Interdiffused SbN-based Quantum Well on GaAs for 1300-1550 nm Diode Lasers

  • PDF / 292,584 Bytes
  • 6 Pages / 612 x 792 pts (letter) Page_size
  • 21 Downloads / 165 Views

DOWNLOAD

REPORT


0891-EE11-09.1

Interdiffused SbN-Based Quantum Well on GaAs for 1300-1550 nm Diode Lasers Ronald A. Arif and Nelson Tansu Center for Optical Technologies, Department of Electrical and Computer Engineering, Lehigh University, Sinclair Laboratory, 7 Asa Drive, Bethlehem, PA 18015, USA ABSTRACT A new method to realize InGaAsSbN quantum well (QW) structures on GaAs substrate is presented. This approach combines the established growth technique of InGaAsN and InGaAsSb QWs by metal organic chemical vapor deposition (MOCVD), with a post-growth thermal interdiffusion to achieve high quality interdiffused InGaAsSbN QW for diode lasers emitting at 1300-1550-nm. In addition to presenting the optimized interdiffused SbN-based QW design at 1550-nm, strain-compensated interdiffused InGaAsSb-GaAsN QW structure is also presented. Preliminary experimental findings of N- and Sb-diffusivities in GaAs matrix show good agreement with theory, indicating the feasibility of realizing interdiffused InGaAsSbN QW. INTRODUCTION Diode lasers emitting in the 1300-1550-nm regime are of great interest for optical communications. GaAs-based lasers not only exhibit superior lasing performance at elevated temperature in comparison to InP-based ones, but also they can be more practically implemented into VCSELs. Several approaches to realize 1300-1550 nm emission on GaAs have been attempted as follow: InGaAsN QWs [1-7], InAs quantum dots (QDs) [8-9], GaAsSb-(In)GaAsN type-II QWs [10-11], and InGaAsSbN QWs [12-14]. Although excellent InGaAsN QW and InAs QD lasers have been demonstrated at 1150-1300-nm, as emission wavelength is pushed beyond ~1400-nm, threshold current density starts to suffer significantly [1]. InGaAsSbN QWs have recently received a sizable interest as an alternative for 1300-1500nm gain media on GaAs. However – as of today – InGaAsSbN QW lasers can only be realized by molecular beam epitaxy (MBE). Recent work on MBE-grown InGaAsSbN has resulted in promising devices at 1498-nm [12] and 1466-nm [14]. Incorporation of Sb and N into InGaAs QW also offers a much improved electron and hole confinement [15], which leads to improved carrier injection efficiency at high temperature. The pursuit of InGaAsSbN systems by MOCVD however, is still immature due to the challenges in incorporating both Sb- and N-species into InGaAs simultaneously under optimum growth conditions. Optimum MOCVD growth of InGaAsN typically requires [DMHy]/[AsH3] of ~ 20-25 resulting in [V]/[III] of ~ 180-200 [3]. In contrast, optimum growth of Sb-containing materials (with TMSb source) only requires [V]/[III] of ~ 3 [16]. This large discrepancy in the optimum growth conditions presents great difficulty in realizing high-quality MOCVD-grown InGaAsSbN QW. Here we present a new method to realize InGaAsSbN QW, which combines the established MOCVD growth of InGaAsN and InGaAsSb QWs, with a post-growth annealing that leads to interdiffusion of Sb and N-species. As the InGaAsN and InGaAsSb are grown in separate layers, optimization of their individual MOCVD growth parameters c