The operation and the theory of the effective-index method for the analysis of a wide range of optical waveguides are reviewed. The asymptotic results that characterize the performance of the method for rectangular waveguides and waveguide arrays are discussed. An accurate version of the method, namely the effective-index method with a built-in perturbation correction, is described. The applications of the effective-index method to nonrectangular waveguides and nonlinear waveguides are also discussed.

This paper presents concepts behind the use of the beam propagation method (BPM) to model the vector nature of optical waves, and for wave propagation at wide angles off-axis. The two enhancements have been combined.

The propagation of electromagnetic fields in straight and circularly curved channel waveguides can elegantly be described in terms of guided modes. They represent non-vanishing solutions of the homogeneous Maxwell equations that propagate without deformation through a waveguide. The source-type integral equation method has already proven to be a powerful method for determining the guided modes of a straight channel waveguide configuration. The method is full-vectorial and mathematically rigorous. In this paper, the method is extended to the circularly curved channel waveguide, as encountered in integrated-optical and opto- electronic devices. A multitude of numerical results is presented.

A novel analytic method for analyzing hybrid DFB structures containing multiple linear and nonlinear sections is presented. The technique is illustrated by considering uniform and phase- shifted nonlinear devices that play the roles of an amplifier, a filter, and an all-optical switch simultaneously.

A new numerical method is suggested for wide-angle beam propagation and transverse pattern formation in nonlinear media described by the Helmholtz scalar wave equation. The method is based on the Gauss-Laguerre decomposition of the transverse beam profile and solution of the set of ordinary differential equations for modal coefficients. The total field including the forward and backward waves is calculated. The role of the backward waves in wide-angle Kerr self-focusing is investigated. It is shown that within the range of parameters used in previous numerical studies no evidence for the backward reflection of the beam by the focal region arises. This agrees with the estimates of the refraction index variation near the self- focus.

A scheme of a very fast way to make all-optical addition and subtraction realized in a nonlinear planar waveguide is presented. The scheme is based on interaction properties of solitons.

We describe simulation of photon-induced modulation to understand the interface effects on the optical response of a low doped GaAs layer. We also present measured photoreflectance lineshapes from low doped (1 to 3 X 10¹⁶ cm ^-3) GaAs layers with different interfaces. These layers are GaAs; with a semi-transparent gold overlayer, with a heavily doped underlayer and an air exposed single layer. A comparison between the simulation and the experimental lineshape indicates three possible photomodulation mechanisms, each with its own characteristic lineshape. This work shows that it is essential to use a multilayer model in simulation in order to accound for the slight gradient in optical response of successive nanometer scale sublayers within a single layer of the material. In this study the gradient is caused by changes in the electric field within the space-charge region, which affects the optical response of the sample. We found that the details of photoreflectance lineshape of each layer depends on the electrical and structural parameters of its immediate interfaces and the electrical characteristics of its neighboring layers.

The electromagnetic characteristics of quantum-well structures with periodic dielectric perturbations are investigated numerically. The numerical solution of the differential equation is improved by using the implicit Range-Kutta method. By studying the mappings in the complex plane, the rule of choosing the sign of the transverse wave vector in the material with complex dielectric constant is established. In the discussion of the propagation characteristics, a comprehensive explanation of the resonance phenomena in periodic dielectric waveguides is presented. The main aim of the study is to determine grating strengths by analyzing the reflection and transmission characteristics of a quantum-well structure with a finite length grating. The power reflection and transmission are related to the grating strength in the various periodic dielectric waveguides.

A S-matrix propagation algorithm for multilayer planar structure, refractive index in each layer being a function of one lateral dimension in layer's plane is developed. The algorithm is described in symbolic operator form not tied to specific representation in which boundary value problem for Maxwell equations might be solved in the layers. In the case of grating layers where refractive indices are periodic with the unique period two Fourier transform based techniques of numerical solution are implemented with S-matrix propagation algorithm. The examples of the optimal design simulations: Si grating-based zero reflection surfaces and perfect reflection layer/substrate structures are considered.

A new coupled mode theory is formally developed for analyzing the grating-assisted codirectional couplers. We show its superiority and inferiority with respect to existing theories. It belongs to the orthogonal category of coupled mode theories.

Interaction of electromagnetic guided modes with index/gain Bragg gratings in step-index round optical fibers is considered theoretically. The analysis includes the construction of exact solution of Maxwell equations in core and cladding regions, which strictly obeys the electromagnetic boundary conditions on core/cladding interface and on Bragg grating boundaries. The grating profiles, for which the exact solutions of Maxwell equations can be expressed in the closed analytical form, are determined and the corresponding expressions for scattered electromagnetic fields are presented. The constructed solution enlarges the class of exactly solvable diffraction problems. It can be used to examine and to control numerical theories of light propagation and scattering optical fibers with intracore Bragg gratings.

