硅
杂质
扩散
热扩散率
氮气
化学
化学物理
半导体
氧气
Crystal(编程语言)
碳纤维
热力学
工作(物理)
材料科学
物理
光电子学
复合材料
有机化学
复合数
程序设计语言
计算机科学
作者
Jinping Luo,Chenyang Zhou,Qihang Li,Yunjie Cheng,Lijun Liu
标识
DOI:10.1016/j.jcrysgro.2021.126476
摘要
The carbon, oxygen and nitrogen are very important light impurities affecting the quality of silicon crystal used in photovoltaic and semiconductor industry. The diffusion coefficients are key parameters in determining the impurity distribution in the melt and crystal. Unfortunately, these parameters are rarely measured by experiments due to challenges of high temperature and complex flow in the melt during crystal growth. Nonetheless, the liquid diffusion coefficients exist as necessary input parameters in the continuum modelling work of mass transport. Here, we first reviewed the liquid state diffusion coefficients used in the continuum models and found out that they are far from consistency. Then, molecular dynamics simulations are employed to calculate the diffusion coefficients and migration barriers of the carbon, nitrogen and oxygen impurities in silicon melt based on various empirical interatomic potentials. The nitrogen diffusivity in silicon melt is reported for the first time. The calculated diffusion coefficients are generally lower than the values used in most continuum models although consistent across different potentials. This discrepancy is discussed in detail.
科研通智能强力驱动
Strongly Powered by AbleSci AI