Experimental analysis and modeling of CO2 solubility in AMP (2-amino-2-methyl-1-propanol) at low CO2 partial pressure using the models of Deshmukh–Mather and the artificial neural network

The equilibrium solubility data for CO2 in aqueous solution of AMP have been determined at temperatures from 293 K to 323 K, partial pressures from 17.47 kPa to 69.87 kPa and concentrations of AMP from 1 M to 4 M. The experimental results show that the solubility of CO2 in AMP increases with partial pressure and decreases with temperature and concentration of solvent. Two different mathematical models have been used to analyze the solubility of CO2 in AMP including those of Deshmukh–Mather and the artificial neural network. The modeling results indicate that the neural network modeling provides a better prediction of experimental CO2 loadings than the Deshmukh–Mather model when compared with experimental results in this work. Therefore, this new modeling method can be useful in predicting the results of CO2 absorption and its accuracy is comparable with those of thermodynamic models which are used widely.

► The absorption of CO2 into aqueous solution of AMP at low CO2 partial pressure was experimentally and analytically investigated by two different mathematical models which are Deshmukh–Mather model and artificial neural network.
► In experimental section it was observed that the solubility of CO2 in AMP increases with partial pressure and decreases with temperature and concentration of solvent.
► This study indicates that the neural modeling has a better prediction of experimental CO2 loadings than Deshmukh–Mather model when is compared with experimental results in this work.