Experimental and theoretical results for gain in bulk and multiquantum well active layer 1.3 micrometers InGaAsP based lasers are reported. Gain, loss, transparency energy, and carrier density have been measured in the subthreshold regime at room temperature and elevated temperatures. Gain has been calculated using an eight band k(DOT)p model for the electronic structure and a conventional density matrix formulation. The calculated and experimental results for the gain spectra, the gain versus density, the chemical potential (quasifermi level separation) versus density, and the gain versus chemical potential are compared at room temperature and 85 C. There is aagreement at several points, but the model substantially underestimates the temperature sensitivity of the gain which has been found in the experiments to be an important factor in the overall temperature sensitivity of threshold current.

The gain dynamics of bulk laser amplifiers is discussed on the basis of simple rate equations, and comparison is made with experimental pump-and-probe measurements employing ultrashort optical pulses. The rate equation model is derived from semi-classical density matrix equations, and is analyzed in different limiting cases, i.e., an adiabatic limit corresponding to nonlinear gain in laser diodes, a small-signal limit yielding an analytical repsonse function, and a strong-signal regime, where numerical simulations are carried out.

In this paper we present a small signal equivalent circuit model for two axially coupled semiconductor lasers. Instead of using the usual approach through the linearization of the device rate equations to obtain the circuit parameters, we postulate an equivalent laser with the same dynamic response as the coupled-cavity laser. We start charcterizing the small signal repsonse of two axially coupled lasers in terms of their oscillation relaxation frequency response through numerical integration of their rate equations. Then we compare it to that obtained from our first circuit model.

The Hopf bifurcation points for single mode solutions of the Lang and Kobayashi equations are determined using asymptotic methods. The approximation is based on a small parameter (epsilon) which is defined as the ratio of the photon and carrier lifetimes. The remaining parameters are scaled with respect to (epsilon) . The critical feedback rate for a Hopf bifurcation is studied in terms of the pump parameter and is either an 0((epsilon) ) or an 0((epsilon) ^1/2) quantity. At a fixed value of the pump parameter, we obtain an expression of the Hopf bifurcation point in terms of the effective feedback strength and the feedback phase. In addition, we investigate the Hopf bifurcation point near and at the lasing threshold.

In a free-running broad-area semiconductor laser, complex spatio-temporal patterns are observed in the near-field output intensity as a result of self-focusing, filamentation, and transverse modulational instabilities. In this paper, results from extensive numerical simulations on the basis of the microscopic semiconductor-Maxwell-Bloch model equations are presented and discussed. The physical processes which lead to the formation of the spatio- temporal patterns and which are involved in the mutual interactions between the light field and the active semiconductor medium on microscopic scales manifest themselves, e.g., in the simultaneous relevance of spectral and spatial hole-burning effects in the charge carrier distributions of the broad-area laser. The dynamic behavior and the complexity in space and time are analyzed with methods from nonlinear dynamics by various theoretical tools: transversely dependent cumulants of the bit-number as well as characteristic cross-correlation functions are computed and the transverse generic eigenmodes of the running laser showing individual dynamics are determined.

We show theoretically, that the detuning between the stop bands of the two DFB sections gives rise to different self pulsation types: i) dispersive self Q-switching of a single mode (around 10 GHz repetition rate), or ii) beating oscillations between two modes of equal threshold gain (more than 100 GHz). Our analysis is based on the dynamic coupled wave equations, together with carrier rate equations. Using a mode expansion of the optical field, we perform a stability and bifurcation analysis as well as a direct numerical integration of the full equations, which demonstrates the stability and bifurcation analysis as well as a direct numerical integration of the full equations, which demonstrated the appearance of the two self pulsation types. In the examples investigated, we find finite self pulsation islands in the plane spaned by the two currents, which is in qualitative agreement with published experimental results.

In this paper we study the behavior of a single mode semiconductor laser under strong signal modulation. To describe the semiconductor laser, we employed the standard rate equation model, commonly used for this type of analysis. A continuation method is used in order to study the evolution of the behavior of the laser as the modulation index is increased. The bifurcation points are determined using Floquet Theory. In order to introduce the parasitic effects, we have used the equivalent circuit model. First, the intrinsic model is used to verify the results obtained with the continuation method and after, we introduce the laser parasitics to obtain new results on the dynamic behavior.

This communication deals with optical noise measurements of an erbium-doped fiber amplifier. The output quantum noise depends on the amplified spontaneous emission power because no input signal is applied. Thanks to a noise simulation, we have been able to deduce the internal gain G of the doped fiber and the spontaneous emission factor n_sp.

