A simple method for detection and quantitative estimation of deep levels in a barrier layer of AlGaN/GaN HEMT structures by analysis of light induced threshold voltage shift

Maciej Matys, Atsushi Yamada, Yoichi Kamada, Toshihiro Ohki · 2025

Read A simple method for detection and quantitative estimation of deep levels in a barrier layer of AlGaN/GaN HEMT structures by analysis of light induced threshold voltage shift by Maciej Matys, Atsushi Yamada, Yoichi Kamada, Toshihiro Ohki (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-08625.txt.

Introduction

Maciej Matys, Atsushi Yamada, Yoichi Kamada, and Toshihiro Ohki offer a simple way to find traps in transistor layers. Their paper, from twenty twenty-five, deals with gallium nitride power devices. Deep traps are flaws that hold charge and hurt how the device switches. Finding and counting them is a key task for makers.

Their method shines light below the band gap on the device. The light frees electrons from the traps and leaves holes behind. A shift in the switch-on voltage then reveals the traps. The paper gives a clear test for whether holes were truly made. Trials on devices grown on two base materials confirm the method works.

Catalog record

CREATORMaciej Matys, Atsushi Yamada, Yoichi Kamada, Toshihiro Ohki
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-08625

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A simple method for detection and quantitative estimation of deep levels in a barrier layer of AlGaN/GaN HEMT structures by analysis of light induced threshold voltage shift Preprint: AIP/123-QED Maciej Matys, Atsushi Yamada, Yoichi Kamada, Toshihiro Ohki Abstract The characterization of deep levels in AlGaN/GaN heterostructures is one of the most important problems in GaN high electron mobility transistors (HEMTs) technology. This work reports on a technique for determination of deep level concentration in AlGaN/GaN HEMT structures. The proposed method is relatively simple, and it is based on the detection of free holes created by optically induced transitions of electrons from the deep levels to the conduction band. The developed method can detect and provide quantitative estimation of deep level traps in a barrier layer of AlGaN/GaN HEMT structures. Furthermore, it provides a framework for analysis of light induced threshold voltage shift, which includes an important experimental criterion of determination whether the holes are generated or not in the AlGaN/GaN HEMT structures by sub-band gap illumination. The method was verified by applications it to a study of the deep levels in GaN HEMTs grown on various substrates, i.e. SiC and GaN. I Introduction Gallium nitride (GaN) high-electron mobility transistors (HEMTs) demonstrate outstanding performance in next-generation high-voltage and power applications KA ; KB ; K1 ; K2 ; K3 ; K4 ; K6a ; K6a1 ; K6b ; K6c . However, despite the impressive progress which was made in improvement of performance of these devices, the reliability issues related to the charge trapping phenomena are still a challenging problem K5 . The typical effect of charge trapping phenomena is called ”current collapse”, i.e. a decrease of the drain current under the large drain voltage operation K6 ; K7 . According to several studies, this undesirable effect is caused by the electron trapping at the deep levels located in the buffer layers or/and at the surface K8 ; K9 ; K10 ; K11 ; K12 . Overall, the deep levels in GaN based HEMTs are present due to epitaxial growth of the HEMT structure K6 ; K7 ; K8 ; K13 ; K14 and process condition (for example surface etching K15 ; K16 ; K17 ) or they can be induced by electrical stress K6 . In order to control the deep levels in GaN based HEMTs, the characterization of their properties is one of the fundamental aspects in the study of the GaN HEMTs. For quantitative analysis of the deep levels in GaN based devices the deep-level transient spectroscopy (DLTS) K17a and deep-level optical spectroscopy (DLOS) K17b are mainly applied K18 . These methods can provide information about the basic properties of the deep levels like their concentration, activation energy or capture cross sections. However, they have one important limitation, namely they cannot clearly indicate the