Nitride VCSEL design for continuous-wave operation of higher-order optical modes

Anticipated performance of possible designs of nitride vertical-cavity surface-emitting lasers (VCSELs) has been analyzed and compared with the aid of an advanced 3D optical–thermal–electrical self-consistent simulation. It has been revealed in the simulation that, to achieve room-temperature (RT) continuous-wave (CW) VCSEL threshold operation, uniformity of both carrier-concentration and temperature-increase distributions within VCSEL active regions are equally important. Besides, mostly because of problems with an effective confinement of an optical field in a radial direction, RT CW lasing in standard nitride VCSELs is usually easier to achieve on higher-order transverse modes than on lower-order ones. Therefore, a novel design of nitride VCSELs with double (annular and central dot) p-side contacts has been intentionally constructed to enhance an excitation of low-threshold high-order transverse optical modes. Anticipated RT CW performance characteristics of the novel design have been proved to be definitely more promising than those of other possible configurations of nitride VCSELs. In particular, our calculations have confirmed that RT CW lasing is possible in this new nitride VCSEL. Besides, this device has been found to be less sensitive (than other nitride VCSELs) to possible crystal imperfections created around an active region while high-resistive areas are formed. PACS 42.55.Px; 85.30.De; 85.60.Jb     

Transverse-mode selectivity in possible nitride vertical-cavity surface-emitting lasers

Excitation of various transverse modes in possible nitride vertical-cavity surface-emitting lasers (VCSELs) is investigated and compared using the effective frequency optical model. In the comparative analysis of laser mode selectivity, two distinctly different configurations of possible nitride VCSELs are considered: the traditional VCSEL design with both (n-side and p-side) ring contacts as well as the uniform-current-injection (UCI) VCSEL design. Our simulation reveals that, during the continuous-wave device operation at room temperature, a multi-mode operation dominated by higher-order transverse modes is typical for traditional nitride VCSEL configurations whereas a desirable single-mode (based on the fundamental LP₀₁ mode) operation turns out to be characteristic for the wide current range in UCI ones. The above different threshold device behaviours are an immediate consequence of essentially different current-spreading phenomena in both VCSEL designs, resulting in completely different not only gain profiles but also temperature distributions within the laser active regions of both VCSELs. Seemingly similar behaviour has been also reported in arsenide VCSELs but it is expected to be much more severe in the case of nitride ones as a result of much higher both electrical resistivities of p-type nitrides and their temperature derivatives of refractive indices.     

Current spreading modification to enhance single-fundamental-mode VCSEL operation at higher temperatures

Continuous-wave (CW) performance of modern oxide-confined (OC) vertical-cavity surface-emitting diode lasers (VCSELs) at room and elevated temperatures is investigated with the aid of the comprehensive fully self-consistent optical-electrical-thermal-gain model. A standard OC GaInNAs/GaAs double-quantum-well VCSEL emitting the 1.3-μm radiation is used as a typical modern VCSEL structure. The oxide aperture is placed at the anti-node position of an optical standing wave within a VCSEL cavity. The desired single-fundamental-mode (SFM) operation has been found to be expected only in VCSELs equipped with relatively small active regions of diameters equal or smaller than 10 μm. Therefore a proton implantation used as an radial additional confinement of the current spreading from the upper annular contact towards the centrally located active region is proposed and its impact on the VCSEL performance is investigated. The above structure modification has been found to enable a radical improvement in the VCSEL performance. In particular, in this case, the SFM VCSEL operation is possible even in VCSELs with quite large active regions and for much wider ambient-temperature range than in the standard OC VCSELs.     

Diode pumped 850 nm vertical-cavity surface-emitting laser

Very little, if any, has been published on optically pumped 850 nm vertical-cavity surface-emitting lasers (VCSELs), particularly for doped structures. This paper investigates GaAs-based VCSELs which have not been optimized for optical pumping work. Characterisation was carried out for both pulsed and continuous wave (CW). Pulsed operation causes a lower rise in temperature, thus postponing the onset of thermal rollover, and allowing the device to be operated at higher powers. A threshold of ∼160 kW/cm², and single mode output with incident power density of up to 225 kW/cm² were obtained. From the simulation work done, it has been observed that for optically pumped VCSELs, at higher pump density, there was faster turn on and higher output power, and that dilute nitride active material give better output performance compared to GaAs.     

