development of efficient group contribution models for carbon dioxide absorption by aqueous amine solutions

In order to design new solvents such as amine solutions, for absorption of carbon dioxide from natural gas, efficient models are necessary. in this work, new easy-to-use group contribution models (gc) were developed for determination of the absorption capacity and the absorption heat of carbon dioxide by aqueous amine solutions. by using a method combining modified particle swarm optimization (mpso) and multiple linear regression (mlr) the most relevant substructures (functional groups) of amines were selected and the linear gc models were developed. the proposed models were assessed by external validation method based on various statistical parameters. the squared correlation coefficient values of 0.825 for the absorption capacity model and 0.881 for the absorption heat model show good predictive ability of both gc models. as a result, the simple models proposed here is accurate and can be used to design of new amine solutions for carbon dioxide absorption.