In this work a new current-sweep stress methodology for quantitatively assessing the mixed-mode reliability (simultaneous application of high current and high voltage) of advanced SiGe HBTs is presented. This stress methodology allows one to quickly obtain the complete damage spectrum of a given device from a particular technology platform, enabling better understanding of the complex voltage, current, and temperature interdependence associated with electrical stress and burn-in of advanced transistors. We consistently observed three distinct regions of mixed-mode damage in SiGe HBTs, and find that hot carrier induced damage can be introduced into SiGe HBTs under surprisingly modest mixed-mode stress conditions. For more aggressively scaled silicon-germanium technology generations, a larger percentage of hot carriers generated in the collector-base junction are able to travel to and hence damage the EB spacer, leading to enhanced forward-mode base current leakage under stress. A new self-heating induced mixed-mode annealing effect was observed for the first time under fairly high voltage and current stress conditions, and a new damage mechanism was observed under very high voltage and current conditions. Finally, as an example of the utility of our stress methodology, we quantified the composite mixed-mode damage spectrum of a commercial third-generation (200 GHz) generation SiGe HBT. It is found that if devices are stressed with either voltage or current alone during burn-in, they can easily withstand extreme over-stress conditions. Unfortunately, devices were easily damaged when stressed with a combination of stress voltage and current, and this has significant implications for the device and circuit lifetime prediction under realistic mixed-signal operating conditions.
| Identifer | oai:union.ndltd.org:GATECH/oai:smartech.gatech.edu:1853/19756 |
| Date | 15 November 2007 |
| Creators | Cheng, Peng |
| Publisher | Georgia Institute of Technology |
| Source Sets | Georgia Tech Electronic Thesis and Dissertation Archive |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0024 seconds