Ultrafast Structural Dynamics Dataset for Photoexcitation-Induced Polarization Reversal in BaTiO3

The dataset systematically records time-resolved electronic structures, electronic state occupations, atomic displacement trajectories, key structural order parameters, and polarization responses, and stores them in a standardized manner using a unified data format and metadata framework. Furthermore, this work systematically validates the data quality in terms of computational-parameter consistency, dynamical stability, and numerical accuracy, thereby ensuring the reliability, traceability, and reusability of the dataset.

This dataset provides a high-quality benchmark for understanding electron-phonon coupling, transient potential-energy-surface reconstruction, and nonthermal phase-transition pathways in photoinduced ferroelectric polarization reversal. It also lays a data foundation for machine-learning-assisted prediction of ultrafast dynamics and the design of light-controlled ferroelectric functional devices.