colabfit/Si_JCP_2017

Si JCP 2017数据集包含四种相态下的64个硅原子的配置：钻石立方相、锡的β相、R8相和液态相。在固体相中，分子动力学模拟的温度为300K、600K和900K；液态相的模拟温度高达2500K。所有松弛过程均在零压力下进行。通过随机扭曲晶体结构准备了额外的配置。使用VASP软件，采用PAW伪势能和PBE交换相关作用。k点网格针对能量收敛0.5 meV/atom和应力收敛0.1 kbar进行了优化。平面波能量截止设置为300 eV。为了减少数据点之间的相关性，分子动力学数据通过在300K时每100个连续结构取一个，以及从较高温度的分子动力学模拟中每20个结构取一个进行了筛选。数据集中的额外详细信息存储在以"dataset_"为前缀的列中。

The Si JCP 2017 dataset contains configurations of 64 silicon atoms in four phases: cubic-diamond, (beta)-tin, R8, and liquid. MD simulations are performed at 300K, 600K, and 900K for solid phases; up to 2500K for the L phase. All relaxations are carried out at zero pressure. Additional configurations are prepared by random distortion of crystal structures. VASP is used with a PAW pseudopotential and PBE exchange correlation. The k-point mesh is optimized for energy convergence of 0.5 meV/atom and stress convergence of 0.1 kbar. The plane wave energy cutoff is set to 300 eV. To reduce correlation between data points, MD data are thinned by selecting one out of every 100 consecutive structures from the MD simulations at 300K, and one out of every 20 structures from higher temperature MD simulations. Additional details in the dataset are stored in columns prefixed with "dataset_".