Laser phase noise deteriorates the high sensitivity of heterodyne optical receivers. To reduce phase noise influence, the intermediate frequency signal resulting from the coherent detection is filtered by a narrow bandpass filter (BPF). The phase noise at the input of the BPF generates an amplitude and phase noise at the output of the BPF. The joint probability density function of these noises is evaluated in the case of a first order filter by numerical resolution of a Fokker-Planck equation. A finite difference operator splitting scheme is used. The accuracy of the numerical solution is checked comparing numerically and analytically calculated moments. In addition, a new very efficient method for the analytical calculation of moments is developed. Contour plots of the probability density for both a finite time integrator and a first order filter are compared in order to show the impact of different filter types on phase noise filtering. The marginal pdf of the amplitude and phase noise at the output of the above filters are also calculated.

System synchronization is an important technology in constructing all-optical signal processing systems. To realize this, an all-optical timing extraction circuit is required, which recovers timing information from an incoming optical data stream, and produces an optical clock without intermediate electric stage. A potentially simple method of all-optical clock recovery uses self-pulsating laser diodes (SP-LDs), and many experimental studies have been reported so far. However, there is no report on theoretical works to our knowledge except the perturbation analysis by Lee and Shin. In this paper, we report numerical analysis of all- optical clock extraction using the SP-LD based on rate equations. The model used in this analysis is two-section SP-LD. The laser is divided into a gain region (region I) and a saturable absorber region (region II). The carrier lifetime of region II is much shorter than that of region I. THe photon density of injection signal is coherently coupled to the SP-LD light. To obtain the time response of the laser, we solved rate equations numerically. It is shown that we can extract clock pulses from signals which contain regulated digital patterns or a pseudo- random binary sequences (PRBS) data pattern. The results will explain the experimental results reported so far. Relative phase of the extracted clock with respect to the input signal varies when the power of the input optical signal fluctuates or a PRBS data pattern is used. However, this inherent effect can be minimized by using low input power level, although the locking range becomes narrow.

The dynamic response of a multiple-quantum well (MQW) laser is investigated assuming that it can possess multiple levels. The results reveal that the lowest heavy-hole and the lowest light-hole transitions are dominant. In the low-power range the heavy-hole transition dominates, while in the high-power range there is a mixing of the heavy-hole and the light- hole transitions. In the last case the frequency response is improved and it can be correctly predicted only treating the MQW as a three-level system.

Opto-electronic signal in the injection laser diode arises on the frequencies of mode beating of external cavity when internal multimode coherent radiation is detected by the active laser chip. Dynamic coefficient of transformation from radiation to injection current is measured in the range 150 MHz to 5 GHz. Fall down of frequency response in the range higher than 1 GHz was discovered, which is connected to finite life time of carriers in active channel.

The transmission-line laser model (TLLM) is a well established method for the modeling of semiconductor lasers. In this paper we present an improvement of this method by using digital signal processing. The advantages provided by this technique are analyzed in this paper, which include a flexible approach to the gain shape and a better control of the computing-time versus accuracy balance. The new model is well prepared to include complex structures based on semiconductor emitters such as external cavities, coupled-cavity structures or even diode arrays.

Additional development of the four areas identified in this paper would provide useful information to those modeling existing semiconductor lasers and to those that hope to design future high performance semiconductor lasers. These four problem areas are by no means unique, and there are numberous problems to fill any void left by thorough solutions to the issues raised herein.

The stripe area of a semiconductor laser with a diverging lenslike medium and a rectangular current-injection stripe is limited by efficiency considerations and by design constraints required to provide a coherent output beam. A larger stripe area is practical in a compound resonator with a lenslike medium in one portion and free expansion of the mode in the other portion. The stripe edges along the direction of propagation should match light rays of the mode on the output pass, at least approximately. Equations for the geometric properties of such a laser are derived, and a numerical simulation indicates that output power in excess of 4 W should be practical impulsed operation. The code did not include thermal effects, so CW operation is not simulated.

It has recently become apparent that a comprehensive quantum well (QW) laser design strategy must address the complicated distribution of carriers both in real and in momentum space throughout the laser structure. Therefore, in addition to appropriate material systems, optimized QW width, and proper light guiding structures, one must carefully optimize the width and the doping of the separate confinement heterostructure (SCH), and tailor the cladding layer/SCH and the SCH/QW bandgap discontinuities. In this talk, we describe new simulation techniques to address these issues and illustrate the importance of optimizing bandgap discontinuities through a study of hot phonon effects on laser performance.

Experimental evidences and modeling calculations are given for the existence of negative differential mode gain in ridge-waveguide laser diodes on the base of strained-layer InGaAs/GaAs quantum-well structures. The phenomenon is found to be related to a mode formation in an active 2D waveguide with monotonic increase of the material gain. The mode gain is calculated in single- and double-QW laser structures at various waveguide parameters including variation of the lateral built-in index step.

Threshold characteristics of quantum-well (QW) lasers have been investigated theoretically. The rates of radiative and Auger nonradiative carrier recombination processes in QW have been calculated. Auger rate dependence on the parameters of QW and the temperature has been investigated. It is shown that Auger rate is a nonmonotonic function of QW thickness a; it has a maximum at small a. Auger rate increases with the barrier height and decrease rapidly with band gap; Auger rate is a power function of temperature. It was shown that the carrier concentration on the generation threshold is a linear function of temperature. Threshold current density is shown to be a power rather than exponential function of temperature.

