Radical stability and redox potential calculations for 89,320 organic radicals in the water phase

Optimized structures for 89,320 organic radicals at the M06-2X/def2-TZVP level of theory in an implicit water solvent using Gaussian 16. Cartesian coordinates of the optimized molecules are included in the SDF file format. Fractional spin of the unpaired electron and buried volumes has been provided in a CSV file for the water-phase calculations of the computationally optimized molecules. In the spin and buried volume CSV, atom_index corresponds to the canonical atom order assigned by RDKit and is also represented by the atom ordering for the corresponding entry in the SDF file. The redox potential database contains the redox potentials (in V relative to SHE) calculated with Guassian 16c at the M06-2x/def2-tzvp level of theory in an implicit water solvent by finding the enthalpy of both the positive and negatively charged states for each radical.<br>In addition to the database of radicals derived from PubChem molecules, we also include 1078 radicals generated by an RL algorithm for potential application in redox flow batteries. In the rl candidates CSV, we include enthalpy and 3D coordinates for the oxidized, radical, and reduced forms of each species (where converged). We also include raw Gaussian 16 output logs for each calculation in rl_logfiles.tar.gz, where the filenames are referenced in the rl candidates CSV.<br>

本数据集包含89320个有机自由基，在隐式水溶剂环境下，采用Gaussian 16软件，于M06-2X/def2-TZVP理论级别下完成结构优化。优化后分子的笛卡尔坐标以结构数据文件（SDF）格式存储。针对经计算优化后的分子开展水相计算，其未成对电子的分数自旋值与掩埋体积数据已通过逗号分隔值（CSV）文件提供。在该自旋与掩埋体积CSV文件中，原子索引对应由RDKit所指定的规范原子顺序，且与SDF文件中对应条目的原子排序完全一致。本数据集附带的氧化还原电势数据库，收录了通过Gaussian 16c软件，于M06-2X/def2-tzvp理论级别、隐式水溶剂环境下计算得到的氧化还原电势（单位为V，相对于标准氢电极（SHE））；该计算通过求取每个自由基的带正电与带负电状态的焓值实现。
除源自PubChem分子的自由基数据集外，本数据集还包含1078个由强化学习（RL）算法生成的自由基，可潜在应用于氧化还原液流电池领域。在rl_candidates.csv文件中，收录了各物种（收敛计算结果）的氧化态、自由基态与还原态的焓值及三维坐标。此外，所有计算的原始Gaussian 16输出日志均打包于rl_logfiles.tar.gz中，文件名可在rl_candidates.csv文件中进行索引查询。