Supplementary data for "Lithium and Oxygen Dimer Diffusion in Solid-State Oxygen-Redox Cathode Materials: Insights from Neural Network Molecular Dynamics"

The following Supplementary Data were obtained using the neural network hpotential molecular dynamics simulation using Matlantis Preferred Potential (PFP) [1] in the study:<b>Title</b>: <i>"Lithium and Oxygen Dimer Diffusion in Solid-State Oxygen-Redox Cathode Materials: Insights from Neural Network Molecular Dynamics"</i><br><b>Authors</b>: Takuhiro Sasadaira, Naoto Tanibata, Hayami Takeda, Naoaki Yabuuchi, Masanobu Nakayama<br><b>DOI</b>: 10.1080/27660400.2025.2602236<br><br><b>Supplementary Data 1</b>:<br>Final crystal structures obtained after 1 ns of NNP-MD simulation at 773 K. The data are provided in VASP POSCAR format. (For computational details, refer to the main text.)<br>Files are named as <code>CONTCAR_x###</code>, where the three-digit integer <code>###</code> corresponds to 100 times the x value in the composition Li_1.2-xMn_0.4Ti_0.4O₂.<br><b>Supplementary Data 2</b>:<br>Representative trajectory files from NNP-MD simulations at 773 K for 1 ns. The data are provided in ASE <code>.traj</code> format. While the analyses in this study were performed using trajectory data saved every 100 fs, the data shared in the repository are downsampled to every 1000 fs due to file size limitations. The selected compositions are x = 0.24, 0.36, and 0.60, corresponding to the data shown in Figures 5 through 9 of the main text.<br><b>References</b><br>[1] Takamoto, S. et al. Towards universal neural network potential for material discovery applicable to arbitrary combination of 45 elements. <i>Nat. Commun.</i> <b>13</b>, 2991 (2022). https://doi.org/10.1038/s41467-022-30687-9