A simple model is developed to describe hot carriers, hot phonons, carrier transport effects on intensity modulation response of quantum well lasers. It is shown that for modulation frequencies smaller than the inverse of phonon lifetime, the contribution to the response function from phonon heating can be characterized by a time constant for electron temperature relaxation and phonon number modulation factor. At small photon densities analytical expressions for the K-factor and the nonlinear gain coefficient are derived. A new mechanism is described that may limit modulation bandwidth due to multiplicative effect of carrier heating and carrier overflow. It is demonstrated that in the absence of carrier overflow, the dependences of the K-factor on the carrier density are determined mainly by hot phonon effects. At high modulation frequencies, carrier injection heating influence strongly on the intensity modulation response resulting in an increase of the modulation bandwidth under large band-gap offsets. Electron capture time is calculated taking into account intersubband transitions. Using our approach decreased amplitude of the oscillations in the capture time versus the well thickness and depth was obtained.

For lasers based on asymmetric quantum-well heterostructures comprising three different quantum wells and barrier layers with a complex potential profile, current carrier injection and radiation emitting processes have been considered. The analysis performed by using rate equations has shown that a regime of regular pulse generation at remote wavelengths is practicle. The radiation pulsation process is accompanied by oscillations of the carrier injection efficiency into quantum wells. At certain parameters of the laser structures it is possible to realize radiation bistable switching-on or the regime where relative powers at different wavelengths change with increasing the pump current. The injection efficiency has been determined by solving the Poisson equation and continuity conditions for electron and hole currents. Carrier tunneling through potential barriers between the quantum wells has been taken into account. Calculations have been performed for the GaAs - Al_xGa_{1- x}As system.

We present a comprehensive numerical model for vertical-cavity surface-emitting lasers that includes all major processes effecting cw operation of axisymmetric devices. In particular, our model includes a description of the 2D transport of electrons and holes through the cladding layers to the quantum well(s), diffusion and recombination processes of these carriers within the wells, the 2D transport of heat throughout the device, and a multilateral-mode effective index optical model. The optical gain acquired by photons traversing the quantum wells is computed including the effects of strained band structure and quantum confinement. We employ our model to predict the behavior of higher-order lateral modes in proton-implanted devices, and to provide an understanding of index-guiding in devices fabricated using selective oxidation.

Two models have been developed to simulate a vertical-cavity surface emitting laser. The first model is a 2D time-dependent bulk dielectric and absorption coefficients. These bulk coefficients depend upon the material, lattice temperature, and carrier concentration. This field model is coupled with a frequency-dependent gain model that describes the quantum well regions in the time domain. Treatment of frequency-dependent media in a finite-difference time-domain code is computationally intensive. On the other hand, because the volume of the active region is small relative to the volume of the distributed laser cavity, the computational overhead is reasonable. A semi-empirical transport model is used to describe the bult transport, which drives the quantum well transport. In addition, the semi-empirical model provides a spatial distribution for the lattice temperature and carrier concentrations. The second model is a 3D solution of Maxwell's equations. The 3D model can be used for cold cavity calculations. The 2D code generates the dielectric and absorption coefficients assuming azimuthal symmetry, providing the initial conditions for the 3D calculation.

A comprehensive quasi-3D model is presented to calculate the CW output characteristics of planar vertical-cavity surface emitting laser diodes. The model contains the determination of the longitudinal and transverse mode structure and accounts for carrier diffusion effects in the active layer in a self-consistent manner. The use of a multiwave propagation technique for the identification of resonator eigenmodes or the computation of diffraction limited mirror reflectivities is explained in detail. A simplified approach is proposed for VCSELs exhibiting strong thermally induced index guiding. Thermal lensing and spatial hole burning effects leading to the onset of higher transverse order mode oscillation are theoretically explained. Quantitative agreement with experimental values for differential quantum efficiency and maximum single-mode output power is achieved by including size-dependent parasitic effects.

A self-consistent analytical thermal-electrical model is developed to simulate thermal properties of etched-well InGaAsP/InP buried-heterostructure VCSELs with dielectric mirrors. The model is then used to investigate the influence of various design parameters on device performance. In particular, we examine the effects of varying the P-cladding doping level, active-region-, mirror-, and well-diameters, solder material, and mirror materials. We find that the dielectric mirrors are the most critical elements of the device. To increase the output power/operation temperature of the device, both mirrors must have high thermal conductivity and minimal scattering loss.

We propose a novel design of semiconductor lasers operating at 1.3 micrometers and 1.5 micrometers . A distinctive attribute of the proposed design is that the AlInGaAs active region is surrounded by an electron stopper layer on the p-side and a hole stopper layer on the n-side. The stopper layers do not impede the carrier injection into the active region and at the same time reduce the thermionic emission of carriers out of the active region. Utilization of stopper layers allows to increase the value of internal quantum efficiency and select the waveguide material corresponding to the optimum optical confinement factor value.

