A Korea University (KU) Research Team led by Professor Park Sungnam (From the left: Park Sungnam, Joung Joonyoung, and Han Minhi). Provided by KU.

By Jiin Ryu (jiinryu@korea.ac.kr)

A research team led by Professor Park Sungnam developed an artificial intelligence (AI) that can be utilized to tremendously reduce the time and cost spent on creating color and fluorescent molecules needed for organic light emitting diodes (OLED) development. The research was published on March 13 in the internationally renowned scientific journal, American Chemical Society.

There are numerous characteristics to be considered when developing color and fluorescent molecules, such as different wavelengths, widths, and efficiency. ‘Deep Learning Optical Spectroscopy’ made predicting photometric characteristics of various organic molecules easier, therefore making the searching process for appropriate molecules drastically more efficient. Calculation time using the technology compared to the existing method was reduced by 63,000 times and prediction error of absorbent and luminous wavelengths by 2.2 and 1.5 times each.

Professor Park’s team developed Deep Learning Optical Spectroscopy by organizing reported cases of 7 optical characteristics of molecules, forming a database that consists of almost 30,000 data. Using this technology, the predictions can be made in different types of environments and the compatibility can be examined before having to synthesize the actual molecules. It is predicted that the technology will speed up the process of ruling out unnecessary candidates, which can take up a big portion of the production cost.

A newly developed web application by the research team will allow anyone to access Deep Learning Optical Spectroscopy at http://deep4chem.korea.ac.kr/.

출처 : 고려대학교 고파스 2025-07-25 10:56:45: