GaN-on-silicon present challenges and future opportunities

Gallium Nitride, in the form of epitaxial HEMT transistors on silicon carbide substrates is now almost universally acknowledged as the replacement for silicon bipolar, power MOSFET, high power devices in the RF, microwave, and mmW arenas. This is particularly true for GaN-on-SiC based MMIC's which enable state-of-the-art high frequency performance and bandwidth to be extended into Ku-Band and Ka-Band applications. The challenge for GaN-on-Silicon technology is to take advantage of these industry accepted GaN-on-SiC results and leapfrog not only the high frequency/high power performance but also drive GaN into a new cost paradigm, enabling the opening of applications currently beyond the reach of silicon carbide based systems. The design and development of basic GaN-on-Silicon structures and devices will be presented. In this discussion comparisons will be made with alternative substrate materials with emphasis on contrasting the inherent advantages of a silicon based system. Theory of operation of microwave and mmW high power HEMT devices will be presented with special emphasis on fundamental limitations of device performance including limitations on the required impedance transformations, internal and external parasitic reactance, thermal impedance, and optimization, and challenges involved by full integration into monolithic MMICs. Lastly, future directions that will enable the scaling of GaN-on-Silicon production into large wafer diameter, mainstream, CMOS silicon semiconductor technologies and marry CMOS digital control with high power/high frequency devices to create the next generation of monolithic ICs will be discussed.