Removal of Congo red by magnetic nano-alumina using response surfacemethodology and artificial neural network

摘要

Easily separable magnetic nano alumina was synthesized for the removal of the toxic dye Congo red from aqueous solution. Magnetic nano alumina particles were characterized by SEM-EDX. The magnetic property of the sorbent was evaluated by the VSM method. The obtained saturation magnetization of 15.88 emu g(-1) showed facile separation of magnetic nano alumina. The effect of influential parameters; pH, temperature, time, initial dye concentration and amount of sorbent on the removal (%) were investigated. The removal (%) was mathematically described as a function of experimental parameters and central composite design (CCD) was applied to estimate the process. The same design was used for a three layer artificial neural network (ANN) model. The predicted CCD data vs. ANN showed a regression value of 0.9999. This linear agreement indicated that the CCD and ANN could ideally predict the process. The results of the two models were compared in terms of coefficient of determination and mean absolute percentage error to indicate the prediction potential of CCD and ANN. The obtained results confirm higher capability and accuracy of ANN in prediction compared with CCD. The experimental data were found to be properly fitted to the Langmuir and Freundlich model which indicates that the sorption takes place on a heterogeneous material. A sorption capacity of 27.397 (mg g(-1)) was achieved for Congo red.