Development, application, and evaluation of artificial neural network in investigating the removal efficiency of Acid Red 57 by synthesized mesoporous carbon-coated monoliths

摘要

Acid Red 57 (AR57) has been successfully removed from aqueous solution by adsorption on our synthesized mesoporous carbon-coated monolith (MCCM). For the first time, a powerful artificial neural network (ANN) model has been developed to predict the removal efficiency of AR57 on MCCM. Three critical parameters in adsorption systems, that is, solution's initial pH, initial dye concentration, and contact time were incorporated in the ANN model in order to optimize the observed adsorption process. Langmuir and Freundlich adsorption models were then fitted to the adsorption data to estimate the adsorption capacity. It was concluded that Langmuir isotherm was best-fitted to the data showing a maximum monolayer adsorption capacity of 1,162.7mg/g. The pseudo-first-order and pseudo-second-order kinetic models were subsequently tested to evaluate the kinetics of the adsorption process. It was revealed that the adsorption kinetics could be better represented by the pseudo-second-order model. A comparison was finally drawn between ANN and pseudo-second-order kinetic models. Based on the error analyses and determination of coefficients, ANN was the more appropriate model to describe the studied adsorption process.