Abstract

Nowadays, with the increase in the power and transistor density of microelectronic devices, the working temperature is becoming a big issue in the performance and reliability of devices. The thermal characterizations of experimental measurement and numerical simulation are essential in the device design and test. In this work, the thermal characterization technique, combining experimental and numerical methods, is performed on high-power microelectronic devices. The thermoreflectance-based experimental thermal characterization method is presented, which has both high spatial and temporal resolutions. In addition, the high-speed self-adaptive full 3D simulation tool is used to make 3D models of high-power devices. This work explores high-power devices under different conditions, including steady-state, short-term transient, and long-term stress. The aging behavior is studied during the long-term stressing of the measured devices, and an unexpected temperature behavior is found and identified. The transient temperature behavior in the short periods, between 4 μs to 1 ms, is studied. In addition, 3D simulation models are built based on the measurement results. Several facts, such as power droop, are considered to improve the 3D model. Two reduced calculation models are created by the guidance of the experimental data and simulation results. To calculate the temperature of actual devices, both calculation models have the function of temperature-dependent thermal conductivity. In addition, the long-term transient electroluminescence (EL) measurement method is presented, which is the EL mapping coupled with a spectrum scan. The process of this method, as well as its results, are shown in this study.

Degree Date

Spring 5-13-2023

Document Type

Dissertation

Degree Name

Ph.D.

Department

Mechanical Engineering

Advisor

Peter Raad

Subject Area

Mechanical Engineering

Number of Pages

249

Format

.pdf

Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Available for download on Wednesday, May 03, 2028

Share

COinS