Paul A. Koning

Thermal Interface Material Technology Advancements and Challenges: An Overview

With an increase in the number of transistors (higher power), shrinking processor size (smaller die), and increasing clock speeds (higher frequency) for next generation microprocessors, heat dissipation at the silicon die level has become a critical focus area for microprocessor architecture and design. In addition, power removal at low cost continues to remain the key challenge as we develop the next generation packaging technologies. Novel Thermal Interface Materials (TIM) are required to be designed and developed to meet these new package thermal targets. This paper presents an overview of the novel TIM technologies developed at Intel including greases, phase change materials (PCM), gels, polymer solder hybrids, and solder TIM for multiple generations of desktop, server and mobile microprocessors. The advantages and limitations of these TIM technologies in the thermal management of flip chip packaging are reviewed for Intel’s microprocessors.Copyright

Thermal Resistance of Particle Laden Polymeric Thermal Interface Materials

Particle laden polymers are one of the most prominent thermal interface materials (TIM) used in electronics cooling. Most of the research has primarily dealt with the understanding of the thermal conductivity of these types of TIMs. For thermal design, reduction of the thermal resistance is the end goal. Thermal resistance is not only dependent on the thermal conductivity, but also on the bond line thickness (BLT) of these TIMs. It is not clear which material property(s) of these particle laden TIMs affects the BLT and eventually the thermal resistance. This paper introduces a rheology based semi-empirical model for the prediction of the BLT of these TIMs. BLT depends on the yield stress of the particle laden polymer and the applied pressure. The BLT model combined with the thermal conductivity model can be used for modeling the thermal resistance of these TIMs for factors such as particle volume faction, particle shape, base polymer viscosity, etc. This paper shows that there exists an optimal filler volume fraction at which thermal resistance is minimum. Finally this paper develops design rules for the optimization of thermal resistance for particle laden TIMs.Copyright

Rheological Study of Micro Particle Laden Polymeric Thermal Interface Materials: Part 1 — Experimental

Particle laden polymers are one of the most prominent thermal interface materials (TIM) used in electronics cooling. Most of the research groups have primarily dealt with the understanding of the thermal conductivity of these types of TIMs. Thermal resistance is not only dependent on the thermal conductivity but also on the bond line thickness (BLT) of these TIMs. It is not clear that which material property(s) of these particle laden TIMs affects the BLT. This paper discusses the experimental measurement of rheological parameters such as non-Newtonian strain rate dependent viscosity and yield stress for 3 different particle volume fraction and 3 different base polymer viscosity materials. These rheological and BLT measurements vs. pressure will be used to model the BLT of particle-laden systems for factors such as volume fraction.Copyright

Rheological Study of Micro Particle Laden Polymeric Thermal Interface Materials: Part 2 — Modeling

Currently there are no models to predict the thickness or the bondline thickness (BLT) of particle laden polymeric thermal interface materials (TIM) for parameters such as particle volume fraction and pressure. TIMs are used to reduce the thermal resistance. Typically this is achieved by increasing the thermal conductivity of these TIMs by increasing the particle volume fraction, however increasing the particle volume fraction also increases the BLT. Therefore, increasing the particle volume fraction may lead to an increase in the thermal resistance after certain volume fraction. This paper introduces a model for the prediction of the BLT of these particle laden TIMs. Currently thermal conductivity is the only metric for differentiating one TIM formulation from another. The model developed in this paper introduces another metric: the yield stress of these TIMs. Thermal conductivity and the yield stress together constitute the complete set of material parameters needed to define the thermal performance of particle laden TIMs.Copyright