Oxide-confined Vertical-cavity Surface-emitting Diode Lasers Research Articles

Room-temperature continuous-wave operation of the In(Ga)As/GaAs quantum-dot VCSELs for the 1.3 µm optical-fibre communication

Efficient room-temperature (RT) continuous-wave (CW) lasing operation of the 1.3 µm MBE (molecular-beam epitaxy) In(Ga)As/GaAs quantum-dot (QD) top-emitting oxide-confined vertical-cavity surface-emitting diode lasers (VCSELs) for the second-generation optical-fibre communication has been achieved. In their design, a concept of a QD inside a quantum well (QW) has been utilized. The proposed In(Ga)As/GaAs QD active region is composed of five groups of three 8 nm In0.15Ga0.85As QWs, each containing one InAs QD sheet layer. In each group located close to successive anti-node positions of the optical standing wave within the 3λ cavity, QWs are separated by 32 nm GaAs barriers. Besides, at both active-region edges, additional single InGaAs QWs are located containing single QD layers. For the 10 µm diameter QD VCSELs, the RT CW threshold current of only 6.2 mA (7.9 kA cm−2), differential efficiency of 0.11 W A−1 and the maximal output power of 0.85 mW have been recorded. The experimental characteristics are in excellent agreement with theoretical ones obtained using the optical-electrical-thermal-recombination self-consistent computer model. According to this, for the 10 µm devices, the fundamental linearly polarized LP01 mode remains the dominating one up to the current of 9.1 mA. The lowest RT CW lasing threshold below 5 mA is expected for 6 µm devices.

Comparison of Usability of Oxide Apertures and Photonic Crystals Used to Create Radial Optical Confinements in 650-nm GaInP VCSELs

Threshold characteristics of GaAs-based 650-nm gallium indium phosphide (GaInP) vertical-cavity surface-emitting diode lasers (VCSELs) with two different optical confinement structures are investigated with the aid of a self-consistent, fully- physical VCSEL model. Efficacy of the optical confinement introduced by the oxide aperture is compared with an alternative single-defect photonic-crystal design with holes etched throughout the whole VCSEL. Initially, photonic-crystal VCSEL reveals 10% lower threshold electrical power than that of the analogous oxide-confined VCSEL. Further optimization of the current injection allows for an additional 20% threshold reduction. The photonic-crystal confinement concept appears to be a very prospective solution for VCSEL configurations, for which oxidation is unfeasible, e.g., for possible nitride or phosphide VCSELs.

Comparison of Exactness of Scalar and Vectorial Optical Methods Used to Model a VCSEL Operation

In the paper, a comparison between results of the optical scalar and vectorial optical methods, using the scalar effective frequency method and the vectorial method of lines, respectively, as their typical examples, applied for simulating an operation of vertical-cavity surface-emitting diode lasers (VCSELs) is presented. Both optical approaches has been applied to simulate an operation of the standard 1.3-mum GaAs-based oxide-confined GaInAsN double quantum-well VCSEL, but comparison conclusions seem to be more general and concerns all VCSELs. Validity limits of the simplified scalar approach have been determined by comparing its approximate results with more exact ones obtained with the aid of the accurate vectorial model. The scalar model has been found to be quite exact in the case of a determination of the wavelength of emitted radiation, whereas its exactness in the lasing threshold analysis is much worse, especially in the case of higher order modes, smaller aperture diameters and/or weak radial optical guidance. Our analysis leads also to more general designing conclusions concerning the optimal position and the thickness of the oxide layer assuring, respectively, the stable single mode operation and the lowest threshold gain

Optimization of oxide-confined vertical-cavity surface-emitting diode lasers

The performance of modern oxide-confined (OC) vertical-cavity surface-emitting diode lasers (VCSELs) is distinctly different than that of earlier VCSEL designs. As a result, hitherto known optimal VCSEL designs cannot be considered as the best ones any longer; they need new analysis to determine their optimal OC designs for their various applications. Therefore, in this paper, a comparative analysis of the room-temperature (RT) continuous-wave (CW) threshold operation of various designs of OC VCSELs is carried out using the 1.3 µm OC intra-cavity-contacted GaAs-based quantum-well (GaIn)(NAs)/GaAs VCSEL as a typical example. The analysis has been made with the aid of a comprehensive fully self-consistent three-dimensional optical-electrical-thermal-gain model. Depending on an application, different OC VCSEL designs have been found to be optimal. The lowest RT CW lasing threshold has been found to be achieved in the 1.5λ-cavity design with only one oxide aperture (for smaller active regions) or in the 3λ-cavity design with two oxide apertures (for larger active regions). On the other hand, to reach the single-fundamental-mode operation and/or longer lasing wavelengths, the 3λ-cavity OC VCSEL with one oxide aperture happened to be the best lasing design, especially for relatively large active regions.

Separate-confinement-oxidation vertical-cavity surface-emitting laser structure

In the present paper, a comprehensive self-consistent three-dimensional model is used to analyze physical aspects of the operation of oxide-confined vertical-cavity surface-emitting diode lasers (VCSELs) and to optimize their structures. The impact of the built-in radial confining mechanisms created by oxide apertures, i.e., the influence of their diameters and localizations on radial confinements of both the current injection into VCSEL active regions and electromagnetic fields of successive cavity modes, has been investigated. Basically, there are two extreme cases: the index-guided (IG) VCSELs with the aperture localized at the antinode position of the optical standing wave, characterized by a very low lasing threshold but exhibiting rather poor mode selectivity, and the gain-guided (GG) VCSELs with the aperture shifted to the node position, ensuring usually the single-fundamental-mode operation, but at the expense of much higher lasing threshold. In the present paper, the separate-confinement-oxidation VCSEL structure has been proposed exhibiting simultaneously advantages of both the IG VCSELs (the low lasing threshold) and the GG ones (the single-fundamental-mode operation).

Usability limits of the scalar effective frequency method used to determine modes distributions in oxide-confined vertical-cavity surface-emitting diode lasers

The accuracy of the commercially used, simplified scalar optical effective frequency method (EFM) is verified by comparing its results with those obtained using more accurate vectorial models represented by the method of lines (MoL). Both EFM and MoL methods are applied to the modern telecommunication oriented GaAs-based oxide-confined vertical-cavity surface-emitting diode laser (VCSEL) emitting 1.3 µm radiation from the InGaAsN/GaAs double quantum well active region. Although both methods are applied to the specific laser structure, the conclusion remains valid for a broad range of different VCSEL designs. The scalar EFM approach has been found to remain surprisingly often quite exact; nevertheless, it fails for structures violating its fundamental assumption of plane wave propagation within the laser cavity. Therefore, a noteworthy discrepancy between modes calculated with the aid of scalar and vectorial approaches occurs for weaker guidance, i.e. for higher-order modes, smaller aperture diameters, thinner oxidation layers and/or apertures shifted from the antinode position. These cases need an exclusive use of the more exact but also more time-consuming vectorial algorithms.