Structure optimisation of modern GaAs-based InGaAs/GaAs quantum-dot VCSELs for optical fibre communication

Room-temperature (RT) continuous-wave (CW) performance of modern 1300-nm oxide-confined In(Ga)As/GaAs quantum-dot (QD) vertical-cavity surface-emitting diode lasers (VCSELs) taking advantage of many QD sheets is investigated using our comprehensive self-consistent simulation model to suggest their optimal design. Obviously, quantum dots should be as uniform as possible and as dense as possible to ensure high enough optical gain. Besides, our simulation reveals that efficient and uniform current injection into VCSEL active regions necessary to enhance excitation of the desired fundamental LP₀₁ mode is accomplished in the VCSEL configuration with the broad-area bottom contact and the ring upper one as well as with the oxide aperture localized within the first period of the upper p-type DBR. The doping of the DBR mirrors is chosen as a compromise between their high enough electrical conductivity and low enough free-carrier absorption. The oxide aperture is additionally introducing the radial optical waveguiding. Moreover, our analysis has been concluded that VCSEL active regions should be composed of at least 9 QD sheets to acquire efficient RT CW operation. Furthermore, rather longer optical cavities are recommended in this case because localization of QD sheets should be adjusted to the anti-node positions of the optical cavity standing wave.     

Thermal aspects of designing CW-operated nitride VCSELs

A room-temperature (RT) continuous-wave (CW) operation of possible nitride vertical-cavity surface-emitting lasers (VCSELs) is considered in the present paper using a simple VCSEL simulation to give some essential guidelines for their proper designing. It is revealed that a substrate material has a critical influence on a possibility of reaching RT CW thresholds which practically excludes currently used sapphire substrates of relatively low thermal conductivity from this application. SQW nitride VCSELs are found to be very sensitive to an increase in temperature, which is followed by their inappropriate CW-operation characteristics. A moderate increase in a number of quantum wells in VCSEL active regions significantly improves their CW achievements, i.e. decreases RT CW thresholds as well as considerably widens their CW threshold ranges.     

Comprehensive self-consistent three-dimensional simulation of an operation of the GaAs-based oxide-confined 1.3-μm quantum-dot (InGa)As/GaAs vertical-cavity surface-emitting lasers

In the paper, a comprehensive fully self-consistent three-dimensional simulation of an operation of the GaAs-based oxide-confined long-wavelength 1.3-m quantum-dot (QD)(InGa)As/GaAs vertical-cavity surface-emitting diode lasers is demonstrated. The model has been intentionally prepared for the PC-class microcomputers to enable its easy application in designing optimal structures of the above devices with desired performance characteristics. An impact of some structure parameters on QD VCSEL room-temperature (RT) continuous-wave (CW) lasing thresholds is discussed. A stable RT CW operation on a single fundamental mode has been found to be possible in modern QD VCSELs with active regions containing more uniform and more dense QDs in stacks of QD layers. The desired single fundamental-transverse-mode operation is possible for smaller active regions of diameters not exceeding 8 m. In the case of larger active regions, on the other hand, higher-order transverse modes of an increasing order are excited first because of increasingly more non-uniform optical-gain distributions.     

Enhanced single-fundamental LP01 mode operation of 650-nm GaAs-based GaInP/AlGaInP quantum-well VCSELs

Minimal optical attenuation of plastic (polymer) optical fibres (POFs) corresponds to the 650-nm wavelength. Currently the GaInP/AlGaInP quantum-well (QW) oxide-confined (OC) vertical-cavity surface-emitting diode lasers (VCSELs) are undoubtedly the laser devices most suited to be used in 650-nm POF optical communication, for which the stable single-fundamental-mode LP₀₁ emission (SFM) is definitely the one most desired. In the present paper, the comprehensive fully self-consistent VCSEL model is used to examine mode selectivity of the above VCSELs. An increase in the VCSEL active-region diameter leads to a gradual modification of the current injection into this region and subsequent carrier radial diffusion within it before their recombination, which is followed by an essential transformation of active-region optical-gain profiles deciding upon an excitation of successive transverse modes. In standard arsenide OC VCSELs, SFM operation is usually limited to relatively small active regions. But for a room-temperature continuous-wave operation of the GaInP/AlGaInP VCSELs, the fundamental LP₀₁ mode remains surprisingly the lowest-threshold one up to relatively large active regions of 9-μm diameters. Nevertheless, in such VCSELs, thresholds of many LP modes become very similar to one another, which leads to their relatively poor mode selectivity and an unwanted multi-mode operation for higher output powers.     

Comparative analysis of lasing performance of oxide-confined and proton-implanted vertical-cavity surface-emitting diode lasers

The comprehensive optical-electrical-thermal-recombination self-consistent VCSEL model is used to compare the performance of oxide-confined (OC) and proton-implanted (PI) VCSELs and to optimise their structures. Generally index-guided (IG) OC VCSELs demonstrate lower lasing thresholds whereas both gain-guided (GG) OC and PI ones manifest much better mode selectivity. Therefore, their either low-threshold IG or mode-selective GG versions may be intentionally used for different VCSEL applications. Lasing thresholds of OC IG VCSELs have been found to be very sensitive to the exact localisation of their thin oxide apertures, which should be shifted as close as possible towards the anti-node position. PI VCSELs, on the other hand, are simpler and cheaper in their manufacturing than OC ones. Although lower threshold currents are manifested by PI VCSELs with very thick implanted regions, lower threshold powers are achieved in these devices with much thicker upper unaffected layer used for the radial current flow from the ring contact towards the laser axis. Paradoxically poor thermal properties of PI VCSELs enable lower lasing thresholds of slightly detuned devices. To conclude, cheaper and mode-selective PI VCSELs may be used instead of OC ones in many of their applications provided ambient temperatures and laser outputs are not too high.     

