The semiconductor laser has been taking an important role in optoelectronics and IT backbones and will be keeping its position for the coming at least 10 years. There would still be a lot of chances to meet really novel subjects in research and development, e.g. the coverage of full spectral ranges, temperature insensitive operation, ultra-high power capability, frequency control, large scale integration, and so on. In this paper, we like to discuss of the semiconductor laser for tomorrow based upon the experience of founding research of vertical cavity surface emitting laser.

Recent manifestations of apparently faster-than-light effects confirmed our predictions that the group velocity in transparent optical media can exceed c. Special relativity is not violated by these phenomena. Moreover, in the electronic domain, the causality principle does not forbid negative group delays of analytic signals in electronic circuits, in which the peak of an output pulse leaves the exit port of a circuit before the peak of the input pulse enters the input port. Furthermore, pulse distortion for these 'superluminal' analytic signal scan be negligible in both the optical and electronic domains. Here we suggest an extension of these ideas to the microelectronic domain. The underlying principle is that negative feedback can be used to produce negative group delays. Such negative group delay scan be used to cancel out the positive group delays due to 'transistor latency' as well as the 'propagation delays' due to the interconnects between transistors. Using this principle, it is possible to speed up computer systems.

The GaInNAs is an attractive material for long wavelength lasers on a GaAs substrate and the GaInNAs vertical cavity surface emitting laser (VCSEL) is a viable candidate for low cost and high performance lasers of 1.3micrometers wavelength networks due to excellent temperature characteristics and manufacturing capability of VCSELs. We have successfully grown GaInNAs quantum wells by chemical beam epitaxy and investigated the growth condition toward better crystal quality by employing a radical nitrogen source and thermla annealing. Lasing characteristics of CBE grown 1.2 micrometers GaInNAs lasers are a threshold current density of less than 1kA/cm², and high temperature operation up to 170 degrees C with an excellent slope efficiency change below -0.004dB/K. A characteristic temperature of 270K is also demonstrated. GaInNAs quantum dots were also investigated for the further progress of GaInNAs lasers. The growth of self-organized Qds and a lasing operation at 77K was demonstrated.

We calculate microscopically the gain and absorption, linewidth enhancement factor and carrier capture times for a GaInNAs/GaAs quantum-well laser operating in the 1.3 micrometers wavelength regime. The results are compared to those for an InGaAsP/InP and an InGaAlAs/InP structure with similar fundamental transition energies. The much higher confinement for carriers in the GaInNAs quantum well is shown to lead to larger gain bandwidths and, for low to moderate carrier densities, to lower linewidth enhancement factors than for the later two material systems. On the other hand, the high depth of the wells leads to longer carrier capture times in GaInNAs/GaAs.

We present calculations of the optical properties of GaInNAs/GaAs quantum wells. In particular, we have estimated the spectral form of the material gain based on the Band Anti-Crossing (BAC) model. The electron effective mass and the conduction band density of states are calculated from the dispersion relation derived in the context of the BAC model. The effect of nitrogen on the valence band is considered to be minimal. Based on these gain calculations, we have derived the dispersion of the linewidth enhancement factor. In the light of these results we discus the limitations of the BAC model.

A novel ridge structure fabricated by selective-area epitaxial growth is proposed for InGaN MQW laser diodes (LDS). This technique is capable of precisely controlling the active ridge width and height, thus enabling stable single transverse-mode operation. Together with a backside n-contact on a low-dislocation-density GaN substrate, this structure provides high productivity and performance for GaN-based blue-violet LDs. The LDs fabricated by this technique have achieved continuous-wave operation at more than 30 mW up to a temperature of 90 degrees C, with a characteristic temperature T₀ of 105 K from 20 to 90 degrees C. The laser design and characteristics are discussed in detail.

Transverse mode in GaN-based violet laser diodes for both vertical and horizontal directions was investigated. In order to achieve stable fundamental mode operation in vertical direction, thick AlGaN contact layer is found to be effective. For the stabilization of a transverse-mode in horizontal direction of the conventional ridge-waveguide structure, it is necessary to precisely control the remaining thickness of p-AlGaN cladding layer. In comparison, inner stripe structure using AlGaN current blocking layer has wide feasibility of the device parameter, excellent stability of large optical confinement, and small aspect ratio of beam divergence, under the condition of the precise control of the AlN molar fraction in AlGaN current blocking layer.

The electroluminescence efficiency of InGaN LEDs is surprisingly high for structures which have high defect concentrations due to growth on mismatched substrates. We have measured the high-injection non-radiative lifetime in InGaN LEDs by analysis of the light current characteristics. We find that the values of (tau) _nr decrease from 18ns at 200K to 5ns at 400K. This behavior is thermally activated with an activation energy of 40 meV which is compatible with the hypothesis that the temperature dependence is due to thermal delocalization of carriers form potential minima caused by modest fluctuations in In composition in the quantum well. We determine the internal quantum efficiency to lie between 52 percent and 65 percent at room temperature over the current range employed.

The origin of the strong polarization fields in nitride heterostructures is discussed by comparing the symmetry properties of zincblende and wurtzite structures. Some peculiar effects in nitride heterostructure electronic properties, induced by the polarization fields are considered, like the determination of the valence band offset by x-ray photo emission spectroscopy (XPS) and the formation of 2D electron gases in AlGaN/GaN. The Fermi level position at MBE GaN, AlGaN and AlN surfaces has been measured in-situ by XPS. The role played by surface states has been emphasized, experimentally and through self consistent calculations.

Using 6H-SiC(0001) and (gamma) -LiAlO₂(100) substrates, identically designed GaN/(Al,Ga)N heterostructures with (C- plane)- and (M-plane)-orientation, respectively, are grown by plasma-assisted molecular beam epitaxy. The latter case is of special interest, because such structures with the hetero-interface parallel to the hexagonal c-axis are free of electrostatic fields. This fact leads to distinct differences in the spontaneous emission. While the photoluminescence from the conventional oriented wells is unpolarized, a strong polarization anisotropy of over 90 percent is observed for the M-plane sample. The data are in excellent agreement with the expectations from the known valence band structure of wurzite GaN, taking into account a confinement induced admixture of the different subvalence bands in the M-plane quantum well. Secondly, a distinctly enhanced recombination rate in the electrostatic-field free M-plane sample with respect to the C-plane MQW is seen in time-resolved photoluminescence studies proving the increased electron-hole overlap in the M-plane case. Photoluminescence studies at high excitation densities are carried out to uncover the physical gain mechanism for both kinds of heterostructures.

In this work the optical cavity of a vertical-cavity surface-emitting laser (VCSEL) is analyzed with the goal of performing a coupled electro-optical simulation of the device. For this simulation, the eigenmodes and the eigenvalues of the optical cavity have to be obtained. A common approach is to treat Maxwell's equations in the frequency domain and to solve the resulting algebraic eigenvalue equation. As an alternative, the electromagnetic problem is solved in time-domain. The response of the optical cavity is calculated by the finite-difference time-domain (FDTD) method. The optical wave propagation is modeled rigorously, including evanescent and propagating waves. From the FDTD simulation, a steady state optical intensity pattern is extracted. The eigenvalues of the dominant modes are determined using a Pade type approximation.

