We discuss the conditions to be satisfied for the growth of cubic GaN films on GaAs(001) and Si(001) by plasma-assisted MBE. A stoichiometric group V/III flux ratio is required for single-phase growth, which is controlled via the surface reconstruction in real time by means of RHEED. While phase purity of cubic GaN-on-GaAs is induced by a well-defined orientation relationship, the growth on Si(001) often results in a phase transition due to the formation of amorphous Si_xN_y inclusions at the GaN-on-Si interface, which act as nucleation centers for the growth of the hexagonal GaN phase. A suitable template, such as a thin GaAs or SiC insertion layer, prevents the formation of the Si_xN_y inclusions. Cubic GaN grown under optimized conditions exhibit a narrow excitonic luminescence line at 3.272 eV at 5 K and high luminescence efficiency at room temperature. In addition, we discuss the origin of the n- type background doping in the as-grown layers and we present the new concept of codoping to achieve high p-type conductivity in GaN at room temperature.

An additional approach to further improve the reliability of ZnSe based devices is to use beryllium containing II-VI compounds. BeS, BeSe and BeTe are characterized by a considerable amount of covalent bonding and a high bond energy. This distinguishes these materials from the conventional ionic wide gap II-VI semiconductors like ZnSe, ZnTe or CdTe. Recently, thin film structures using Be- compounds have been fabricated and characterized. It became clear that--besides the application aspects--these materials are also very interesting from a more fundamental point of view. Using Be-containing II-VI compounds, ionic and covalent lattice matched II-VI materials can be combined in quantum well structures. The type II band alignment of BeTe and ZnSe gives additional freedom in the band gap engineering, and it is possible to grow lattice method quaternaries of low polarity onto silicon. Here, basic principles of Be containing II-VI compounds will be described, and the potential of these novel materials will be discussed.

ZnSe-based laser diodes have recently encountered strong competition from those grown from GaN related materials. These two material systems behave in a very different way as far as defect generation and propagation are concerned. For ZnSe-based materials the lifetime of a laser-diode is very sensitive to the density of pre-existing extended defects in the epitaxial material. Therefore, fabrication of a long- lived ZnSe-based laser diode requires an elimination of extended defects as well as making low-resistivity components in order to minimize device heating. We discuss the molecular beam epitaxy growth and characterization of ZnSe-based epitaxial structures on various III-V buffer layers lattice matched to GaAs. The status of our ZnSe-based laser diodes and microcavity LEDs will also be discussed.

We describe the use of first-principle calculations to address problems related to defects, doping, and band- structure engineering in III-V nitrides. For n-type doping it is found that nitrogen vacancies are too high in energy to be formed during growth, but silicon and oxygen readily incorporate as donors. The properties of oxygen, including DX-center formation, support it as the main cause of unintentional n-type conductivity. For p-type doping we find that the solubility of Mg is the main factor limiting the hole concentration in GaN. We discuss the beneficial effects of hydrogen during acceptor doping. Compensation of acceptors by nitrogen vacancies may occur, becoming increasingly severe as x increases in Al_xGa_1-xN alloys. Acceptors other than Mg are also investigated. Finally, we discuss our first-principles investigations of the atomic and electronic structure of heterojunction interfaces between the III-nitrides, and provide values for band lineups.

Microscopic deep defect characteristics of the blue laser materials in the II-VI and III-V wide-bandgap semiconductors have been studied from the view point of the device performance and degradation. The deep defect centers, monitored by transient capacitance spectroscopy technique, have revealed quite different properties in electrical activity, defect density and carrier injection enhanced generation of the defects for the two different wide-bandgap blue laser materials.

A unified theoretical model for the degradation of II-VI blue-green laser diodes and light-emitting diodes (LEDs) is presented. The model is based on the recombination-enhanced defect generation. Based on this model and the corresponding rate equations for the laser diodes and LEDs, we show a general picture to explain the aging behaviors of both the optical output power of the LEDs under a constant current aging condition and the increase of the operation current of laser diodes under constant optical power operation. We found a universal curve for the optical output decay of the II-VI LEDs as a function of the aging time. The power decay is associated with the increase of defect density caused by electron-hole recombination enhancement. Our model also explains very well the increase of the operation current of laser diodes under the constant optical output aging condition.

p-type doping of wide band gap II-VI semiconductors is a key technology to form the ohmic contact with metal electrodes in device fabrications. Using an alkaline metal compound such as K₂S, Na₂Se or Na₂Te which contains dopant atoms, excimer laser doping experiments were carried out for ZnSe and CdTe. Influence of the electrical properties on this treatment was mainly measured by means of the Hall measurement. The resistivity of ZnSe drastically decreased from 10⁵ to 10^-2 Ohm cm and the value of hole carrier concentration increased up to 4.8 X 10¹⁹ cm^-3. For CdTe also resistivity decreased from 10⁵ to 10^-1 Ohm cm. Formation of p-type ohmic contact in ZnSe p-n diode was also investigated.

The purpose of this work is to study the effect of carrier dynamics on quantum-dot laser performance and to clarify how the carrier relaxation lifetime and the bi-exciton effect manifest themselves in the quantum-dot laser operation. We derive carrier-photon rate equations for quantum-dot lasers and simulate the output power vs. injected current relationship and the small-signal modulation response. We clarify criteria on the carrier relaxation lifetime as well as the inhomogeneous broadening linewidth, the dot density and the crystal quality to achieve high-performance. Then, we develop an optical gain formula for the bi-exciton- exciton transition and modify rate equations to describe bi- exciton lasing in quantum dots. We show spontaneous emission and lasing properties of bi-excitons, and discuss the effect of carrier relaxation lifetime and oscillator strength on the laser operations.

We have evaluated the optical gain of GaN/AlGaN quantum well structures with localized states, taking into account the Coulomb interaction. The localized states are introduced in the well as quantum dot-like subband states. We have used the temperature Green's function formalism to treat the many-body effects and have found a new excitonic enhancement of the optical gain involved the localized states. This enhancement is stronger than the conventional Coulomb enhancement. It might play an important role to reduce the threshold carrier density.

The gain spectra and the dynamics of semiconductor lasers are investigated on timescales from continuous wave down to the 100 fs regime. First, we measure continuous wave gain spectra over a wide spectral range with a transmission experiment using the ultrabroad spectrum of a 10 fs Ti:sapphire laser. The agreement between our experimental results and a parameter-free microscopic theory is excellent, only if carrier scattering and polarization are treated correctly. Accordingly, carrier thermalization and material polarization dynamics have to be considered in order to achieve a fundamental understanding of the dynamics of semiconductor lasers. The ultrafast carrier and polarization dynamics are experimentally investigated by femtosecond heterodyne pump-probe and four-wave-mixing experiments. Finally, we study the influence of carrier thermalization on the emission dynamics of a vertical-cavity surface-emitting laser (VCSEL). The VCSEL is optically pumped with femtosecond pulses from a Ti:sapphire laser and the temporal evolution of its emission is investigated as a function of the initial excess energy of the photoexcited carriers. The dynamics become slower for larger excess energies since carrier cooling is slower. The fastest response is obtained for resonant pumping.

A non-Markovian model for the optical gain of strained-layer quantum-well lasers is studied taking into account the valence-band mixing, strain effects, many-body effects and the non-Markovian relaxation using the time-convolutionless reduced-density operator formalism for an arbitrary driven system coupled to a stochastic reservoir. Many-bode effects are taken into account within the time-dependent Hartree- Fock approximation and the valence-band structure is calculated from the 6 X 6 Luttinger-Kohn Hamiltonian. The optical gain with Coulomb (or excitonic) enhancement is derived by integrating the equation of motion for the interband polarization. It is predicted that the Coulomb enhancement of gain is pronounced in the cases of compressive and unstrained quantum wells while it is negligible in the case of tensile strained quantum well.

A nonequilibrium Green's functions approach is presented for the consistent computation of semiconductor quantum well optical spectra including strong Coulomb correlations within the coupled photon and carrier system. BetheSalpeter-like equations are given for the optical response and recombination rates in the excited medium. Bandstructure, quantum-confinement, many-body and cavity resonator effects are included in the microscopic approach. The theory is applied to the description of absorption/gain, luminescence, single and two-beam photoluminescence excitation spectroscopy for arbitrary temperatures and carrier densities. Numerical results, showing good agreement with recent experiments, are presented for ITT-V and Il-VI materials, from the linear regime, characterized by excitonic effects to the high density case in which a strongly interacting electron-hole plasma is proposed as the dominant mechanism. Keywords: Many-Body Effects; T-matrix; Bethe-Salpeter Equation; Nonlinear Absorption; Strongly-Correlated Electron-Hole Plasma; Coupled Valence-Band Multiple Quantum Wells.

