Heat conduction properties of graphene: Prospects of thermal management applications

摘要

As the electronic industry moves towards few-nanometer-scale CMOS and 3D IC designs thermal management becomes crucially important for achieving high performance and reliability of advanced electronic chips [1]. One approach for mitigating the self-heating problems is finding materials with very high thermal conductivity, which can be integrated with Si ICs or used as fillers in the next generation of the thermal interface materials (TIMs). In 2008, we discovered that graphene reveals extremely high intrinsic thermal conductivity, which can exceed that of bulk graphite [2–3]. To measure the thermal conductivity of an object with a thickness of just one atomic layer, we developed an original experimental technique and applied it to graphene flake suspended across trenches in Si wafers. In this technique, the micro-Raman spectrometer performed the function of a thermometer measuring the local temperature rise from the shift in the spectral position of the Raman G peak. We explained the fact that the intrinsic thermal conductivity of graphene can be larger than that of graphite by the fundamental difference in the low-energy phonon transport in 2D graphene and 3D graphite [4–6]. The extremely high thermal conductivity of “free” suspended graphene does not mean that it will be automatically preserved when graphene is incorporated inside semiconductor chips or composite TIMs. Thermal conductivity of graphene layers depends strongly on their geometrical size, coupling to the adjacent substrate or capping layers, edges roughness and defect concentration. I will overview the experimental and theoretical results for the thermal conductivity evolution of the few-layer graphene (FLG) considering two limiting cases of the phonon transport limited by the intrinsic and extrinsic effects. The use of graphene as interconnects and heat spreaders in advanced 2D and 3D computer chips will also be discussed. The last section of the talk will have a descr- - iption of the data for graphene TIM materials. We found that thermal conductivity of several types of epoxy TIMs can be significantly increased by an addition of the chemically derived graphene even at very small graphene's loading fractions. The increase in the effective thermal conductivity of graphene TIMs is much stronger than that for conventional filler materials [7]. A general outlook at the prospects of graphene electronics will conclude the talk.