<p>The state of polarization of photoluminescence that is emitted along (001) and (110) crystallographic orientations has been studied to obtain quantitative information on the interfacial strain, thickness and biaxial strain in InP/InGaAs(P)/InP quantum wells. It is demonstrated that the edge emission from quantum wells is a sensitive function of the strain and thickness of the quantum well. The anisotropic polarization of photoluminescence that is emitted normal to the plane of the quantum well provides information regarding the interfacial features. This anisotropy has been correlated to the anisotropic strain field that is associated with the strained bonds at the interfaces of the quantum wells.</p> / Doctor of Philosophy (PhD)

<p>Quantum wire (QWR) lasers are of interest because of predicted improvement of performance compared to quantum well lasers. For fiber optical communication, lasers are required that operate at wavelengths of 1.3 μm or 1.55 μm. This requires lasers grown in the InP-based materials. In the GaAs-based system, it has been shown that good QWR lasers can be achieved through epitaxial growth onto GaAs substrates having etched V-grooves. However, to date, no InP-based QWRs have been successfully grown in V-grooves. This arises from two problems: (i) the tendency for InP to planarize during growth, in contrast to the resharpening effect of AlGaAs; (ii) the potential for lattice-mismatch between InGaAs(P) and InP, which could result in strain relaxation and defect formation. Therefore, the purpose of this thesis is to establish the conditions whereby InGaAs/InGaAsP QWRs can be achieved and to confirm and characterize the one-dimensional behaviour. V-grooves with both(111)A and(111)B sidewalls are used in this study. They are obtained using chemical etching. The etching process is explained and conditions necessary to achieve each type of V-groove are established. Gas source molecular beam epitaxy has been used to grow various epitaxial structures. For InP layers grown under different growth conditions, it is found that the V/III flux ratio significantly affects the shape of a V-groove bottom and the roughness of both the V-bottom and the sidewalls. With the growth conditions optimized, InP layers can be grown which retain the sharpness and the smoothness of the V-groove. The sharpness of the bottom of the groove is related to growth conditions that decrease the growth rate at the bottom by limiting the supply of the group V component. InGaAs/InP quantum well structures have been grown using the optimized conditions. For (111)A V-grooves, transmission electron microscopy shows that all epilayers are defect-free and that InGaAs/InP quantum wires are successfully obtained with well thickness variation as high as a factor of 6. Lateral subband separations are estimated by a simple one-dimensional parabolic potential model with the thickness determined by TEM. Photoluminescence emission from the InGaAs quantum wires is spatially resolved with a spatially selective etch technique. For (111)B V-grooves, defects such as dislocations are observed in the bottom. However, the growth of bulk InGaAsP and InGaAs/InGaAsP yields different results. No extended defects have been observed in the InGaAsP layer grown in either (111)A or (111)B V-grooves. This is because, at the groove bottom the layer is both P and In rich, as analyzed by using energy dispersive X-rays. This tends to reduce the strain compared to InGaAs deposition, where the absence of P leads to In-rich, strained material. The growth of an InGaAsP layer in a (111)A V-groove results in a flat and wide bottom, which excludes the possibility of forming InGaAs/InGaAsP quantum wires. However, the growth of an InGaAsP layer in (111)B V-grooves results in a sharper bottom, such that crescent-shaped InGaAs/InGaAsP structures are formed when the InGaAs layer is very thin. Quantum-wire behaviour has been confirmed through the observation of lateral subbands in the photoluminescence (PL) spectra. This is further supported by polarization measurements. Also, the subband separation observed in PL spectra is consistent with a calculated value using the one-dimensional parabolic potential model. The InGaAs/InGaAsP QWRs should be able to be incorporated into laser structures.</p> / Doctor of Philosophy (PhD)