We experimentally measure the first-order spatio-temporal characteristics of filamentation and discover effects of the stripe width. We use an analytic theory to explain and reproduce these results through an expression for the filament gain, in which contributions of various mechanisms can clearly be seen. Through this model and computer simulations, we determine the stability boundaries of the material parameters for which the device will not exhibit filamentary tendencies. We then propose a new method of controlling filamentation using below-bandgap semiconductor nonlinearities. With simulations, we determine under what conditions this imposed nonlinearity can counteract the carrier-induced self-focusing inside the active region.

We present a systematic characterization of the gain, refractive index, and linewidth enhancement factor in a semiconductor laser and a semiconductor optical amplifier. We compare the refractive index shift and linewidth enhancement factor due to electrical pumping alone and then with electrical and optical pumping. Changes in carrier density are estimated throughout with a many-body optical gain model taking into account the valance-band mixing. The spectral characteristics of the refractive index shift show potentially ideal conditions for broadband cross-phase modulation. Further measurements of the refractive index shift and linewidth enhancement factor in a semiconductor optical amplifier show operating conditions under which the refractive index shift and linewidth enhancement factor change sign, revealing potentially ideal conditions for cross gain modulation.

Physical and numerical self-consistent models of the semiconductor optical amplifier (SOA) accounting for effects of the carriers diffusion and the heat emission are developed. Transient characteristics of SOA are obtained for short (below 50 ns) and for long (up to 20 mks) periods in quasi-stationary non-coherent approximation with regard to processes of the heat diffusion. Frequency dependent nonlinear optical reply to the bias current modulation is observed. The physical interpretation of the observable in the SOA phenomena is given, and the main parameters, which determine the regime of the device operation, are exposed. Obtained results are in qualitative agreement with the experimental data observed.

Quantum well composition intermixing is a thermal induced interdiffusion of the constituent atoms through the heterointerface. The intermixed structures created by both impurity induced and impurity free or vacancy promoted processes have recently attracted high attention. The interdiffusion mechanism is no longer confined to a single phase diffusion for two constituent atoms, but it can now consist of two or multiple phases and/or for multiple species, such as three cations interdiffusion and two pairs of cation-anion interdiffusion. A review on the impact of intermixing on device physics is presented with many interesting features. For instance, both compressive or tensile strain materials and both blue or red shifts in the bandgap can be achieved depending on the types of intermixing. The recent advancement in intermixing modified optical properties, such as absorption, refractive index as well as electro-optics effects are discussed. In addition, this paper will place a strong emphasis on the device application of the intermixing technology. The advantage of being able to tune the material provides a way to improve the performance of photodetectors and modulators. Attractive distributed-feedback and vertical cavity laser dynamics have been shown due to some unique device physics of the quantum well intermixing. Several state-of-the-art results will be summarized with an emphasis on its future development and directions.

Bandgap energy renormalization by many-body interactions has been studied in a series of n-type 8-nm-wide GaAs/AlGaAs single quantum wells using magnetoluminescence spectroscopy at 1.4 K and for magnetic fields up to 30 T. The 2D-carrier densities varied between 1 and 12 X 10¹¹ cm^-2. At the maximum 2D-carrier density, the bandgap energy difference between the doped and undoped samples was about 34 meV.

Multiple cations intermixed In_0.53Ga_0.47As/In_0.52Al_0.48As quantum well structure with 60 angstroms well width is investigated by using the expanded form of Fick's second law. It was found that a maximum compressive strain of 0.64% is obtained when annealing time reaches 3 hours at 812 degree(s)C in the indium sublattice. For a small interdiffusion, i.e. 1 to 1.5 hrs, the subband separation between the lowest heavy and light hole states is at its greatest. This is a major contribution to the band structure and averaged density of states, thus enhancement in optical gain up to 40% is obtained. For a large interdiffusion, i.e. up to 6 hrs, a large blue shift of the peak gain from 0.842 to 1.016 eV is observed.

We discuss the use of ensemble Monte Carlo techniques for the simulation of some opto-electronic devices. This approach has been reasonably successful for device simulation, and the Monte Carlo approach has worked very well for sub-picosecond time scales, but computational time becomes excessive for multi-picosecond simulations. Various nonlinear effects such as carrier-carrier scattering, non- equilibrium phonons, quantization in low-dimensional systems, and finite collision duration have all successfully been incorporated into the Monte Carlo method.

A microscopic model for the operation of multi-quantum well laser diodes is described. It includes bulk transport of carriers modeled by drift-diffusion equations, confined carriers in the quantum wells modeled by the Schroedinger equation, photon modes modeled by a Helmholtz equation and couplings described by rate equations. Application of this model shows that the carrier distribution in the active layer of the laser can not be described by quasi-equilibrium conditions. One consequence is the substantially non-uniform distribution of carriers among the quantum wells when the laser is biased above threshold. Another consequence is the observation of photoluminescence in wide area devices under short circuit conditions.

An integrated device and package 3D model is developed to computationally investigate the thermal crosstalk in arrays of proton-implanted top-surface emitting lasers. A self- consistent electro-thermo-opto model is employed for the device. The anisotropic thermal property is considered for the package model. Temperature dependency of critical device and material properties is included, as well as multiple heat generation mechanisms. Effects of spacing on lasing performance and non-uniformity of VCSEL arrays are found significant. Thermal crosstalk becomes worse for increased sizes and packaging densities of laser arrays. Degraded laser performance is found due to the thermal crosstalk, especially for the lasers closest to the center of the array package.

A new method which is very stable and fast, has been developed, to self-consistently solve the electric equations, scalar Helmholtz equation and photon rate equation. This method is based on a new concept, control equation, so that high efficient algorithms for the non- linear single-variable equation can be applied to control the solution procedure. Above threshold, tremendous improvement in speed and stability was obtained. Multiple spectral mode behavior can be simulated with the same efficiency of single-mode simulation, owning to the inclusion of the control equation concept. This 2D simulator works well even for very high bias. This method has the potential to be applied to the simulation of a wide variety of semiconductor lasers. In addition, the carrier energy transport and lattice thermal diffusion are accounted for. The finite element method is applied to the Helmholtz equation. The generalized eigenvalue problem thus generated is solved directly and quickly by a kind of Lanczos algorithm. The discretization of all fundamental equations is on the same triangular element mesh. Simulation of a GRIN-SCH BH SQW laser shows good agreement with the experimental data.

The lasing mode characteristics of planar and dielectrically-apertured one-wave and two-wave Fabry-Perot microcavities are analyzed theoretically using a semiclassical model, in the limit that the dielectric apertures are optically thin. Idealized cavities based on single interface and constant reflectivity mirrors are studied, though the model can be applied to more realistic cavities based on distributed Bragg reflectors or vertical cavity surface emitting lasers. The analysis shows that shortening the cavity length can lead to better confined modes with lower threshold, especially for very small aperture sizes.

A theoretical analysis of the polarization properties of Vertical-Cavity Surface-Emitting Lasers is presented. A model is derived which contains the influence of material (strain in cavity layers, scattering rates, refractive indices of layers) and cavity design (size, number of layers) parameters on input/output characteristics of the system. The electromagnetic field is represented as a superposition of cold-cavity vectorial eigenmodes of the VCSEL cavity. Each eigenmode is characterized by its resonant wavelength, threshold gain and 3D vectorial eigenmode profile. The electromagnetic field is then coupled to semiconductor Bloch equations describing the active quantum well material. The theory is evaluated for the case of two orthogonal HE₁₁-like modes and a six-band model of quantum well.

We study the polarization behavior of birefringent multi- transverse mode index-guided Vertical-Cavity Surface- Emitting Lasers during CW and modulation operation. The model we use incorporates the (alpha) -factor, different carriers populations associated with different magnetic sublevels of the conduction and heavy hole valence bands in quantum well media, spin-flip relaxation processes and cavity anisotropies. The effect of carrier diffusion and spatial hole burning on the polarization behavior within the fundamental mode regime is also considered. We find that spin-flip dynamics and the (alpha) -factor are crucial for the selection of the polarization state corresponding to a given injection current. We also analyze the direct-current polarization modulation response for both the small and large signal modulation regimes. In both cases the current is modulated around the polarization switching (PS) current, but in the large signal modulation regime the current goes from below threshold to above the PS current. A frequency doubled polarized output is obtained under large signal modulation for modulation frequencies smaller than 10 MHz. At greater frequencies (around 100 MHz) transient mode polarization partition noise becomes important. At the greatest considered frequency (1 GHz) no PS is observed. Finally the effect of multiple transverse mode operation is addressed by considering the two lowest order modes: LP₀₁ and LP₁₁.

