可靠性(半导体)
模式(计算机接口)
失效模式及影响分析
材料科学
光电子学
可靠性工程
电子工程
计算机科学
物理
工程类
复合材料
操作系统
功率(物理)
量子力学
作者
Yongkun Sin,Jesse Theiss,Michael Huang,Emily Tang,Jeffrey Childs,In‐Tae Bae,Adam Bushmaker
摘要
Future space satellite systems will use high-speed 25 Gbps 850-nm multi-mode (MM) vertical cavity surface emitting lasers (VCSELs) for communications and 795-nm single-mode (SM) VCSELs for atomic clocks applications. The main advantage of deploying these VCSELs in space satellites over 850- or 795-nm edge emitting lasers is the absence of COMD (catastrophic optical mirror damage). In particular, VCSEL-based atomic clocks have a potential to significantly improve timekeeping accuracy and navigation positioning errors. Space satellite systems require stringent reliability of these VCSELs, but they are significantly lacking reliability data. We assess suitability of these VCSELs for high reliability applications through the physics of failure investigation to study reliability, failure modes, and degradation mechanisms. Also, this work is part of our efforts to understand the physical origin of degradation in oxide confined VCSELs under high current density operation. For the present study, we investigated reliability and failure modes of two state-of-the-art VCSEL types – 25 Gbps 850- nm MM oxide confined VCSELs and 795-nm SM oxide confined VCSELs. Accelerated life-tests of both VCSEL types were performed under varying stress conditions to model wear-out failures for reliability assessment. These life-tests were performed under ACC (automatic current control) mode. We also studied VCSELs that were exposed to ESD (electrostatic discharge). We employed optical beam induced current (OBIC), photocurrent spectroscopy, and electron beam induced current (EBIC) techniques for failure mode analysis (FMA). FMA was performed on life-tested VCSEL failures as well as on ESD tested VCSEL failures. Lastly, we employed plasma focused ion beam (PFIB) for removal of portions of top-DBR mirrors for EBIC and for slice-and-view techniques.
科研通智能强力驱动
Strongly Powered by AbleSci AI