Sensitivity of a VCSEL threshold performance to inaccuracies in its manufacturing

The paper describes an impact of various possible inaccuracies in manufacturing of verticalcavity surface-emitting diode lasers (VCSELs), like thicknesses and compositions of their layers different from assumed ones, on VCSEL room-temperature (RT) continuous-wave (CW) threshold performance. To this end, the fully self-consistent comprehensive optical-electrical-thermal-recombination VCSEL model has been applied. While the analysis has been carried out for the 1.3-μm oxide-confined intra-cavity contacted GaInNAs/GaAs VCSEL, its conclusions are believed to be more general and concern most of modern VCSEL designs. As expected, the VCSEL active region has been found to require the most scrupulous care in its fabrication, any uncontrolled variation in compositions and/or thicknesses of its layers is followed by unaccepted RT CW lasing threshold increase. Also spacer thicknesses should be manufactured with care to ensure a proper overlapping of the optical standing wave and both the gain and lossy areas within the cavity. On the contrary, less than 5% thickness changes in distributed-Bragg-reflectors are followed by nearly insignificant changes in VCSEL RT CW threshold. However, exceeding the above limit causes a rapid increase in lasing thresholds. As expected, in all the above cases, VCSELs equipped with larger active regions have been confirmed to require more careful technology. The above results should enable easier organization of VCSEL manufacturing.     

光电负反馈下单向耦合注入垂直腔表面发射激光器的混沌同步特性研究

基于自旋反转模型（SFM）,对光电负反馈下单向耦合注入垂直腔表面发射激光器（VCSEL）的混沌同步特性进行了数值仿真和理论分析.研究结果表明：随着反馈强度的增加,激光器输出功率呈现两个不同的混沌区域.在单一偏振模区域,能实现很好的同步性,但接近两模共存时,同步质量急剧下降;在两模共存区域,偏振模的连续性差,同步质量不理想.两模共存时的同步质量比单一偏振模差的这一特性,与已报道的基于光反馈的情景相反.产生这一特殊现象的原因是光电负反馈诱发了偏振模竞争效应.此外,还发现内部参数的变化对两个混沌区域的选取有较大 关键词： 线性偏振模 光电负反馈 垂直腔表面发射激光器（VCSEL） 混沌同步     

Design of an InGaAs/InP 1.55 μm electrically pumped VCSEL

The design of an electrically pumped InGaAs quantum well based vertical cavity surface emitting laser (VCSEL) on InP substrate is presented. Such optically pumped VCSELs have already been demonstrated. To design electrically pumped VCSEL, three simulations steps are needed: optical simulation gives access to the standing-wave electric field distribution, to design the active region and the Bragg mirrors. Thermal simulation is helpful to design metallic contacts while the energy band diagram is obtained by electrical simulation to design the buried tunnel junction useful for carrier injection. All these simulations are compared to experiment.     

Modal behavior of photonic-crystal vertical-cavity surface-emitting diode laser analyzed with Plane Wave Admittance Method

A three-dimensional, full vectorial model has been applied for investigation the modal characteristics of a phosphide photonic-crystal vertical-cavity surface-emitting diode laser. The photonic crystal has been defined as a regular, hexagonal net of holes, etched in the upper, p-type Distributed Bragg Reflector of the laser. The electromagnetic field has been confined to the active region by a single defect of the photonic crystal providing overlapping with carriers funneled by a tunnel junction. We have determined the range of design parameters, which provide single mode operation.     

VCSEL structures used to suppress higher-order transverse modes

Currently unwanted excitation of higher-order transverse modes is the most serious drawback of vertical-cavity surface-emitting diode lasers (VCSELs) limiting their possible applications. In the present paper, various methods used to suppress those modes are described and their effectiveness is compared. It is well known that, because of a nearly uniform current injection into their active regions, small-aperture VCSELs without any modification offer quite high single-fundamental-mode (SFM) output. However, their series resistance is often too high, which aggravates their high-modulation performance. Similarly uniform current injection may be also achieved with the aid of a tunnelling junction. Generally, methods suppressing higher-order modes take advantage of higher optical gain within the central part of the active region, higher radiation losses outside this region and/or higher central mirror reflectivity. Currently, applications of a tunnel junction, an impurity-induced disordering or an inverted shallow surface relief seem to be the simplest and the most effective methods. The deep etched holey structure or the ARROW structure enable obtaining similar single-mode output powers but they may be used in special cases only because of their complex technology. Photonic crystals may probably enable more advanced mechanisms of suppressing higher-order modes in future because currently their application seems to be still far from being optimised.     