This paper describes crystal growth techniques for achieving good electroluminescence efficiency and narrow linewidth from a GaAs-based light emitting device emitting at 1.3 micrometers . The long wavelength emission is achieved using a quantum dot active region grown by sub-monolayer In, Ga and As. Low threshold lasing at a shorter wavelength of 1.15 micrometers is achieved in a GaAs-based oxide-confined vertical- cavity laser using alternating single monolayer growth of InAs/GaAs QDs.

We provide an overview of the issues and progresses of semiconductor based vertical-cavity surface-emitting lasers (VCSELs) for optical information processing applications. Targeting several important VCSEL structures which have been reported to be effective in improving the characteristics of threshold, transverse modes, polarization and high speed modulation, we discuss physical origin and design parameters for the control of these characteristics.

The hot Electron Light Emission and Lasing in Semiconductor Heterostructures devices (HELLISH-1) is novel surface emitter consisting of a GaAs quantum well, within the depletion region, on the n side of Ga_1-xAl_xAs p- n junction. It utilizes hot electron transport parallel to the layers and injection of hot electron hole pairs into the quantum well through a combination of mechanisms including tunnelling, thermionic emission and diffusion of `lucky' carriers. Super Radiant HELLISH-1 is an advanced structure incorporating a lower distributed Bragg reflector (DBR). Combined with the finite reflectivity of the upper semiconductor-air interface reflectivity it defines a quasi- resonant cavity enabling emission output from the top surface with a higher spectral purity. The output power has increased by two orders of magnitude and reduced the full width at half maximum (FWHM) to 20 nm. An upper DBR added to the structure defines HELLISH-VCSEL which is currently the first operational hot electron surface emitting laser and lases at room temperature with a 1.5 nm FWHM. In this work we demonstrate and compare the operation of UB-HELLISH-1 and HELLISH-VCSEL using experimental and theoretical reflectivity spectra over an extensive temperature range.

A versatile electro-thermal and optical numerical simulation tool has been developed that can handel various types of vertical-cavity surface-emitting laser designs. As an example, we consider a cylindrically symmetric top-emitting mesa laser with an oxide window. The control-volume-based model provides spatial distributions of carrier density in the active region plane as well as electrical potential, current density and temperature distributions in the entire device. We focus our attention on current redistribution processes, and find out that the current crowding effect is strongly affected by the presence of oxide window. In addition to the geometry-related crowding, we describe the current self-distribution (CSD) effect related to spatially non-uniform stimulated recombination. The CSD is shown to counteract the spatial hole-burning caused by carrier consumption through intense stimulated emission.

The Finite Difference Time Domain method has been used to analyze the dispersion diagram of a photonic crystal comprised of a perforated dielectric slab and the properties of a micro-cavity formed by introducing a defect into such a crystal. Computational requirements of the method, its advantages and disadvantages, and results for the structure analyzed are discussed.

A numerical method for the calculation of the local density of states of the classical electromagnetic modes existing within a finite periodic dielectric material is proposed. This information allows for a computation of the modified spontaneous emission rate inside the structure, for weak coupling of a dipole transition to the electromagnetic modes. The approach is based on a calculation of the electric field Green function, and includes the contribution arising from surface guided modes as well as from scattering modes of the system. The required computational effort is enormously reduced if the necessary integration over a 2D Brillouin zone is deformed into the complex plane. An implementation of the approach utilizing a transfer matrix approach for the calculation of the Green function is reported, and used in the calculation of the local density of states within a finite 2D periodic structure. The method can also be used to find the spatial distribution of emitted light in both the near and far fields.

A novel pump-noise-suppressed semiconductor laser scheme is proposed. In this scheme, electrons injected into a stable intersubband state dressed by strongly resonant optical couplings stimulate amplitude-squeezed light with extremely low-threshold. Because of its high conversion efficiency and short stimulated emission lifetime at low injection current regime, this scheme can generate weak broad-band amplitude- squeezed light. We numerically estimate the characteristics of gain coefficient and output noise spectrum of this scheme.

A microscopic interpretation of spontaneous and stimulated emission in semiconductor microcavities is developed using Semiconductor Luminescence Equations obtained from a quantum theory for the interacting electron-hole system and microcavity photons. The properties of bare quantum well luminescence as well as nonlinear photoluminescence of microcavity systems showing a threshold-like transition are consistently explained.

We report experimental studies of laser emission from a microcavity embedded with inhomogeneously broadened GaAs quantum wells. Extremely small lasing threshold is observed when the cavity resonance is tuned to the low energy tail of the inhomogeneous exciton distribution. Detailed experimental studies further show that the excitation level at the minimum lasing threshold is below the exciton Mott density and that the lasing process changes qualitatively when the excitation level is increased from below to above the exciton Mott density. A preliminary analysis indicates that the small lasing threshold observed is due to the enhanced dipole coupling rate in the microcavity as well as the very small number of localization sites resonant with the cavity when the cavity is tuned to the low energy tail of the inhomogeneous exciton distribution. The results also suggest the possibility of achieving laser emission from a few or even a single localized exciton by using cavities with higher Q-factor and greater dipole coupling rate.

Spontaneous emission control has been achieved in GaAs/AlGaAs quantum well lasers by the use of Distributed Bragg microcavity perpendicular to the quantum wells. The room temperature emission is inhibited whilst below 130 K there is an enhancement. These changes to the spontaneous emission process directly effect the threshold current producing a 25% reduction at room temperature. Theoretical modeling of the lasers is in agreement with the experimental results and highlights the effect of the microcavity on the field-dipole interaction.

We report for the first time angle-dependent multiple- wavelength emissions over 30 nm range from a toroidal cavity, which is a photonic quantum ring laser naturally formed in a VCSEL structure, with an ultralow threshold approaching the (mu) A-range consistent with Yariv's quantum wire view.

Fabry-Perot microcavities are used for the alteration of photoluminescence in hydrogenated amorphous silicon nitride grown with and without ammonia. The photoluminescence is red-near-infrared for the samples grown without ammonia, and blue-green for the samples grown with ammonia. In the Fabry- Perot microcavities, the amplitude of the photoluminescence is enhanced, while its linewidth is reduced with respect to the bulk hydrogenated amorphous silicon nitride. The microcavity was realized by a metallic back mirror and a hydrogenated amorphous silicon nitride--air or a metallic front mirror. The transmittance, reflectance, and absorbance spectra were also measured and calculated. The calculated spectra agree well with the experimental spectra. The hydrogenated amorphous silicon nitride microcavity has potential for becoming a versatile silicon based optoelectronic device such as a color flat panel display, a resonant cavity enhanced light emitting diode, or a laser.

We study the temperature effects on gain, refractive index, and the linewidth enhancement (or alpha) factor in a 2D semiconductor. We consider both the standard alpha factor due to carrier density variation and that due to temperature variation. The so-called total alpha factor due to both carrier density and temperature is also studied. As a result of temperature change, the alpha factors can be positive, negative, or zero.

A self-consistent rate equation analysis of the direct current modulation response is performed through a calculation of the relevant carrier and tunneling lifetimes for a prototype triple quantum well structure suitable for incorporation in intersubband lasers. Opportunities for accessing predicted THz modulation response capabilities in such devices are examined. The intersubband optical gain spectra of these devices are investigated.

Temperature dependencies of the threshold current, device slope efficiency and heterobarrier electron leakage current from the active region of InGaAsP/InP multi-quantum-well lasers with different profiles of acceptor doping were measured. We demonstrate that the temperature sensitivity of the device characteristics depends on the profile of p- doping, and that the variance in the temperature behavior of the threshold current and slope efficiency for lasers with different doping profiles cannot be explained by the change of the measured value of the leakage current with doping only. We show that doping of the p-cladding/SCH layer interface in InGaAsP/InP multi-quantum-well lasers leads to improvement of the device temperature performance. We also show that doping of the active region increases the value of the optical loss without degradation of characteristic temperature T₀.

