Research on the Prediction of Nano-Organic Synthesis Reaction Pathways Based on Graph Neural Networks

At present, due to the potential biological toxicity risks and economic considerations in the preparation of nanomaterials, their green synthesis and application in environmental governance have attracted much attention.However, this field still faces core challenges: the molecular mechanism of the green synthesis pathway is not yet clear, and the pollutant removal efficiency still has a significant gap compared with traditional methods.In this study, the synthesis conditions of nanometers were optimized through the graph neural network model, and an improved graph neural network algorithm based on molecular segmentation was proposed.Based on the composition mechanism of material molecules and the division of functional groups, an unsupervised learning method is constructed to segment the graph data structure composed of molecules.Combined with the structure after molecular segmentation, a new graph neural network is designed to pay more attention to the local effects of functional groups.Through experiments in databases such as solubility, the improved graph neural network has better prediction performance.Meanwhile, the combination of molecular segmentation and graph interpretation algorithms guides the graph interpretation algorithms to search for substructures containing complete functional groups.For the interpretation of structure-performance, it is more in line with the mechanism of molecular composition and has more practical significance for performance analysis and the design of new materials.