The effective frequency method (EFM) is generalized to the case of vector analysis of LP modes in VCSELs and applied to proton-implanted and oxide-confined VCSEL structures. Resonant wavelengths and mode profiles of a number of vector LP modes are calculated for different values of active region radius. The vector EFM is shown to provide important information about the vertical and radial components of the energy flux inside the laser structure. The maximum values for 'weak' E_z and 'strong' E_x components of the electrical field within the laser structure are calculated. Their intensity ratio in the 'near-cutoff' situation is suggested as a quantitative verification of weak-guiding assumption in VCSELs, on which the derivation of the EFM was based.

We present detailed yet largely analytical models for gain, optical bandwidth, and saturation power of vertical-cavity laser amplifiers (VCLAs) in reflection and transmission mode. VCLAs are potential low-cost alternatives to in-plane laser amplifiers and they have the inherent advantage of polarization insensitivity, high fiber coupling efficiency, and low noise figure. Simple formulas for the optical gain-bandwidth product are derived which are valid for any type of Fabry-Perot amplifier. Our saturation model is based on rate equations for electrons and holes. It considers a sub-linear material gain, gain enhancement by the standing wave effect, Auger recombination, defect recombination, and spontaneous emission. Common linear approximations are avoided to correctly predict performance limits. Excellent agreement with measurements on novel 1.3-micron VCLAs is obtained. The models are used to analyze device performance and to investigate optimization options. With reduced top mirror reflectivity and increased pump efficiency, substantial and simultaneous improvements of optical bandwidth and saturation power are predicted without sacrificing gain. Parameter plots are given which allow for an easy exploration of the VCLA design space, matching desired performance goals with the required mirror reflectivity and pump current.

We develop a mesoscopic model of semiconductor dynamics for vertical-cavity surface-emitting lasers which allows us to describe polarization and transverse mode dynamics simultaneously. Within this model, we study the selection processes and the turn-on delay for the switch-on of different transverse modes in gain-guided VCSELs. We consider different active-region diameters, excitation conditions and current shapes. Following the application of the current pulse, transverse modes become excited in a quite definite sequence. After the turn-on, the VCSEL initially switches-on in the fundamental transverse mode, but higher-order transverse modes become excited later. In general, the results obtained are in qualitative agreement with experiments reported recently. Finally, we discuss the current shape dependence on the transverse mode selection at threshold.

We present a microscopic theory of the coherent third order optical response of semiconductor quantum well micro cavities, specialized to the four-wave-mixing configuration in the spectral vicinity of the lowest exciton frequency. The theory is that of a quantum mechanical many-electron system dipole-coupled to a classical radiation field. The many-electron dynamics is treated within the dynamics- controlled-truncation formalism restricted to the 1s-exciton subspace. Within this limitation, al Coulomb correlation effects are included, resulting in an effective theory of exciton-polariton scattering. The theory is evaluated for various polarization configurations each of which depends differently on the underlying many-body effects, such as phase-space filing, Hartree-Fock exchange, and two-exciton correlations.

We present a microscopic model of emission in a series of strain-compensated GaInAs/GaAsSb type-II superlattice structures with infrared applications. The need for an improved understanding of the optoelectronic characteristics of these systems, both in terms of basic physics and technological applications, is identified. The band lineup in heterostructures containing alloys is frequently determined using the Model Solid theory with linear interpolation of input parameters between those of the constituent compounds. However, for the present superlattices, this approach did not provide a description of the band lineups which was consistent with experimental data. Band lineups were subsequently fitted to achieve spectral cutoff measurements, and we found that these offsets were in better agreement with experimental data than those predicted using the above method. On using these lineups as input to our empirical pseudopotential model, lineshapes exhibiting good agreement with experiment were computed. We analyze the role played by wave-function confinement in determining spectral features and investigate the potentially degrading effects of Auger recombination on device performance. The results of this study advance the characterization of these systems, indicating links between their microscopic properties and optical spectra.

The electronic structures of (InAs)_m/(GaSb)_n short- period superlattices are investigated by the empirical pseudo potential method and multiband kp envelope function approximation with the same underlying bulk band structures. The calculated result are compared to each other and with experiment. Generally, the results of the empirical pseudo potential method are in better agreement with the experimental measurement. The superlattice active region of an optically-pumped type-II laser are investigated by the empirical pseudo potential method. The calculated result predict that its lasing wavelength is about 3.5 micrometers at 80K and internal losses due to intervalence absorption have been suppressed.

Remarkably few applications of boundary element techniques to the solution of Schroedinger's equation have been reported. However, use of boundary elements can reduce the dimensionality by one, and the increased computational efficiencies enable one to compute eigenstates and eigenvalues of 3D quantum dots on desktop PCs. In this work, we introduce the boundary element technique and describe the single band quantum mechanical properties of various quantum dot and quantum wire configurations. The observed behavior of coupled quantum structures results in the equivalent of molecular bonding and antibonding states. Extensions of the method are developed with a numerical perturbation technique for spatially varying potentials, such as the influence of an electric field on quantum wires. Excellent agreement with an exact solution for a quantum wire is reported.

The properties of quantum dot intersubband light emitters and the unique carrier dynamics in the quantum dots that lead to the realization of these long wavelength devices are described. The favorable relaxation times can be exploited to realize far IR emission and detection based on intersubband transitions in the dots.

The optical cross-section is calculated for intra-band transitions in the semiconductor quantum dot (QD). Structures with QDS are promising for IR photodetectors and also for unipolar semiconductor lasers. Atom-like dot is considered in the harmonic oscillator approximation. The cross-section is derived for ground state-to-first-excited states transitions in the peak of the spectral band. It appears to be not sensitive to the wavelength of the transition, but is influenced by the bandwidth. The latter is determined by the scattering of dots in size and in composition. Whenthe bandwidth is 10 meV, the expected absorption peak is of order of > 10³ cm^-1 at the bulk density of dots of 10¹⁷ cm^-3. A comparison is given with available data on the unipolar absorption in semiconductor QD structures.

We validate a microscopic laser model based on the quantum kinetic equations using experimentally determined threshold current, gain, spontaneous emission and quasi-Fermi level separation data taken on GaInP/AlGaInP lasers. By comparison of further experimental and calculated optical properties we find that there is a significant contribution to the threshold current from non-radiative recombination within the quantum wells.

High indium content InGaAs/GaAs laser structures have been grown on (111)B GaAs substrates by Molecular Beam Epitaxy (MBE). The laser devices showed room temperature CW emission, low threshold current densities and emission wavelength up to 1100 nm. The influence of the internal piezoelectric field on the emission properties is studied theoretical and experimentally. A self consistent model was developed in order to simulate the gain and spontaneous emission spectra. Modeling results were compared with measurements of the spontaneous emission spectra, and a good qualitative agreement was obtained. By analyzing these results, we conclude that the piezoelectric field is not completely screened out even for high injection currents, and that the screening level is strongly dependent on the In content. Spontaneous emission measurements in two different configurations (top and edge emission) were compared, yielding similar results for a range of experimental conditions.

