This PDF file contains the front matter associated with SPIE Proceedings Volume 6889, including the Title Page, Copyright information, Table of Contents, and the Conference Committee listing

Strain-compensated InGaN quantum well (QW) active region employing tensile AlGaN barriers was analyzed as improved gain media for laser diodes emitting at 430-440 nm by taking into account the carrier screening effect. The use of strain-compensated InGaN-AlGaN structure is advantageous because the tensile barriers compensate the compressive strain in the InGaN QW and the larger band offset allows high temperature operation. The band structure is calculated by using a 6-band k times p formalism, taking into account valence band mixing, strain effect, spontaneous and piezoelectric polarizations. The carrier screening effect is incorporated in the calculation, which is solved self-consistently. The spontaneous emission spectra show a significant improvement of the radiative emission for strain-compensated structure. The optical gain analysis shows enhancement in the peak optical gain for the strain-compensated QW structures. Threshold analysis of both the conventional InGaN-GaN QW and strain-compensated InGaN-AlGaN QW active regions indicate a reduction in the threshold carrier density and threshold current density for diode laser employing the strain-compensated QW as its active region.

InGaN epitaxial film growth has been performed on (0001) c-, (1120) a- and ₍₁₁₀₀₎ (see manuscript for full text) m-plane ZnO substrates by metalorganic vapor phase epitaxy in the temperature range of 550°C - 680°C. The grown layers were confirmed to be single crystalline by X-ray diffraction.

Most III-V nitride light emitting diodes have an n-down structure with Ga polarity. In such a device, the active layer is grown on top of the n-cladding layer and the p-type cladding layer is grown on top of the active layer. We have analyzed the band structure of such a device and found a reduced effective conduction band barrier due to the positive spontaneous and piezoelectric polarization charge, resulting in large electron overshoot and necessitating the introduction of the commonly employed electron blocking layer. On the other hand, the polarization charge at the corresponding interface for a p-side down device with Ga polarity is negative, resulting in a significant enhancement of the electron barrier and the existence of a 2D hole gas near the interface. These are beneficial to the performance of single heterojunction LEDs.

While single-junction solar cells may be capable of attaining AM1.5 efficiencies of up to 29%, multi-junction (MJ, Tandem) III-V compound solar cells appear capable of realistic efficiencies of up to 50% and are promising for space and terrestrial applications. In fact, the InGaP/GaAs/Ge triple-junction solar cells have been widely used for space since 1997. In addition, industrialization of concentrator solar cell modules using III-V compound MJ solar cells have been announced by some companies. This paper presents principles and key issues for realizing high-efficiency MJ solar cells, issues relating to development and manufacturing, and applications for space and terrestrial uses.

All-optical noninvasive control of a multi-section semiconductor laser by means of time-delayed feedback from an external Fabry-Perot cavity is realized experimentally. The role of the optical phase as a specific new control parameter of this type of delayed-feedback control is stressed. Using phase-dependent feedback from a resonant plane Fabry-Perot interferometer, the stabilization of unstable steady states and unstable self-pulsations is achieved, including experimental demonstration of all-optical chaos control. In the latter experiment, optical chaos is transformed into a regular 12.9 GHz self-pulsation. This result is the fastest realization of chaos control ever reported. The control is noninvasive, only less than one per mille feedback keeps the stabilized states stable.

A semiconductor laser subject to delayed optical feedback is investigated in the limit of intense feedback power. Back-injection of light with variably rotated polarization reveals a symmetry breaking in laser emission spectra and output power when the rotation angle is changed in the vicinity of the orthogonal orientation. To explain the observed asymmetry we propose a simple geometric model which includes the relative contributions of both TE and TM lasing modes into the feedback light. In a range of feedback polarization rotation angles the emission spectra of the laser reveal a gap with width of more than a terahertz. The position of the gap and its width are shown to be regulated by means of feedback polarization rotation angle. We demonstrate that a theoretical approach, based on carrier density grating induced potential, explains our experimental results.

