DFT dataset for "Symmetry Breaking in the Superionic Phase of Silver-Iodide"

This dataset for the bulk AgI system, linked to the publication "Symmetry Breaking in the Superionic Phase of Silver-Iodide" [Phys. Rev. Lett. 134, 026306 (2025)], includes calculations performed using Quantum ESPRESSO with the PBE-D3 functional. It explores temperatures up to 600 K (800 K in some cases) at selective pressures through active learning. The data is categorized into training and testing sets, with configurations derived from MD trajectories of different phases (wurtzite, zincblende, rocksalt, etc.). This data is used for generating a machine learning force-field which is used for large scale simulations of the superionic transition in silver-iodide. Additionally, short ab initio trajectories are provided, integrated using the Verlet algorithm, which are used for validation of the machine learning force-field. The checkpoints for the committee of the machine learning force field are available at: https://gitlab.com/amirhajibabaei/bulkagi本数据集针对体相碘化银（bulk AgI）体系，关联于发表于《物理评论快报》（Phys. Rev. Lett.）134卷第026306号（2025年）的论文《碘化银超离子相中的对称性破缺》。数据集采用搭载PBE-D3泛函的Quantum ESPRESSO程序完成计算，并通过主动学习方法，在特定压强条件下探索了最高600 K的温度区间（部分案例中最高可达800 K）。数据被划分为训练集与测试集，其构型源自纤锌矿、闪锌矿、岩盐矿等不同晶相的分子动力学（Molecular Dynamics, MD）轨迹。本数据集可用于构建机器学习力场，以实现碘化银超离子转变的大规模模拟。此外，数据集还附带了采用韦尔莱算法（Verlet Algorithm）积分得到的短从头算轨迹，用于对所构建的机器学习力场进行验证。该机器学习力场委员会的检查点文件可在以下链接获取：https://gitlab.com/amirhajibabaei/bulkagi