Numerical simulation of granular silicon growth and silicon fines formation process in polysilicon fluidized bed

Abstract

Operating conditions strongly affect the yield and quality of polysilicon in a polysilicon fluidized bed. In this study, a new model of polysilicon fluidized bed was established using the Euler–Euler model coupled with population balance model (PBM), which was combined with fluid flow, heat, and mass transfer models, while considering the scavenging effect of silicon fines. The effects of different operating conditions on the deposition and formation rates of silicon fines were investigated. Results show that the model can correctly describe the particle growth process in the fluidized bed of polysilicon. The silicon fines and the interphase velocity difference show “N”- and “M”-shaped distributions along the axial direction, respectively. The particle temperature and concentration near the wall are higher than those in the central region. The decomposition of silane in the bottom region of the bed is dominated by heterogeneous deposition. The scavenging of silicon fines occurs in the dilute-phase region. The effects of operating conditions, i.e. inlet gas temperature, silane composition, and gas velocity, on the reactor performance were also explored comprehensively. Increasing the inlet gas composition and velocity enhances the formation rates of solid silicon and fines. Increasing the inlet gas temperature promotes the growth of solid silicon and inhibits the formation of silicon fines. High fluidization ratio, low inlet silane concentration, and high inlet gas temperature enhance the selectivity of silicon growth.