Chemical Research in Chinese Universities ›› 2025, Vol. 41 ›› Issue (5): 1014-1020.doi: 10.1007/s40242-025-5170-1

• Articles • Previous Articles     Next Articles

Darwin4Matter: A Platform Integrating Machine Learning and Quantum Chemistry for New Materials Design

RONG Hui2, CHEN Yili2, ZHANG Shubo2, CHEN Yue2, SHEN Lin1,2, FANG Wei-Hai1,2   

  1. 1. Key Laboratory of Theoretical and Computational Photochemistry of Ministry of Education, College of Chemistry, Beijing Normal University, Beijing 100875, P. R. China;
    2. Yantai-Jingshi Institute of Material Genome Engineering, Yantai 265505, P. R. China
  • Received:2025-08-13 Accepted:2025-09-11 Online:2025-10-01 Published:2025-09-26
  • Contact: CHEN Yue, E-mail: chenyue@xhtechgroup.com;SHEN Lin, E-mail: lshen@bnu.edu.cn;FANG Wei-Hai, E-mail: fangwh@bnu.edu.cn E-mail:chenyue@xhtechgroup.com;lshen@bnu.edu.cn;fangwh@bnu.edu.cn
  • Supported by:
    This work was supported by the Natural Science Foundation of China (Nos. 22193041, 22573008, 22288201), the Robotic AI-Scientist Platform of the Chinese Academy of Sciences and the Fundamental Research Funds for the Central Universities, China.

Abstract: Machine learning (ML) has emerged to play a more and more important role in material science. Here, we develop a platform named Darwin4Matter that integrates machine learning and quantum chemistry for new materials discovery. The framework consists of six steps: quantum chemistry prediction, Δ-machine learning correction, molecular augmentation, machine learning prediction, molecular production, and verification. Using this platform, we start from a very small dataset and design three new functional molecules with high refractive indexes in the visible spectrum, which serves as the capping layer of organic light-emitting diode devices for improving light extraction efficiency. The superior performance over currently used materials has been verified experimentally, exhibiting significant commercial value in the field of advanced display materials.

Key words: Data-driven materials design, Machine learning, Quantum chemistry, Organic light-emitting diode