The use of a multilayer perceptron (MLP) for modelling the phenol removal by emulsion liquid membrane

Abstract Extraction of phenol from aqueous solution was studied using emulsion liquid membrane (ELM). In this study, due to the extreme complexity and non linearity of ELM process, a multilayer perceptron (MLP) was developed to predict the extraction efficiency of phenol. The effect of operational parameters such as: the ratios of volume ratio of internal phase to organic phase, volume ratio of emulsion to aqueous external phase, the emulsification speed and time, the surfactant concentration, the extractant and sodium hydroxide concentrations were studied to optimize the conditions for maximum removal of phenol. The results showed that a network with 3 hidden neurons was highly accurate in predicting the extraction efficiency (more than 98%). This accuracy was reflected by high correlation coefficient R = 0.99 and a root mean square error below 0.5. The result indicated that the MLP model explained in this study is an applied tool to predict the extraction efficiency of phenol by ELM.