We present a study of barrier height effects on the high-temperature performance of 1.3 micron strained layer InGaAsP/InP quantum well lasers. Broad-area Fabry-Perot lasers were fabricated and their light-current characteristics were measured at temperatures between 20 degrees C and 80 degrees C. Based on our experimental results we analyze the effect of the barrier bandgap using the commercial laser simulation software LASTIP. The simulator calculates all relevant physical mechanisms, including their dependence on temperature and local carrier density, self-consistently. The strained quantum-well optical gain computation is based on the 4 x 4 kp method considering valence-band mixing effects. A drift-diffusion model including thermionic emission at hetero-interfaces is used for the calculation of the carrier transport. Careful adjustments of material parameters, in agreement with data reported in the literature, are performed in order to reproduce the measurements. Lowering the barrier height in the active region leads to an improved performance of our laser with respect to threshold current and slope efficiency. An optimum barrier bandgap range of 1.21 - 1.24 eV is identified for our laser. This is partially attributed to the non-uniform carrier-distribution across the quantum-wells.

Experimental measurements of threshold current density as a function of temperature have been analyzed in terms of the characteristic temperature, T₀, and temperature gradient (Delta) _TJ_th equals (delta) J_th/(delta) T, for a number of semiconductor laser device structures. These include AlInGaAs/InP, InGaAsP/InP, and AlGaAs/GaAs. A theoretical model is used to investigate the possible loss mechanisms in laser diodes that cause the superlinear increase of threshold current with temperature. The characteristic temperature T0 is found to vary with temperature and device length, thus making it somewhat misleading when quoted without qualification. A different approach based on plotting ln((Delta) _TJ_th) vs. ln(J_th) shows a linear relationship that is dependent on device structure only, allowing the use of a new figure of merit for the temperature performance of semiconductor lasers.

InGaAs/AlGaAs large optical cavity (LOC) single quantum well (SQW) lasers emitting at 980nm were grown incorporating an AlGaAs/GaAs short-period superlattice layer next to the quantum well in order to improve the carrier confinement and thus high-temperature operation. Symmetric and asymmetric structures have been realized. High characteristic temperatures T0 above 300K between 20#C and 50#C operating temperature were measured for the symmetric structures without deterioration of the internal quantum efficiencies (> 90%) and low intrinsic losses (about 1cm-1). The improvement in the characteristic temperature is mainly attributed to the reduced thermionic emission of the carriers out of the quantum well due to the large effective barrier height of the short-period superlattice. Caused by the incorporation of the short-period superlattice the devices showed a higher series resistance, which could be lowered by switching to asymmetric structures. These asymmetric devices had unchanged high internal quantum efficiencies and low intrinsic losses but lower characteristic temperatures of about 250K.

The emission dynamics of an optically pumped 1.3 +m (GaIn)(NAs)/GaAs vertical-cavity surface-emitting laser is investigated. We achieve room-temperature operation at 1285 nm with a low optical pumping threshold and fast emission dynamics: A minimum peak delay of 15.5 ps and a minimum pulse width of 10.5 ps are observed after excitation with 100 fs pulses. Laser operation with picosecond emission dynamics is demonstrated over a wide temperature range from 30 K to 388 K. We explain this extraordinarily large temperature operation range on the basis of measurements of the optical gain in (GaIn)(NAs)/GaAs. We find a gain broadening at elevated carrier densities due to contributions of higher subband transitions.

Carrier diffusion and thermal conduction play a fundamental role in the operation of high-power, broad-area semiconductor lasers. Restricted geometry, high pumping level and dynamic instability lead to inhomogeneous spatial distribution of plasma density, temperature, as well as light field, due to strong light-matter interaction. Thus, modeling and simulation of such optoelectronic devices rely on detailed descriptions of carrier dynamics and energy transport in the system. A self-consistent description of lasing and heating in large-aperture, spatially-inhomogeneous edge- or surface-emitting lasers (VCSELs) require coupled diffusion equations for carrier density and temperature. In this paper, we derive such equations from the Boltzmann transport equation for the carrier distributions. The derived self- and mutual-diffusion coefficients are in general nonlinear functions of carrier density and temperature including many-body interactions. We study the effects of many-body interactions on these coefficients, as well as the nonlinearity of these coefficients for large-area VCSELs. The effects of mutual diffusions on carrier and temperature distributions in gain-guided VCSELs will be also presented.

Current driven switching between two orthogonal linear polarizations in the fundamental mode of VCSELs is the object of intensive experimental and theoretical work. We developed a model based on the experimental evidence that the gain/loss difference between the two modes is an important factor in polarization switching. We analytically and numerically study a nearly-degenerate two-mode intensity rate equation model. The gain coefficients we use are current dependent and gain saturation with increasing optical power is also taken into account. Two types of stationary solutions emerge: pure-mode solutions (one mode lases) and mixed mode solutions (both modes lase). Stability analysis shows that when the gains equalize, switching between the two pure-mode solutions occurs. The nonlinear gain induces a region of bistability around the switching point. Taking into account the different time scales present in the model and using asymptotic techniques, we can further reduce the model to a single dynamical equation, which can be solved analytically. Stochastic effects (e.g. due to spontaneous emission) can be incorporated in this 1D Langevin-type equation. This allows us to explain the mode hopping in terms of a First Passage Time over the potential barrier in a double potential well.

Numerical simulations of the spatial dynamics of the light output of 2D arrays of VCSELs are presented. The cases presented include square and circular arrays of nine elements. For both configurations the spacing between elements is varied to study the effects on the interaction between elements. In addition, the effects of index guiding on the supermodes of the arrays will be shown. It was found that, with only a small amount f index guiding, the interactions between elements of the VCSEL array are effectively eliminated for all spacings between elements. The time evolutions of the spatial profiles of the laser intensity and carrier density are obtained by solving the Effective Semiconductor Bloch-Maxwell equations by a finite- different algorithm. The algorithm can handle devices with multiple active regions of nay shapes or pattern. There is no a priori assumption about the type or number of modes.

In this paper we present numerical and experimental investigations on the synchronization of the instabilities originated by the mutual coupling of two semiconductor lasers in face to face configuration. We have restricted ourselves to the analysis of two lasers with identical parameters and operating at the same frequency. Numerical simulations are based on standard rate equations for each semiconductor laser whereas the mutual injection is modeled by including delayed optical fields. Experiments are performed using almost identical Fabry Perot lasers coupled through the TE component. As soon as the coupling strength is increased we observe fluctuations in the power dynamics that appears synchronized except for a small time lag. This asymmetric operation of the perfectly symmetric system allows to differentiate between leader and laggard lasers. Synchronization properties are studied making use of the synchronization plots and cross-correlation measurements. Extensive investigations of the dependence of the time traces and correlation degree on the coupling strength and current level demonstrate good agreement between numerical and experimental observations.

Laterally coupled laser diodes are devices in which several lasing stripes are laid side to side, allowing light from one emitter to couple into its neighboring ones. Coupling has been found to cause serious disadvantages such as turning the laser arrays into intrinsically unstable devices. In this paper we show that coupling is also responsible for some unique features that can be used to generate short pulses.

Laterally coupled diode lasers emitting at 1.3 um are presented. Devices were fabricated with distances between ridges varying from 2.76 um to 8.32 um. Electronic coupling effects are investigated by individually varying the currents in each ridge while monitoring output power. It is observed that for devices with 8.32 um separation between ridges there is minimal current sharing, whereas for 2.76 um separation there is considerable current sharing. Optical coupling is measured via the far-field, where most devices show out-of-phase locking, although in-phase locking is observed in a minority of cases. Devices therefore show conditions necessary for the observation of high speed dynamics.

