Latest research directions: 2023–2026
Where GaN HEMT physics is moving now: reliability, gate stacks, vertical devices, UWBG materials, and thermal integration.
p-GaN gate reliability
p-GaN gate devices are commercially important, so reliability dominates recent research. Current questions include forward gate stress, dynamic gate breakdown under switching, frequency-dependent degradation, hole accumulation/depletion in p-GaN, Mg-related defects, p-GaN/AlGaN junction behavior, VTH drift, and how to accelerate tests without creating unrealistic failure modes.
2024–2025 work includes reviews of p-GaN gate characteristics, frequency-dependent gate breakdown studies, and proposals such as fully depleted p-GaN and GaN/AlN/AlGaN composite barrier structures.
Research links
MIS/MOS-HEMTs
MIS and MOS gate stacks reduce leakage and can widen gate-drive margin, but introduce dielectric/interface traps, hysteresis, bias-temperature instability, and oxide TDDB. Research is active in Al2O3, HfO2, SiO2, SiN, AlN, surface nitridation, oxygen plasma treatments, and in-situ dielectric deposition.
Research links
Vertical GaN
Vertical GaN aims to move beyond lateral-device scaling limits for kV-class power. The promise is higher voltage/current density with a true vertical drift region. The bottlenecks are native substrate cost/wafer size, drift-region impurity control, trench gate reliability, avalanche behavior, field crowding, and leakage through defects. 2024 IEDM work explicitly frames vertical GaN reliability in terms of lessons learned from Si and SiC.
Research links
AlN and Al-rich UWBG HEMTs
Related wide- and ultra-wide-bandgap materials
AlN and β-Ga₂O₃ extend the wide-bandgap roadmap, but practical devices are limited by thermal transport, contacts, defects, and p-type doping.
AlN and Al-rich AlGaN push toward wider bandgap, higher breakdown field, stronger polarization, and better thermal conductivity. 2025 XHEMT work on single-crystal AlN substrates points to a possible RF path beyond conventional GaN-on-SiC. The big open problems are ohmic contacts, substrate cost, epitaxial defect control, strain/cracking, and manufacturable reliability.
Research links
β-Ga₂O₃ and related UWBG materials
β-Ga2O3 is attractive because it has an ultra-wide bandgap and melt-grown bulk substrates, but its poor thermal conductivity and lack of robust p-type doping are severe limits. It is more directly competitive in diodes and MOSFET-like power devices than in RF HEMTs. The broader lesson for GaN is that breakdown field alone is not enough: heat removal, contacts, doping, and reliability decide the usable device.
Research links
Thermal integration
GaN-on-diamond, top-side diamond, high-conductivity packages, substrate thinning, flip-chip, and electrothermal compact models are active because thermal resistance now limits both RF power density and power-switching reliability. Recent work focuses less on the ideal thermal conductivity of diamond and more on the reproducible thermal boundary resistance of real integrated stacks.