Kim Ji-hwan, a master’s student in the Department of Chemistry at the Graduate School of Kookmin University (President Jeong Seung Ryul; advisor: Professor Joung Joonyoung), received the “Outstanding Oral Presentation Award” in recognition of the creativity and excellence of his research at the “149th Summer Symposium of the Physical Chemistry Division of the Korean Chemical Society,” held in Busan from June 21 to 23.

Kim Ji-hwan, a member of the Artificial Intelligence Chemistry Laboratory, presented his research findings at this summer symposium under the title “MEMo: Mixture Embedding via Mole Fraction.”

This methodology is drawing attention because it overcomes the limitation of existing models—which treat the properties of mixtures as mere simple weighted averages of their pure components and thus fail to capture non-ideal mixing behavior—by directly incorporating composition information from the learning stage, thereby enabling the prediction of nonlinear changes based on composition.

These results demonstrate consistent performance improvements over existing methods across five key physical properties—density, sound speed, vapor pressure, refractive index, and surface tension—and are significant in that they enable the prediction of physical properties for new mixtures without experimentation, thereby enhancing the efficiency of materials and process design.

Student Kim Ji-hwan stated, “Starting from the premise that actual mixtures are not simply the sum of their individual components, we sought to demonstrate that artificial intelligence can learn even the subtle nonlinear behaviors resulting from composition.” He added, “Going forward, we aim to expand the model to include a wider variety of component systems and conditions, developing it into a predictive tool that can make a practical contribution to the development of new materials.”