Correlations between the constituent molecules, crystal structures, and dielectric constants in organic crystals

Organic crystals are crystals composed of molecules of organic compounds. Materials made from organic crystals are used in devices, such as capacitors and ferroelectric memories. It is desirable to develop new materials with improved physical properties such as the dielectric constant (DC). However, the relationship between the constituent molecule (CM), crystal structure (CS), and DC of the organic crystals is not clearly understood. In this study, we investigated the relationship between CM, CS, and DC with existing data and machine learning. Using regression analysis, we could construct machine learning models between CM and DC, between CS and DC, and between CM and CS, and could predict DC from CM, DC from CS, and CS from CM using the constructed models. • Organic crystals are crystals composed of molecules of organic compounds. • The relationship between molecules, crystal structure, and dielectric constant of organic crystals was modeled. • Prediction models between each of the physical properties were constructed. • The information on crystal structures could be predicted from molecular information. • Contributions to material design considering molecular, crystal structure and dielectric constant are expected.