Research data for "Full-cycle device-scale simulations of memory materials with a tailored atomic-cluster-expansion potential"

This is a dataset related to the publication "Full-cycle device-scale simulations of memory materials with a tailored atomic-cluster-expansion potential". The .zip files contain two main folders: 1. Device-scale molecular dynamics (MD) trajectories: this folder contains the device-scale MD trajectories generated in this work. 2. Atomic-cluster-expansion (ACE) potential model and training dataset: this folder includes the ACE model and its corresponding training dataset (GST-ACE-24). Detailed Contents: 1. The device-scale MD trajectories have been categorised according to the main text figures: "Simulations_of_cross-point_device": full-cycle simulations of cross-point device-scale model of Ge1Sb2Te4 (532,980 atoms) as shown in Fig. 3 of the paper. "Simulations_of_mushroom-type_device": crystallisation simulations of mushroom-type device-scale model of Ge1Sb2Te4 (794,808 atoms) as shown in Fig. 4 of the paper. 2. The ACE models trained in this work and the corresponding reference dataset: "GST-ACE-24.yace": the ACE potential file, ready for use in LAMMPS for ACE-MD simulations. "GST-ACE-24_training_dataset.xyz": the training dataset, provided in (extended) XYZ format.