多晶硅还原炉内化学反应与流动耦合机制的数值研究

Numerical Study on Coupling Mechanism of Chemical Reaction and Flow in Polysilicon Reduction Furnace

摘要

改良西门子法是多晶硅工业生产的主流技术，还原炉作为核心反应装置，其内部硅烷或三氯氢硅的热分解与化学气相沉积反应，与流体流动、传热传质过程存在强烈耦合效应，直接决定多晶硅沉积速率、产品纯度及能耗水平。由于还原炉内耦合过程涉及多物理场相互作用，实验测量难以全面揭示其内在规律，数值模拟成为研究该耦合机制的有效手段。本文系统分析多晶硅还原炉内化学反应与流动的耦合特性，梳理数值模拟的基础理论与关键模型，深入探讨流速分布、温度场与反应速率的核心耦合机制，总结模拟方法的优化策略，结合实践案例验证模拟的有效性，最后展望未来发展方向，为还原炉结构优化与工艺参数调控提供理论支撑。

Abstract

The modified Siemens process remains the dominant technology in polycrystalline silicon production. As the core reactor, the reduction furnace exhibits strong coupling effects between the thermal decomposition of silane or trichlorosilane and chemical vapor deposition reactions, fluid dynamics, and heat/mass transfer processes. These interactions directly determine the deposition rate, product purity, and energy consumption. Given the complex multiphysics interactions within the furnace, experimental measurements cannot fully reveal the underlying mechanisms, making numerical simulation an essential tool for studying this coupling. This paper systematically analyzes the coupling characteristics between chemical reactions and fluid flow in polycrystalline silicon reduction furnaces. It reviews fundamental theories and key models for numerical simulation, explores core coupling mechanisms involving flow velocity distribution, temperature fields, and reaction rates, summarizes optimization strategies for simulation methods, validates simulation effectiveness through practical case studies, and ultimately outlines future research directions to provide theoretical support for furnace structure optimization and process parameter control.

关键词

多晶硅还原炉 / 化学反应 / 流动耦合 / 数值模拟 / 传质传热 / 沉积特性

Key words

Polysilicon reduction furnace / Chemical reaction / Flow coupling / Numerical simulation / Mass and heat transfer / Deposition characteristics

参考文献

[1] 赵程文轩, 徐燕燕, 侯雨, 等. “双碳”目标下多晶硅生产节能优化研究[J]. 陕西煤炭, 2026, 45 (01): 157-162. DOI:10.20120/j.cnki.issn.1671-749x.2026.0129.

[2] 王俊华, 王阳, 赵云浩, 等. 电子级芯片用多晶还原工序质量管控要点[J]. 中国集成电路, 2025, 34 (05): 80-84.

[3] 石何武, 张升学, 杨永亮. 多晶硅还原生产系统节能提质探讨[J]. 绿色矿冶, 2025, 41 (01): 55-59+70. DOI:10.19610/j.cnki.cn10-1873/tf.2025.01.009.