In recent years, Gallium Nitride (GaN)-based semiconductor devices have attracted significant attention in the field of power electronics due to their intrinsic advantages over traditional semiconductor materials such as silicon. In particular, GaN High Electron Mobility Transistors (HEMTs) offer higher channel conductivity, high breakdown voltage, and low on-resistance R_DSON, making them ideal for high-frequency applications and high-performance power converters. However, the full integration of GaN devices into industrial systems requires a thorough understanding of reliability issues and the mechanisms that influence long-term device performance. A critical aspect in the development of GaN devices is the choice of substrate. Commonly used substrates, such as silicon (Si), silicon carbide (SiC), and sapphire exhibit distinct properties that affect lattice quality, charge distribution, and parasitic trapping phenomena. Notably, buffer-free technology enables the growth of GaN on SiC with reduced parasitic trapping sites, leading to improved 2DEG (Two-Dimensional Electron Gas) uniformity. In contrast, GaN-on-Si devices experience more pronounced parameter degradation due to strong capacitive coupling with the conductive substrate, while GaN-on-Sapphire combines a highly insulating substrate with trapping dynamics comparable to GaN-on-Si. Another important factor is the behavior of devices under vertical bias, known as back-gating effects. Studies on ohmic structures and HEMTs have revealed internal charge redistribution, which can be modeled using capacitive-resistive networks according to the Maxwell-Wagner approach. This modeling allows accurate prediction of the free charge distribution in the 2DEG over time under different back-gating conditions, highlighting the influence of substrate quality and lattice matching on device performance. Beyond understanding the underlying physics, characterizing trapping sites has been addressed through innovative techniques, such as analyzing the capacitive coupling between field plates and the 2DEG. This approach enables spatial localization of active traps and the development of mathematical models describing 2DEG retraction under various off-state bias conditions. Finally, the impact of device instabilities on power converters has been explored through “on-site” measurement techniques capable of monitoring on-resistance variations during actual device operation. This method allows observation of self-heating effects and long-term degradation without interrupting device operation, providing a valuable tool for reliability testing of HEMTs and advanced power devices. In summary, this dissertation aims to investigate degradation mechanisms and trapping phenomena in GaN power HEMTs by comparing different substrates and device technologies, developing accurate physical and mathematical models, and introducing innovative experimental methods for characterizing and monitoring device reliability under realistic operating conditions.

Reliability of GaN-HEMTs for Power Conversion on Alternative Substrates

CAVALIERE, ALBERTO
2026

Abstract

In recent years, Gallium Nitride (GaN)-based semiconductor devices have attracted significant attention in the field of power electronics due to their intrinsic advantages over traditional semiconductor materials such as silicon. In particular, GaN High Electron Mobility Transistors (HEMTs) offer higher channel conductivity, high breakdown voltage, and low on-resistance R_DSON, making them ideal for high-frequency applications and high-performance power converters. However, the full integration of GaN devices into industrial systems requires a thorough understanding of reliability issues and the mechanisms that influence long-term device performance. A critical aspect in the development of GaN devices is the choice of substrate. Commonly used substrates, such as silicon (Si), silicon carbide (SiC), and sapphire exhibit distinct properties that affect lattice quality, charge distribution, and parasitic trapping phenomena. Notably, buffer-free technology enables the growth of GaN on SiC with reduced parasitic trapping sites, leading to improved 2DEG (Two-Dimensional Electron Gas) uniformity. In contrast, GaN-on-Si devices experience more pronounced parameter degradation due to strong capacitive coupling with the conductive substrate, while GaN-on-Sapphire combines a highly insulating substrate with trapping dynamics comparable to GaN-on-Si. Another important factor is the behavior of devices under vertical bias, known as back-gating effects. Studies on ohmic structures and HEMTs have revealed internal charge redistribution, which can be modeled using capacitive-resistive networks according to the Maxwell-Wagner approach. This modeling allows accurate prediction of the free charge distribution in the 2DEG over time under different back-gating conditions, highlighting the influence of substrate quality and lattice matching on device performance. Beyond understanding the underlying physics, characterizing trapping sites has been addressed through innovative techniques, such as analyzing the capacitive coupling between field plates and the 2DEG. This approach enables spatial localization of active traps and the development of mathematical models describing 2DEG retraction under various off-state bias conditions. Finally, the impact of device instabilities on power converters has been explored through “on-site” measurement techniques capable of monitoring on-resistance variations during actual device operation. This method allows observation of self-heating effects and long-term degradation without interrupting device operation, providing a valuable tool for reliability testing of HEMTs and advanced power devices. In summary, this dissertation aims to investigate degradation mechanisms and trapping phenomena in GaN power HEMTs by comparing different substrates and device technologies, developing accurate physical and mathematical models, and introducing innovative experimental methods for characterizing and monitoring device reliability under realistic operating conditions.
9-mar-2026
Inglese
MENEGHINI, MATTEO
Università degli studi di Padova
File in questo prodotto:
File Dimensione Formato  
tesi_definitiva_Alberto_Cavaliere.pdf

embargo fino al 08/03/2029

Licenza: Tutti i diritti riservati
Dimensione 12.2 MB
Formato Adobe PDF
12.2 MB Adobe PDF

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/362020
Il codice NBN di questa tesi è URN:NBN:IT:UNIPD-362020