We have investigated the spectral characteristics of semiconductor lasers locked to the external light injected from a modulated laser. The numerical model for semiconductor lasers under the external optical injection is based on the Lang's equation and has been extended in order to take into account the simultaneous injection of the multiple sidebands of the current-modulated laser. The numerical simulation results show that the unselected sidebands will affect the optical and RF-spectral characteristics even when the semiconductor laser is locked to the target sidebands. The simulation results are confirmed by the experimental results.

For chaotic optical communications using semiconductor lasers, synchronization of the chaotic waveforms of two semiconductor lasers, one functioning as the transmitter and the other as the receiver, has to be accomplished. Two systems of single-mode semiconductor lasers that exhibit chaotic behavior are considered. A single-mode semiconductor laser with optical injection follows a period-doubling route to chaos, whereas one with delayed optoelectronic feedback follows a quasiperiodic route to chaotic pulsing. The strategy for devising a scheme for synchronizing two semiconductor lasers is to make the coupled transmitter and receiver lasers to be described by two sets of dynamical equations of identical form. The transmitter and receiver are coupled through a signal and are both driven by a force that is a function of the signal. Any message to be communicated can be encoded in the signal. Schemes for synchronization with unidirectional coupling are devised, modeled, and studied for both optical injection and optoelectronic feedback systems. Experimental data on the synchronization of both systems are presented. Methods for encoding and decoding messages are also discussed for both systems.

Single-frequency diode lasers have been frequency stabilized to 200 Hz at 1.5 microns and to 20 Hz at 793 nm with 10-100 ms integration times using narrow spectral holes in the absorption lines of Er3+ and Tm3+ doped cryogenic crystals. The narrow spectral holes are used as frequency references, and this laser performance was obtained without requiring vibrational isolation of either the laser or frequency reference. Kilohertz frequency stability for 100 s integration times is provided by these techniques, and that performance should be improved to the Hertz level and should be extended to longer integration times with further development. Miniaturized external cavity diode lasers and 5 mm-sized reference crystals will provide compact portable packages with a closed cycle cryocooler. The achieved frequency stabilization provides lasers that are ideal for interferometry, high-resolution spectroscopy such as photon echoes, real time optical signal processing based on spectral holography, and other applications requiring ultranarrow-band light sources or coherent detection.

This paper is concerned with the phenomenon of excitability in semiconductor lasers consisting of a DFB section and a passive dispersive reflector (PDR). We assume that the PDR section contains a Bragg grating and/or a passive Fabry- Perot filter guaranteeing a dispersive reflection of the optical field. We investigate a single mode model for PDR lasers and derive conditions under which excitable behavior can be demonstrated. Especially, we show the existence of a threshold, that is, only perturbations above the threshold imply a large excursion from the steady state, and where the response is almost independent of the strength of the perturbation; moreover we establish the existence of a refractory period, i.e., if a second perturbation is applied before the refractory time has passed, then the system does not respond.

HDBR (Hybrid Distributed Bragg Reflector) laser (also called FGL - Fiber Grating Laser), has been recently demonstrated as low-chirp and potentially low cost lasers emitting at predetermined (ITU-T grid) wavelengths, as well as high quality picosecond pulses laser source. WDM transmission of three densely spaced directly modulated HDBR lasers at 2.6 Gbit/s over 117 km of standard fiber and the direct modulation at 10 Gbit/s have been demonstrated. By using appropriate cavity and grating design, we have realized, for DWDM applications, an HDBR laser that can be directly modulated at 10Gbit/s, while for OTDM systems, we have recently demonstrated a Mode-Locked HDBR laser source, for picosecond optical pulse generation at 10 GHz repetition rate.

This paper presents an overview of work undertaken and directed at the utilization of chaotic laser diodes in secure optical communications systems. Particular emphasis will be given to experimental work using external cavity laser diodes.

We demonstrate numerically a secure communication scheme based on the synchronization of two chaotic laser diodes that are respectively subjected to incoherent optical feedback and incoherent optical injection. In this scheme, the optical fields emitted by the two lasers and the fields that are fed back and injected into these two lasers have orthogonal polarizations. Consequently, the external fields do not coherently interact with the lasing fields but only act on the population inversions. Synchronization of both lasers does not require fine tuning of their optical frequencies neither accurate control of the external cavity lengths contrary to the cryptographic systems based on conventional optical feedback. The message encoding/decoding is achieved by chaos shift keying.

Period doubling is an irregular phenomenon exhibited by semiconductor lasers under high frequency modulation. When it occurs, the repetition frequency of the laser power becomes half of that of the modulation signals, which is undesirable to most applications. This paper reports the control of period doubling in modulated semiconductor lasers by external optical injection and demonstrates for the first time that such nonlinear dynamics can be used advantageously in realizing all-optical clock frequency division. We show that period doubling in modulated semiconductor lasers can be either suppressed or enhanced by providing external CW optical injection. The dependence on injection wavelength, and injection power has a been investigated systematically to establish an improved understanding towards the control of period doubling in modulated semiconductors lasers. The fact that period doubling can be enhanced by optical injection is further used to realize all-optical clock frequency division. To demonstrate this, an optical clock signal at 19.6 GHZ was injected into a semiconductor laser. By adjusting the resonance frequency of the laser to around 9.8 GHz through dc bias, strong period doubling was observed, which resulted in a frequency-halved optical clock signal at 9.8 GHz with a remarkably low level of phase noise. Our investigations have also shown that without changing the biasing condition of the laser this low level of phase nose can be maintained within 1-dB range over an input frequency range of 400 MHz, which is a distinct advantage over other techniques.

The polarization dependence of 1550-nm SOAs based on tensile strained bulk InGaAsP is analyzed numerically, focusing on their wavelength and gain dependence. We demonstrate that strained bulk SOAs are applicable for a wide range of carrier density and wavelength. The gain spectra are calculated based on the k.p method, and the carrier-density and wavelength dependence of the gain is evaluated. We demonstrate that the optimization enables us to make SOAs whose gain polarization sensitivity is within 1 dB under a 20-dB gain in a 60-nm bandwidth in real devices.

We present results on the fabrication and characterization of two and one dimensional photonic crystals for optoelectronic device applications. By using high resolution electron beam lithography 2D and 1D photonic crystals structures are defined on GaAlAs/GaAs and InP/InGaAsP waveguide layers. A crucial step of the patterning is the dry etching, in which structures on a (sub) 100 nm scale with aspect ratios (width to height) of ten or more have to be obtained. We have realized straight waveguides, waveguides with sharp bends, waveguides with build - in cavities as well as lasers with 1D and 2 D photonic crystal mirrors. By using a build in quantum dot layer, optical modes in the passive structures can be investigated. In the InGaAsP as well as in the InGaAs material system ridge waveguide lasers with photonic crystal mirrors have been realized. For the InGaAs system 1D Bragg reflectors with reflectivities above 95 percent have been obtained. These mirrors are essential for mircolasers with active resonator length down to 12 micrometers . These are the shortest edge emitting lasers realized to date.