The analysis of the non linear dynamics observed in lateral couple diode lasers (LCDL) is a key issue on the study and understanding of these devices. In this work an analysis of the different nonlinear regimes observed in devices with different lateral distances is made. By the observation of both the RIN spectrum and the filtered optical spectrum, a clear identification of the different nonlinear regimes observed in these devices is made, as well as an identification of the main variables responsible for the appearance of these nonlinear regimes.

The optical response of semiconductor quantum wells is investigated theoretically to explain nonlinear transients generated via intense terahertz (THz) fields. A microscopic description of THz-induced interaction processes is developed while several numerical examples are presented to illustrate properties in a typical THz-pump and optical-probe configuration. The results identify signatures of the ac-Stark effect, ponderomotive contributions, and extreme-nonlinear dynamics.

Fully microscopic many-body models are used to ivestigate the temperature dependence of radiative and Auger losses in semiconductor lasers. Classical estimates based on simplified models predict carrier density independent temperature dependencies, 1/T for the radiative losses and a temperature activated exponential dependence for the Auger losses. Instead, the micorscopic models reveal for the example of a typical InGaAsP-based structure a 1/T³-dependence for the radiative losses at low carrier densities. For high densities this dependence becomes much weaker and deviates from a simple power law. Auger losses can be described by an exponential dependence for limited temperature ranges if a density dependent activation energy is used. For the threhsold carrier density a temperature dependence close to T² is found instead of the linear temperature dependence assumed by the simplified models.

We have studied the modulation properties of a vertical cavity surface-emitting laser (VCSEL) coupled to an electrooptical modulator. It is shown that, if the modulator is placed in a resonant cavity, the modulation of the light output power is governed predominantly by electrooptic, or electrorefraction effect rather than by electroabsorption. A novel concept of electrooptically modulated (EOM) VCSEL based on the stopband edge-tunable distributed Bragg reflector (DBR) is proposed which allows overcoming the limitations of the first-generation EOM VCSEL based on resonantly coupled cavities. A new class of electrooptic (EO) media is proposed based on type-II heterostructures, in which the exciton oscillator strength increases from a zero or a small value at zero bias to a large value at an applied bias. A EOM VCSEL based on a stopband-edge tunable DBR including a type-II EO medium is to show the most temperature-robust operation. Modeling of a high-frequency response of a VCSEL light output against large signal modulation of the mirror transmittance has demonstrated the feasibility to reach 40 Gb/s operation at low bit error rate. EOM VCSEL showing 60 GHz electrical and ~35 GHz optical (limited by the photodetector response) bandwidths is realized.

The coupling of electrothermal and optical models in multidimensional semiconductor laser simulation is a non-trivial task: While electrothermal models are conveniently formulated in a system of partial differential equations, the optical problem necessitates the solution of Maxwell's equations with results in eigenvalue problems with the eigenvalues representing lasing wavelengths and losses of the laser's modes. The state-of-the-art approach to achieve a self-consistent electro-optothermal solution os a Gummel-type iteration where the electrothermal equations (using a Newton scheme), and the optical eigenvalue problems are solved iteratively until a convergence criterion is reached. This is extremely time-consuming and unstable, especially for simulation of devices featuring closely spaced multiple modes (for example broad area lasers, VCSELs). In this work, we present a novel numerical electro-optothermal coupling scheme for semiconductor lasers which is based on the integration of the optical eigenvalue into a single global Newton formulation, The necessary derivatives are obtained by perturbation theory. The proposed scheme is more robust and decreases the computational burden (simulation time) by more than an order of magnitude compared to Gummel-type interations. This novel coupling scheme allows to investigate the influence of the electro-optothermal coupling on the device characteristics and internal physics. The effects of bias conditions on the modal dynamics, optical wavelengths, losses, and far-field patterns are analyzed. For a VCSEL, we quantify the role of gain guiding and thermal lensing.

Solid-state lasers that can generate optical pulses in the picosecond and femtosecond domains have progressed rapidly over the past decade from laboratory systems to an impressive range of commercial systems. Novel materials, notably quantum-dot semiconductor structures, have enhanced the characteristics of such lasers and opened up new possibilities in ultrafast science and technology. In our most recent work we have shown that quantum-dot devices can be designed to provide efficient means of generating and amplifying ultrashort optical pulses at high repetition rate rates.

