Toward increased throughput of femtosecond laser surface processed silicon: The effect of increased thermal loading on microstructure self-organization

With increased transistor densities and the push to 3D chip architecture, thermal management in high power electronics has become an important and rapidly advancing area of research. One method for managing extreme heat fluxes generated by chip stacks is direct-to-chip two-phase heat transfer, where excess heat created by the high power electronics is used to force a change of state of a coolant from liquid to vapor. In the past, permanent, self-organized femtosecond laser-induced micro- and nano-scale surface features have shown to enhance the performance of two-phase heat transfer systems by enhancing the efficiency at which the surfaces boil, leading to decreases in device temperatures and overall power consumption, and limiting thermal strain on the devices. Previously, the industrial viability of the femtosecond laser surface processing (FLSP) methods was limited due to low throughput. However, the new generation of high repetition rate (>10 kHz), high pulse energy (>1 mJ) femtosecond lasers provide a path toward the integration of the FLSP of materials into industrialized devices. As the average incident power increases, the formation of laser-induced micro- and nano-scale surface features can change. In this study, the effect that increased thermal accumulation has on the self-organization of microstructures facilitated on Si using pulse energies up to 1500 μJ at repetition rates up to 50 kHz is correlated with direct, spatially resolved thermal measurements and postprocessing physical surface analysis.