Using the 3-D finite-difference time-domain (FDTD) method, we analyzed various designs of optical microcavities based on a thin semiconductor membrane perforated with a hexagonal lattice of air holes. The microcavities consisted of single or multiple lattice defects formed by increasing or decreasing radii of air holes, thereby producing acceptor or donor-like defect states. The analyzed geometries include single acceptor or donor defects, as well as ring, hexagonal and triangular cavities. Properties of these structures (excluding single donor-type defect) have not been analyzed previously in a slab of finite thickness. Frequencies, quality factors and radiation patterns of localized defect modes were analyzed as a function of parameters of photonic crystal (PC) and defects. Finally, we discuss possible applications of these structures in active or passive optical devices.

A technique to control growth area of self-assembled quantum dots is required to avoid optical absorption in fabricating quantum dot lasers with optical passive devices such as photonic crystals in the same growth plane as the active layer. We propose and demonstrate a scheme (Area-Controlled Growth) for controlling growth area of self-assembled InAs quantum dots using selective metalorganic chemical vapor deposition (MOCVD). Using this growth technique, the amount of material deposited within mask windows is controlled by varying the width of the mask. In the growth of self-assembled quantum dots, the density of quantum dots largely depends on the quantities of group III sources used. Therefore, by optimizing the growth conditions and mask pattern, quantum dots can be formed in only selected areas of a growth plane. However, in the regions where dots are formed there is variation of dot density and size along the mask stripe direction because of the diffusion of species in the vapor phase, which is peculiar to selective MOCVD. We achieve more uniform distributions of dot density and size by improving the mask pattern. This growth technique can be also applied to fabricate integrated devices for optical communication system containing an external modulator such as photonic crystals together with quantum dot lasers lasing at 1.3 micrometers .

Photonic crystals have seen major advances in the past few years in the optical range. The association of in-plane waveguiding and 2D photonic crystals in thin-slab or waveguide structures leads to overall good 3D confinement with easy fabrication. We present experiments on a variety of structures and devices, as well as modeling tools, which show that 2D PCs etched through waveguides supported by substrates are a viable route to high-performance PC-based photonic integrated circuits. In particular, they exhibit low out-of-plane radiation losses. Low-loss waveguides, high finesse micro cavities, and their mutual coupling are demonstrated.

Using a 2D photonic-crystal slab structure, we have demonstrated a strong 2D photonic band gap with the capability of fully controlling light in all three dimensions. Our demonstration confirms the predictions on the possibility of achieving 3D light control using 2D band gaps, with strong index guiding providing control in the third dimension, and raise the prospect of being able to realize novel photonic-crystal devices. Based on such slab structure with triangular lattice of holes, a 60 degree photonic-crystal waveguide bend is fabricated. The intrinsic bending efficiency is measured within the photonic band gap. As high as 90 percent bending efficiency is observed at some frequencies.

Using full three-dimensional finite-difference time-domain (FDTD) simulations of Maxwells' equations we investigate the mode and photonic band structure and far-field pattern of distributed feedback resonators used in organic lasers. The distributed feedback structure under investigation consists of a two layered system. The first layer is a thick substrate that has a one or two dimensionally corrugated (nano-patterned) surface structure. The active material is a thin layer on top of the corrugated structure. The whole structure is assumed being surrounded by vacuum. Our FDTD calculations are carried out by applying mixed uniaxial perfectly matched layers (UPML) and periodic boundary conditions. This new technique allows us to investigate both guided and leaking modes of dielectric periodic systems. The far-field is obtained by a near-field to far-field transformation. The mode pattern is calculated by spatially resolved discrete temporal Fourier transformation of a particular frequency of interest. A similar computation reveals the frequencies that form the photonic band structure. Our computations show the characteristics of the DFB resonator and explain central aspects of the lasing process in these devices, such as the position and width of the band gap. These results are in good qualitative and quantitative agreement with experimental results.

Uni-traveling-carrier photodiode (UTC-PD) is a newly developed high-speed photodiode for generating high output current. The high saturation current is realized by its operation mode in which only electrons are used as active carriers. The bandwidth of a UTC-PD increases with increasing photocurrent level, which results from the fact that electron transport changes from diffusive to drift/diffusive motion due to the self-induced field in the absorption layer. In this report, photoresponse characteristics of InP/InGaAs UTC-PDs, which depend on device parameters and operation conditions, are discussed. Theoretical analysis for calculating 3-dB-down bandwidth as a function of operation current is also presented. The charge-control model for photodiodes, used in this study, allows us to predict small-signal bandwidth by numerical calculation.

InP/In_0.53Ga_0.47As avalanche photodiodes (APDs) have been widely deployed in high-bit-rate, long-haul fiber optic communication systems due to the higher sensitivity, relative to a PIN photodiode, afforded by internal gain of the APD. Owing to their materials and structural limitations it is uncertain whether the performance of InP-based APDs will be adequate for 10 GB/s systems and subsequent higher- speed systems. One of the impediments for the InP-based APDs is the fact that InP has roughly equal electron and hole ionization rates. This result in a symmetric multiplication process with relatively high multiplication noise and the gain-bandwidth product of an APD are primarily determined by the structure of the multiplication region. Recently, it has been reported that submicron scaling of the multiplication region thickness leads to lower multiplication noise and higher gain-bandwidth products. This is due to the nonlocal nature of impact ionization, which can be neglected if the thickness of the multiplication region is much greater than the 'dead length', the distance over which carriers gain sufficient energy to impact ionize. The advantage of thin multiplication regions, i.e., those for which, the dead space accounts for a significant portion of the total thickness, is that the number of ionization chains that result in multiplication greatly in excess of the average gain is reduced, which in turn yields lower noise for a given gain. In this paper we describe materials and structural modifications to the thin multiplication regions that result in even lower excess noise. For gains <EQ 20 APDs with thin Al_xGa_1-xAs multiplication layers have achieved excess noise factors less than twice the shot noise. We have also shown that ultra low noise can be achieved with an Impact-Ionization-Engineered approach that utilizes heterojunctions to incorporate adjacent regions with low and high ionization rates.

The characteristics of a superlattice infrared photodetector, which has a 20-period GaAs/AlGaAs superlattice embedded between two AlGaAs current blocking layers are investigated. We propose a model to explain the conduction of dark current and photocurrent. In particular, the interesting feature of this detector, which can also be explained by our model, is bias redistribution due to the background photocurrent. With the background radiation incident upon the detector, the blocking layers are tilted up for 126meV at zero bias. The electrons are all confined by the tilt-up blocking layers and cannot tunnel through them. As the external bias increases, 70 percent of the voltage is added on the rear blocking layer and attempts to decrease its barrier height. Especially at -0.18V, the barrier is flat and the barrier height is 16.5meV higher than the bottom of the second miniband. In this case, only those photoelectrons with high enough energy can pass the blocking layer. As the external bias increases up to -0.33V, the rear blocking layer is tilted down for 105meV. The blocking layer seems to be transparent to all photoelectrons to tunnel through. In addition, the photo responsivity due to the short wavelength excitation indicates that the photoelectrons do have the inelastic relaxation of the energy along the growth direction.