A set of single 10 nm Ga_xIn_1-xP/(Al_xGa_1-x)_0.52In_0.48P quantum well lasers with compressively strained active regions have been grown by low pressure metal-organic chemical vapor deposition with the intention of examining the physical mechanisms which inhibit successful laser operation at high strain. In this paper we show that strain induced structural non-uniformity similar to Stranski-Krastanov growth is the primary mechanism for both the partial relief of strain and an increase in the optical losses which combine to cause an increase in the threshold current density as the compressive strain increases from 1% to 2%. In this paper we present a detailed analysis of the characteristics of the laser devices and using photoluminescence, electron microscopy, and x-ray microanalysis we relate the laser performance to the structural features we observe at high levels of mismatch between the active region and the material surrounding it.

We present a structure which is capable of being fabricated into two distinct devices, both with considerable potential in the field of optical communications in particular with reference to wavelength domain multiplexing. The structure is based on two back to back p-i-n Ga_xAl_1-xAs structures with a single quantum well of GaAs in each intrinsic region. The light emitter device operates by forward biasing either of the p-i-n elements. In forward bias holes flood into the quantum well in the intrinsic region. Electrons are prevented from doing so by a potential barrier. A longitudinal electric field applied along the central n-doped region heats the electrons in this region and gives them sufficient energy to overcome the barrier and flood into the quantum well and hence recombine with holes which are already present. The wavelength converter device operates with one p-i-n structure forward biased and one reverse biased. The forward biased element has a quantum well positioned near the p-doped region. Light of the appropriate wavelength is absorbed in this quantum well. The holes scatter out of the quantum well and drift into the p- doped region. The electrons are scattered out of the quantum well and drift towards the n-doped region, creating additional carriers through impact ionization, thereby creating gain. The electrons flooding over the n-doped region, must overcome a potential barrier to enter the forward biased element, therefore cold electrons are prevented from entering this region. Electrons which are able to overcome the barrier fall into a quantum well positioned near the barrier, where holes are already waiting, as in the light emitting device.

The physics and applications of interband nonlinearities in semiconductor lasers and amplifiers are reviewed. Although these nonlinearities are said to be `slow' (with responses of order nanoseconds), several methods have been reported where ultrafast switching speeds (up to 40 Gb/s and above) have been achieved. Potential network applications for components operating at these speeds abound in areas such as demultiplexing, wavelength conversion and signal regeneration.

Nearly degenerate four-wave mixing in above-threshold laser diodes with symmetric or asymmetric facet reflectivities, subject to strong probe injection power, is investigated theoretically, taking into account the effects of pump depletion, carrier diffusion, gain saturation, gain compression, total power dependence of the gain and coupling coefficients as well as the longitudinal dependence of the nonlinear interaction. It is shown that the pump depletion effect plays an important role for strong input probe power. The reflectivity efficiencies of probe and conjugate wave demonstrate significantly different behaviors for strong probe injection power case, and a greater than approximately 3 dB enhancement of reflectivity efficiencies can be achieved in an asymmetric laser diode, compared to a symmetric laser.

We have demonstrated the highly nondegenerate four-wave mixing among sub-picosecond optical pulses in a 1.3 micrometers multiple-quantum well semiconductor optical amplifier (SOA). We could directly measure the four-wave mixing signal in the output spectrum by current pulse pumping of the device. We achieved a high conversion efficiency of over 10% at a frequency conversion range of less than 1 THz. The limitation of conversion efficiency due to the gain saturation of the SOA was effectively overcome by using the short optical pulses. Preliminary results of optical sampling with picosecond time resolution are also described.

In the recent established theories of semiconductor lasers, the gain is analyzed basing on a third order perturbation theory in which the gain suppression is approximated by the third order. The limitation of applying such theories stands on the limited range of the applied injection current. To extend the application to higher ranges of the injection current, we go beyond the third order towards an infinite gain expansion, presenting full analysis and discussion of the gain suppression basing on a two-mode model. Formulas for general terms of the density matrix and gain are reported. A general formula for the gain of a non- oscillating mode is presented and is shown to reduce to the previously reported formulas in cases of homogeneous and inhomogeneous broadening in a single mode operation. Numerical discussion of the unsuppressed, linear, and suppressed gain as well as the contribution of the higher order gain terms is presented. Furthermore, simplified and convenient expressions of the different orders of the gain expansion are investigated basing on the numerical results. A criterion to truncate the expansion is then investigated basing on the magnitude of the applied range of the injection current. The gain can be represented only by a third order expansion in the conventional range of the injection current up to about 7 of its threshold value I_th within accuracy less than 0.1%, while the 5-th order expansion gives accurate description at higher current values up to about 25 I_th. The higher order terms have to be included in the gain expansion at the higher ranges of the injection current to get more exact gain analysis.

In this paper, a new model of self-pulsating semiconductor lasers (SPLs) is proposed based on the analogies between a SPL and a laser oscillator. In fact a SPL is first of all an oscillator even though it is difficult to identify amplifiers and feedback elements. The use of such analogies allows in some conditions to establish a very simple rate equation model of SPLs. One of the advantages of the model is the ability to analyze the injection locking properties of SPLs under external optical signal injection. In particular, it has been found that (1) the frequency looking range is directly proportional to the power ratio, (2) the spectral linewidth is equal to that of the incoming signal in the locking state. These points have been verified experimentally on a distributed feedback SPL. Nevertheless, a SPL is different from a laser in that the self-pulsation signal can be short pulses as well as sinusoidal. The short pulses nature means that there are higher order harmonics created by nonlinear effects related to the laser operation. These higher order harmonics can be taken into account in our new model, which becomes however more complex.

This paper presents the dynamics of a semiconductor laser subject to instantaneous phase-conjugate feedback. Recently, the behavior of such a laser have been explored by means of bifurcation diagrams. However, the exact nature of the involved dynamics and bifurcations remained unclear. Here we present a detailed study of the changes of the dynamics as the feedback strength is varied.

For a diode laser subjected to feedback from a phase- conjugating mirror we present (1) the first exact stability- analysis and (2) various spectra. The stability properties are intermediate between those of the injection laser and the laser with conventional optical feedback. The role of a finite response-time is to drastically enhance the steady- state stability. For moderate feedback the frequency noise is suppressed by several orders of magnitude, and the main relaxation frequency of the laser shows a crossover from the usual relaxation oscillation frequency to a new frequency determined by the amount of feedback. This may be of technological importance since it improves the modulation bandwidth.

Numerical studies of the dynamic transfer of one chaotic oscillation appeared in a laser diode, which is subject to weak optical feedback, to a variety of periodic oscillations by modulating the injection current of the laser diode, are described. We evaluate also the response of periodic attractors in the stabilized oscillations by calculating statistic distributions of the transient time among those attractors. The average transient time is approximately 21 ns, which is longer than the full round time in the controlled periodic attractors.

The dynamics of two mutually coupled semiconductor lasers is studied experimentally and numerically for weak coupling. The two lasers are pumped at different levels and have dissimilar free-running relaxation oscillation frequencies and intensities. We have found that the two lasers exhibit large amplitude intensity oscillations for one laser and small amplitude intensity oscillations for the other laser. In our previous work, we studied the effect of the coupling strength. In this paper, we keep the coupling strength fixed and examine the effect of the detuning between the optical frequencies. As the detuning comes close to a multiple of the relaxation oscillation frequency of one of the two lasers, we observe the amplification of the intensities due to resonance and synchronization. Our experimental observations of the laser spectra are in good agreement with the numerical spectra and bifurcation diagrams computed from the semiconductor laser rate equations.

In this work it is studied the dynamics of a single-cavity semiconductor laser device, able to work in self-pulsating regime as a consequence of a nonlinearity distributed in the whole active region and driven by the electromagnetic field intensity in the laser cavity. Numerical values of the nonlinearity degree necessary to establish the self- pulsating regime are given which can be used for a comparison with experimental measurements of semiconductor materials properties. Optical feedback influence on the self-pulsating device is also studied showing how pulsation characteristics are affected by the presence of a reflected signal and pointing out feedback ability to cause their generation or quenching. The possibility of using current and optical feedback for the pulsation characteristics control and for the switch of the laser output is discussed.

A multimode iterative scheme to model the dynamic and noise properties of semiconductor laser diodes is presented and results obtained for semiconductor lasers under continuous wave conditions.