<p>InP and related materials have become important materials for applications in many electronic and photonic devices because of their high electron mobilities and high saturation electron drift velocities. InGaAs and InGaAsP, grown lattice matched to InP, are the materials of choice for making light sources and detectors in the wavelength range of 1.3-1.6 μm for the present day high data rate, long haul, fibre-optic communication systems. InGaAs metal-semiconductor-metal (MSM) photodetectors, InGaAs high-electron mobility transistors (HEMTs) and their monolithic integration as photoreceivers are promising devices for opto-electronic integrated circuits (OEICs). In the fabrication of the InGaAs MSM-PDs and HEMTs, good Schottky contacts are critical to achieve high performance, such as low dark current and high breakdown voltage, of the devices. However, the Schottky barrier height of metal/InGaAs contacts is only about 0.2-0.3 eV, which is not high enough for device applications. InP was chosen as the contact material in our fabrications of InGaAs MSM-PDs and HEMTs because of its compatibility in growth and lattice matched to In₀₅₃Ga₀₄₇As. In this research, the following investigations were carried out to understand the physics of the InP Schottky barrier, to obtain good Schottky contacts to InP, and to improve the performance of the InP/InGaAs MSM-PDs. (1) It was found that the current-voltage characteristics of metal(s)-InP Schottky contacts depart from the behavior predicted for the thermionic emission model. To explain these barrier anomalies. a modified current-voltage relationship was developed based on the inhomogeneous Schottky barrier. (2) A systematic investigation on metallization schemes shows that the Schottky barrier height of metal(s)-InP ranges from 0.38 to 0.5 eV, depending on the metal(s) and deposition techniques and the surface conditions of the InP. Among them, Au/Ti/Pt produces the highest barrier height while Au/Ni/Pt has the best uniformity in metal(s)-InP Schottky contacts. To further increase the Schottky barrier height, a thin In1-xGxP (x ranges from 0.1 to 0.3) layer of 100-200 Å grown on top of InP was employed. The barrier height obtained is found between 0.56 to 0.65 eV. The dark current is significantly reduced while the reverse breakdown voltage is increased from about 15 V to 20 V. (3) High-performance InGaAs MSM-PDs with InP as the barrier layer was fabricated. The MSM-PDs with a finger width of 2μm, finger spacing of 2μm, and an active area of 50 x 50 μm², have a dark current of about 200 nA at a bias of 10 V, a low capacitance of 0.2 pF, and breakdown voltages of about 15 V. A fairly high responsivity of 0.75±0.05 A/W at λ=1.3 μm was obtained. The temporal response is characterized by a rise time of about 4.0±0.2 ps, fall time of 8.0±0.2 ps, and FWHM of 8.5±0.2 ps which corresponds to a 3-dB frequency (unit current gain cut-off frequency), f3dB, of 20 GHz. The performance is comparable to the best of those InGaAs MSM-PDs with InAlAs as the contact layer. (4) Because of the lateral conduction in MSM-PDs and the Fermi-level pinning in InP due to surface states, the surface conditions play an important role in device performance. Surface passivation using (NH₄)₂Sx solution was successfully implemented to improve the performance of InGaAs/InP MSM-PDs.</p> / Doctor of Philosophy (PhD)

<p>We present a study of non-uniform pumping effects in semiconductor lasers. The first portion of this work is a theoretical undertaking, in which a physical model of a split-electrode, ridge-waveguide, InGaAsP/InP laser is developed. This model is based on the time-dependent solution of the carrier and photon rate equations, and is capable of describing such device features as multiple electrical contacts, illumination of the facets via an external light source, multiple optical cavity modes, and continuous variations in the carrier and photon concentrations. The simulations generated from this model demonstrate: (1) output power vs. bias characteristics that include threshold control, discontinuities, and bistability, (2) wavelength tuning via gain peak shifts and varying refractive index, (3) self-sustained pulsations caused by repetitive Q-switching, (4) external injection effects such as injection locking, all-optical switching, and optical self-pulsation control, and (5) current modulation characteristics. The effects of varying the device bias and geometric parameters on these phenomena are presented. The second part of the thesis consists of an experimental investigation of a series of multi-quantum well InGaAsP/InP split-contact lasers. By varying the bias currents and contact lengths, we have studied their effects on the output power and spectra of these devices. The experimental results we present are in good agreement with the simulations based on our theoretical model.</p> / Doctor of Philosophy (PhD)

