Germanium for next generation photonic and microelectronic devices

点击翻译按钮获取中文摘要

With sustained exponential growth, global internet traffic is expected to reach 2.3 zettabytes (2.3x270) by 2020 [2016 Cisco]. However, mainstream short-reach communications and on-chip interconnects have been dominated by metal wires, which are much slower, less energy efficient and hard to scale in size. Optical interconnections via silicon (Si) photonics have been widely recognized as a potential solution to overcome this bottleneck. Germanium (Ge) as the most Si-compatible semiconductor has been the underlying and enabling material for Si photonics. Ge has been widely used in photodetectors and modulators providing a data rate of > 50 Gbps [2015 Chen, 2016 Srinivasan]. For Si-compatible lasers, Ge can be used as 1) transition layers between lasing materials such as InGaAs and AlGaAs and Si [2012 Lee, 2016B Lin, 2016 Liu, 2016 Nakao] due to its small lattice mismatch to them and the ease of integration with Si and 2) a lasing material thanks to bandgap engineering [2010 Liu, 2012 C-A]. On the microelectronics side, Ge has been widely used in SiGe heterojunction bipolar transistors (HBTs) for applications in wireless communications.******We propose the following topics on Ge in Si photonics and microelectronics.******1. It is highly desired to have low defect density Ge films on Si to serve as III-V and Si transition layers. Aspect ratio trapping technology (ART) can produce high quality Ge. However, it needs additional fabrication steps and is inferior in thermal conduction. A low/high temperature (LT/HT) growth method is advantageous over ART in these two aspects. However, the Ge quality is not as good. Arsenic doping has been shown to greatly improve Ge quality [2016 Lee], while impacts from other dopants have not studied. We propose to study doping impacts on Ge quality using LT/HT method for high quality Ge film growth on Si.******2. We propose to study the potential and the optimizations of Ge-on-Si lasers by device modeling and simulations.******3. As higher concentration of Ge is used in HBT base layer, Si-Ge interdiffusion is becoming more problematic. We propose to study the interdiffusion behavior in PNP type HBTs, especially the impacts from phosphorus and carbon and the modeling of these impacts for faster and more energy efficient wireless communication systems.*********The proposed research will enable optoelectronic integrated circuit (OEIC) on Si platforms such as a single-chip optical transceiver, which provides the ability to download movies in seconds and are much cheaper and smaller than the current technology with external lasers. The research outcomes can lead to deeper penetration of optical fiber communications, faster wireless communications and significant advancements in the current information technology hardware industry. We truly believe that the research proposed is at the research frontier and will benefit Canada as a world leader in optical communications and information technology greatly.**