The origin of Low Frequency Fluctuations (LFF), occurring for semiconductor lasers with delayed optical feedback from an external mirror, has been intensively discussed within the last ten years. One of the key questions has been whether the LFF are a deterministic or a noise induced phenomenon. We present experimental evidence for a deterministic mechanism by demonstrating the two major predictions of this model: irregular pulsing of the intensity on picosecond timescales underlying the LFF dynamics, and the existence of stable compound cavity modes with high gain within the LFF regime. We present a systematic investigation of the dynamical regimes with respect to both, feedback rate and injection current. We demonstrate that stable emission can indeed be achieved for a wide parameter range within the LFF regime.

We analyze the microwave dynamics of a semiconductor lasers with optical feedback in the Low Frequency Fluctuation regime. Our experimental measurements, which are based on a statistical approach, are at variance with the solution of the model commonly used to describe this instability. We then discuss the possible reason for the differences and develop an analogy with the turn on transient of a semiconductor laser.

We show that there exists phase-locked motion on a torus in the full single-mode rate equations of a diode laser subject to weak optical injection. The stability of the torus depends on the linewidth-enhancement factor (alpha) . The case of unstable torus motion leads to a bistability, which gives rise to hysteresis when the detuning of the input signal is varied for suitable injection strength. The torus becomes unlocked in a heteroclinic bifurcation. Our results confirm the validity of a previous analysis of an averaged phase/amplitude equation by means of bifurcation theory. We present the relevant dynamics by several panels, which include plots of the time series of the power, phase plots, and field and RIN spectra.

Electron-spin relaxation has been investigated in a variety of quantum confined structures including GaAs-based quantum wells and quantum wires as well as InP-based quantum wells, and it has been shown to exhibit ultrafast relaxation in the picosecond, and even in the femtosecond, range. This ultrafast relaxation has been used together with the exciton absorption nonlinearity as an novel principle of all-optical switching device which can avoid the speed limit due to the slow carrier lifetime in conventional switching devices. Experiments on GaAs-based quantum-well spin switches have shown 4-ps gate switch operation. Also, experimental analyses have shown the feasibility of these devices at a high contrast ratio (13 dB) and high repetition rate (40 GHz). The application of this device to a demultiplexer has been proposed.

The feasibility of the inter-subband transition (ISBT) in Al(Ga)N/GaN quantum wells (QWs) as a device mechanism for ultrafast optical switches is theoretically investigated. The 1.55-micrometers ISBT is shown to be feasible because of its large conduction band discontinuity. The inter-subband relaxation time at 1.55 micrometers is estimated to be about 100 fs, which is 20 - 30 times shorter than that in InGaAs QWs. A large electron-electron scattering rate causes a short dephasing time (about 10 fs), which reduces the peak value of the third-order nonlinear susceptibility. At a high carrier density, however, the dephasing time is increased because of the screening and the exclusion, which enhances the nonlinear susceptibility. Ultrafast relaxation of the inter-subband optical nonlinearity in GaN QWs is little affected by the delay in intra-subband energy relaxation caused by non-equilibrium phonons. These characteristics suggest that the ISBT in nitride QWs is a promising mechanism for multi-terabit/s optical switches.

We report the observation of exciton blue shift in a DC- biased GaAS single quantum well with off-resonant pumping which is distinguished from the ac Stark shift by polarization selective pump probe spectroscopy with small detuning condition. This blue shift is interpreted as a result of internal field screening by virtual excitons.

There have recently been significant advances in compact sources of ultrafast optical pulses. We review our work on developing fiber lasers that are modelocked using a saturable Bragg reflector. These lasers exhibit interesting polarization dynamics due to the lack of any explicit intracavity polarizers. The use of these lasers as a source in a wavelength division multiplexing transmitter is discussed.

Erbium-doped fibers are ideal as the basic components for lasers and amplifiers operating near 1550 nm. Consequently, considerable interest and research in the past few years has focused on the use of Erbium-doped fibers in actively and passively modelocked fiber lasers operating in both the normal and anomalous dispersion regimes. This paper presents a variety of analytic and numerical techniques which can be used to quantitatively describe the dynamic pulse formation in a passive laser cavity which includes dispersion, nonlinearity, loss, and bandwidth limited gain-saturation. The mechanisms for modelocking (e.g. quantum-well semiconductor structures, APM action, etc.) are incorporated into the modelocking models via various appropriate approximations. The models capture the fundamental intensity-dependent loss required to achieve stable modelocked operation. This generic feature of modelocked lasers will be discussed in detail and illustrated in various modelocking devices and configurations. Additionally, the characteristic differences between modelocking in the normal dispersion regime, for which highly chirped pulses are observed, and the anomalous dispersion regime, for which soliton pulses can be achieved, will be discussed. Finally, stability of the modelocked pulses will be investigated. Here we will emphasize the range of stable one-pulse per round trip operation and the instability to multi-pulse operation. For multi-pulse operation, mechanisms for harmonic modelocking such as gain- relaxation, dispersive radiation, and the acoustic effect will be discussed. In most instances, analytic methods can be utilized to gain significant insight into the laser operation and stability.

Sol-gel process appears as an interesting alternative to realize optical integrated devices working at 1.55 micrometers wavelength in reason of the low cost of the materials and equipment used and the good performances realized. The principle of fabrication is to deposit a photocurable sol- gel layers on a substrate like silicon and to imprint waveguides through a predefined mask with an UV light. The buried quasi-symmetrical waveguides achieved are well adapted to the use of the Galerkin's method associated with Hermite-Gauss basis functions to resolve scalar, semi- vectorial and vectorial waves equations. Optical integrated devices can be simulated in many steps. In a first time, the device is dividing into several cross sections. Transverse field is, then, calculated by series expansion to determine the mode amplitudes of each section. The field propagation is determined in establishing a relation between the coefficients of two consecutive sections. The description of the transverse sections is easily realized region after region. Each region presenting a constant refractive index, an analytical formulation is given. From this calculation method, the optimizations of a transverse structure achieved by sol-gel process is made according with the fabrication parameters. Propagation method is, then, applied an can be generalized to study optical integrated passive devices.

GaInP is a very important material for opto-electronic components operating in the visible spectra range ((lambda) equals 630...690 nm). Depending on MOVPE growth conditions, spontaneous ordering occurs which is established in form of a monoatomic superlattice arrangement of the group III components in {111}_B planes. The reduced symmetry in ordered crystals leads to a polarization anisotropy of the absorption coefficient, which was investigated with transmission experiments. The polarization anisotropy was utilized to realize polarization threshold switches and polarization detectors. The polarization threshold switches consist of a n-i-p photodiode with an intrinsically integrated junction field-effect-transistor amplifier. The polarization anisotropy of the ordered GaInP absorption layer results in a polarization dependent source- gate photocurrent. This finally determines the working point of the device and yields in a very sensitive control of the source-drain conductivity G_SD by the polarization of the incident light. First measurements reveal a switching contrast of 50 dB between light linearly polarized along the [011] and [01-1] crystal direction, respectively.

We study the wave propagation through a semiconductor optical amplifier. It is investigated how the anisotropy caused by the quantum wells and the waveguiding together determine the polarization sensitivity of the amplification. We also analyze how the tensile strain in the quantum well can be optimized such that the resulting amplifier operates independent of the state of polarization of the incident field. The promising properties of non-(001) grown quantum wells are discussed.

The unification of transverse electric (TE) and transverse magnetic (TM) beam propagation algorithms is made possible through a transformation which converts the wave equation for TM fields in planar waveguides into a form identical to the corresponding TE wave equation. The transformation can be applied to any smoothly varying waveguide. This transformation can be made independently of any paraxial or other approximations. Thus, any TE propagation algorithm can also be applied immediately to TM fields without additional approximations. This includes the classical fast Fourier transform beam propagator, which has not previously been applied successfully to TM propagation. We also specifically develop a Finite Difference Beam Propagation Method that applies to both TE and TM propagation in 1D (planar) geometry. Previous implementations for the TM case involve an approximation that in certain circumstances leads to severe errors (including the totally unphysical occurrence of field amplification). This is the first TM propagator which exactly conserves power. We also investigate the role of the reference background wavenumber (or index) and clarify its role as it is dynamically adapted. The algorithms proposed are easily adaptable to wide-angle beam propagators and to modern transparent boundary conditions. The extension of these ideas to rapidly varying structures (such as Bragg gratings) is also briefly discussed.

