Lead removal from water using DES functionalized CNTs: ANN modeling approach

The removal of lead ions from water is investigated in this study, the non-linearity relationship between the parameters used in the adsorption process were modeled. The artificial neural network modeling method has been used to reflect the non-linear relationship between the variables of the process. A novel functionalized carbon nanotube is used to adsorb Pb(II) ions from water. A total of 158 runs were performed at a laboratory scale and used for modeling. The adaptive neuro fuzzy inference system (ANFIS) model more accurately described the adsorption process, the ANFIS R2 was 0.998, and a feed-forward neural network (FF-NN) R2 was 0.993. Various indicators were used to determine the model’s efficiency and accuracy. To classify the adsorption kinetic order, three kinetic models were evaluated. The pseudo-second-order model was well fitted to the simulated data. The maximum relative error was 7.078% for the ANFIS model and 11.62% for the FF-NN, indicating the higher accuracy of the ANFIS model.