We examine carrier relaxation and radiative recombination in AlGaAs-based near IR and AlGaInP-based visible fractal quantum well heterostructures. Through temperature dependent photoluminescence, we demonstrate that enhanced population of higher lying energy levels can be achieved by varying the thickness of the layers in the fractal heterostructure. This distribution of carriers results in room temperature emission over a relatively broad range of wavelengths: approximately 700-855 nm for AlGaAs structures and 575-650 nm for AlGaInP structures. Spectra are compared to theoretical calculations to evaluate the nonequilibrium nature of the carrier distributions. Time resolved photoluminescence measurements demonstrate an approximately linear relationship between the radiative decay time and the layer thickness of the structure. Correspondingly, integrated luminescence measurements at room temperature reveal a factor of four increase in the light output efficiency of the structure as the fractal layer thickness is increased from 50 angstrom to 400 angstrom. The applicability of these heterostructures to broadband LEDs is discussed.

Nanostructures based on III-V semiconductor materials have reached a status which enables basic physical studies on size effects in device and in nanostructures. The expected benefits of high modulation bandwidth, low laser threshold, and improved linewidth enhancement factor in DFB lasers, to say only a few, which are believed to be based mainly on the changed density of states (DOS) function in low dimensions might be counterbalanced by altered carrier energy relaxation and k-space filling in those structures. To investigate systematically size effects and device aspects, a continuous change of structure and active device size is needed from 2D to 0D dimensions. This requirement can be met by high resolution electron beam lithography in conjunction with low damage etch processes and epitaxial overgrowth. In this presentation we discuss the technology and design considerations of lasers with low dimensional active regions as well as DOS effects and device relevant carrier relaxation effects. The technology part will focus especially on low damage etch processes such as RIE- ECR. Nearly damage free structuring processes can be demonstrated. Based on this low damage dry etch process we obtained electrically pumped wire DFB lasers with relatively high output power (up to 6 mW) and operation temperature (60 degrees C). Time resolved optical ps-spectroscopy as well as high excitation spectroscopy on wire and dot nanostructures demonstrate strongly changed k-space filling and carrier relaxation mechanisms in low dimensions and represent a serious limitation of device speed. Results obtained from electrically pumped wire DFB lasers confirm the carrier relaxation and k-space filling effects in device structures which have been observed by optical pump experiments in nanostructures. Despite the band filling effects in low dimensional structures, the wire DFB lasers show clearly the expected feature of gain coupling and enhanced differential gain which might demonstrate the applicability of mesoscopic laser devices in common data communication approaches.

Theoretical analysis of the gain and threshold current of a quantum dot (QD) laser is given taking into account the linewidth broadening caused by fluctuations in QD sizes. The following processes are taken into consideration together with the main process of the radiative recombination of carriers in QDs: the band-to-band radiative recombination of carriers in the waveguide region, carrier capture into QDs, and thermally excited escape from QDs. Expressions for the threshold current density depends on the root mean square of QD size relative fluctuations, surface density of QDs, thickness of the waveguide region, and total losses that have been obtained in an explicit form. The minimal threshold current density and optical values of the structure parameters are calculated as univeral functions of the main dimensionless parameter of the elaborated theory. This parameter is the ratio of the simulated transition rate in QDs to the spontaneous transition rate in the waveguide region. Theoretical estimations presented in the paper confirm the possibility of a significant reduction of the threshold currents of QD lasers as compared with conventional quantum well lasers.

Electromagnetic field inside a broad-area diode laser is considered as a system of counterpropagating waves. A study of the lateral field instability was performed by using a 6- wave mixing theory. It is shown that perturbation waves with the tilted angle (phi) <EQ 1.2 degrees inside the active region, and respectively, with the side lobes of the far-field pattern of less than 4 degrees, have the greatest growth increment. These waves produce lateral intensity modulation with the period 10 divided by 30 micrometers for the 0.85 micrometers lasing wavelength. The appearance of such waves corresponds to the instability threshold of a homogeneous lateral distribution of optical power in a diode laser. The calculation results explain an experimental observation of the chaos in lateral transversal intensity distribution in usual broad-area diode lasers. The theory allows a stability analysis of the homogeneous field distribution in diode lasers of a novel design.

A new version of the Minilase laser simulator is presented. This version self-consistently calculates the bulk carrier transport, quantum capture, spectral hole burning, and radiative processes present in quantum well laser diodes. Minilase is used to investigate the effects of carrier capture and spectral hole burning on the modulation response and to determine their relative significance. Also, a Monte Carlo simulation of carrier capture and scattering within the quantum well is presented. This simulator includes all relevant carrier-photon, carrier- phonon, and carrier-carrier interactions. A comparison of the carrier distributions calculated by the Mont Carlo and the Minilase simulators is used to validate the methods used in Minilase.