Conventional models of the time response of avalanche photodiodes (APDs) assume that carriers travel uniformly at their saturated drift velocity, vsat. To test the validity of this drift velocity assumption (DVA) the model was used to compute the distribution of exit times of electrons generated in an avalanche pulse and the results were compared with those of Monte-Carlo (MC) simulations. The comparison demonstrates that, while the DVA is valid for thick (1um) avalanching regions, it does not take account of non-equilibrium effects which occur in thin avalanching regions, nor of the effects of diffusion. As a consequence, the DVA model may increasingly underestimate the speed of APDs as the width of the avalanche region is reduced.

Conventional models of the time response of avalanche photodiodes (APDs) assume that carriers travel uniformly at their saturated drift velocity, vsat. To test the validity of this drift velocity assumption (DVA) the model was used to compute the distribution of exit times of electrons generated in an avalanche pulse and the results were compared with those of Monte-Carlo (MC) simulations. The comparison demonstrates that, while the DVA is valid for thick (1um) avalanching regions, it does not take account of non-equilibrium effects which occur in thin avalanching regions, nor of the effects of diffusion. As a consequence, the DVA model may increasingly underestimate the speed of APDs as the width of the avalanche region is reduced.

The dead-space carrier multiplication theory properly predicts the reduction in the excess noise factor in a number of APDs. The theory is applied to measurements, obtained from J. C. Campbell and collaborators at the University of Texas, for InP, InAlAs, GaAs, and AlGaAs APDs with multiplication-region widths ranging from 80 nm to 1600 nm. A refined model for the ionization coefficients is reported that is independent of the width of the device multiplication region of each device. In addition, in comparison to predictions from the conventional multiplication theory, the dead-space multiplication theory predicts a reduction in the mean bandwidth as well as a reduction in the power spectral density of the impulse response. In particular, it is shown that the avalanching noise at high-frequencies is reduced as a result of the reduction of the multiplication region width.

High speed, high efficiency, low noise and high saturation power are the characteristics desired for detectors in high bit-rate long-haul optical communication systems. We present the modeling of traveling-wave application photodetectors. These novel monolithic devices combine optical gain and absorption in a distributed fashion along a traveling-wave structure, providing high-responsivity and high-speed performance, without sacrificing saturation power. We present the models used to simulate the behavior of these devices, as well as their result. We show that TAP detectors have higher saturation power than other detectors with the same bandwidth-efficiency product, at the price of a small noise penalty, which is also calculated. The result is a net increase in the dynamic range.

Based on the statistical properties of the amplified spontaneous emission, the photo detection noise spectrum at the output of an optical amplifier is studied. General expressions of noise corresponding to the beating of the amplified spontaneous emission with itself and with a deterministic signal are derived. The result revealed that the signal-spontaneous emission beating noise spectrum is a transposition of the spontaneous emission spectrum to the low frequency domain, and the spontaneous emission- spontaneous emission beating noise spectrum is the autocorrelation function of the spontaneous emission spectrum. A filtered non-uniform spontaneous emissions spectrum resulted in a non-white photo detection noise spectrum. These theoretical results are verified experimentally on a erbium-doped fiber amplifier.

We study the formation of self organized light peaks, in GaAs microcavities. By means of analytical and numerical techniques, experimentally accessible parametric domains can be found, where stable and robust CS can be addressed, shifted and brought to interaction ranges, thus realizing some basic schemes for optical information treatment. A Fourier-Newton approach is applied to gain quantitative information on CS's dynamical response to external control fields or on CS pair interaction.

Theoretical studies on cavity transverse nonlinear dynamics have shown the possibility of exploiting the self-organizing properties of light response in the form of periodic patterns and cavity solitons. While very few experimental confirmations exist - only in macroscopic systems - we report on the first experimental results obtained in bulk and multi-quantum-well AlGaAs microresonators. These systems combine the advantage of : a monolithic character deriving from their well-controlled epitaxial growth conditions, a variety of nonlinear optical properties near the band gap edge, and a high Fresnel number. We review the general properties of semiconductor microresonators that lead to optical self-organization, emerging from the interplay between the dispersive or saturable absorptive nonlinearities and transverse mechanisms such as light diffraction and carrier diffusion. We show the first observation of periodic rolls, rhombs and hexagons patterns, stress the strong interaction of these patterns with the transverse fluctuations of the cavity thickness. Finally, we present the observation of precursor forms of cavity solitons, and evidence the strong thermal contribution they involve.

The dead-space carrier multiplication theory properly predicts the reduction in the excess noise factor in a number of APDs. The theory is applied to measurements, obtained from J. C. Campbell and collaborators at the University of Texas, for InP, InAlAs, GaAs, and AlGaAs APDs with multiplication-region widths ranging from 80 nm to 1600 nm. A refined model for the ionization coefficients is reported that is independent of the width of the device multiplication region of each device. In addition, in comparison to predictions from the conventional multiplication theory, the dead-space multiplication theory predicts a reduction in the mean bandwidth as well as a reduction in the power spectral density of the impulse response. In particular, it is shown that the avalanching noise at high-frequencies is reduced as a result of the reduction of the multiplication region width.

We demonstrate that light propagation in strongly-modulated 2D/3D PhCs becomes refraction-like in the vicinity of the photonic bandgap, which is contrary to the fact that light propagation in weakly-modulated photonic crystals is very different from refraction and thus the definition of refraction index becomes meaningless. Such a crystal behaves like a material having an effective refractive index controllable by the band structure. This situation is analogous to the effective-mass approximation in electron-band theory. The propagation states having a negative effective index exhibit unusual properties, such as mirror-like imaging effect, image-transfer effect. These properties are confirmed by finite-difference time domain simulations.

One of the key topics in today's semiconductor laser development activities is to increase the brightness of high-power diode lasers. Although structures showing an increased brightness have been developed specific draw-backs of these structures lead to a still strong demand for investigation of alternative concepts. Especially for the investigation of basically novel structures easy-to-use and fast simulation tools are essential to avoid unnecessary, cost and time consuming experiments. A diode laser simulation tool based on finite difference representations of the Helmholtz equation in 'wide-angle' approximation and the carrier diffusion equation has been developed. An optimized numerical algorithm leads to short execution times of a few seconds per resonator round-trip on a standard PC. After each round-trip characteristics like optical output power, beam profile and beam parameters are calculated. A graphical user interface allows online monitoring of the simulation results. The simulation tool is used to investigate a novel high-power, high-brightness diode laser structure, the so-called 'Z-Structure'. In this structure an increased brightness is achieved by reducing the divergency angle of the beam by angular filtering: The round trip path of the beam is two times folded using internal total reflection at surfaces defined by a small index step in the semiconductor material, forming a stretched 'Z'. The sharp decrease of the reflectivity for angles of incidence above the angle of total reflection leads to a narrowing of the angular spectrum of the beam. The simulations of the 'Z-Structure' indicate an increase of the beam quality by a factor of five to ten compared to standard broad-area lasers.