We report on the recent advances in InP-based Quantum Dashes (Qdashes) material for 1.55μm optoelectronic devices. We achieve highly uniform, reproducible and wavelength-controlled Qdashes, with a length ranging from 50nm to 500nm depending on the growth conditions. These Qdashes lead either to high modal gain distributed feedback (DFB) lasers or low chirp semiconductor optical amplifier (SOA). Moreover, we demonstrate that Qdashes are compatible with buried ridge stripe and shallow ridge technology and lead to very reliable lasers. Directly modulated lasers with 10GHz bandwidth are demonstrated in continuous wave mode operation. 10Gb/s transmission over 25km in semi-cooled operation is achieved using DFB buried lasers. Qdashes optimization leads to SOA with internal gain of 10 dB and a -3dB optical bandwidth of 120 nm at 50°C, paving the way for semi-cooled CWDM optical sources. Furthermore, low chirp Qdashes SOA are evaluated as optical boosters after a modulated source. Although we still observe overshoots on the amplified signal, the chirp, even in their saturation regime, is low enough to allow for 50 km of transmission at 10Gb/s.

The linewidth enhancement factor (LEF) and nonlinear gain coefficient of an InAs/AlGaInAs quantum dot (QD) laser are measured using an injection locking technique. The nonlinear gain coefficient was found by curve-fitting the measured LEF as a linear function of the output power. The LEF of the InAs/AlGaInAs quantum dot laser was measured to be 1.2 to 8.6 at output powers from 2 to 10.2 mW, leading to a corresponding nonlinear gain coefficient of 1.4 x 10^-14 cm³. This value for the nonlinear gain coefficient is three orders of magnitude higher than the typical quantum well nonlinear gain coefficient of 10^-17 cm³. Consequently we expect that the dynamics under optical injection and external feedback of this type of quantum dot laser will be dramatically different than in quantum well lasers, suggesting that a careful re-examination of the dynamics of this type of laser is needed.

We report on observation of spontaneous emissions from a single quantum dot after resonant excitation. To capture weak emissions from a single quantum dot, reflection of an excitation laser was reduced by applying obliquely incident geometry and crossed polarization configuration. In addition, collected emissions were temporally resolved to be separated from residual reflection. These allowed us to realize a simple manipulation and read out of an exciton qubit. We observed an exciton Rabi oscillation in excitation amplitude dependence of the emission intensity. We determined an exciton dipole moment from the Rabi oscillation and from the emission decay time, independently. Comparison between the two values shows reasonable agreement.

Theoretically predicted "dipole lasing" as spontaneous excitation of coherent oscillations of dipole momentum of metal nano-particles placed inside or near the surface of the medium with optical amplification. It has close analogy with ordinary lasing, but the polarization of nano-particles stands for the optical cavity mode. Oscillations of polarization cause coherent radiation from nano-particles acting as "nano-antennas". Optical cavity is not necessary, so that the minimum size of the dipole laser can be on the nano-scale. Dipole laser frequency corresponds to the localized plasmon resonance of nano-particles. The manifestation of the dipole lasing is in the divergence of the nano-particle polarisability, that is a second-order phase transition. Threshold conditions, enhancement of the spontaneous emission and optical bistability is dipole lasers are found.

Direct laser diodes can typically provide only a limited single mode power, while ultrahigh-brightness is required for many of the market-relevant applications. Thus, multistage power conversion schemes are applied, when the laser diodes are used just as a pumping source. In this paper we review the recent advances in ultra-large output aperture edge-emitting lasers based on the photonic band crystal (PBC) concept. The concept allows near- and far-field engineering robust to temperature and strain gradients and growth nonuniformities. High-order modes are selectively filtered and the effective optical confinement of the fundamental mode can be dramatically enhanced. At first, we show that robust ultra-narrow vertical beam divergence (<5 deg. FWHM) can be achieved simultaneously with ultrahigh differential efficiency (80-85%) and significant single mode power for several wavelengths of the key regions. A maximum single mode power of 1.4 W is achieved for 980 nm lasers. At second we extend the PBC concept towards the 2D photonic crystal. A significant field extension in the vertical direction allows a robust fabrication of the field-coupled lateral multistripe PBC arrays with a total multistripe width of 0.2 mm. We also demonstrate that the concept of high-order modes filtering works well also in the lateral direction. Finally, we address possible options for 3D managing of light towards wavelength stabilized laser operation by processing of the multistripe arrays along their lengths. The concept opens a way for 3D photonic crystal edge emitting lasers potentially allowing scalable single mode power increase to arbitrary high levels.