Intervalence band coherences in semiconductor quantum wells induced by femtosecond light pulses are investigated. Especially, the possibility of creating dark states, known from atomic physics, in semiconductor quantum wells, is discussed. The analysis includes many-body effects due to the Coulomb interaction and bandstructure effects appropriate for GaAs quantum wells.

It is shown theoretically and experimentally that the spectral mode gain curve of a 0.98 micron InGaAs/GaAs laser may be periodically modulated with the period of 2-3 nm and the modulation depth up to several cm^-1. A distinctive feature of such heterostructure lasers is that a GaAs substrate is transparent for a 0.98 micron laser wavelength. If the cladding layers are not very thick, then radiation may be tunnelled considerably from a waveguide through the cladding layers, and then propagate in the substrate and the cap layer. As a result of radiation reflection from upper and bottom contacts an additional cavity is formed in the direction normal to the layers, and it modulates the spectral mode gain of such lasers. An analytic expression for the mode gain has been obtained for the case of practical interest where the radiation leakage from the waveguide is small. The mode gain calculation results are in a satisfactory agreement with the experimentally measured values. The leakage effect is analyzed from the viewpoint of constructing new types of a laser with a 2D cavity.

First-principles density-functional calculations utilizing ab initio pseudopotentials and plane- wave expansions are used to determine lattice parameters, bulk moduli, and band structures for AlN, GaN, and InN. It is found that large numbers of plane waves are necessary to resolve the nitrogen 2p wave functions and that explicit treatment of the gallium 3d and indium 4d electrons is important for an accurate description of GaN and InN. Several properties of ternary zinc-blende alloys are determined including their bond-length and bond-angle relaxation and their energy-gap bowing parameters. The similarity of the calculated zinc- blende and wurtzite direct gaps also allows estimates to be made of the energy gap versus composition for wurtzite alloys.

We present magnetoluminescence data which provides a quantitative measure of the energy- band dispersion curves of novel compound semiconductor optoelectronic materials. Data for a n-type strained-layer InGaAs/GaAs (quantum-well width approximately 8 nm) and a n-type 4.5 nm-wide GaAs/AlGaAs lattice-matched single-quantum well are presented. We find that the conduction-bands are almost parabolic, with a mass of about 0.068m₀ for the GaAs/AlGaAs structure. The valence-bands are nonparabolic with wave vector dependent in- plane valence-band masses varying from about 0.1m₀ at zone center to about 0.3m₀ for 20 meV energies.

The temperature dependence of threshold current and quantum efficiency for Ga_xIn_{1- x}P (x equals 0.4, 0.6; (lambda) equals 680, 633 nm) single 80 angstrom quantum well lasers is analyzed using a model for the electron leakage current. This model fits the experimental data, correctly describing the rapid increase in threshold and drop in quantum eficiency as temperature increases. Also it indicates that the drift component of the electron leakage current is important, because of the poor p-type conductivity in AlGaInP. In addition, a single quantum well Ga_0.5+(delta )In_0.5-(delta )P/(AlGa)_0.5P laser structure is demonstrated, which can provide similar gain in both polarizations. The slightly-tensile- strained quantum well has the light hole ground state, which gives the lowest transparency current for TM-mode gain. However, the TE-mode gain is dominant at high drive currents. The gain-current relationships have been characterized for each polarization, and found to cross at a modal gain value of 25 cm^-1. Lasers whose threshold gain is near this crossover value were found to emit in either one or both polarizations, with a very wide range of polarization assymetry possible. A simple QW gain model can be used to describe this behavior.

Microcavity lasers have been predicted to offer low threshold current, high quantum efficiency and high modulation bandwidth. In this report we review the physics underlying microcavity device behavior. Specifically we cover dipole-field coupling for both localized (point) dipoles and extended dipoles. In general, optical pumping of the devices is required to create extended dipoles. We also outline the difference between the weak (irreversible) coupling regime and the strong (reversible) regime. For photonic application the intermediate, superradiant regime is perhaps more interesting than the strong coupling regime. Finally, we describe our recent experimental efforts to make high quantum efficiency devices by creating extended excitonic dipoles in electrically pumped devices.

Dependence of the intensity noise on structure of microcavity laser is theoretically examined, by help of a generalized presentation of the quantum noise in semiconductor lasers where character of multimode operation is taken into account. Key idea reducing the noise is to restrict the field components only to useful modes for the lasing operation. Precise selection of the cavity length is very important for reduction of the noise in microcavity structure.

