A paper from the Laboratory of Data and Chemical Engineering (headed by Associate Professor KANEKO Hiromasa) graced the cover of Volume 7, Issue 5 (2025) of the international academic journal ACS Applied Polymer Materials.
In the electronics market, there is a demand for materials with high insulation and low dielectric loss, and there is a growing need for materials with low dielectric constant. Researchers repeatedly design, synthesize, and measure materials to develop low-permittivity materials using their knowledge and experience, but the development period is long and costs such as personnel, reagent, and equipment are also incurred. The purpose of this study is to propose a monomer structure of low dielectric constant epoxy resin focusing on a typical epoxy resin with good moldability. They built a dielectric constant prediction model using machine learning and generated a number of hypothetical chemical structures of epoxy resins using a method based on backbone and side chain bonds and a method based on a hypothetical chemical reaction, respectively. Then, the structure of epoxy resin with low dielectric constant was successfully searched by predicting the dielectric constant of the structure generated by the constructed model.
≪Japanese version≫
