Design of polyimides with targeted glass transition temperature using a graph neural network. Issue 8 (8th February 2023)

• Record Type: Journal Article
• Authors: Qiu, Haoke; Qiu, Xuepeng; Dai, Xuemin; Sun, Zhao-Yan
• Is Part Of: Journal of materials chemistry. Volume 11:Issue 8(2023)
• Page Start: 2930
• Page End: 2940
• Publication Date: 2023-02-08
• Subjects: Materials -- Periodicals; Chemistry, Analytic -- Periodicals; Optical materials -- Research -- Periodicals; Electronics -- Materials -- Research -- Periodicals; 543.0284
• DOI: 10.1039/d2tc05174e
• Languages: English
• ISSNs: 2050-7526

Abstract

Abstract : Models were obtained to predict the T g of polyimides and compared with the experimental results. 110 polyimides are screened through high-throughput calculations and chemical knowledge. Novel molecular design guidelines are provided. Abstract : Polyimide substrates used in flexible display devices need to withstand very high temperatures and be highly thermally stable. The discovery of polyimides that satisfy these requirements, especially with high glass transition temperatures ( T g ), has been a hot research topic. However, conventional structural design of polyimides requires domain knowledge and expertise. In addition, the latent polyimide structures can reach millions, making it impractical to navigate their structure–property landscapes under conventional experimental screening. To address this matter, we utilized a graph neural network to build a regression model for predicting polyimides' T g and a classification model for two-step ensemble molecular screening. Eight polyimides were synthesized with T g measured using differential scanning calorimetry (DSC) and these values were well matched to the predicted values of the regression model. Unlike common machine learning models, our models substantially learn the effect of specific functional groups on polyimides T g and thus offer visual interpretation of the effect of functional group variations on T g, providing an unknown performance metric (Rot2Ring) for designing molecules with specific T g . Meanwhile, the model predicted T g values of 8 205 096 polyimides and a comprehensive screening has been implemented. Finally, combined with chemical knowledge, we derived 110 alternative polyimide structures with T g exceeding 400, which can be used by scientists for structure filtering. Furthermore, we propose an accessible toolkit – polyScreen, which enables predicting T g from structures in one step even without programming experience and more prediction modules will be updated in the next releases. This work demonstrates the feasibility of accurate prediction of material properties, providing potential polyimides with excellent properties for flexible displays, membranes and fibers utilizing graph neural networks, and provides quantitative guidance for the design of polyimides.