厚度偏差对VCSEL的反射谱和反射相移的影响

采用光学传输矩阵方法分析了厚度偏差对VCSEL的反射谱和反射相移产生的影响。结果表明,反射镜和VCSEL中各层厚度的偏大,将使反射镜的中心波长以及VCSEL的模式波长向长波方向移动,而反射镜和VCSEL中各层厚度的偏小,将使反射镜的中心波长以及VCSEL的模式波长向短波方向移动。将键合界面离有源区稍微远一些,有利于减小其厚度偏差对VCSEL的模式波长的影响。     

Fracture mode transitions during indentation of columnar TiN coatings on metal

Pyramidal indentations and focused-ion-beam machining have been used to study the damage during contact loading of textured columnar titanium nitride hard coatings on three steels and aluminium. The influence of the substrate hardness and coating thickness on the type and extent of fracture have been examined. In particular, the primary mechanism of deformation is shown to undergo a change from inter-columnar shear to other forms of subsurface damage. A physical model is proposed to explain this transition in the fracture mode. The implications of such behaviour for coating design are discussed.     

Current spreading and series resistance of proton-implanted vertical-cavity top-surface-emitting lasers

In this paper, the current flow through the whole volume of the proton-implanted Vertical-Cavity top-Surface-Emitting Lasers (VCSELs) is analysed in detail. A simple approximate analytical relation was derived for a radial distribution of the current density entering active regions of those lasers. This distribution is nearly uniform in the case of VCSELs with a very small active region, but is becoming more and more non-uniform with an increase in its size. In VCSELs with very large active regions, current is flowing practically only within a narrow annular area close to the active-region perimeter. The VCSEL series electrical resistance is determined as a function of its active-region radius.     

Management of the electrical injection uniformity in broad-area top-emitting VCSELs

The electrical properties of broad-area 850 nm top emitting VCSELs have been investigated in order to improve carrier injection uniformity in their active zone. First, we have demonstrated using an electrical simulation tool that a multi-point localized injection design associated with a spreading layer at the top of the device (ITO) can lead to a significant improvement of carrier injection and on its spatial distribution. Secondly, the electrical contrast achievable by applying this method with localized etchings has been experimentally measured. Finally, stripe-shaped devices with output power up to 50 mW in a continuous-wave operation at room temperature have been demonstrated.     

新型多有源区隧道再生光耦合大功率半导体激光器

针对大功率半导体激光器面临的主要困难,提出并实现了一种隧道再生多有源区耦合大光腔 高效大功率半导体激光器机理.该机理能有效地解决光功率密度过高引起的端面灾变性毁坏 、热烧毁和光束质量差等大功率激光器存在的主要问题.采用低压金属有机化合物气相淀积 方法生长了以碳和硅作为掺杂剂的GaAs隧道结、GaAs/InGaAs 应变量子阱有源区和新型多有 源区半导体激光器外延结构,并制备了高性能大功率980nm激光器件.三有源区激光器外微分 量子效率达2.2,2A驱动电流下单面未镀膜激光输出功率高达2.5W. 关键词： 半导体激光器 大功率 金属有机化合物气相沉积     

Design and Fabrication of GaInAsP/InP VCSEL with Two a-Si/a-SiN x Bragg Reflectors

We report on the design and fabrication of a 1.55 μm wavelength Vertical Cavity Surface Emitting Lasers (VCSELs) which consists of two dielectric Bragg mirrors and a InGaAsP-based active region. The dielectric materials are amorphous silicon and amorphous silicon nitride. Layers of such materials have been deposited by magnetron sputtering and analyzed in order to determine their optical properties. A large refractive index difference of 1.9 is found between these materials. Distributed Bragg Reflectors (DBRs) based on these dielectric materials quarter wave layers have been studied by optical measurements and confronted to theoretical calculations based on the transfer matrix method. A maximum reflectivity of 99.5% at 1.55 μm and a large spectral bandwidth of 800 nm are reached with only four and a half periods of a-Si/a-SiN_x. The VCSEL was fabricated by metallic bonding process. This method allows to bond an InP-based active region as the gain medium on a Si substrate thanks to the formation of a Au–In alloy. This process is performed at a low temperature of 240°C without damaging the optical properties of the microcavity. This VCSEL has been characterized by an optical pumping experiment with a low and a high-density optical power and a laser emission has been obtained at room-temperature.