Python-based high-throughput extraction of void information, solvent accessible volume and adsorbate molecules from MOF for adsorption-separation applications.

With the rapid growth of chemical data and information, there is an increasing need for analyzing large chemical datasets and extracting key or feature information. Currently, more than 100,000 types of metal-organic frameworks (MOFs), as the material recently awarded the Nobel Prize in Chemistry, have been experimentally synthesized. The performance of MOFs in adsorption-separation applications depends on their specific void characteristics, including void count, spatial distribution, and volume size. This study presents the entire process including data collection, recognition of key or feature information, the workflow of using Python tools, and the automatic output of results for void information, solvent accessible volume (SAV) and adsorbate molecules. By processing 219 CIF files collected from open-access publications, CCDC, and supporting information files, we successfully extracted 498 total blocks, including 259 blocks with void information, 157 blocks with SAV data, 286 blocks with squeeze details, and 1573 individual voids. In addition, we identified adsorbate molecules (diethyl ether, chloroform, water, ethanol, toluene, carbon dioxide) in MOFs. The method demonstrates computational efficiency, requiring only standard CPU resources to process large datasets.