We demonstrated four-wave mixing (FWM) with two pumps in a wavelength selectable laser to achieve a wide conversion bandwidth. Non-degenerated FWM with dual pumps in SOAs has been proposed to flatten the conversion efficiency variation accompanied with the variation of detunings between input signal and output conjugate. By utilizing a wavelength selectable laser, this complex scheme can be easily realized by a single device. We used a wavelength selectable laser, which consists of an 8-channel DFB laser array connected to an multi-mode interference (MMI) combiner with an SOA at its output. The DFBs had different lasing wavelengths with 3.18 nm spacing. We introduced a signal wave through one of the DFB lasers, and two of the lasers were used as pump sources. The wavelength of the signal and one of the pumps were fixed. The wavelength of the output conjugate was changed by switching to a different second pump laser. Conversion efficiencies between -14 and -16 dB were observed with wavelength detunings between 8 and 27 nm. An efficiency variation as small as 2 dB over a 2 THz frequency detuning range was achieved, in contrast to more than 5 dB variation in single pump schemes.

A finite difference beam prop0agation mode is used in conjunction with laser rate equation simulations to study the inclusion of an intracavity lens in a high-power tapered ridge waveguide laser diode emitting at 980 nm. A parabolic lens is introduced in the top of the ridge near the front facet via a change in the waveguide effective refractive index profile. The inclusion of the lens has led to 13 percent reduction in the threshold current and an improved power slope efficiency from 0.4 W/A up to 0.8 W/A. The lens has caused near field broadening of 2 micrometers at full width half maximum power, indicating more efficient use of the cavity. The far field has narrowed by 1 degree indicating higher brightness. The model uses a mesh for reach of its points the standard carrier rate equation is solved across the active layer. The 2D wave equation is solved for the two counter propagating fields using a finite difference algorithm. The result of the does how good agreement with experiment.

We fabricated optical waveguides with a relative refractive index difference of maximally 45 percent on a silicon-on-insulator (SOI) substrate. We designed the Si channel using the finite difference time domain simulation and fabricated using electron beam lithography and inductively coupled plasma etching. The single or quasi-single mode propagation was observed for channel width of 0.3 - 1.0 microns and thickness of 0.32 microns at a wavelength of 1.55 microns. Propagation loss evaluated using the Fabry-Perot resonance method was of 10 inverse cm order for these channel widths. The large effective index of the guided mode over 4.5 was also observed, which well agreed with the simulation. When channel width was 0.5 microns, the bend loss of 0 - 1 dB was roughly evaluated even though the bend radius was 0.5 - 9.5 microns. The similar low loss was also confirmed for half circular disk waveguides butt-joined to use as 90-degree-bends. These results suggest the potential of an ultra-high density optical wiring in lightwave circuits.

High-speed optoelectronic devices must be interconnected with optical waveguides. Furthermore, it is quite likely that optical waveguides will be required to interface with superconducting devices that operate at cryogenic temperatures. Thus, it is important to examine femtosecond pulse propagation in cryogenic waveguides. Our approach uses the time domain susceptibility by Fourier transforming frequency domain models of semiconductors developed by Adachi. The resulting time domain susceptibility functions are closed form and cover the temperature range from 0 to 800 K. With this new model, time domain solutions to Maxwell's equations are now obtained for a femtosecond pulse propagating in a GaAs waveguide. The effects of linear dispersion and attenuation are examined for waveguides at different temperatures. However, the solution is limited to one spatial dimension in the direction of propagation to simplify the numerical algorithm.

We have calculated the refractive index of a Al_xGa_1-xN/GaN square quantum well (QW). The imaginary part of the dielectric function has been obtained by summing up the contributions of the dominant interband transitions, excitonic contributions, and the continuum contribution, obtained by weighting the well's and the barrier's continuum contributions. In the calculation of the contribution of the conduction-valence band bound-state effect without electron- hole interaction, conduction bands are assumed to be parabolic and valence bands have been calculated using Chuang's model, but with Chan's basis expansion method instead of finite-difference scheme. Excitonic contribution has been described with an expression derived by the density-matrix approach at the subband edge without the influence of band mixing. The continuum contributions have been described with the modified Adachi's model. The effects of the aluminum model fraction x and the width of the well on the refractive index are analyzed and discussed.

We report on experimental characterization of GaInNAs and GaInAs semiconductor materials grown on GaAs by Chemical Beam Epitaxy. The optical characterization of the samples has been carried out by orthodox photoluminescence and ellipsometry measurements. We show that even a small amount of nitrogen added to GaInAs has a considerable effect on the optical properties, causing a red shift in the emission wavelength, a reduction in PL efficiency and an increase in the refractive index.

We present the results of our studies concerning the pulsed operation of a bulk GaInAsP/InP vertical cavity surface emitting laser (VCSEL). The device is tailored to emit at around 1.5 micrometers at room temperature. The structure has a 45-period n-doped GaInAsP/InP bottom Distributed Bragg Reflector (DBR), and a 4 period Si/Al₂O₃ dielectric top reflector defining a 3-(lambda) cavity. Electroluminescence from a 16micrometers diameter top window was measured in the pulsed injection mode. Spectral measurements were recorded in the temperature range between 125K and 240K. Lasing threshold current density has a broad minimum at temperatures between 170K-190K. In order to establish the relationship between the temperature dependence of the threshold current and the gain peak, we investigated the spectral dependence of edge and surface emission from optically pumped structures without the top reflector. Experimental result are compared with existing theories concerning the temperature dependence of threshold current.

Self-consistent numerical calculation and photoluminescence (PL) measurements have been used to investigate the temperature dependence of the optical Stark effect in n-doped GaAs/AlGaAs single asymmetry quantum wells (SAQWs), grown by molecular beam epitaxy. In the low-temperature regime (5 to 40 K) a remarkable blue shift (9.8 meV) is observed in the PL peak energy, as the optical excitation intensity increases from 0.03 to 90 W/cm2. The blue shift is well explained by the reduction of the two-dimensional electron gas (2DEG) density, due to the charge-transfer mechanism. At about 80 K, however, an anomalous behavior of the PL peak energy was found, i.e. a red shift has been observed as the optical excitation intensity increases. This anomalous behavior has been explained by combining the effects of band gap renormalization, band bending, temperature dependence of the band gap, temperature dependence of the 2DEG density, and temperature dependence of the fundamental energy position.

Electron-determined nonuniform carrier distribution inside multiple quantum wells (MQW) is experimentally discovered. Two groups of mirror-imaged nonidentical quantum well InGaAsP/InP lasers diodes are designed, fabricated, and measured. Measured characteristics of both groups show that electron, instead of hole, is the dominant carrier affecting carrier distribution. Carrier transport effects including carrier diffusion/drift and capture/emission processes inside MQW are described to explain the nonuniform carrier distribution. The reason for the electron dominated carrier distribution is because electron takes less time to be capture into QW 2D states than hole does. The sequence of the nonidentical QWS is also shown to have significant influence on device characteristics.

A new approach for operation of tunable lasers based on quantum well wires is proposed. The laser operation uses the effect of an intense, long-wavelength laser field radiation applied to the semiconductor device. Different geometry's concerning the shape and size of the semiconductor quantum structure as well as the orientation of the applied laser field with respect to the quantum device was considered. We calculate the laser-dressed quantum well wire (QWW) potential energy, in the frame of the nonperturbative theory and finite difference method. Then, we show that when the intense, long-wavelength laser field radiation, is applied to the semiconductor-based quantum well wire device a significant optical Stark effect is observed for the bound state energy. Furthermore, under the action of the laser-dressed potential, a strong enhancement of the blue shift occurs for the electron-heavy hole recombination processes as the QWW lateral size is reduced. This effect may provide full control of the frequency operation of QWW-based lasers and thus would be of great help in tailoring the physical parameters of the semiconductor QWW device.