The cutoff wavelengths of the higher-order supermodes of symmetric square-core and rectangular-core step-profile couplers are determined in order to interpret the spectral response of a practical single-mode buried-channel waveguide coupler. Our results suggest the possibility of simple diagnostic tests for such devices and highlight the design constraints imposed by second order supermodes. We also compare the behavior of these couplers with that of circular-core couplers to illustrate how the geometry of each core affects the coupler cutoff characteristics.

To circumvent the numerous performance-limiting effects of hot-carrier phenomena in semiconductor quantum-well lasers, we have demonstrated the innovative approach of directly injecting carriers into the lasing subband by tunneling. These lasers, made with a variety of material systems, have shown evidence of reduced hot-carrier effects. Specifically, measured small-signal modulation bandwidth of approximately 50 GHz and maximum intrinsic bandwidth of 110 GHz have been achieved with 0.98 micrometers lasers. These are the highest measured modulation bandwidths in any laser. Auger recombination has been virtually eliminated in 1.55 micrometers lasers and reduced chirp and temperature dependence are also demonstrated. Significant reduction of hot-carrier and carrier leakage effects have also been recently demonstrated in small-area vertical-cavity surface-emitting lasers. These experimental results are supported by recent simulations that identify gain suppression in high speed lasers to be caused by a coupling between the electron temperature and the quasi Fermi level. Lasers with quantum dots as gain media promise high differential gain, very low threshold current, temperature-insensitive operation and high modulation bandwidth. We have investigated room-temperature single-mode ridge-waveguide quantum box lasers in which the quantum box gain regions are realized by self-organized growth and carrier injection is achieved by conventional means over hetero-barriers and by tunneling. The tunneling injection quantum dot lasers show improvements in both differential gain (6 X 10^-14 cm²) and modulation bandwidth. These results will be presented and discussed.

Light-current characteristics for the 1.55 micrometers side- injection light-controlled bistable laser diode measured for a range of operating characteristics show good agreement with corresponding results calculated from steady-state rate equations. Voltage measurements show characteristic shifts at switch-on and switch-off which are attributed to carrier redistribution between the gain and absorber sections. All- optical wavelength conversion is demonstrated over an input wavelength range of 56 nm, and the dependence of optical switching power on wavelength shows less than 3 dB sensitivity over a range of 31 nm.

In this paper, modeling of semiconductor diode lasers by sampling the cavity of the laser using digital techniques is presented. These models, that sample the cavity of the semiconductor laser simulating the spread of the field in the active zone, are the only methods that permit evaluation of the spectral behavior of the device. Throughout this work, we present the study of the temporary evolution of the spectrum of the longitudinal modes of a semiconductor laser by using digital simulation techniques. Due to the digital nature of the filter and algorithm, we can adjust any given frequency response and reduce CPU time in computing the results. We present two digital models, DLM and LDDM, which improve the behavior of the traditional methods used in order to obtain temporary spectral response of semiconductor laser diodes.

In this paper we study the influence of random noise on route to chaos followed by directly modulated laser diode as we found that is more complex than just a Feigenbaum sequence. We have already shown that when the laser is modulated at twice its resonance frequency and the modulation index is increased, a period tripling stable solution coexists with the Feigenbaum sequence for certain values of the modulation index, giving rise to hysteresis loops and chaotic bifurcations. This coexistence gives the noise a relevant role in determining the dynamic behavior. We have studied the route to chaos of the laser diode when the random noise fluctuations are introduced in the model. For this study, we have used three different numerical methods. We have employed a single step fourth order Runge- Kutta algorithm, which has been commonly used in previous studies, and its results have been compared to those provided by two stochastic integration methods: Euler- Maruyama and a stochastic Heun method. In all three cases, the behavior evolves from the single periodic response through period doubling, period quadrupling and period tripling in accordance with recent experimental studies, showing that in the route to chaos, the period doubling sequence is effectively truncated due to random noise. The reason for the truncation is found in the nearby coexisting period three solution. Nevertheless, the Runge-Kutta algorithm smoothes off the random noise fluctuations, probably due to its large correlation time.

We report on the excitation and propagation of an exciton density front inside a semiconductor that is characterized by high absorption and large optical nonlinearity. Femtosecond optical pulses are used for both the excitation of the density front and the probe of the front propagation. We analyze in detail the spectra of reflected probe pulses that carry information about the propagating density front due to partial internal reflection of light at the boundary between regions of high and low excitation density. Time resolved data show that the Doppler shift of the internal reflection is limited to the duration of the pump pulse indicating the highly transient character of the exciton front propagation.

An accurate numerical approach has been developed and is presented, based on the use of finite element method (FEM) to study the nonlinear effect of Second Harmonic Generation (SHG) in optical waveguides. The evolution of the fundamental and second harmonic fields is followed by the implementation of a beam-propagation-type program using a split-step Crank Nicholson procedure based on the FEM. For the purpose of comparison, second harmonic generation using the Cherenkov radiation scheme in a planar waveguide is first considered and the results obtained in this research show very close agreement with those of earlier published work. Also presented are results obtained on frequency doubling for guided modes in planar waveguides using the quasi-phase matching scheme. Whilst planar waveguides allow for a comparatively easy analysis and modeling, practical waveguides are 2D devices with arbitrary diffusion profiles, and in this paper the work is extended to SHG in channel waveguides, with confinement in two transverse dimensions. Results are presented for both Cherenkov and quasi-phase matched devices. This method could prove to be very useful in the design and optimization of optical guided-wave devices and the present approach can be extended to the study of cascaded effects in nonlinear devices.

In this paper, we demonstrate that eigen mode expansion can be used successfully to find the propagation constants and field distributions for the modes in an optical fiber. Instead of using sine-functions, we use Bessel functions as the basis for the expansion of the modes. Comparison between the results of this method and the exact analytical results for a single-mode, step-index optical fiber is shown. Also, theoretical and experimental results for a commercial boron co-doped germano-silicate fiber are compared.

We show how the concepts central to object oriented programming--inheritance, polymorphism and encapsulation-- can be used efficiently to implement extensible and reusable code that can be readily applied to integrated optics modeling. In particular, we will show how object oriented programming can lead to elegant and adaptable simulation code that can be easily tailored for specific needs. This approach leads to small and elegant code that can--in certain instances--be preferable to fully mouse driven commercial programs. Examples are given. We have developed a collection of modern modeling techniques using object oriented programming aimed at integrated optics. This readily available code library encompasses beam propagation methods, mode solvers, gratings and coupled mode analysis. Being based on a general programming language, namely C++, it is possible to use the library to automate tedious calculations and/or optimization problems. The library can also be used to rapidly develop custom made software for rapid prototyping and/or to apply to specific research problems. Moreover, due to its object oriented nature, the library pieces fit together naturally and versatile simulation code can be written.

We discuss optical properties of self-assembled CdSe-based quantum dots (QDs) studied by photoluminescence (PL), selective PL, PL excitation, and magneto-PL in our laboratory as well as by other groups. Our own QD specimens were fabricated by depositing approximately 3 monolayers of CdSe on ZnSe using MBE. Samples consisting of capped QDs showed intense excitonic transitions from the QD ensemble. The PL spectra were strongly inhomogeneously broadened, but survived to relatively high temperatures without further broadening of the PL line. Selective PL showed clear LO- phonon resonances which--together with the observed temperature stability of the PL band--clearly indicate that the broad band is a superposition of atomic-like spectra emitted by strongly localized excitons. In addition to CdSe QDs on ZnSe, we also studied magneto-PL of QD systems involving diluted magnetic semiconductors (DMSs), in an attempt to determine the degree of exciton localization in the dots. While these experiments are still inconclusive, the results--and particularly the observation of a dramatic magnetic-field-induced enhancement of PL intensity in DMS based systems--are extremely interesting in their own right, and will also be discussed.

We report photoluminescence (PL) studies of zero-dimensional excitonic confinement in InAs/GaAs self-assembled quantum dots (SAQDs) fabricated by Stranski-Krastanow growth. A low- temperature near-field optical microscope is employed to identify individual PL peaks due to single SAQDs. Highly spatial and spectral evaluation of each PL peak identify the Zeeman spin splitting with the magnetic field up to 10 T. The comparison of different confinement structures was also discussed.

Although porous silicon has received the most attention over the last 5 years, other structures containing Si nanocrystallites have recently been shown to emit strong luminescence at room temperature. This presentation reviews the state-of-the-art in the preparation and properties of silicon quantum dot structures, and their use in electroluminescent devices. The materials science and the structural, chemical, electrical, and optical properties of both porous silicon and recrystallized silicon superlattices are discussed. The fabrication of light-emitting devices consisting of these two types of nanoscale silicon materials is then described. Finally, the present and projected stability, efficiency, and speed of these LEDS is reviewed and their integration with silicon microelectronic driver circuits demonstrated.