We have recently demonstrated an ultrafast photonic crystal laser and cavity coupled laser array with modulation rates of 1THz at room temperature, a 20 GHz optical modulator with activation energies of 60 fJ and a quantum dot photonic crystal laser with large signal modulation rates of 30GHz. These devices are enabled by the enhanced light-matter interaction in photonic crystals, and serve as the building blocks of on-optical information processing circuits.

Photonic crystal membrane microcavities (PCMC) exhibit modes with highest quality factors and ultrasmall volume at the same time. This makes them the ideal solid state implementation for studying cavity quantum electrodynamics, as a quantum emitter such as a quantum dot can be placed at an electric field maximum with only moderate technological effort. Ultimately, this shall lead to novel classes of light emitters, such as highe efficiency LEDs or devices for quantum information processing. This paper discusses PCMC's operating in the weak coupling regime, shows an efficient and realistic simulation method based on the finite element method, and the design trade-offs for cavities used as light emitters. Finally, a comparison to measured spectra illustrates technological aspects.

We present a theoretical thermal analysis of mid-infrared quantum-cascade lasers (QCLs) using a two-dimensional anisotropic heat diffusion model. Several InP-based devices are simulated over a range of operating conditions in order to extract temperature-dependent thermal resistances. These thermal resistances are used to compare the effectiveness of various heat management techniques. Finally, heat flow analysis is performed in order to understand the internal thermal dynamics of these devices.

A design strategy for short wavelength (equation) quantum-cascade lasers (QCLs) on InP substrates is discussed and the performance of these lasers evaluated. The QCLs are based on strain-compensated AlAs-In_0.73Ga_0.27As heterostructures grown using gas-source molecular-beam epitaxy on InP substrates. Both composite barriers based on AlAs-In_0.55Al_0.45As and composite wells based on In_0.73Ga_0.27As-In_0.55Al_0.45As are used to achieve laser emission at wavelengths as short as 3.05 μm by avoiding leakage from the upper laser state into either the higher-lying miniband or into indirect states within the heterostructure.

We propose and theoretically investigate mid/far-infrared photodetectors based on frequency up-conversion in a near-resonant cascade of interband and intersubband transitions in high optical nonlinearity asymmetric quantum well structures. Such structures can yield high detectivity and responsivity in the bandwidth of the order of 30% of a central frequency in the mid-infrared range. Resonant up-conversion detectors can be designed for both collinear and perpendicular pump and signal beams. They can be integrated with semiconductor pump lasers to yield compact devices. We present specific device designs based on GaAs/AlGaAs and InGaAs/AlInAs heterostructures and calculate their expected figures of merit.

In this paper an overview is given of the results we have obtained at the COBRA Research Institute in our work on passively modelocked semiconductor lasers operating in the 1.5 μm wavelength region. Most results concern modelocked ring lasers that are realized monolithically in the InP/InGaAsP materials system as well as simulations using lumped element and traveling wave type models. The experimental results show that the ring lasers appear as the more stable type of lasers. The modeling results show the importance of using a symmetrical configuration in the ring laser for stable operation. Most recent results on linear modelocked quantum dot lasers at 1.5 μm indicate the improvements possible using these materials.

Semiconductor ring lasers show great promise for rotation sensing through the Sagnac effect of frequency shifting. Ensuring a controlled unidirectional operation of ring lasers can greatly benefit this application. An S-section racetrack design for semiconductor ring lasers was previously developed with the goal of favoring the waves traveling in a preselected direction and suppressing the counterpropagating waves. However, that design turned out to be not as effective as expected, with bistable behavior and directional switching over a wide range of pumping currents. We report on design, fabrication and characterization of Y-junction S-section InAs/InGaAs/GaAs/AlGaAs quantum dot ring lasers with improved unidirectionality. The new design suppresses the unwanted counterpropagating waves more effectively than it was possible in the previous S-section-racetrack design.