<p>An experimental and theoretical study of the dynamics of NH₃ lasers is presented. The significant achievements of this work are summarized below. In initial experiments, a pulsed transversely-excited CO₂ laser operating on the 9R(30) transition is used to optically pump mixtures of NH₃ in buffer gas. A simple oscillator/amplifier system allows the performance of the NH₃ amplifier in the 11 μm region to be characterized. Small-signal gain coefficients of >10%/cm are measured on the aQ(3,3) transition at 10.8 μm, while pump conversion efficiencies of ~50% are shown to occur under saturation conditions. The NH₃ laser system is described by a rate-equation model, which is validated by comparison with experiment over a wide range of operating conditions. Measurements are made for NH₃ concentrations ranging from 0.05 to 0.2%, for Ar, N₂ and He buffer gas pressures from 170 to 700 Torr, and for gas temperatures from 200 to 300 K. Optically pumped NH₃ is shown to be a versatile and efficient system for the amplification of mid-infrared radiation. The rate-equation model is used to aid in the design of a simple and efficient NH₃ laser. This laser is tuned over more than 70 vibrational band transitions between 10.08 and 14.14 μm. Output energies greater than 1 J per pulse are achieved on several of the strongest lines. In a non-selective cavity an energy conversion efficiency of greater than 35% is obtained with a maximum output energy of 4.6 J. Optically pumped NH₃ is shown to be a flexible and efficient system for the downconversion of CO₂ radiation to the 10-14 μm region. In other experiments, it is shown that optically pumped high pressure mixtures of NH₃ in N₂ are efficient, broadband amplifiers of pulsed CO₂ radiation. In a dilute NH₃ mixture at 6 atmospheres and 200 K, a single pass gain of 150 (21.8 dB) is measured for the 10P(34)CO₂ transition. Gain is observed in NH₃ at pressures as high as 10 atmospheres. Experimental measurements are made for a range of wavelengths in the 10.7 μm region, and the results compared with calculations based on a simplified rate-equation model. The operation of a two-step optically pumped NH₃ laser (using two CO₂ lasers) is examined. Output is obtained from 16-22 μm. Experimental measurements made at 200 K and 300 K are compared to calculations based on an extended rate-equation model. As a result of this work it was shown that a single CO₂ pump laser can be used to obtain 16-22 μm lasing at 200 K.</p> / Doctor of Philosophy (PhD)

<p>For decades, microwaves have been used to create gas discharges for many applications. This thesis deals with the use of microwaves to excite gas discharges for incoherent optical sources, with particular emphasis on excimer systems. In addition, microwave excitation of a gas laser is considered. A novel apparatus was designed and built to couple 2.45-GHz microwave radiation into a gas discharge. The microwave resonator is the essential part of this equipment, and a detailed discussion of its design and performance is given. The resonator is characterized both theoretically and experimentally in order to determine the coupling efficiency and peak electric-field strength. Specialized theory is developed in order to evaluate many parameters of a microwave-excited discharge. The phenomenon of skin effect is investigated quantitatively and expressions for the plasma frequency and electron density are developed in terms of collision frequency and observable parameters (e.g., skin depth). Expressions for peak electric-field strength, ionization coefficient and collisionless electron energy are also developed. The results of an extensive investigation of continuous-wave microwave-excited excimer fluorescence are reported. Rare-gas halide, homonuclear halogen and heteronuclear halogen systems are examined and the corresponding ultraviolet spectra are presented. Truly continuous excimer emission has been achieved (for the first time) on several transitions. For systems of particular interest (e.g. XeCl and KrCl), the effects of total pressures and gas composition on fluorescence output are investigated, and the appropriate spectra are presented. Finally, the potential operation of microwave-excited carbon dioxide and argon-ion gas lasers is investigated, and upper limits are deduced for the small-signal gain under various conditions.</p> / Doctor of Philosophy (PhD)

<p>A technique for obtaining maps of the spatially-resolved and polarization-resolved electroluminescence of 1.3 μm semiconductor diode lasers has been developed. It reveals information about the strain, scattering and absorption which exist in the active region of lasers. These data have been correlated with the spectral output of the lasers, and models have been developed which explain the correlation.</p> / Doctor of Philosophy (PhD)

