First Operation of AlGaN Channel High Electron Mobility Transistors

Abstract

A channel layer substitution of a wider bandgap AlGaN for conventional GaN in high electron mobility transistors (HEMTs) is one possible method of enhancing the breakdown voltage for higher power operation. Wider bandgap AlGaN, however, should also increase the ohmic contact resistance. We utilized a Si ion implantation doping technique to achieve sufficiently low resistive source/drain contacts, and realized the first HEMT operation with an AlGaN channel layer. This result is very promising for the further higher power operation of high-frequency HEMTs.

AlGaN channel HEMTs on AlN buffer layer with sufficiently low off-state drain leakage current

“…In the AlGaN channel HEMTs initially reported [9], however, we could not realise breakdown voltage enhancement because of the relatively high off-state drain leakage current. To extract the attractive potential of the higher electric breakdown field in the AlGaN and enhance the breakdown voltage of HEMTs, it is important to suppress this off-state drain leakage current.In this Letter we report growing an Al 0.39 Ga 0.61 N/Al 0.16 Ga 0.84 N epitaxial layer on the AlN buffer layer instead of the GaN buffer layer [9]. In the HEMTs fabricated on this epitaxial layer the off-state drain leakage current was sufficiently suppressed.…”

Formation of AlGaN and GaN epitaxial layer with high p-carrier concentration by pulse supply of source gases

“…On the AlGaN, GaN:Mg or AlGaN:Mg (0.16 μm) is grown in the pulse growth mode. Figure 3 shows the experimental results for the pulse and continuous modes of source gas supply for GaN, in which the temperature dependences of the hole concentration for each mode are shown with the hole mobility as a function of the Mg dopant of (Cp) 2 Mg in the case of the pulse supply mode. The carrier concentration is measured using a conventional Van der Pauw Hall measurement system and a temperature-dependence measurement system.…”

Si implant-assisted Ohmic contacts to GaN

“…The problem with ion implantation is that the GaN has to be annealed at temperatures above which the rate of the preferential evaporation [10] becomes significant, which creates hexagonal etch pits. Researchers have attempted to deal with this problem by annealing for a short time (30 s) in flowing NH 3 [6,7], placing the samples face-to-face [2,3], under high N 2 pressures [4,5], or with sputtered Si 3 N 4 [8,9] or AlN [4,5]. However, face-to-face capping does not hermetically seal the surface, the partial pressure of N 2 greatly exceeds the N 2 over pressures that were used [10], and sputtered Si 3 N 4 [11] or AlN [12] does not stick very well and is relatively weak so it can be punctured by the large N 2 vapor pressures at the annealing temperatures that were used.…”