the transition from the high moment (HM) FM phase to the NM phase which occurs, according to our calculations, at 74.5 at. % Fe. Our calculations of the DLM and FM states show that in Fe-rich Invar alloys, the FM order contributes to stabilizing large moment magnitudes. Our systematic study of the hyperfine fields (HF) by ESCs supports the phenomenological model of Dang and Rancourt (1996a). Calculations on a supercell of Fe11Ni5 show the effects of nearest neighbour (NN) chemical and magnetic environments on the local magnetic moment magnitude and the HF distribution, thereby explaining the main observed features in Fe-rich alloys. Using the IS as a direct probe of the electronic structure, a HM/LM (low moment) transition has been unambiguously observed to occur at ∼70 at. % Fe. Our results have allowed us to conclude that the Invar effect in Fe-Ni is not directly caused by HM/LM admixture or so-called 2-gamma-state like excitations. (Abstract shortened by UMI.)

Experimental observations are presented on Bose-Einstein condensation of excitons in Cu2O at low temperatures. These observations deal with traveling exciton packets in a high quality natural crystal. The properties of the exciton packets are examined using two different excitation depths of formation of the excitons (affecting the size of initial exciton cloud). The properties of the condensate include for example: the amplitude, the velocity of the exciton packet, the critical conditions for condensation and phase diagram. Interference effects of two coherent packets are also studied using these different initial excitation depths. Stimulated emission of particles into the exciton packet by injection of thermal excitons in the crystal volume is investigated in greater detail as well as the attenuation of the lateral laser beam (used to create the thermal excitons that trigger the amplification phenomenon).

This thesis studies several aspects of the interdiffusion of InGaAs(P)/InP quantum well (QW) heterostructures, from the fundamental defect mechanisms, through optimization of processing parameters, to novel device applications. Conclusions from each of these areas have been drawn which further the scientific understanding and the manufacturability of the technique. The thermal stability of a series of different wafers is studied to highlight how poor quality of growth can cause increased interdiffusion, and to review the requirements for achieving repeatable annealing. Purposeful and controlled interdiffusion is accomplished through the introduction of excess defects into layers above the QWs, which during a subsequent anneal, diffuse through the QWs and enhance interdiffusion of atoms of the QWs with atoms of the barriers. These excess defects are introduced using two different techniques, via growth at low temperatures (LT) using chemical beam epitaxy (CBE), and via implantation of phosphorus ions. The CBE LT growth technique is new, and reported for the first time in this thesis. Characterization of the as-grown layers leads us to believe that they have an excess of phosphorus. The diffusion rate of the mobile defects which cause the intermixing is also measured, and the interdiffusion is shown to occur predominantly on the group-V sublattice. Due to many similarities between this and the results of the implantation technique, it is proposed that these mobile defects are the same for both intermixing approaches, and that the behaviour can be explained by a phosphorus interstitial mechanism. Annealing recipes for the implantation-induced technique are optimized, and the sample-to-sample reproducibility of the blueshift for this method was found to be quite good (standard deviations of ∼6 meV on blueshifts of ∼70 meV). The lateral selectivity and refractive index changes are characterized, and used in combination to create novel buried waveguide devices.

Four specific studies are presented in this thesis, regarding the high-pressure phases of the three compounds. ZrO$\sb2$ was studied at pressures of up to 73 GPa using energy dispersive X-ray (EDX) diffraction techniques. The quenched phase was analysed using angle dispersive X-ray (ADX) diffraction with imaging plates, and was identified as having a cotunnite-type structure (space group Pnma) by Rietveld refinement. HfO$\sb2$ was also studied to 72 GPa using EDX diffraction measurements, and its quenched phase was also identified as having a cotunnite-type structure, based on the simularities of its EDX and Raman spectra with those of the quenched phase of ZrO$\sb2$. TiO$\sb2$ was involved in two separate studies. The first is a Raman spectroscopic study up to 73 GPa, starting with anatase. Some EDX diffraction measurements were also taken to pressures up to 105 GPa. Additionally, EDX diffraction measurements of TiO$\sb2$ up to 28 GPa were taken to determine the equations of state of its high-pressure phases, starting with anatase and furtile at atmospheric pressure. (Abstract shortened by UMI.)

