Big graphene dataset
收藏资源简介:
Density functional theory simulations of graphite structures with random (in size and location) structural defects. Dataset contains coordinates of carbon atoms and the total energy computed using the PBE functional as implemented in the software VASP. Input parameter files are also included so that the calculations can be reproduced. NOTE: This dataset is useful for training machine learning models of carbon materials, as there are more than 500,000 unique calculations of graphene (3.5 nm x 3.5 nm unit cell). The dataset was used as training data for an extensive deep neural network which was used to simulate a micron of graphene.
本数据集为针对存在尺寸与位置均随机分布的结构缺陷的石墨结构开展的密度泛函理论(Density functional theory, DFT)模拟数据集。数据集包含碳原子坐标,以及通过VASP软件中实现的PBE泛函(PBE functional)计算得到的体系总能量。此外还附带了全部输入参数文件,可完整复现所有计算过程。 注意:本数据集包含超过50万次针对3.5 nm×3.5 nm晶胞石墨烯的独立计算结果,可用于训练碳基材料的机器学习模型。该数据集曾作为大规模深度神经网络的训练数据,该网络被用于模拟微米级尺度的石墨烯体系。




