Temperature impact and analytical modeling of the AlGaN/GaN-on-Si saturation drain current and transconductance

Abstract

The saturation drain current and the gate saturation transconductance for AlGaN/GaN on silicon (1 1 1) high-electron mobility transistors (HEMTs) have been experimentally investigated in the temperature range of 25-300 • C. An analytical physical-based closed-form is proposed for modeling the gate transconductance taking into account the polar-optical phonon scattering of the electrons in the two-dimensional electron gas HEMT channel. It is suggested that the experimental temperature dependence of T −1.1 is due to the electron channel mobility dependence coupled with the effect of the access resistances and the channel self-heating due to power dissipation.

Modeling the effect of thin gate insulators (SiO2, SiN, Al2O3 and HfO2) on AlGaN/GaN HEMT forward characteristics grown on Si, sapphire and SiC

“…The simulated large channel HEMT (based on the experimental results of Ref. [22]) has a conventional layout of 4 mm, 4 mm and 5 mm for L gs , L and L gd , respectively. Insulator thickness is varied in the range of 0-20 nm for SiO 2 , SiN, Al 2 O 3 and HfO 2 .…”

“…4(b) are from the experimental data of ref. [22]). However, the introduction of series resistance is still unable to explain the maximum transconductance experimentally observed (the g m vs V gs curve usually has a parabolic-like curve with a maximum).…”

“…[22]. We have included this experimental curve to benchmark the reference Schottky HEMT g m , assuming the conservation of the electron mobility and the polarization fields when it is moved to a MIS-HEMT approach.…”

“…We have included this experimental curve to benchmark the reference Schottky HEMT g m , assuming the conservation of the electron mobility and the polarization fields when it is moved to a MIS-HEMT approach. From this fit, (the parameters and fabrication details of which are well reported in [22]), we can theorize as to what the effect would be of varying the gate insulator and the substrate. The accuracy of the most simplified approach may be questionable in some cases, (we have discussed briefly before the n s , 0 variation due to the passivation of surface states reported by several authors), but it still could bring a qualitative insight into the effect of different insulators on the on-state of the HEMT.…”

Contact engineering of GaN-on-silicon power devices for breakdown voltage enhancement

“…Also, the growth of the high-quality GaN epitaxial layer on a large-scale silicon substrate has shown substantial progress in recent years, which allows the production of GaN-based devices taking advantage of modern silicon technology with low cost. Recent advances have demonstrated high-performance GaN-based power devices realized on the silicon substrate, while the V BK of these devices are still below the theoretical limit of GaN [2][3][4][5] and could further be improved. It has been reported 1 Author to whom any correspondence should be addressed.…”