Effect of Schottky barrier height on quantitative analysis of deep-levels in n-type GaN by deep-level transient spectroscopy

Keito Aoshima,Masahiro Horita,Jun Suda

doi:10.1063/5.0073747

Keito Aoshima, Masahiro Horita + Show 1 more

Open Access

https://doi.org/10.1063/5.0073747

Copy DOI

Journal: AIP Advances	Publication Date: Nov 1, 2021
Citations: 3	License type: cc-by

Affiliation: Nagoya University

Abstract

Deep-level transient spectroscopy (DLTS) using Schottky barrier diodes (SBDs) is widely used for quantitative analysis of deep levels. This study focuses on the dependence of Schottky barrier height on apparent time constants and concentrations of electron traps in n-type GaN. DLTS using SBDs with various barrier heights was carried out. Experimental data show that large reverse leakage currents due to low barrier heights resulted in underestimation of time constants and concentrations. Theoretical calculations considering the impact of leakage currents reproduced experimental results well. Based on the calculations, we suggest a minimum required barrier height where accurate time constants and concentrations can be evaluated.

Highlights

The impact of barrier height on trap concentration evaluation is widely known, and systematic investigations through quantitative analysis were reported for GaAs3 and SiC.4 Ma et al demonstrated that the observed concentration of the energetically deep EL2 defect with an energy level of 0.83 eV below the bottom of the conduction band (EC − 0.83 eV) in GaAs is sensitive to barrier height.3 Reshanov et al showed change in peak heights of defects EH6 (EC − 1.39 eV) and EH7 (EC − 1.53 eV) in 4H–SiC by using different metals for Schottky electrodes.4 A large reverse leakage current due to a low barrier height affects evaluation of an emission time constant through Deep-level transient spectroscopy (DLTS) measurement5–7 because this current induces an electron capture process under reverse bias voltage
We have reported that three electron traps (G2a, G2b, and G2c) at around 1 eV from the conduction band minimum are formed in n-type GaN by gamma-ray irradiation
In the case of PR ∼ 0 at x < WR − λ, under the reverse bias voltage, the accurate trap concentration can be evaluated by DLTS

Summary

Introduction

The impact of barrier height on trap concentration evaluation is widely known, and systematic investigations through quantitative analysis were reported for GaAs3 and SiC.4 Ma et al demonstrated that the observed concentration of the energetically deep EL2 defect with an energy level of 0.83 eV below the bottom of the conduction band (EC − 0.83 eV) in GaAs is sensitive to barrier height.3 Reshanov et al showed change in peak heights of defects EH6 (EC − 1.39 eV) and EH7 (EC − 1.53 eV) in 4H–SiC by using different metals for Schottky electrodes.4 A large reverse leakage current due to a low barrier height affects evaluation of an emission time constant through DLTS measurement5–7 because this current induces an electron capture process under reverse bias voltage (during measurement). To obtain accurate energy levels, capture cross sections and scitation.org/journal/adv concentrations of each electron trap level, isothermal capacitance transient spectroscopy (ICTS)15 was performed using gamma-ray irradiated GaN p+-n diodes in which reverse leakage currents were negligible. We discussed the dependence of Schottky barrier height on apparent emission time constants and concentrations of the traps quantitatively.

Results

Conclusion