ACGC plus: a general group contribution framework for diverse properties prediction

Abstract

Atomic connectivity group contribution (ACGC) is a method developed by atomic adjacent group (AAG), shape factors and atomic connectivity factors (ACFs) for property prediction. As a crucial parameter for dealing with the challenge of accurately predicting properties of isomers, ACFs are defined for describing the global position of each group in a molecule. In this work, ACFs plus (ACFs+) is proposed to describe the local position of a group by considering the contribution of the core atom in AAG and nearby atoms. As such, ACGC plus (ACGC+) models are developed with ACFs+ to predict key phase transition properties of organic compounds (i.e., Δ_fusS, Δ_vapH°, Δ_subH°, Δ_fusH, T_b, T_m, T_c, P_c, and V_c). Both predictability and robustness are rigorously validated using external validation and cross-validation. The R²_test values for phase transition entropy and enthalpy range from 0.906 to 0.992, the R²_test values for critical properties are greater than 0.989, and the R²_test values for T_b and T_m are 0.979 and 0.845, respectively, which indicate high predictability of ACGC+ models. The R² values for all properties are close to the R²_train values, which further validates the stability of the ACGC+ model. Furthermore, the mean absolute errors of the ACGC+ models decreased by 1.44%–7.91% compared to the ACGC models. These results demonstrate that the ACGC+ method provides high accuracy in predicting the properties of phase transitions.

Keywords

group contribution method / atomic connectivity group / shape factor / phase transition properties / thermodynamic properties

RIGHTS & PERMISSIONS

Higher Education Press 2026

Just Accepted

This article has successfully passed peer review and final editorial review, and will soon enter typesetting, proofreading and other publishing processes. The currently displayed version is the accepted final manuscript. The officially published version will be updated with format, DOI and citation information upon launch. We recommend that you pay attention to subsequent journal notifications and preferentially cite the officially published version. Thank you for your support and cooperation.