Subject：Artificial Intelligence Methods to Assist in the Design and Manufacture of Fine Chemicals
Speaker：Lei Zhang
Time：14:00-15:00 p.m., 6th Jul.,2023
Place：Room 307, Laboratory Building No. 19
Abstract：China's fine chemical industry is big but not strong, and the manufacturing of high-end fine chemicals is still the key technology that restricts the development of China's chemical industry. Therefore, product engineering with the goal of intelligent design is the key to realize the intelligent manufacturing of high-end fine chemicals. In this paper, a mechanism-data fusion-driven approach is used to study the design and inverse synthesis methods of fine chemicals, and good results are achieved in the design problems of flavors, rubber auxiliaries, and small molecule drugs. First, the construction of accurate universal conformational relationships is a key scientific problem in the prediction of fine chemical properties, and the fusion of quantum chemistry and deep learning is used to develop a high-throughput generation method of molecular charge density distributions, reveal the mechanism of molecular conformational relationships, and realize high-throughput prediction of physical properties, which provides a theoretical basis for molecular intelligent design. Secondly, efficient molecular assembly strategy and reverse design method are the key scientific issues for intelligent molecular structure design of fine chemicals. A molecular assembly strategy of skeleton-group coupling is proposed, and a mixed-integer nonlinear planning model is constructed to realize the goal of reverse molecular design from properties to structure, which improves the molecular design efficiency by 10 times and the novelty of molecular design by more than 95% compared with the high-throughput virtual screening. Finally, the agile development of complex molecular synthesis paths is a key scientific issue in the intelligent synthesis of fine chemicals. Combining thermodynamic and kinetic mechanisms, we constructed a fully mapped reaction template database based on reaction mechanisms, and established a quantitative prediction model of reaction rate in solvent environment, and optimized the design of reaction paths in synergy with homogeneous catalysts, so as to achieve the agile development of synthesis routes. The proposed method can provide theoretical basis and technical guidance for efficient and intelligent design of fine chemicals.
Introduction：
Lei Zhang, graduated from the Department of Chemical Engineering, Tsinghua University with a Ph.D. in 2014, and post-doctoral fellowship at the Technical University of Denmark and the Hong Kong University of Science and Technology, is currently an Associate Professor and Ph.D. Supervisor at the School of Chemical Engineering, Dalian University of Technology. His main research interests are process systems engineering and chemical product engineering. He has been awarded the Young Science and Technology Award in Process Systems Engineering, the Best Paper Award in Comput. Chem. Eng. and the Outstanding Young Scientist Award in Results in Engineering. He has published more than 20 SCI papers as the first and corresponding author. He has chaired two top-level grants and one youth grant from the National Natural Science Foundation of China (NSFC). He is a member of the Process Systems Engineering Committee of the Systems Engineering Society, and a young member of the Information Technology Application Committee and the Simulation and Emulation Committee of the Chemical Society.