In situ Optical Observations of Degradation of Thermal Greases with Thermal Cycling

In situ Optical Observations of Degradation of Thermal Greases with Thermal Cycling

Event Date: May 31, 2023
Authors: R. Kulkarni, P. Nagrani, and A. Marconnet
Journal: 2023 Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)
Paper URL: Link to Full Text
2023 Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm), Orlando, FL, May 30 - June 2, 2023. DOI: 10.1109/ITherm55368.2023.10177650

Power requirements and cycling requirements of electronic devices have increased significantly over the past several years. Within electronics packages, thermal greases or thermal interface materials (TIMs) are extensively used to minimize contact resistance and enable effective heat flow paths for heat dissipation. However, due to time-varying heat loads from semiconductor chips as system use varies, thermal greases undergo thermal (and thermomechanical) cycling and often experience deterioration over their lifetime. The degradation can appear as regions with a deficit of the material or the separation of the matrix polymer from particles that give the material high thermal conductivity. In either case, this can lead to local hotspots with excessive junction temperatures. Past work on reliability of thermal greases documents a range of methods to accelerate the deterioration process but general focus on analyzing the system after failure. Here we develop a system for observing the redistribution of the thermal grease during thermal cycling with fixed bondline thickness (BLT). Briefly, the thermal grease is squeezed between a heater and an optically transparent substrate with shims to define the BLT during cycling. In this work, we evaluate degradation of two TIMs, differing in viscosity and thermal conductivity, for four fixed BLTs and two power cycle times. Periodically throughout the cycling process optical images are taken to observe the evolution of the thermal grease. Metrics for analyzing pump-out data include analysis of the boundary migration, as well as the evolution of the void fraction and TIM area. This in situ analysis of TIM deterioration will help to identify the governing factors for TIM pump-out and potential mitigation strategies.