Applying a magnetic field along principal axes, the fine structure of the groundstate exciton caused by the electron- hole exchange interaction is studied for (Zn,Cd)Se/ZnSe quantum wells and CdSe quantum dots on ZnSe. In quantum wells, the zero-field splitting between optically allowed and forbidden (dark) states is increased by a factor of about two with respect to the bulk ZnSe value (0.25 meV). The localization of excitons gives rise to a finite, but minor contribution from the non-analytical part. For quantum dots, as a result of the 3D confinement, the splitting is further enhanced and reaches values ranging between 1.5 and 2.0 meV. Its weak size dependence is a consequence of the finite barrier height at the CdSe/ZnSe hetero-interface. Our findings signify that dark excitons do not profoundly affect the laser action of these quantum structures.

Detailed theoretical analysis of the temperature dependence of threshold current density of a semiconductor quantum dot (QD) laser, j_th, is given. Temperature dependences of the components of j_th associated with the radiative recombination in QDs and in the optical confinement layer (OCL) are calculated. Violation of the charge neutrality in QDs is shown to give rise to the slight temperature dependence of the current density component associated with the recombination in QDs. The temperature is calculated (as a function of the parameters of the structure) at which the components of j_th become equal to each other. Temperature dependences of the optimum surface density of QDs and the optimum thickness of the OCL, minimizing j_th, are obtained. The characteristic temperature of QD laser, T₀, is calculated considering carrier recombination in the OCL (barrier regions) and violation of the charge neutrality in QDs. The inclusion of violation of the charge neutrality is shown to be critical for the correct calculation of T₀. The characteristic temperature is shown to fall off profoundly with increasing temperature. A drastic decrease in T₀ is shown to occur in passing from temperature conditions wherein j_th is controlled by radiative recombination in QDs to temperature conditions wherein j_th is controlled by radiative recombination in the OCL. The dependences of T₀ on the root mean square of relative QD size fluctuations, total losses and surface density of QDs are obtained.

Threshold currents in small-aperture VCSELs are likely to be dominated by diffraction losses. We have developed a semianalytic technique to estimate the lasing mode energies, field profiles and cavity losses--including absorption, mirror, and diffraction losses--in oxide-apertured VCSELs. By coupling these modes to the full, nonparabolic electronic bandstructure, and by solving the resulting multimode related rate equations, we can model the light versus current characteristics of microcavity VCSELs. We apply our model to a low-threshold VCSEL structure and calculate threshold currents of 30 - 40 (mu) A, in qualitative agreement with experiments. We estimate spontaneous emission factors (beta) as large as 1.7 X 10^-2 for a 1.5 micrometers radius device.

A comprehensive numerical model for investigating the thermal, electrical and optical characteristics of vertical cavity surface emitting lasers with a diffused quantum wells structure is presented. In general, this model performs a self-consistent calculation of quasi 3D distribution of temperature, voltage and optical field. The quasi 2D diffusion and the recombination of carrier concentration inside the quantum well active layer are also introduced into the model. In particular, this model includes the calculation of the quasi 3D ion-implantation profile. In addition, the influence of impurity induced compositional disordering on the optical gain and refractive index of the quantum wells active layer is also taken into consideration. Using this model, the steady state characteristics of diffused quantum wells vertical cavity surface emitting lasers are studied theoretically. It is shown that significant improvement of stable single-mode operation can be obtained using diffused quantum wells structure.

We perform a theoretical study of carrier transport effects in the frequency modulation response of single-mode quantum well (QW) lasers under both small and large signal modulations. The lasers considered are ultra-high speed InGaAs/GaAs MQW lasers with intrinsic modulation bandwidth of 40 GHz. A rate-equation model, including spontaneous emission noise, for the carriers in the active region (quantum wells) N_w, the carriers in the separate confinement heterostructure and in the barriers (core) N_c and for the optical field is used. Carrier transport effects are included by considering two effective time constants: the transport/capture time, and the re-emission time out of the QW's (escape time). We also consider the change in the laser frequency due to the refractive index changes in both the wells and the core. In the wells the laser frequency increase with N_w is described by the linewidth enhancement factor. Since the laser frequency is below the bandgap in the core, the refractive index change with N_c, dn_c/dN_c, can be positive or negative depending on the doping level. When dn_c/dN_c > 0 the magnitude of the FM response decreases when the capture (escape) time increases (decreases), but the shape is flat. We have also studied the frequency statistics under large signal modulation. A linear relationship between frequency chirp range and turn-on time is obtained, as in bulk lasers. For QW lasers biased below threshold the chirp range is found to increase (decrease) for large capture times when dn_c/DN_c is positive (negative) due to the contribution of the carriers in the core. The opposite behavior is found for small escape times.

A simple optical rate equation for a distributed feedback laser is derived following an analytical procedure that is based on transforming a known integral equation into an equivalent differential equation. Using this equation to model fluctuations in the phase and photon number, when supplemented with a rate equation for the carrier number expressions for the relaxation oscillation characteristics and linewidth are derived. We find that in a simple distributed feedback structure the relation for the relaxation oscillation frequency is identical to that of the Fabry-Perot laser. An effective linewidth-broadening factor is predicted showing strong dependence on longitudinal hole- burning. Power re-broadening of the linewidth and a nearly vanishing power-independent component are predicted. Finally, rate equations for injection-locking are derived, and a symmetric dynamically stable locking band predicted.

In this paper the possibility of controlling steady states in a model of an external cavity laser diode with optimized impulsive delayed feedback is demonstrated. An examination is made of the application of such feedback via modulation of the laser drive current. Account is taken of practical constraints arising from technical delay and the frequency in application of the control signal. It is demonstrated that control of both unstable periodic orbits and steady states is achievable in the fully developed coherence collapse regime with or without preliminary targeting.

The reverse full-depletion voltages (V_RD's) of PnpN depleted optical thyristors (DOTs) are obtained as a function of semiconductor parameters by using the finite difference method. For the full-depletion, the DOT requires a reverse bias, which determine the clock voltage level. The required magnitude of clock pulse should be as low as possible for the fast operation and the integration with electronics. In this study, we define the condition for the full-depletion and calculate V_RD's varying semiconductor parameters, such as layer thickness and doping concentration. As a result, V_RD is found to be very sensitive to the device parameters. As the worst case, V_RD is changed by 34% for 10% variation of the low doped center layer thickness.

Merging optical technology with fast packet switching will develop generation-after-next advanced technologies scalable to a Tbit/second. Serious challenge of such an implementation is the lack of an optical switch that would have a fast tuning speed, wide tuning range, low insertion loss, and be highly selective with a narrow passband at the same time. For this purpose we propose using a recent experimental Los Alamos National Laboratory and Micron Optics Co. break through--building an optical switch that works three orders of magnitude faster than its commercial predecessors, has a wide tuning range and low insertion loss, and is highly selective. Fast optical tuning in several microseconds is necessary to perform high-speed optical packet switching. One approach to achieve fast wavelength tuning is to use high-speed piezoelectrically- driven Fiber Fabry-Perot tunable filters. A special controller has substantially improved the shape of the driving signal and the response of the filter. The fastest switching time achieved without ringing is 3.0 microseconds. At the same time, implementation of high-speed optical packet switching rises the challenge of the bit synchronization of the packets. Conventional circuits are inadequate when a fast clock recovery for the short length packets is required because they need thousands bits to lock on. To deal with it, we propose to implement a bit clock synchronization technique for high-sped packet-switched optical network on a packet-per-packet basis. As a result, a new packet-switched media access control protocol can be designed to minimize the searching time.

Tunable distributed Bragg reflectors with comb-like reflection spectrum have been shown to allow semiconductor lasers tuning over about 100 nm wavelength range. Recently we proposed a concept of BSG as a multiwavelength reflector and looked into a number of key design issues for widely tunable lasers. In this paper we analyze the technical aspects of designing and implementing the BSG. We show that the calculated performance of the binary grating is at least as good as that of the existing approaches. However, BSG offers additional design freedoms in positions and amplitudes of reflectance peaks, which in turn allows extension of the tuning range of the laser or relaxing the tuning current requirements. The principles of widely tunable lasers have been adapted to a practical design which should significantly simplify the control requirements for the tuning. We show that a BSG reflector for 1.55 micrometers tunable laser can be successfully implemented with the smallest feature size as large as 0.12 micrometers , which is well within the capacity of electron beam lithography. A simulated tuning curve for a BSG laser predicts a high side- mode suppression ratio (> 35 dB).

