随时间变化的栅氧化层击穿
介电强度
材料科学
电介质
电场
量子隧道
凝聚态物理
泊松方程
泄漏(经济)
电压
机械
光电子学
电气工程
栅极电介质
物理
宏观经济学
工程类
经济
晶体管
量子力学
作者
Qingqing Zhang,Lishuai Yu,Zhengpin Bian,Dong Yuan,Hailing Sun,Biao Tang,Xubing Lu,Feilong Liu,Guofu Zhou
出处
期刊:Physical review applied
[American Physical Society]
日期:2023-02-02
卷期号:19 (2)
被引量:2
标识
DOI:10.1103/physrevapplied.19.024008
摘要
Time-dependent dielectric breakdown (TDDB) is a crucial issue for the dielectric reliability. In this work, we present a full three-dimensional mechanistic model for calculation of the TDDB process in polycrystalline thin films. The model is based on the multiphonon trap-assisted tunneling theory and takes into account the intrinsic three-dimensional discreteness of traps at the dielectric grain boundaries. The leakage current density is calculated by solving coupled three-dimensional master equation and Poisson equation. The net phonon emission associated with each charge trapping and release event is treated as a local point heat source, which then enters the Fourier heat equation for three-dimensional temperature distribution calculation. The generated trap is determined by local temperature and electric field, which is subsequently included in the next round of calculation of electric and thermal properties. A positive feedback loop gradually leads to an increase of trap density, temperature, and leakage current density, and finally the dielectric breakdown. Our model can, to a good approximation, reproduce the experimental leakage current density-voltage characteristics and the Weibull distribution of time to breakdown at different dielectric thicknesses, stress voltages, and environmental temperatures. We find that in realistic devices, the three-dimensional trap-to-trap transport of electrons contributes a non-negligible part to the leakage current when the dielectric approaches breakdown. Our approach of three-dimensional mechanistic simulation is computationally efficient such that evolution of ${10}^{3}$ traps during the TDDB process can be easily performed on a standard desktop computer.
科研通智能强力驱动
Strongly Powered by AbleSci AI