Vibrational Properties of Metastable Polymorph Structures by Machine Learning
Vibrational Properties of Metastable Polymorph Structures by Machine Learning
复制标题
DOI：
10.1021/acs.jcim.8b00279
复制
发表时间：
2018-12-01
影响因子：
5.6
通讯作者：
Mingo, Natalio
中科院分区：
文献类型：
--
作者：
Legrain, Fleur;van Roekeghem, Ambroise;Mingo, Natalio
尽管振动性能是至关重要的有限温度稳定性以及导热性和其他输运性质的固体的从头算预测，其列入从头算材料库已被昂贵的计算要求所阻碍。在这里，我们通过展示使用机器学习从原子平衡位置的知识中可以快速实现力常数和振动特性的良好估计来应对挑战。在KZnF 3的121种不同的机械稳定结构上训练的随机森林算法达到了0.17 eV/angstrom（2）的原子间力常数的平均绝对误差，并且它比训练此类化合物的完整力场更便宜。预测的力常数，然后用来估计声子的光谱特征，热容量，振动熵，和振动自由能，比较以及与从头算。该方法可用于有限温度下稳定性的快速估计。
Despite vibrational properties being critical for the ab initio prediction of finite-temperature stability as well as thermal conductivity and other transport properties of solids, their inclusion in ab initio materials repositories has been hindered by expensive computational requirements. Here we tackle the challenge, by showing that a good estimation of force constants and vibrational properties can be quickly achieved from the knowledge of atomic equilibrium positions using machine learning. A random-forest algorithm trained on 121 different mechanically stable structures of KZnF3 reaches a mean absolute error of 0.17 eV/angstrom(2) for the interatomic force constants, and it is less expensive than training the complete force field for such compounds. The predicted force constants are then used to estimate phonon spectral features, heat capacities, vibrational entropies, and vibrational free energies, which compare well with the ab initio ones. The approach can be used for the rapid estimation of stability at finite temperatures.