Data-driven Discovery of Polar Organic Cocrystals: Integration of Machine Learning and Automated Screening

Polar organic cocrystals hold great promise for various advanced technological applications. However, their relatively low prevalence highlights the challenges of achieving the desired polar packing arrangements, making their discovery complex and demanding. Here, we present a data-driven approach that integrates machine learning (ML) with high-throughput (HT) automation to accelerate the discovery of polar organic cocrystals. Using ML methodologies, we identified key parameters governing polar cocrystal formation, enabling the targeted selection of molecular candidates. A total of 20 cocrystal combinations with chloranilic acid (CA) were explored, with 20 solvent systems screened for each combination, enabling a highly efficient selection process across a vast chemical space. HT automation further streamlined the synthesis and characterization processes through rapid screening and accurate structural validation, while comprehensively exploring the chemical landscape. Experimental validation yielded 16 new hydrogen-bonded cocrystals, 8 in polar space groups, achieving a polar cocrystal discovery rate (50%) over three times higher than the Cambridge Structural Database (CSD) average (~14%). his integrated approach represents a new approach in polar organic cocrystal research. The findings highlight the potential of this approach for advancing functional molecular materials, paving the way for next-generation applications using polar organic cocrystals.