Removal of Cu(II) using three low-cost adsorbents and prediction of adsorption using artificial neural networks

Abstract

Adsorption of copper using groundnut seed cake power, sesame seed cake powder and coconut cake powders as bioadsorbents was optimized at a pH of 5, temperature of 40 °C, initial metal concentration of 10 mg/L, contact time of 30 min and adsorbent dosage 0.75 g for groundnut seed cake powder and 1.0 g of sesame seed cake powder and coconut cake powder. From the results of kinetic studies, it was concluded that the adsorption process followed pseudo-second-order kinetics. Langmuir adsorption isotherm fit perfect for the adsorption of Cu(II) using the three adsorbents. Maximum adsorption capacity was found to be 4.24 mg/g. Artificial neural network modeling which was adopted for the predication of adsorption of Cu(II) using groundnut seed cake powder, sesame seed cake powder and coconut cake powder was carried out by using back-propagation algorithm. Correlation plot drawn for the experimental and ANN predicted values showed a strong correlation coefficient of 0.989, indicating that the network trained fit apt for the prediction of adsorption process.