Ideal Channel Field Effect Transistors

Abstract

Abstract : Narrow bandgap semiconductors offer high carrier mobilities and low contact resistances while wide bandgap semiconductors offer high breakdown voltages. A series of heterojunction transistors have been investigated and proved to be effective for improving both speed and power output in the past two decades. These devices include double heterostructure InP/InGaAs/InP bipolar transistors and composite channel InAlAs/InGaAs/InP/InAlAs high electron mobility transistors (HEMTs), which have taken the full advantage of the matched lattice constant (or pseudomorphic growth). However, for the most popular wide bandgap semiconductor GaN and SiC, the lattice mismatch between GaN and semiconductors with a reasonably small bandgap (including InGaN) is so large that pseudomorphic growth is very difficult. For instance, the critical thickness of InN on GaN is about one monolayer. To marry the advantages offered by both narrow bandgap and wide bandgap semiconductors, we explored direct wafer bonding for ideal channels made of extremely mismatched materials for field effect transistors. Toward this target, we have performed the following studies. This investigation demonstrated it is feasible to fabricate composite channel transistors, however, more experiments are necessary to understand the effects of the interface between the mismatch materials. A. Theoretically calculate scattering rates in composite channels. B. Theoretically calculate breakdown voltage in heterostructures containing wide bandgap material. C. Experimentally form mismatched GaAs/GaN heterostructures through direct wafer bonding and study the breakdown improvement compared with GaAs/GaAs homojunctions. D. Experimentally form InGaAs channel on mismatched GaN substrate and fabricate MISFETs on InGaAs/GaN.