The precise interferometric systems employed in today's artificial satellites require semiconductor lasers of the highest callibur. But, one particularly large obstacle has stood in the way of their broad application; the stabilization of their oscillation frequencies. While a number of different approaches have been tested, none have provided overall, long-term stability. Most recently, we used a Doppler-free absorption line of Rb atoms with a precision temperature controller and an improved laser mount; in this instance, relative optical frequency stability rated 9.07×10^-13 ≤ σ(2,τ) ≤ 7.54×10^-10, in averaging time for 0.01s ≤ τ ≤ 23s. Furthermore, we heated the Rb cell to up to 313K, in order to enhance the control signal and improve oscillation frequency stability.

External cavity diode lasers (ECDL) are presently experiencing a surge in popularity, as laser light-sources for advanced optical measurement systems. While these devices normally require external optical-output controls, we simplified the setup, a bit, by adding a second external cavity. This technique boasts the added advantage of having a narrower oscillation-linewidth than would be achievable, using a single optical feedback. Because drive-current and atmospheric temperature directly impact the ECDL systems' oscillation frequency, during frequency stability checks, it was necessary, in this instance, to construct a slightly smaller ECDL system, which we mounted on a Super-Invar board, to minimize the influence of thermal expansion. Taking these and other aggressive and timely measures to prevent atmospheric temperature-related changes allowed us to achieve an improvement in oscillation-frequency stability, i.e., to obtain the square root of Allan variance σ =2×10^-10, at averaging time τ =10^-1. We introduced a vertical-cavity surface-emitting laser (VCSEL) to the setup, for the simple reason that its frequency is far less susceptible to changes in temperature, than other lasers of its type. And, because VCSELs are widely available, and the ECDL systems that use them improve frequency stability, we replaced the Fabry-Perot semiconductor laser with a VCSEL.

In our work, we considered theoretically 1D and 2D photonic bandgap (PBG) systems containing nonlinear covers with short response time in femtosecond area. The short signal passing through the PBG system the angular total reflection area was calculated by FDTD, transfer matrix and perturbation theory methods. The photonic structure vs intensity behaviour was investigated for a few systems consisting of periodically layered structures covered with an optically nonlinear material. Theoretical estimations of the logical gate parameters were made for linear 2D Si-SiO₂ , Si-air and Ge-Se photonic crystals covered with the nonlinear doped glass. It was shown that the beam angular-frequency diagrams contain extremely sensitive areas inside the total reflection range, where the weak nonlinearity leads to dramatic change in light reflection and transmission. Two principal schemes of all-optical logical devices were analyzed and possible applications in all-optical adders and logical gates were discussed.

The design and performance characteristics of a novel Acousto Optic Tunable Filter (AOTF) are presented. Particular attention has been paid to the reduction of optical side lobes, maximising the light throughput and achieving efficient wideband RF matching of a device for use in hyperspectral imaging systems. Conventional AOTFs are known to yield an optical pass band with side lobes at unacceptable levels of ~-10dB relative to the transmission peak. It is known that shaping the acoustic beam ("apodisation") can suppress the side lobe transmission of the AOTF and improve its imaging capabilities. Results of a novel electrode apodisation pattern are presented, reducing sidelobes to ~-25dB. This produces an AOTF which is capable of being placed in a diffraction limited optical system and introduces negligible amounts of image degradation. The large transducer area (associated with the large optical aperture) and acoustic impedance mismatch between the AO substrate (TeO₂) and transducer (LiNbO₃) pose a challenge in achieving wideband RF performance. Acoustic mismatch between substrate and transducer has been addressed by the introduction of a special acoustic matching layer in the bond. The layer reduces dispersion in the transducer impedance easing broadband matching. The transducer has a low (<1 Ohm) radiation resistance which must be matched to the RF driver (typically 50 ohms). This very low impedance may be swamped by the parasitic impedances of the electrode, bond layers and wire bonds used for electrical connection. Thus, the transducer is split into series-connected sections to increase the "bare" impedance. We present results to show the performance increase that can be obtained this way.