A Korea University research team has developed an artificial intelligence (AI) system that explains why new materials change color and brightness. This breakthrough is expected to accelerate the development of organic light-emitting diodes (OLEDs) for smartphone displays, fluorescent materials for cancer detection, and solar cell components.
Korea University announced on September 4 that Professor Seongnam Park’s chemistry department team created an AI model that analyzes molecule-solvent interactions to pinpoint the causes of color shifts.
OLED materials and fluorescent diagnostic agents emit light in liquid solvents, with their color and intensity varying based on the solvent used. While existing AI could predict the resulting color from specific molecule-solvent combinations, it couldn’t explain why these changes occurred.
The team developed an AI technique that calculates color change effects by breaking down molecules into functional groups (atom clusters that determine molecular properties). They dubbed this method Solvatochromic Subgroup Contribution (SSC).
Their analysis revealed that amine functional groups significantly influence color changes depending on the solvent, while methyl groups have minimal impact. By quantifying the effects of different functional groups, researchers can now verify the reasoning behind AI predictions.
This new AI model outperforms previous versions in prediction accuracy and aligns with established chemical principles. The team believes this technology will streamline the development of fluorescent diagnostic tools and OLED and solar cell materials by identifying optimal molecular structures before synthesis.
Currently, the application is limited to common organic molecules and faces challenges with complex structures like polymers or nanoparticles.
The study was published online in the prestigious journal Advanced Science on August 27.