To overcome the limitations of empirical synthesis and expedite the discovery of new polymers, this work aims to develop a data-driven strategy for profoundly aiding in the design and screening of novel polyester materials. Initially, we collected 695 polyesters with their associated glass transition temperatures (Tgs) to develop a quantitative structure-property relationship (QSPR) model. The model underwent rigorous validation (external validation, internal validation, Y-random and application domain analysis) to demonstrate its robust predictive capabilities and high stability. Subsequently, by employing an in-silico retrosynthesis strategy, over 95000 virtual polyesters were designed, largely expanding the available space for polyester materials. External assessments highlight the good extrapolation ability of the QSPR model. Furthermore, we experimentally synthesized diverse virtual polyesters with Tgs covering a sufficient large temperature range. It is believed that this data-driven approach can drive future product development of polymer industry.

The thermodynamic properties at variable temperature and pressure, such as density (ρ) and viscosity (η) are necessary in chemical process design. The quantitative structure-property relationship (QSPR) is a quick and accurate method to obtain the properties from a large number of potential ionic liquids (ILs). The QSPR models for ρ and η may have “pseudo-high” robustness validated by leave-one-out cross-validation (LOO-CV) and weakened stability with the unbalanced data point distribution. A rigorous model evaluation method named the leave-one-ion-out cross-validation (LOIO-CV) was proposed to evaluate robustness of ILs QSPR models. Balancing the distribution of data points in ILs, two f(T,P,I)-QSPR models were developed with norm index (I) to predict ρ and η of ILs at variable temperature and pressure. LOIO-CV method can enhance the stability QSPR model in predicting the properties of ILs with new cations and anions, which is essential for data driven design of ILs.

In this work, atomic connectivity group contribution (ACGC) method is developed for predicting critical properties of organic compounds. Herein, a new group defining method, namely atomic adjacent group (AAG) method, is proposed to describe the relationship between core atom and its adjacent atoms. For distinguishing isomers effectively, the shape factor (SF) is used to describe the effect of molecular shape on group, and atomic connectivity factors (ACF) are defined for describing the position of each group in a molecule. The external and internal verification methods were utilized during the modelling process. Compared with AAG model, ARE decreased by 6.82-42.57 % when SF was considered and, ARE decreased by 24.19-62.25 % when both SF and ACF were applied as using the ACGC method. Accordingly, SF and ACF are effective in improving the group contribution method and ACGC method is accurate in calculating the properties of organic compounds.

ILs thermodynamic properties at variable temperature and pressure, such as density, viscosity and thermal conductivity, are necessarily basal parameters of chemical engineering process. In this paper, some norm descriptors-based QSPR models are established to predict the properties of ILs at variable temperature and pressure. The f-T-P models are developed with 9020 data points of 314 ILs for density, 7342 data points of 351 ILs for viscosity and 608 data points of 87 ILs for thermal conductivity. These models have satisfactory statistical results for the calculation of ILs properties. The validation analysis shows that these QSPR models have good stability and predictability. And the norm descriptors are universal for predicting the properties of ILs. Moreover, these QSPR models are applied to predict parameters of f-T-P models for 16329 ILs generated by combing the cations and anions in the dataset, which might be valuable and further used handily by other chemists.