Integrated cooling devices in silicon technology

Silicon technology has become a good alternative to copper for the elaboration of efficient cooling devices required in power electronics domain. Owing to its high degree of miniaturization, it is expected to provide suitable microchannels and other inlets holes that were not achievable by copper micromachining. Besides, the use of silicon technology provides a variety of bare materials (silicon dioxide, silicon nitride, silicide, etc.) which may be either insulator or conductive, with a good or bad thermal conductivity. This large choice makes it possible to built up rather complex multilayer devices with mechanical properties good enough in comparison with hybrid copper technology heat sinks. Nevertheless, the use of silicon technology, where the microchannel width may reach few tens of microns, raises fundamental features concerning the fluid displacement within such small sections. More precisely, fundamental fluid mechanics studies have to be conducted out in order to get an accurate description of the fluid boundary layers and to provide basic data on the exchange mechanisms occurring at these surfaces. In this paper, we review the operation principles of both single- and double-phase heat exchange devices elaborated in silicon technology. Forced-convection heat sinks as well as integrated micro heat pipes are analyzed. An analytical approach is adopted to evaluate their total thermal resistances as a function of several geometrical parameters. Numerical simulations are then used in order to assess the accuracy of the analytical approach and to evaluate the impact of the fluidic aspects on the whole performance. The optimum devices are then conceived thanks to an appropriate optimization procedure taken into account the several experimental constraints. Reference values of similar copper devices are reminded and the advantages of the silicon integrated approach are highlighted.