The variety and complexity of fiber Bragg gratings have grown tremendously in recent years. Advances in fabrication have increased interest in techniques to design gratings with specific spectral characteristics. Existing mathematical design algorithms require both the amplitude and phase response of the grating in reflection to be specified. However, sometimes other quantities are specified in design requirements, such as transmission coefficients or group delay characteristics. The spectral quantities associated with gratings are not all independent and there are dispersion relations connecting them. New relationships between the reflectance, transmittance and various group delays are presented. These relationships provide constraints on the possible gratings that can be designed, and provide a means for designing gratings from different specifications. The group delay in transmission of an arbitrary grating is always determined uniquely from the amplitude response. The corresponding quantities in reflection are determined uniquely if the grating is symmetric. For non-symmetric gratings only upper and lower bounds on the group delay in reflection can be obtained. A better physical intuition of these mathematical design algorithms is provided by a resonance mode formalism. Resonance modes play an important part in understanding linear nonuniform gratings, analogous to the role played by waveguide modes in waveguide theory. Using resonance mode expansions, exact expressions are obtained for the fields, the grating profile and the spectral characteristics for a large class of nonuniform gratings. The exact solutions can be used to investigate designs for grating structures.

Recently developed version of the modified generalized method of moments for misaligned astigmatic beams in nonlinear waveguides is compared with the direct numerical scheme making use of Gauss-Laguerre decomposition of the asymmetric beam profile. It is demonstrated that the approximation of the beam by a flexible generalized Gaussian function provides a correct description of the global beam behavior even for relatively strong nonlinearities.

Based on the Galerkin's method, a numerical method is developed to analysis the cutoff frequencies of guided modes of optical waveguides with arbitrary refractive index profile. Solutions are presented in the quasi-vector regime. Optical waveguide structures with single core of arbitrary shape are considered in this paper. The calculated results are compared favorably with exact vector solution and circular-harmonic expansion method.

Continuous field transformation of the HE_1m modes in various tapered waveguide structures is considered depending on the losses (gain) distribution in the waveguide material and on the behavior of dispersion curves in the plane of transverse wavenumber in cladding. Numerical sweep method is used to plot and analyze radial distribution of cylindrical components of the Poynting vector. Power flows of radiation propagating along the fiber axis in core and cladding are shown to depend on the tapered structure configuration.

Aluminum nitride (AlN) thin film is a wide bandgap semiconductor which shows great potential for opto- electronic applications in blue and ultraviolet regions. A prism-thin film coupling technique, and an automatic temperature controlling device have been used to study the thermo-optic properties of single crystalline aluminum nitride thin films grown on sapphire substrates by low pressure metal organic chemical vapor deposition. The temperature coefficient of refractive index n_e for AlN is determined to be 6 X 10^-5/ degree(s)C at a wavelength of 632.8 nm.

We present the results of analysis of microstructural and optical properties of composite nonlinear optical waveguides made by ion implantation of LiNbO₃ with MeV ions of silver. Light guiding properties were studied by the prism coupling method, and nonlinear optical susceptibility was characterized using the Z-scan method. Special attention was paid to the influence of a host material and heat treatment on light guiding, optical absorption, and third order susceptibility.

Experimental responsivity data for a two quantum well QWIP is explained using electric field induced redistribution of density of states. The redistribution of density of states is introduced in the calculation of electron transition rate. It is observed that though the eigen energies of the individual wells are uncorrelated, there exists a strong correlation in the absorption coefficient of the individual quantum wells of a multiple quantum well systems. This redistribution of the density of states is considered as an additional source of broadening in the absorption spectrum of a QWIP.

We report calculations of the collection current of interdigitated InGaAs metal-semiconductor-metal photodetectors. We show how interdigital spacing and thickness of the semiconductor layer influence the collection current. Both front and back illumination of devices carried on thin film membranes by means of epitaxial liftoff are examined.

In this paper we outline work we have recently completed on the novel high speed photodetector which has come to be known as the `back-gated metal-semiconductor-metal' photodetector or BG-MSM for short. This device was first realized by E. Greger et. al. and us. Recent theoretical work by Hurd et. al. provided a sound basis by which to test the device against a model. The work presented here clearly indicates that the main conclusion of Hurd et. al. is sound. That conclusion was that the main response speed gain obtained by the back-gating is obtained by simply earthing the back contact. As well as this confirmation we have found strong dependence of the peak height of the response on the back gating voltage and a very strong position dependence of the excitation spot on the response particularly at low applied fields. We conclude that the back gating is an advantage over the conventional metal-semiconductor-metal photodetector and needs to be studied more extensively both theoretically and experimentally using devices capable of operation at above 100 GHz.

We present the structural analysis of prechirping bulk electroabsorption modulators (EAM's) considering the effective chirp parameter ((alpha) _eff). For an accurate estimation of (alpha) _eff, we propose a novel method that uses the modulator output intensity as a weighting factor. The transmission characteristics using (alpha) _eff obtained by the simple method have an excellent agreement with those obtained by a simulation using the dynamic small- signal chirp parameters. With this method, we perform the structural analysis of the prechirping InGaAsP bulk EAM's varying the device structures and operating conditions in terms of power penalty. From the simulation results, the relations between (alpha) _eff and power penalty is found out for the InGaAsP bulk EAM's.

A new numerical method based on a split-step technique to study the time-domain propagation in integrated optical circuits under the paraxial approximation is presented. The algorithm can be efficiently implemented using a finite difference approach such as the well known Crank-Nicolson one. In this paper we consider its application to passive devices with dispersion and Kerr type non-linearity where soliton emission may occur. Examples of analysis are given.

This paper describes an analysis of directly modulated fiber-optic links using an active matching. Two MMICs amplifiers have been used as post-amplifier for transmitter and preamplifier for receiver. A large signal laser diode model is used as transmitter. An active matching is obtained between 50 (Omega) and both a low impedance of laser and a high impedance of photodiode. The simulated performances of fiber-optic links are presented in terms of gain, bandwidth, noise figure and distortion, in the range: 0.1 - 20 GHz. The noise figure has been reduced by using an active matching of fiber-optic links, also the bandwidth and gain have been increased.

A comprehensive model is presented to investigate optical, electrical, and thermal behaviors of large-area high-power diode laser amplifiers. The resistor network model describing lateral current confinement is used to define waveguide structures. Spatial variation of current injection is taken into account through junction voltage variations. Hole burnings on gain coefficient and junction voltage are calculated from a 1D carrier diffusion equation. To study thermal lensing, the finite element thermal analysis is integrated into the electric model and the beam propagation method. The described distributed electric/thermal model result in some new explanations in thermal lensing and beam filamentation.

An analysis is performed of the effects of strong optical feedback on the characteristics of single-transverse mode VCSELs. Consideration is given to the cases of short (1.5 cm) and long (15 cm) external cavities. When the laser is operated well above threshold in a short external cavity with high reflectivity, a strong dependence on the accumulated external cavity phase is observed. Stable operation is found for an optimum phase while for other values of the phase chaotic dynamics is observed. Such behavior is not seen near threshold where the laser output is stable for any accumulated phase. For the longer external cavity the influence of the phase term is found to be insignificant: chaotic behavior is seen in the output over a wide range of operating currents.

InGaN multi-quantum-well structure laser diodes with Al_0.14Ga_0.86N/GaN modulation doped strained-layer superlattice cladding layers grown on an epitaxially laterally overgrown GaN substrate were demonstrated to have an estimated lifetime of longer than 10,000 hours under continuous-wave operation at 20 degree(s)C. The activation energy of the lifetime was estimated to be 0.5 eV. With the operating current increasing above the threshold, a self- pulsation with a high frequency of 3.5 GHz was observed.

The threshold-like growth of the higher energy photoluminescence from a normal-mode-coupling microcavity was previously attributed to exciton polariton lasing (boser) based on Bose condensation into the upper polariton branch. Experimental evidence is presented here showing that this boser crossover occurs just as normal-mode coupling collapses to the perturbative weak coupling, so that boser action is fermionic after all. I.e., it can be understood as electron-hole recombination within a microcavity with density-dependent emission properties.