Simulation of Life Testing Procedures for Estimating Long-Term Degradation and Lifetime of AlGaN/GaN HEMTs

Finite element 3-D thermal simulations of long-term degradation in AlGaN/GaN HEMTs for high-power applications are reported on, in which temperature evolves over time as the local degradation rate varies within the modeled device based on the local temperature of the degrading region (i.e., the channel). Specifically, hotter regions within a device are modeled as degrading faster due to a thermal component to the degradation rate equation. This allows self-consistent simulation of life testing, commonly used to estimate long-term reliability by extrapolating failure times seen at elevated channel temperatures to a lower "use" temperature. We find that it is necessary to consider the entire distribution of temperatures within the device instead of at one characteristic location to get the most accurate estimates for long-term device life. The effect of device geometry, assumed degradation mode, incorrect thermal resistance data, and dissipated power level on this lifetime estimation error is investigated. It is found that the error in the extrapolated failure time is greatly increased when both the thermal resistance is in error and the dissipated power of the life test does not match the expected power during operation, compared to when only one of these is off.