Predicting ionic liquid based aqueous biphasic systems with artificial neural networks

The development of artificial neural networks to describe the formation of [C(4)mim][CF3SO3]-based aqueous biphasic systems (ABSs) for a broad range of salts, composed of diverse combinations of cations and anions is reported here. Three different artificial neural networks were used for predicting the cation effect of the chloride-and acetate-based salts and the anion effects of sodium salts to induce the formation of ionic-liquid-based ABSs. For the development of ANN, the input data set was classified into three sets: seventy five percent of the total data were selected for training and twenty five percent of the data were selected for testing and validation. The results show that using the proposed neural network model a good agreement between the experimental data points and the predicted values was achieved. These results suggest that the ANN method may be a useful tool for predicting the formation of IL-based ABS. (C) 2014 Elsevier B.V. All rights reserved.