Neural Network-Driven Molecular Insights into Alkaline Wet Etching of GaN: Toward Atomistic Precision in Nanostructure Fabrication

We present a Behler-Parrinello-type neural network potential (NNP) for a complex quinary system consisting of Gallium (Ga), Nitrogen (N), Potassium (K), and Oxygen (O). This highly accurate NNP was trained on a diverse dataset to reliably model interactions and dynamics in this multi-component material system, which is crucial for applications in GaN-KOH chemical wet etching.本工作提出了一种适用于由镓（Gallium, Ga）、氮（Nitrogen, N）、钾（Potassium, K）和氧（Oxygen, O）组成的复杂五元体系的贝勒-帕里内洛型神经网络势（NNP）。该高精度神经网络势基于多样化数据集训练得到，可可靠建模该多组分材料体系中的相互作用与动力学过程，其对于氮化镓-氢氧化钾（GaN-KOH）化学湿法刻蚀领域的应用至关重要。