Thermodynamic modeling of the KCl + formamide/glucose/proline + water ternary systems and activity coefficient prediction based on artificial neural network

Abstract

In this research, thermodynamic modeling and activity coefficient prediction of KCl in the (water+formamide/glucose/proline) mixed solvent systems were reported. Thermodynamic study was performed using the potentiometric data based on extended ion interaction Pitzer–Archer model in various mixed solvent systems containing 0, 10, 20, 30 and 40% mass fractions of formamide and glucose at T=298.2K and 0, 2.5, 5.0, 7.5 and 10.0% mass fractions of proline at T=308.2K and ambient pressure over ionic strength ranging from 0.0014 to 3.9579mol·kg−1. The adjustable parameters were determined and the obtained results were then interpreted based on extended ion interaction Pitzer–Archer model. In addition, artificial neural network was used to model and to predict the mean ionic activity coefficients of KCl in the (water+formamide/glucose/proline) mixed solvent systems. The mean absolute deviation of the designed neural network in prediction of the mean ionic activity coefficients was about 0.0083. Therefore, the artificial neural network can efficiently predict the activity coefficients of electrolyte in mixed solvent systems.