spatial location of deep levels in the AlGaN/GaN HEMT structure K6 . Figure 1: Band diagram of the semiconductor device under reverse bias polarization with marked two trap levels A and B. Inset shows the calculated emission time of an electron from a trap with energy E_{T} at RT in the case of Al 0.3 Ga 0.7 N. In this paper, we present a simple method which can provide both ( I ) quantitative estimation of deep-levels and ( II ) precise information on their spatial location in the GaN HEMT structure. The key point of the proposed method is detection of free holes created by optically induced transitions of electrons from the deep levels to the conduction band (CB). We verified the developed technique by applications it to a study of the deep levels in GaN HEMTs grown on various substrates, i.e. SiC and GaN. The paper is organized as follows. In Section II, we describe the theoretical backgrounds of the proposed method. In Section III, we present the experimental results on deep levels obtained from GaN HEMTs on SiC and GaN substrates and in Section IV, we summarize the key points of this paper. II Theory and background II.1 Definition of deep level At first, for the future discussion, we explain on what basis to determine a defect state in the band gap as a deep one in a semiconductor device. Fig. 1(a) shows schematically the band diagram of an electronic device under the reverse bias polarization with marked two traps with energy levels A and B. For the considered reverse bias, we assumed that the position of the Fermi level ( E_{F} ) is below the trap level A (see Fig. 1(a)). The trap A is located at the larger energy distance much further from the conduction band (CB) compared to the trap B. Due to this fact, trap A is fully occupied by electrons while the trap B is empty from electrons. In other words, shifting Ef by the reverse bias below the energy level of trap A will not lead to electron emission from trap A in reasonable time in contrast to trap B for which emission electrons occurs almost immediately after shifting E_{F} by the gate bias ( V_{G} ). Due to this difference, we can call the trap A as ”frozen” or ”deep”. Fig. 1(b) shows the calculated emission time of electron ( \tau ) from a trap with energy E_{T} at room temperature (RT) in the case of the AlGaN material, which is the main subject of this paper. The \tau was calculated from the Shockley–Read–Hall (SRH) statistics according to the following equation K19 : \tau=\frac{1}{v_{th}N_{C}\sigma}\exp(\frac{E_{T}}{kT}) (1) where v_{th}=10{{}^{6}} cm/s, N_{C}=10^{19} cm -3 , \sigma=10^{-14} cm -2 and T are the electron thermal velocity, the density of states at CB, the capture cross section and temperature, respectively K20 . From Fig. 1(b) one can note that for the trap located at 1 eV \tau is approximately 10^{5} s ( \approx 28 hours). Thus, for the AlGaN material (and also GaN) it can be safely assumed that every trap located deeper that 1 eV is ”frozen” at RT, i.e. it does not change the electronic state under shifting E_{F} . This is important point to understand the shift of the threshold voltage ( V_{th} ) of the AlGaN/GaN Schottky barrier diode (SBD) upon light illumination, which will be discussed in the next section. II.2 Light induced V_{th} shift of AlGaN/GaN SBD and hole emission Figure 2: (a) Band diagram of AlGaN/GaN SBD polarized under V_{G1}<V_{th} in the case when only electron emission from the deep-levels E_{1} and E_{2} takes place. (b) Schematic illustration of light induced V_{th} shift and (c) expected dependencies of \Delta V_{th} with V_{GL} when only electron emission from E_{1} and E_{2} occurs. Figure 3: (a) Band diagram of AlGaN/GaN SBD polarized under V_{G1}<V_{th} in the case when both electron and hole emission from the deep-levels E_{1} and E_{2} takes place. (b) Schematic illustration of light induced V_{th} shift and (c) expected dependencies of \Delta V_{th} with V_{GL} when both electron and hole emission via E_{1} and E_{2} levels occurs. Let us consider now the AlGaN/GaN SBD diode polarized under the gate bias ( V_{G1} ) below V_{th} whose band diagram is shown in Fig. 2(a). We assumed that both AlGaN and GaN layers contain some deep levels with energy E_{1} and E_{2} respectively. The

Read the complete text