We demonstrate lasing action in a novel microcavity laser which can be utilized for intracavity spectroscopy as well as high contrast imaging of small (approximately 10 micrometers ) structures. The system can be easily visualized as a Fabry-Perot cavity containing a gain media and an object for study. Since the primary constraint on the object is transparency at the lasing wavelength, investigation of lasing in objects such as microspheres, liquid droplets, and biological cells is possible. The resonator consists of an epitaxial MBE grown mirror and gain region on a GaAs wafer. This is essentially an open-cavity vertical cavity surface emitting laser. The object to be studied is placed on the wafer and covered with a glass dielectric mirror which acts as the output coupler. When the semiconductor gain region is photo- pumped, the object within the cavity provides lateral optical confinement through its index difference with the surrounding media, increases the cavity Q, and thus encourages lasing action. The emitted laser light can be spectrally and spatially resolved. The narrow lasing lines can provide information about the lasing modes supported and the size of the object. The spatially resolved laser light provides high contrast microscopic images of the electromagnetic modes oscillating in the resonator. We present an investigation of stable lasing modes in polystyrene spheres.

Intracavity contacted vertical-cavity lasers are a class of index-guided surface-emitting lasers which use thin p-type and n-type layers within the optical cavity to supply current to the active region. The design allows the use of two ring contacts on the same surface, top or bottom emission, and semi-insulating substrates. A dielectric aperture is used to provide both current constriction and optical confinement. The models used to design the laser's continuous wave characteristics are discussed and the calculated characteristics are compared with experimental measurements. The microwave modulation response of lasers of varying diameter has been measured with wafer-level probing techniques. The very high modulation efficiency of the lasers, up to 5.7 GHz/(root)mA, shows good agreement with theory. The analysis indicates the directions and challenges for the realization of very high-speed, low-power surface-emitting lasers.

A comprehensive, 3D, thermal-electrical self-consistent finite-element model is described and used to investigate thermal properties of GaAs-AlGaAs proton-implanted top-surface-emitting lasers. Special attention is paid to carrier diffusion within the layer containing the active region and to its influence on temperature profiles. In the model, an analytical approximation is used to describe the 3D current spreading between the annular top contact and the broad- area bottom contact. Temperature dependence of many device and material parameters is included. Multiple heat sources are taken into consideration. The carrier diffusion equation, including injection-current generation, ambipolar diffusion as well as bimolecular and spontaneous recombination terms, is solved numerically using the finite-element method for the layer containing the active region. The results indicate that carrier diffusion strongly influences the distribution of main heat sources. As a result, both current-spreading and heat- spreading phenomena are modified.

Long-wavelength vertical-cavity surface-emitting lasers are investigated using electro-thermal, optical, and electronic modeling. The strain-compensated InGaAsP multi-quantum-well active region of the device example is vertically sandwiched between various distributed Bragg reflectors (DBRs). InP/InGaAsP, Si/SiO₂, and GaAs/AlAs mirrors are considered as well as novel combinations like SiC/MgO. The model includes nonuniform current injection, distributed heat sources, temperature dependent material properties, and k(DOT)p band structure calculations. Device parameters such as thermal resistance, threshold current, and external quantum efficiency are compared and heating effects are evaluated. Simulated light power versus current characteristics exhibit the typical thermal roll-over in continuous wave operation. The complex influence of the DBR materials is analyzed in detail.

We report thermal effects revealed by a self-consistent treatment of plasma and lattice heating in vertical cavity surface-emitting lasers (VCSELs). The basic idea of our treatment is to couple the equations for carrier density and field amplitude in the conventional laser theory with those for two additional variables, the plasma and lattice temperatures. The CW operation of the VCSELs is investigated both for a fixed and for a self-consistently determined lattice temperature. In the first case plasma heating results in an increase of carrier density with pumping and thus in a pumping dependent frequency shift. In the latter case, both plasma and lattice heating induce a thermal switch-off of the laser as the pumping is increased. Furthermore, depending on the initial alignment of the cavity frequency and the ambient temperature of the device, heating can introduce a discontinuous threshold, exhibiting a bistability between lasing and nonlasing states. While some of our theoretical predictions are in qualitative agreement with known experiments, others await experimental verification.

A novel approach of on-wafer wavelength control for vertical cavity surface emitting lasers (VCSELs) is proposed using nonplanar metalorganic chemical vapor deposition. The resonant wavelength of 980nm VCSELs grown on a patterned substrate can be controlled in the wavelength range over 45nm by changing the size of circular patterns. We have fabricated linear and 2D multiwavelength vertical surface emitting laser (VCSEL) arrays fabricated by using this technique. The threshold of multi-wavelength VCSELs formed on the patterned substrate is as low as 3 mA. A possibility of an extremely large wavelength span for multi- wavelength arrays will be discussed. The proposed method will be useful for multi-wavelength VCSEL arrays as well as for the cancellation of wavelength nonuniformity across a wafer.

