The possibility of gallium oxide (β-Ga2O3) heterojunction bipolar transistors

Abstract

Bipolar junction transistors have not been viable with β-Ga2O3 due to its poor hole mobility and unavailability of shallow acceptors. Many p-type oxides form high-quality heterojunction diodes (low ideality factor and high breakdown voltage) with β-Ga2O3. We propose using these heterojunctions to make a β-Ga2O3 heterojunction bipolar transistor (HBT). Cu2O/β-Ga2O3 heterojunction is especially promising because of the relatively high electron diffusion length (∼μ m) in Cu2O, a low electron injection barrier at the Cu2O-Ga2O3 interface, and breakdown voltages of >1000 V. Using Silvaco TCAD, we simulate a β-Ga2O3 heterojunction bipolar transistor with a Cu2O base and estimate the power figure of merit (PFOM). We find that the low bandgap of Cu2O severely limits the performance of these HBTs. Reports of Cu2O-Ga2O3 diodes with extremely high breakdown voltage are probably due to heavily doped Cu2O or interface defects, but these effects do not translate to the HBT. For HBTs with PFOM better than the state-of-the-art β-Ga2O3 unipolar transistors, we need alternative p-type oxides with a bandgap E g > 3.4 eV and electron diffusion length >0.4 μ m. We discuss the possible candidates. Using an empirical model for the critical avalanche breakdown field, we estimate the maximum PFOM for possible β-Ga2O3 HBTs.