Thermal Transport in Graphene and its Application in Chip Cooling

Graphene, the latest isolated allotrope of carbon made of individual atomic sheets bound in two dimensions, shows many remarkable properties. Academic and industry research groups around the globe are carrying out theoretical and experimental studies to discover and investigate characteristics of graphene. Due to its outstanding properties, graphene has a potential to revolutionize technology. Particularly, graphene was found to be one of the best known heat conductors, thus it can be used in nanoelectronic and optoelectronic devices as a heat spreader component. It was also recently reported that graphene is an excellent conductor of heat with the room temperature thermal conductivity values on the order of or exceeding those of carbon nanotubes. In the experiment an indirect method of measuring G peak position of the Raman spectrum as a function of both the temperature of the graphene sample and the power of the heat source was used to compute the thermal conductivity. The sample in the experiment had rectangle geometry and the heat source was approximated as a hot line on the surface of the sample. The thermal conduction in graphene flakes was simulated using the finite element method with the help of COMSOL software package, which solves numerically the partial differential equations, in order to investigate how thermal transport is influenced by a surface geometry of the sample and geometries of the heat sources. We found that all mentioned factors impact heat propagation in different ways and have to be included in the experimental data extraction. The results of the simulations proved correctness of the experimental determination of the thermal conductivity in single layer graphene. Numerical procedure was also developed to simulate heat propagation in semiconductor device structures with graphene layers incorporated as heat spreaders. The simulation results showed that the incorporation of graphene and its multilayers with proper heat sinks can substantially lower the temperature of the chip. The developed simulation procedure can provide a necessary input for next experiments on heat conduction in graphene structures e.g., graphene multi-layers and graphene heat sink structures and other device-level thermal management applications.