Investigating the solution behavior of CO₂ in ester based on a hybrid of ANN technology and PS method

Artificial neural networks (ANNs) are artificial intelligence tools, and pattern search (PS) is a method for solving optimization problems. Although these two technologies have been extensively used in the chemical and engineering fields, they have limitations. This study proposes a methodology based on a hybrid of ANN technology and PS method to investigate the solution behavior of carbon dioxide (CO2). Five CO2-ester binary systems are selected as the model systems to demonstrate the point of interest. The results reveal that this methodology can overcome the defects of the two technologies and develop their advantages, thereby providing a satisfactory description of the solubility data of CO2 in these esters. The proposed methodology can be applied to a series of CO2 binary systems and is helpful in selecting a suitable absorbent for CO2 capture technology.