研究目的
To perform a comparative analysis of the impact of the Si3N4 cap versus the GaN cap on the performance of GET-SBDs.
研究成果
Si3N4 cap devices exhibit higher breakdown voltage, lower RON-dispersion, longer lifetime, and lower time-to-breakdown variability compared to GaN cap devices, making them more suitable for high-power and high-temperature applications. The electric field distribution and material properties contribute to the observed differences in reliability.
研究不足
The study is limited to specific cap layer thicknesses (5 nm Si3N4 and 3 nm GaN) and device structures. The influence of buffer leakage and material properties may not be fully isolated. Further investigation is needed to clarify correlations between electric field and breakdown mechanisms.
1：Experimental Design and Method Selection:
The study compares two types of AlGaN/GaN Schottky barrier diodes with gated edge termination (GET-SBDs) fabricated with different cap layers (Si3N4 or GaN). Methods include DC and pulse measurements, temperature-dependent leakage current analysis, time-dependent dielectric breakdown (TDDB) tests, and TCAD simulations to understand electric field distribution and reliability.
2：Sample Selection and Data Sources:
Two wafers with identical epitaxial stacks but different cap layers (5-nm Si3N4 or 3-nm GaN) were processed. Devices were fabricated on Si (111) substrates using MOCVD growth. Data were collected from measurements on these devices and ring capacitors.
3：List of Experimental Equipment and Materials:
Equipment includes high-temperature oxidation (HTO) for SiO2 deposition, atomic layer etching for barrier recessing, plasma-enhanced atomic layer deposition (PEALD) for Si3N4, and metal deposition for contacts. Materials include AlN, AlGaN, GaN, SiO2, Si3N4, TiN-based metal stack, and Ti/Al-based metal stack.
4：Experimental Procedures and Operational Workflow:
Fabrication involved depositing passivation layers, etching, depositing GET dielectric, forming Schottky and ohmic contacts. Characterization included DC reverse sweeps, pulsed IV measurements, Arrhenius plots for activation energy, TDDB tests with constant voltage stress, and simulations for electric field profiles.
5：Data Analysis Methods:
Data were analyzed using Weibull distributions for TDDB, linear fits for activation energies, and TCAD simulations for electric field analysis. Statistical techniques and software tools for simulation and fitting were employed.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
