No description available.

No description available.

No description available.

by Ho Hing Wa. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1999. / Includes bibliographical references. / Abstracts in English and Chinese. / Abstract --- p.i / 摘要 --- p.iii / Acknowledgment --- p.iv / Table of Contents --- p.v / Chapter 1 --- Introduction / Chapter 1.1 --- Background of optical generation of millimeter-waves --- p.1 -1 / Chapter 1.2 --- Application of the two-section DFB laser on optical generation of millimeter-waves --- p.1 -4 / Chapter 1.3 --- Analysis --- p.1-4 / Chapter 1.4 --- Organization of Thesis --- p.1 -5 / References --- p.1-6 / Chapter 2 --- Techniques of Optical Generation of Millimeter-waves / Chapter 2.1 --- Direct modulation and mode-locking of lasers --- p.2-1 / Chapter 2.2 --- Beating of two optical waves --- p.2-3 / References --- p.2-11 / Chapter 3 --- Experimental Results and Discussions / Chapter 3.1 --- Device structure and experimental setup --- p.3-1 / Chapter 3.2 --- Light-Current Characteristics of the two-section laser --- p.3-3 / Chapter 3.3 --- Spectral behaviours of the two-section DFB laser --- p.3-5 / Chapter 3.3.1 --- Linewidth of the two-section DFB laser --- p.3-5 / Chapter 3.3.2 --- Wavelength tuning of the two-section DFB laser --- p.3-5 / Chapter 3.3.3 --- Biasing conditions for the dual-mode oscillations --- p.3-16 / Chapter 3.4 --- Optical generation of millimeter-waves --- p.3-17 / Chapter 3.4.1 --- Mechanism of beating --- p.3-17 / Chapter 3.4.2 --- Generation of millimeter-waves by optical beating --- p.3-20 / Chapter 3.5 --- Optical Transmission of the millimeter-waves --- p.3-22 / References --- p.3-24 / Chapter 4 --- Theory of DFB Laser / Chapter 4.1 --- Development of DFB laser --- p.4-1 / Chapter 4.2 --- Structure of DFB laser --- p.4-2 / Chapter 4.3 --- Model of one-section DFB laser --- p.4-4 / Chapter 4.4 --- Analysis of two-section DFB laser --- p.4-10 / Chapter 4.4.1 --- Introduction of transfer matrix method --- p.4-11 / Chapter 4.4.2 --- Formulation of transfer matrix --- p.4-12 / Chapter 4.4.3 --- Application of the transfer matrix --- p.4-13 / References --- p.4-17 / Chapter 5 --- Numerical Analysis of the Spectral Behaviours of the Two-Section DFB Laser / Chapter 5.1 --- Solving the Transcendental Equation --- p.5-1 / Chapter 5.2 --- Simulation of the spectral behaviour of the two-section DFB laser --- p.5-4 / Chapter 5.2.1 --- Assumptions and Approximations --- p.5-4 / Chapter 5.2.2 --- Parameters --- p.5-5 / Chapter 5.2.3 --- Computer Implementation --- p.5-6 / Chapter 5.2.4 --- Results and Discussion --- p.5-7 / References --- p.5-14 / Chapter 6 --- Conclusion / Chapter 6.1 --- Summary --- p.6-1 / Chapter 6.2 --- Future works --- p.6-2 / Appendices / Appendix A Source code for simulation of spectral behaviours of the two-section DFB laser --- p.A-1

