Predictability of Carbon Dioxide and Ethane Solubility in Ionic Liquids: A Simulation Approach
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Capturing greenhouse gases using solvents is considered an efficient solution to address climate change and surging anthropogenic activity. In an attempt to find efficient solvents and avoid high experimental costs, predicting the solubilty of acid gases in different solvents is attracting attention. Ionic Liquids (ILs) are considered promising solvents for the sweetening of natural gas streams. These ionic liquids can also be used to capture CO2 from flue gases. This study examines the predictability of modeling the solubility of CO2 and C2H6 in ionic liquids based on some intrinsic properties such as critical temperature, critical pressure and acentric factor as well as process operating conditions such as temperature and pressure. Accordingly, recent experimental data has been collected for 18 ILs which have not been investigated in simulation studies, and are modeled using artificial neural network (ANN) and adaptive neuro fuzzy inference system (ANFIS). Performance analysis suggests that the solubility of CO2 and C2H6 in an IL, knowing the five selected parameters, can be predicted with satisfactory precision using either ANN or ANFIS. The models presented in this study outperform, in terms of accuracy, previous models in the literature for the specific ionic liquids selected. More specifically, the mean squared errors (MSE) of the generated functions using ANN analysis are 6.93 × 10-5 and 7.94 × 10-6 for CO2 and C2H6, respectively. The same performance measured for the generated fuzzy inference system gave 6.72 × 10-5 and 1.07 ×10-5for CO2 and C2H6, respectively. Finally, individual and mixed-effects of the five variables on the solubility are assessed. The mixed effect of pressure on other parameters seems to be relatively significant.