Multiscale coupled plasma–feature simulation for 3D profile evolution in Si etching under HBr/Cl2 mixture plasmas

Coupling reactor-scale plasma simulations with device-scale three-dimensional (3D) feature evolution models enables predictive profile simulation under realistic plasma boundary conditions. This coupling is essential for accurate analysis and mechanistic interpretation of etched feature evolution during plasma etching processes. However, quantitatively validated multiscale simulations that reliably couple bulk plasma behavior with surface reaction databases for realistic 3D profile prediction remain limited. This study proposes a multiscale simulation framework that integrates reactor-scale plasma predictions with a device-scale 3D profile evolution platform to investigate crystalline Si etching under HBr/Cl2 mixture plasma conditions at varying gas mixing ratios. Reactor-scale simulations predict plasma properties in a virtual sheath regime, providing the characteristics of ion and neutral particles impinging on 3D surfaces. These characteristics are incorporated into a 3D feature profile simulator that computes real-time shape evolution using an effective 3D mesh, Compute Unified Device Architecture-accelerated ballistic transport, and an OpenMP-parallelized surface reaction model. The plasma–surface reaction set estimates etch yields and surface reaction coverages as functions of the HBr/Cl2 mixing ratio and ion energy over a wide range of plasma process conditions. Realistic transport mechanisms include incident ion and neutral distributions from the bulk plasma, reflected ions, and re-emitted neutrals from feature sidewalls. The predicted profile evolution agrees well with the experimental results. Additionally, the model provides quantitative information on the etch rate, ion flux, ion energy, radical fluxes, and reaction coverages, enabling mechanistic interpretation of etching characteristics in HBr/Cl2 mixture plasma etching of crystalline Si.