We have studied two examples of atomic processes induced by electronic excitation in insulating crystals. One is the bistable defect system in $\rm CdF\sb2{:}M\sp{3+},$ the other is the athermal halogen atom desorption. The bistable (shallow-deep) defect systems associated with trivalent impurities (In, Ga, Y and Sc) in CdF$\sb2$ are examined. The equilibrium lattice relaxation around the defect and the wavefunction of the electron bound to the impurities are determined by minimizing the energy of the defects. The impurity-fluorine interatomic potentials determined using the electron-gas model of Gordon and Kim are used, and the defect electron is treated by the extended-ion method. In order to compare the deep and shallow states using the same discrete lattice model, a very large cluster of atoms is treated. Two groups of trivalent impurity centers are found. With In and Ga, there is a low, but clearly identified, potential barrier which separates the deep level from the shallow one. In Sc and Y, only a simple shallow level state is obtained. The analysis of the results shows that the difference is to be attributed to the short-range potential of the trivalent impurity centers. On the basis of present work, we predict that Tl$\sp{3+}$ would exhibit similar bistable behaviour. Energetic halogen atom desorption observed from certain alkali halide crystal under electron or photon stimulation is being investigated. The relaxed structure of the localized one-center (Frenkel-type free exciton) and two-center (self-trapped exciton: scSTE) electronic excitations are studied. It is shown that the triplet scSTE state undergoes an instability on and near the surface similar to that in the bulk. For the first and second layers' scSTE below the (100) surface, the excited electron localizes with preference closer to the surface. Further below, there is no preference. Only when the off-center relaxation of the scSTE propels $\rm Br\sbsp{2}{-}$ toward the surface can there be a desorption. The third layer scSTE in KBr leads to desorption with a possible kinetic energy of about 1 eV, the trajectory undergoes deviation as the ejected Br$\sp0$ clears the surface. In NaBr, the scAPES of third layer's scSTE encounters a barrier as in the bulk, no energetic desorption is expected in NaBr as the experiment shows. The presence of a In$\sp+$ adjacent to a scSTE in the third layer, on the side closer to the surface, results in the excited electron localizing on the Br site nearer to the surface and on In$\sp+,$ pushing the V$\sb{k}$ center deeper into the bulk. We believe that this causes the observed inhibition of Br$\sp0$ desorption in In-doped KBr. In CaF$\sb2,$ though the excited electron localizing on the surface for the first layer's scSTE below the (111) surface, the F atom desorption is still expected due to the different lattice structure with the ejecta's trajectory undergoing correction to approach the normal of the surface. In all the examples studied here, the large atomic displacement is induced by the excited electron.

Vacancy enhanced compositional mixing of quantum well (QW) structures induced by ion beam implantation has shown itself to be a useful technique for achieving spatially selective tuning of the QW bandgap. As QWs form the active region of most optoelectronic devices, this method is being actively pursued as a means of monolithically integrating many such components onto a single substrate. This work presents a novel variation on this technique allowing one to deliver several times the maximum QW bandgap shift traditionally attributed to a single implant/RTA (rapid thermal anneal) cycle. This is achieved by distributing the total implanted ion dose over multiple sequential ion implantation and RTA cycles. Continuous-wave photoluminescence (CWPL) measurements are used to monitor the extent of QW intermixing taking place, and demonstrate the success of this innovation in various material systems (GaAs/AlGaAs, InGaAs/GaAs, InGaAs/InP). In another experiment, the effectiveness of high energy (8 MeV As$\sp{4+})$ implantation in intermixing GRINSCH (graded-index separate confinement heterostructure) QW laser structures grown in InGaAs/GaAs was demonstrated. Preliminary investigations were carried out on the effects of sample temperature, and ion current density during implantation on the intermixed QWs. Results from CWPL measurements suggest that better quality intermixed material can be obtained by implanting at elevated sample temperatures, and low ion current densities. Finally the effect of ion implantation induced intermixing on the effective radiative lifetimes in GaAs/AlGaAs quantum wells is investigated using the technique of time-resolved photoluminescence (TRPL). Below the critical dose (required to initiate disordering), the carrier lifetimes appear enhanced by the processing although no changes are discernible in the CWPL spectra. Above the critical dose, carrier lifetimes are reduced by residual defects created in the intermixed wells by the implantation procedure. These observations reflect the greater sensitivity of TRPL in detecting residual damage produced by intermixing QW material over CWPL.