Data for: On-the-Fly Non-adiabatic Molecular Dynamics Reveals Dissociation Mechanisms of Multiply-Charged Molecules

Title: Theoretical Trajectories and Experimental Data Overview This dataset contains both the theoretical non-adiabatic molecular dynamics (NAMD) trajectories and the experimental measurements used to investigate the dissociation mechanisms of multiply-charged molecules (e.g., CO2^3+). File Contents The dataset consists of two main files: Data.hdf5 (Theoretical Simulations): Contains the raw NAMD trajectories simulated using the SHARC program. It has been parsed into a highly structured HDF5 file. It includes time-dependent 3D nuclear coordinates (in Å), velocities (in a.u.), and energy tables (in eV) for thousands of trajectories across different initial electronic states. data.xlsx (Experimental Data): Contains the experimental data obtained via COLTRIMS reaction microscope, including the Kinetic Energy Release (KER) spectra and the data points used to construct the experimental Dalitz plots. File Format and Structure The data is provided in standard HDF5 format (Data.hdf5). The file is organized hierarchically by initial electronic state and trajectory ID: /{state_name}/{trajectory_name}/ (e.g., /Doublet_0/traj_0001/). Global Attributes (Root Level) created_at: Timestamp of file creation. units_geometry: Angstrom (Å) units_velocity: Atomic units (a.u.) Group Contents (Per Trajectory) For each individual trajectory group, the following datasets and attributes are provided: 1. Datasets (Time-dependent variables) geometry: 3D nuclear coordinates at each time step. Shape: (N_steps, N_atoms, 3). Unit: Å. velocity: 3D nuclear velocities at each time step. Shape: (N_steps, N_atoms, 3). Unit: a.u. energy_table: Time-evolution of various energies (Kinetic, Potential, Total, and specific electronic states). Shape: (N_steps, N_columns). Unit: eV. The column headers are stored in the columns attribute of this dataset. expec_out_raw: The raw string content of the SHARC expec.out file. 2. Attributes (Metadata & Extracted Properties) elements: Array of atomic symbols (e.g., ['O', 'C', 'O']). atomic_masses: Array of atomic masses. multiplicity: Spin multiplicity of the initial state (e.g., Doublet). nstep: Total number of simulated time steps. delta_energy: The calculated kinetic energy difference between the final and initial steps, used as a proxy for Kinetic Energy Release (KER). Unit: eV. nfrag: Number of molecular fragments at the final simulation step (determined using a distance cutoff of 3.5 Å). frag_list: A simple string representation of the final fragments (e.g., CO, O). frag: Detailed JSON string containing the indices and elements of each final fragment. How to Read the Data You can explore the file structure and contents visually using graphical HDF5 readers such as HDFView. For automated analysis, you can easily load this dataset using the h5py and numpy libraries in Python: import h5py # Open the HDF5 filewith h5py.File("Data.hdf5", "r") as h5: # Access a specific trajectory from a specific state traj = h5["Doublet_0"]["traj_0001"] # 1. Read static attributes elements = traj.attrs["elements"] ker_ev = traj.attrs["delta_energy"] final_fragments = traj.attrs["frag_list"] print(f"Fragments: {final_fragments} | KER: {ker_ev:.2f} eV") # 2. Read time-dependent datasets geom = traj["geometry"][:] # Shape: (Steps, Atoms, 3) in Angstrom vel = traj["velocity"][:] # Shape: (Steps, Atoms, 3) in a.u. # 3. Read the energy table and its column names energy_data = traj["energy_table"][:] energy_columns = traj["energy_table"].attrs["columns"]