材料科学
MOSFET
各向异性
电场
光电子学
热导率
功勋
击穿电压
电压
晶体管
电子迁移率
阈值电压
消散
热的
凝聚态物理
电阻率和电导率
兴奋剂
场效应晶体管
功率半导体器件
宽禁带半导体
频道(广播)
电流密度
电子工程
电气工程
功率密度
功率MOSFET
作者
Yinfei Xie,Yang He,Zhengyue Li,Yongze Xu,Weiye Liu,Huaixin Guo,Huarui Sun
标识
DOI:10.1109/ted.2025.3625537
摘要
The high critical electric field strength and ultra wide bandgap of $\beta $ -Ga2O3 make it promising for power electronics, but the low thermal conductivity necessitates extensive focus on thermal management. This study focuses on enhancing the electrical performance of $\beta $ -Ga2O3 metal–oxide–semiconductor field-effect transistors (MOSFETs) with optimized heat dissipation using a 3-D electrothermal model. We take into account the anisotropic mobility and thermal conductivity, channel orientation, and delta-doping process for electrical optimization. Our results indicate that thermal conductivity anisotropy significantly impacts the electrical characteristics more than mobility anisotropy does. Simulations of MOSFETs with various channel orientations reveal that $\beta $ -Ga2O3 with a 45° (010)-oriented channel exhibits superior electrical performance due to the highest thermal conductivity. The investigation of delta-doping effects shows that higher charge density increases the breakdown voltage by broadening the depletion region, with an optimal value of $5\times 10^{{13}}$ cm ${}^{-2}$ . Excessive doping density, however, risks local breakdown. Positioning the delta-doping layer near the channel interface homogenizes the electric field and enhances the breakdown voltage, but excessive proximity can be harmful. Our optimized $\beta $ -Ga2O3 MOSFET design achieves a high off-state breakdown voltage ( $V$ ${}_{\text {br}} =2284$ V) and a high-power figure of merit (PFOM = 369.5 MV/cm2) at a low specific on-resistance ( $R$ ${}_{\textsc{on}, \text{sp}} =14.1$ m $\Omega ~\cdot $ cm2). This work highlights the necessity for synergistic design practices in $\beta $ -Ga2O3 MOSFETs, which integrates both electrical and thermal considerations.
科研通智能强力驱动
Strongly Powered by AbleSci AI