Correlation of density for binary mixtures of methanol+ionic liquids using back propagation artificial neural network

Abstract

Ionic liquids (ILs) are amazing fluids introduced as a replacement for conventional solvents due to their unique properties. Unfortunately, they have several unfavorable features such as high viscosity, which makes pumping them difficult on industrial scale. In this regard, several researchers mix the ionic liquids with each other or some conventional solvents, organic and inorganic compounds, to eliminate those unfavorable features. So the binary properties of the ILs mixtures have been increasingly measured and correlated through the past years. One of the most widely used solvents and additives in the different chemical industries is methanol. In the present investigation, the capability of artificial neural networks for correlating the binary density of the ILs systems containing methanol as a common part (total of 426 experimental data points) has been examined. The results revealed that the best network architecture obtained in this study was feasible to correlate the binary densities of the ILs mixtures with average absolute relative deviation percent (AARD%), average relative deviation percent (ARD%) and correlation coefficient (R2) values of 0.85%, −0.05 and 0.9948, respectively.