Nondegenerate four-wave mixing in semiconductor optical amplifiers was studied both as a spectroscopic tool for probing semiconductor dynamics and as a wavelength conversion technique. Four-wave mixing spectra were measured at detuning frequencies ranging from GHz to THz rates and ultrasfast intraband mechanisms having relaxation time constants of 650 fs and less than 100 fs were revealed in the measurements. Cross-polarization four-wave mixing was also measured to study the inter quantum-well carrier transport process in quantum-well amplifiers. In addition, broadband wavelength conversion using four-wave mixing in semiconductor optical amplifiers was investigated. Results concerning the conversion efficiency over spans up to 65 nm, as well as a demonstration of wavelength conversion with gain are presented. The issue of converted signal-to-background noise in this process is also addressed.

The semiconductor Maxwell-Bloch equations provide a model that is grounded in the fundamental physics of semiconductors which include a variety of many body effects. Many of these effects are particularly noticeable when the semiconductor is probed with ultrashort pulses. We present computational results describing the computed behavior of model equations which describe the propagation of femto-second pulses in bulk GaAs. It is shown how the inclusion of additional physics modifies the predictions of the model. Among the effects that are discussed are plasma heating, plasma cooling, self-focusing, and memory effects.

The influence of Coulomb effects in the linear and nonlinear optical and electronic properties of semiconductors with variable effective dimensionality is discussed in conjunction with bandstructure engineering techniques. The combined influence of valence bandcoupling, quantum confinement and many body effects are analyzed in lattice-matched, strained, and strain-relived structures. Quasi-analytical results for linear/nonlinear absorption spectra are given for coherently coupled superlatices. Decoupled multiple quantum wells are treated through rigorous numerical methods and the results are compared with analytical solution in limiting cases. The theory is applied to the realistic simulation of lasers and optical computing devices.

In this paper, the various physical mechanisms in low-frequency intensity fluctuations are identified, that occur when a diode laser is subject to moderate optical feedback while operating close to its solitary threshold. The maximum gain mode, which surprisingly often stable, serves as a seemingly unreachable goal for the system. In attempting to reach this mode, the system forms mode-locked pulses. In between pulses the mode-locking is frustrated and inevitably the system passes too close to one of the many saddle points, that will take the system back to the low power solitary laser state.

Injection-locked directly current modulated semiconductor laser transmitters are theoretically investigated with respect to transmission performance. By large signal simulation of laser, standard single-mode fiber propagation and direct detection optically pre-amplified PIN receiver, transmission distances of 80-120 km at 10 Gb/s and 40-60 km at 15 Gb/s have been evaluated with a bit-error-rate < 10^-9 with reasonable power penalty. Exploiting fiber nonlinearities with higher power launched into the fiber is demonstrated to increase the transmission distance by about 40%. Additionally the dynamics of the residual chirp of the laser is shown to act favorably on transmission performance. Guidelines for operation conditions of injection-locked lasers depending on detuning between laser and slave laser and injected power are given. Gain switching including optical feedback and the gain- levering effect have been investigated with respect to pulse production for optical time domain multiplexing. A new method for short pulse generation is presented. It is based on single frequency CW light injection into an unmodulated single mode laser under nonstable locking- conditions. Repetition frequencies larger than 150 GHz can be achieved. By soliton generation in a dispersion shifted fiber pulse widths of less than 3 ps FWHM with a squared hyperbolic cosecans shape can be generated.

We have analyzed the behavior of a diode laser with low-amplitude optical injection. The noisy injection field has phase fluctuations only, with a linewidth larger than both the solitary linewidth and the locking range. Simulations show that the laser output spectrum can be much narrower than the injection spectrum, and may have a dip at the solitary laser frequency. A model describing these features is based on linear superposition of the stationary responses to monochromatic injection. The mechanism of line narrowing is that the response to injection ouside the locking range lies mainly inside.

Phase-conjugate feedback affects a laser in ways which are fundamentally different than conventional feedback. Notably, when the laser oscillates in more than one longitudinal mode, phase-conjugate feedback initiates a novel mode-coupling mechanism which can even lead to mode locking behavior. This paper explores these mode-coupling effects and also summarizes the simularities and differences between the two types of feedback.

The effects of phase-conjugate feedback on a semiconductor laser are studied, beginning with a derivation based on linear response theory of the single-mode field rate equation. For sufficiently high feedback levels and fast-responding PCM, a multimode extermal cavity mode spectrum emerges analogous to that of the conventional feedback case, except that the spacing of the modes is half that of the latter, and tuning is possible by varying the frequency of the PCMs pump laser relative to that of the free-running laser. The latter may also be used as a control parameter to achieve stable operation at high feedback levels, and to maximize the fraction of energy in the central 'spike' of the laser spectrum. The extermal cavity modes also show up as spikes in the RIN and phase noise spectra. Finally, the effects of pump linewidth on laser linewidth reduction and finite PCM response time are described.

A microscopic theory, that is based on the coupled Maxwell-semiconductor-Bloch equations, is used to investigate the effects of many-body Coulomb interactions in semiconductor laser devices. This paper describes two examples where the many-body effects play important roles. Experimental data supporting the theoretical results are presented.