This thesis presents an investigation of the modal gain characteristics of Vertical Cavity Surface Emitting Lasers (VCSELs). This is motivated by the experimental observations which tend to indicate fundamental mode operation at just above lasing threshold and multi-transverse mode operation at higher injection current levels. The complete mode spectrum of the cylindrical cavity is first analysed to illustrate that the original expectations that single (wavelength) mode operation solely due to the large wavelength separation created by short device lengths may not be realisable. The modal gains of different modes are calculated to demonstrate that mainly the difference between the modal gains, and not the separation between the resonances, provides a more satisfactory explanation of the fundamental mode operation at just above lasing threshold. At higher injection current levels, the increase in the modal gains of the higher order transverse modes (due to spatial hole burning) explains the excitation of higher order transverse lasing modes. The model relies on calculating the modes of a cylindrical dielectric resonator and the the corresponding modal gains are obtained from a perturbation analysis which takes into account the gain profile due to the injected inversion population distribution. A self consistent evaluation of the inversion population distribution (which provides the required local gain profile) is derived from the corresponding diffusion equation for the injected carriers in the active layer of the device. The development for obtaining the above lasing threshold inversion population distribution (which includes radial and azimuthal variation) has been done hierarchically such that the numerical procedures developed in the simpler stages of the model directly apply to solve a part of the next level of sophistication. This hierarchy has helped to provide a very efficient and compact numerical procedure and may be seen as an important aspect of the work done in the thesis. Further refinements include the evaluation of the injection current profile dependence on injection contact geometry and current spreading.

Strained layers are incorporated into many electronic devices and particularly into semiconductor lasers. These strained layers can relax, both elastically and plastically, which often impairs the performance of the device. This thesis presents several methods for calculating elastic strain relaxation: a Fourierseries method for stresses imposed on the surfaces of a rectangular block; a Fourier-integral for stress imposed on the surfaces of an infinite layer; and a Green-function method for the stress field about buried inclusions. The methods are used to calculate the strain distributions in a transmission electron microscopy sample, the relaxation at the end facet of a strained-layer laser, and the strain field about a rectangular buried layer. The effects of the strain relaxation on the optical absorption of the laser facet and the zone-centre band structure of the buried layer are discussed. The equilibrium theory of critical thickness is examined in detail and is shown to make unreasonable predictions for highly strained layers; a modification which corrects this behaviour is suggested. The equilibrium theory equates the line tension of a strain relieving dislocation to the strain energy it relieves in the layer. The additional energy corrections which can be included in the line tension are discussed, together with the failure of the equilibrium theories to reliably predict plastic relaxation in all situations.

No description available.

Vertical cavity surface emitting laser (VCSEL) structures have been grown by both metal-organic chemical vapour deposition (MOCVD) and molecular beam epitaxy (MBE). These incorporate 3 strained InGaAs / GaAs quantum wells placed resonantly in a two wavelength long optical cavity, formed between AlAs / GaAs quarter wave dielectric reflector stacks through which current is injected. The reflection spectra of these stacks is studied in detail; the effects on the laser threshold gain of absorption due to impurities and of errors in growth are investigated. Methods of disruption of the AlAs / GaAs heterointerfaces have been used to reduce the operating voltage. The completed designs use 200A intermediate layers containing 30 or 50% aluminium or a superlattice graded region simpler than that used in previous designs. The effectiveness acceptor dopants; Be in MBE, C and Zn in MOCVD; is studied also. Modulation doping was employed to reduce the effects of optical absorption. Devices were fabricated into mesas by SiC14 reactive ion etching or defined by proton implant isolation. MBE grown devices were resonant at wavelengths in the range 950 to 1059mn with essentially constant (at —1020nm) eihhi transition energies in the wells. A detailed study of the wavelength variation of threshold current density Jth (X)was made. A minimum of 366A.cnr2 was measured at 1018nm in mesa devices. A similar relation is found for ion-implanted devices but the minimum is increased to 535A.cm-2 by incomplete isolation. Gain calculations, including strain effects, are used to explain the Jth(X) variation. Implanted devices offer superior c.w. performance due to reduced thermal and ohmic resistances. The relative offset between the gain spectrum and cavity resonance was examined for c.w. operation. It was found that carrier thermal effects limit the output power rather than shifts in the offset. The bias voltage of MOCVD grown devices is as low as 1.7V and the threshold current is as low as 764A.cm-2. This is higher than for MBE grown devices because of growth thickness errors and non-optimal alignment of the gain spectrum and cavity mode. The uniformity in emission wavelength is ±1% over 80% of a 2 inch diameter wafer, offering suitability for very large uniform arrays.

No description available.

Thesis (Ph.D.) -- McMaster University, 1997. / Includes bibliographical references. Also available via World Wide Web.