Graph Dynamical Networks for Unsupervised Learning of Atomic Scale Dynamics in Materials

Authors: Tian Xie1, Arthur France-Lanord1, Yanming Wang1, Yang Shao-Horn1, Jeffrey Grossman1*

  1. Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
  • Corresponding author email: jcg@mit.edu

DOI10.24435/materialscloud:2019.0017/v1 (version v1, submitted on 09 May 2019)

How to cite this entry

Tian Xie, Arthur France-Lanord, Yanming Wang, Yang Shao-Horn, Jeffrey Grossman, Graph Dynamical Networks for Unsupervised Learning of Atomic Scale Dynamics in Materials, Materials Cloud Archive (2019), doi: 10.24435/materialscloud:2019.0017/v1.

Description

Understanding the dynamical processes that govern the performance of functional materials is essential for the design of next generation materials to tackle global energy and environmental challenges. Many of these processes involve the dynamics of individual atoms or small molecules in condensed phases, e.g. lithium ions in electrolytes, water molecules in membranes, molten atoms at interfaces, etc., which are difficult to understand due to the complexity of local environments. We develop graph dynamical networks, an unsupervised learning approach for understanding atomic scale dynamics in arbitrary phases and environments from molecular dynamics simulations. We show that important dynamical information can be learned for various multi-component amorphous material systems, which is difficult to obtain otherwise. We develop a software package "gdynet" at https://github.com/txie-93/gdynet which implements the graph dynamical networks algorithm. This record contains the MD trajectories of a Li-S toy system, a Si-Au binary system, and a polymer battery electrolyte system in a format designed for the "gdynet" package.

Materials Cloud sections using this data

No Explore or Discover sections associated with this archive entry.

Files

File name Size Description
li2s-traj.npz
MD5MD5: 5e935498b94c80c1da6394b06440e14d
23.2 MiB A .npz file containing the molecular dynamics simulation trajectories of the Li-S toy system.
li2s-traj-graph-train.npz
MD5MD5: 944a521962ff413bbb5aa310aaee3c7e
55.0 MiB A .npz file containing the training data for the Li-S toy system after preprocessing.
li2s-traj-graph-val.npz
MD5MD5: f0a22d17649c7881da74dd13636b33e5
13.8 MiB A .npz file containing the validation data for the Li-S toy system after preprocessing.
li2s-traj-graph-test.npz
MD5MD5: c672d02e5a7eaf2a37d311e1f1961c0a
68.7 MiB A .npz file containing the testing data for the Li-S toy system after preprocessing.
siau-traj.npz
MD5MD5: 055a3728605fee44372516f0a5a4cc8c
92.6 MiB A .npz file containing the molecular dynamics simulation trajectories of the Si-Au binary system.
peo-traj-a.npz
MD5MD5: 8fc46b6341f0fd5c1739542b30769572
1.9 GiB A .npz file containing the molecular dynamics simulation trajectories of the PEO/LiTFSI binary system.
peo-traj-b.npz
MD5MD5: 153e52b7cff0f43c6f5aed2a55e5ce59
1.9 GiB A .npz file containing the molecular dynamics simulation trajectories of the PEO/LiTFSI binary system.
peo-traj-c.npz
MD5MD5: ad28b8ff20d1c1afd34ac1fe9a398092
1.9 GiB A .npz file containing the molecular dynamics simulation trajectories of the PEO/LiTFSI binary system.
peo-traj-d.npz
MD5MD5: 2c9abaf1fb491451349cde92b7b8bd4d
1.9 GiB A .npz file containing the molecular dynamics simulation trajectories of the PEO/LiTFSI binary system.
peo-traj-e.npz
MD5MD5: 43eaab6447ec4c32a18fbd25ed88d65e
1.9 GiB A .npz file containing the molecular dynamics simulation trajectories of the PEO/LiTFSI binary system.
README.txt
MD5MD5: 224a53617d0f6d2028b359eedee86479
2.0 KiB README file for this data.

License

Files and data are licensed under the terms of the following license: Creative Commons Attribution Non Commercial 4.0 International.

Keywords

machine learning molecular dynamics polymer deep learning graph neural networks amorphous interface trajectory

Version history

09 May 2019 [This version]