Experimental and theoretical study of carbon dioxide solubility in aqueous solution of potassium glycinate blended with piperazine as new absorbents

• CO2 solubility in potassium glycinate + piperazine + water is measured.
• Solution of potassium glycinate + piperazine has very high loading capacity.
• Deshmukh-Mather model is extended to predict CO2 loading.
• An artificial neural network is developed to predict CO2 loading.
• Suggested solution has suitable stability toward oxidative degradation.

In this work aqueous solution of potassium glycinate, piperazine and potassium glycinate blended with piperazine have been utilized through CO2 solubility measurements in order to investigate the possible use of these salts of amino acid blended with alkanolamine for CO2 absorption. The equilibrium solubility of CO2 for aqueous solutions of potassium glycinate, piperazine and potassium glycinate blended with piperazine at 1.0, 4.0 and 10.0 wt.% overall mass concentration are experimentally measured with an equilibrium cell at CO2 partial pressure ranging from 5.1 to 2508.7 kPa and temperatures between 293.15 and 323.15 K. Obtained data show that loading capacity decreases with increase in temperature and concentration of potassium glycinate and potassium glycinate blended with piperazine. Obtained CO2 loadings are very high at low concentration of potassium glycinate. In order to predict CO2 solubility in aqueous solutions potassium glycinate, piperazine and potassium glycinate blended with piperazine, Deshmkh-Mather model is extended. The values of MSE, ARD and R2 for the extended Deshmukh-Mather model are 0.0741, 8.9582 and 0.9826, respectively. In addition, an artificial neural network (ANN) is developed to predict CO2 solubility in aqueous solutions of piperazine, potassium glycinate and potassium glycinate blended with piperazine. The values of MSE, ARD and R2 for the optimal trained ANN are 0.0179, 4.1528 and 0.9962, respectively.