Three-Dimensional Cellular Automaton Modeling of Annealed Silicon Thin Film Surface Morphology

Controlling the surface morphology of polycrystalline silicon (Si) thin films is essential for enhancing the performance of thin-film transistor (TFT) devices. This study introduces a novel three-dimensional cellular automaton (CA) algorithm for simulating the surface morphology evolution of Si thin films during crystallization. Unlike conventional CA models, this algorithm enforces strict mass conservation and dynamically adjusts the liquid surface to accurately account for density variations between the liquid and crystalline silicon phases. The model's accuracy was validated through comparisons with experimental data, and molecular dynamics simulations were used to investigate the atomic-scale mechanisms driving surface protrusion formation. Key parameters affecting surface morphology were examined, including seed crystal distribution, shape, and initial orientation, nonuniform linear temperature fields, and liquid-to-solid density ratios. Protrusion height was significantly affected by the solid-liquid interface orientation (increasing with near-perpendicular alignment), growth site spacing (following a logarithmic relationship), and hydrogen (H) doping (which reduced the protrusion via density modulation). These findings provide insights for controlling surface morphology and offer a pathway to optimize the fabrication process of TFT devices.