Theoretical and experimental results concerning the study of a novel wavelength converter amplifier, which can be tuned, with the amplification of an external voltage are presented. The device consists of a Ga_1-xAl_xAs graded quantum well, placed on the n-side of the depletion region of a Ga_1-xAl_xAs p-n junction. As a result of the competition between the built-in field and the grading, in the absence of an external bias, the quantum well acts as an isolated well. Forward biasing of the junction reduces the built-in field; thus the field associated with the grading becomes effective. The tuning of the operation wavelength is based on the anti-Quantum Confined Stark Effect and achieved during the forward biasing. In this study we present the numerical results based on a 2D modeling of the device where exciton binding energy, absorption co-efficient and transition energy are obtained as a function of applied field. Experimental results show a tuning range of around 40nm.

A novel single-moded self-sustained pulsation mechanism suitable for single-section long-wavelength semiconductor lasers with wavelength constraints is proposed. It is based on the principles of gain-switching and the dispersive property of the active layer in laser diodes (LDs). Numerical simulations using a carrier heating model illustrate the mechanism. Results indicate that if a LD is engineered such that its lasing wavelength, over a range of bias currents, experiences an optical gain that is slightly above threshold, self-gain-switching occurs giving rise to self-pulsations. The self-gain-switching is brought about by wavelength shifting due to index-change induced by carrier density fluctuations and current heating. Depending on the magnitude of the bias current the self-pulsating waveform can be sinusoidal or pulse-like.

A numerical analysis is carried out to investigate the influence of injecting either unmodulated or modulated light from an external laser diode (LD) into a 1.55 micrometers InGaAsP distributed feedback (DFB) self-pulsating (SP) LD. Injection of weak unmodulated light causes the SP-LD to behave mostly chaotically apart from some regions of self-sustained-pulsations (SSP) and multiple-peak periodic oscillations. Increasing the injected optical power has the effects of changing the behavior of the optically-injected SP-LD from chaos to multiple-peak periodic oscillations followed by high-frequency (HF) self-oscillations before reaching an equilibrium state (i.e. a steady level). The SP-LD is found to exhibit quasi-periodic behavior when injected with lightly modulated light. At relatively higher modulation index, the output of the optically-injected SP-LD becomes chaotic.

Evaluation of the influence of laser noise on the precision of single-mode rate equations in modeling laser diode (LD) behavior is performed. The inclusion of Langevin noise sources in simulations is found to increase the relaxation damping significantly. This phenomenon explains the reported truncations of higher-order bifurcations leading to chaos and the enhancement of period-doubling when laser noise is taken into account in simulations. Numerical analysis of the nonlinear dynamics of directly-modulated LD is performed, and results show that laser noise has been mistaken as an important factor which enables an agreement between calculated and measured results to be achieved. Instead, simulations carried out to investigate the effect of current-dependent gain suppression on the nonlinear behavior of LD reveal that temperature plays an important role in modeling LD behavior. Hence, accurate electro-thermal modeling is critical in reproducing measured behavior of LD in simulations.

This paper presents a work on spatiotemporal behavior and study of design parameters for two laterally coupled semiconductor diode lasers. The structures to be studied are Two Laterally Coupled semiconductor diode Lasers (TLCL) to 1,3um and they are expected to increase the modulation bandwidth to two or three times the modulation bandwidth for a single semiconductor laser, by using the principles of coupling between emitters. We show a study of the modeling used to find the spatiotemporal behavior of these devices when we change some design parameters, which are very important if we want to obtain a high-speed modulation control by current injection, because these devices can present unstable dynamics depending on coupling level. Temporal evolution of the Near-field, Far-field, photon and carrier densities and the phase between the fields inside the two lasers are used to study the dynamic behavior of the TLCL when we change parameters as the distance between emitters, carrier diffusion, current spreading and the level of injected current.

Hot electron light emitting and lasing semiconductor heterostructure (HELLISH) is a novel longitudinal transport, surface emitting device. The operation of the device as a light emitter and vertical cavity surface emitting laser (VCSEL) has been previously demonstrated by us for the GaAs/GaAlAs material system. A basic GaInAsP/InP HELLISH structure and an advanced GaInAsP/InP HELLISH VCSEL structure are described in this work and designed for 1.5 micrometers emission. The basic HELLISH structure consists of a GaInAsP quantum well placed on the n-side of the InP p-n junction. The advanced structure has a similar active region to the basic HELLISH structure, but with a resonant cavity defined by the addition of DBRs, where the current is injected directly into the active region without having to pass through the DBRs. The modeling and design of these structures are described, including self-consisting numerical 1D solutions of the Poisson and Schrodinger equations.

Hot Electron Light Emitting and Lasing Semiconductor Heterojunction device is a novel emitter that utilizes hot carrier transport parallel to the layers of Al_xGa_1-xAs p-n junction containing GaAs quantum well(s) in the depletion region. Electrons and holes drifting in their respective channels are heated up temperatures well above the lattice temperature and consequently transferred in real space over the built-in potential barrier into the GaAs quantum well via phonon assisted tunneling and or thermionic emission. The recombination occurs in the quantum well. The Top Hat structure HELLISH presented in this work provides a new functionality of the device where the n and p-layers are contacted separately but are biased with the same voltage longitudinally. In this configuration hot carrier injection into the active region is further enhanced in the vicinity of the cathode due to the effective forward biasing of the junction. Therefore, the emission intensity is increased compared with the conventional HELLISH device. In the vicinity of the anode, however, there is an effective reverse biasing and in this region the top hat device acts as an absorber. As a result of these two features the device can be operate as a wavelength converter and amplifier. The intensity of the emitted light is independent of the polarity of the applied voltage. However, positions of the absorber and emitter depend on the polarity. The device may offer a wide range of light logic functions. The speed and the efficiency of the device depend on both the longitudinal and the transverse fields. A simple 2D model is developed to explain the device dynamics.

Narrowband optical filters and all-optical Add-Drop Multiplexers (ADMs) are key components in dense-WDM systems. Implementation of these devices in semiconductor waveguide technology allows for integration with passive and active photonic devices for compact, high-functionality modules. This paper describes an approach based on quarter-wave shifted DBR filters. Add-Drop Multiplexers are implemented by integrating of filters with directional couplers, and the devices are analyzed using a combination of transfer matrix methods and commercial beam propagation software. Through analytical and numerical methods we demonstrate the possibility of implementing filters and ADMs with an optical bandwidth less than 25GHz. Channel cross-talk isolation better than -20dB can be achieved, with waveguide propagation loss being the critical issue. Using filter design from the theory of electron filters translated to the optical domain, higher-order optical filters with Chebyshev and Butterworth responses are designed. Reconfigurability is important for network versatility and robustness. Semiconductor waveguide devices can be tuned by injection of carriers. Based on the mode, the tuning range is evaluated and found to span several channel spacings, limited by the added waveguide propagation loss form free-carrier absorption. The design of a 25 GHz filter with 300 GHz tuning range is shown.