Phys.org October 25, 2024
Fast and accurate prediction of dielectric function facilitates the development of novel dielectric materials, an ingredient of many cutting-edge technologies such as 6G networks. Researchers in Japan introduced a versatile machine-learning scheme implemented in Git hub for predicting dipole moments of molecular liquids to study dielectric properties. They attributed the center of mass of Wannier functions (called Wannier centers), to each chemical bond and created neural network models. They applied liquid methanol and ethanol to demonstrate that their neural network models successfully predicted the dipole moment of various liquid configurations in close agreement with DFT calculations. The dipole moment and dielectric constant in the liquids were greatly enhanced by the polarization of Wannier centers due to local intermolecular interactions. The calculated dielectric spectra quantitatively agreed with experiments over terahertz (THz) to infrared regions. They investigated the physical origin of THz absorption spectra of methanol, confirming the importance of translational and librational motions. According to the researchers their method is applicable to other molecular liquids and can be widely used to study their dielectric properties… read more. Open Access TECHNICAL ARTICLE
Schematic image of dipole moments for a single bond and lone pair. represents the position of the BC. Credit: Phys. Rev. B 110, 165159, 25 October 2024