Analysis of indium (III) adsorption from leachates of LCD screens using artificial neural networks (ANN) and adaptive neuro-fuzzy inference systems (ANIFS)

Abstract

Ten different adsorbent materials were tested to adsorb indium (III) from leachates of LCD screens, aiming to concentrate this valuable material. Artificial neural network (ANN) and adaptive neuro–fuzzy inference system (ANIFS) were applied to analyze the indium (III) adsorption. The input variables for the network models were: specific surface area, point of zero charge, adsorbent dosage and contact time. Adsorption capacity (q) was used as output variable. The adsorption capacity values ranged from 8.203 to 1000 mg g–1. The ANN modeling presented the best fit when the Levenberg–Marquardt algorithm was used. The ANFIS modeling presented the optimum performance when the hybrid method was used. Among the tested adsorbents, chitosan presented the best performance; attaining adsorption capacity of 1000 mg g–1 within 20 min. This is an excellent value since the maximum indium concentration in LCD screens is 0.613 mg g–1. This high capacity was attributed to the coordination ligation between chitosan and indium (III).