<p>A chemical kinetic model has been constructed to predict the pressure and gas and electron temperature dependency of the neutral and ionic species composition in Ar-CO₂ mixtures under thermal plasma conditions. Pure Ar and Ar-C models have also been constructed as a part of this investigation. The models include electron impact, thermal impact, ion-molecule, and recombination reactions as well as accounting for diffusion. Important metastable and excited states of species have been accounted for as well as the presence of neutral molecules, radicals, and atoms, positive and negatively charged atoms and molecular ions as well as electrons. All relevant electron temperature, gas temperature and pressure terms have been included, and primarily experimentally derived reaction rate constants were utilized. Electron and gas temperature dependent species concentrations were obtained under both thermodynamic equilibrium and non-equilibrium conditions for gas temperatures from 300 to 15000 K, electron temperatures from 300 to 20000 K, and pressures from 1 Torr to 15200 Torr. Percentage mixtures of C and CO₂ in Ar were varied between 0.1 and 40%. Results indicate that the neutral and excited species Ar, Ar*, Ar**, C, CO, CO₂, O, O₂ and O₃, positive ions Ar⁺, Ar₂⁺, C⁺, CArO⁺, CO⁺, CO₂⁺, CO₄⁺, C₂O₂⁺, O⁺, O₂⁺, O₄⁺, and O₅⁺ and negative ions CO₃⁻, CO₄⁻, O⁻, O₂⁻, and O₃⁻ and electrons are observed. Under thermodynamic equilibrium conditions, Ar, C, CO, CO₂, and O were found to be the dominant neutral species, C⁺, CO₄⁺, O⁺, and O₂⁺ the dominant positive ions, and O⁻ and electrons the dominant negatively charged species under certain gas temperature ranges. The introduction of thermodynamic non-equilibrium conditions, changes in gas pressure, and percentage mixture of C or CO₂ were observed to have a significant influence on the temperature dependent concentrations of these species. The results were found to be in good agreement with previous experimental and theoretical results. Comparison of the present results with the results of the Saha equation seem to indicate that the simple Saha equation type approach may mispredict the plasma density of Ar, Ar-C, and Ar-CO₂ mixtures under thermal plasma conditions.</p> / Doctor of Philosophy (PhD)

The study presented in this thesis was motivated by the large uncertainty on the concentration of atmospheric electrical discharges to the global nitrogen budget. This uncertainty is partly due to the fact that information concerning the NOx production efficiency of electrical discharges having current signatures similar to those of lightning flashes is not available in the literature. Another reason for this uncertainty is the fact that energy is used as a figure of merit in evaluating NOx production from lightning flashes even though insufficient knowledge is available concerning the energy dissipation in lightning flashes. The third reason for this uncertainty is the lack of knowledge concerning the contribution of discharge processes other than return strokes to the NOx production in the atmosphere. Lightning is not the only process in the atmosphere that causes ionisation and dissociation of atmospheric air. Cosmic rays continuously bombard the Earth with high energetic particles and radiation causing ionization and dissociation of air leading to the production of NOx in the atmosphere. The work carried out in this thesis is an attempt to improve the current knowledge on the way in which these processes contribute to the global NOx production. Experiments have been conducted in this thesis to estimate the NOx production efficiency of streamer discharges, laser-induced plasma, laboratory sparks having current signatures similar to those of lightning flashes, alpha particle impact in air and finally with the lightning flash itself. The results obtained from laboratory electrical discharges show the following: (a) The NOx production efficiency, in terms of energy, of positive streamer discharges is more or less similar to those of hot discharges. (b) The NOx production efficiency of an electrical discharge depends not only on the energy but also on the peak and the shape of the current waveform. (c) The current signature is a better figure of merit in evaluating the NOx yield of electrical discharges. As a part of this thesis work a direct measurement of NOx generated by lightning flashes was conducted and the results show that slow discharge processes such as continuing currents could be the main source of NOx in lightning flashes. Concerning NOx production by other ionisation processes such as alpha particle impacts in the atmosphere, the data gathered in this thesis show that each ionising event in air leads to the creation of one NOx molecule. In terms of energy the NOx production efficiency of alpha particles is similar to that of electrical discharges. The theoretical studies conducted within this thesis indicate that M-components contribute more than the return strokes to the NOx production. The calculations also show that the contribution to the global NOx budget by return stroke is not as high as that assumed in the current literature.

<p>In several x-ray related areas there is a need for high-precision elements for x-ray generation and focusing. An elegant way of realizing x-ray related elements with high precision and low surface roughness is by the use of microfabrication; a combination of semiconductor processing techniques and miniaturization. Photolithographic patterning of silicon followed by deposition, etching, bonding and replication is used for batchwise fabrication of small well-defined structures. This thesis describes microfabrication of a miniature x-ray source and a refractive x-ray lens. A miniature x-ray source with diamond electrodes has been tested for x-ray fluorescence. Another version of the source has been vacuum encapsulated and run at atmospheric pressure. One-dimensionally focusing saw-tooth refractive x-ray lenses in silicon, epoxy, and diamond have been fabricated and tested in a synchrotron set-up. Sub-micron focal lines and gains of up to 40 were achieved. The conclusion of the thesis is that the use of microfabrication for construction of x-ray related components can not only improve the performance of existing components, but also open up for entirely new application areas.</p>