Modeling and optimization of experimental parameters for the removal of heavy metal and production of porous adsorbents by RSM and ANN techniques

Abstract

Alongside the swift development of a wide variety of industrial sectors, heavy metals containing wastewater is to pose a grave threat to the environment and human health. Heavy metals, unlike organic wastes and contaminants, are non-biodegradable and highly toxic even in a small amount. Various techniques have to date been proposed and employed to remove heavy metals from aqueous medium. One of these techniques, adsorption is the most preferred wastewater treatment because of its low-cost, high metal adsorption capacity and easy operation. To enhance the heavy metal removal efficiency of adsorbents used in adsorption processes the precursor is usually thermally carbonized. In this study, bovine animal waste was chemically activated and carbonized. The obtained highly porous adsorbents were utilized for the removal of lead heavy metal in an aqueous medium. Experimental parameters were optimized using Response Surface Method (RSM) and Artificial Neural Networks (ANN) for microwave irradiation power, impregnation ratios and carbonization temperatures. To check the accuracy of results, several statistics such as R2, RMSE (root mean square error), mean squared error (MSE), mean absolute error (MAE) and Person’s Chi-square measure were used. To determine input-output behavior in ANN algorithm hidden neuron numbers and tan-sigmoid, log-sigmoid and purelin transfer functions were also optimized. Prediction by both the RSM and ANN models was successfully evaluated and both of them showed a similar performance (R-value ≈ 0.99). Additionally, adsorption mechanism was enlightened by Brunauer-Emmett-Teller (BET), Scanning electron microscopy SEM-EDX, x-Ray diffraction (XRD), and Fourier transform infrared (FTIR) analysis. The adsorbents used in experiments were found to be mainly calcium, phosphorous and oxygen containing mesoporous materials with ≈ 30 g/m2 of surface area.