Adsorption of malachite green dye onto almond peel waste: a study focusing on application of the ANN approach for optimization of the effect of environmental parameters

Almond peel waste was collected, characterized, and then used as an adsorbent for removing malachite green (MG) dye from aqueous solutions. The environmental conditions of MG dye adsorption were pH, 2.5–10.5; almond peel dose, 0.25–1.5 g/L; initial MG concentration, 10–60 mg/L; and adsorption time, 0–180 min. These were optimized by using an artificial neural network (ANN) tool. At pH = 8.5, 96.1% of the MG dye could be removed using almond peels. According to the coefficient of determination results, the Langmuir isotherm proved to be the equation that best fit the data of the isotherm study. Furthermore, the kinetic study showed that the data on MG adsorption of the almond peel waste was consistent with the pseudo-second-order model. The ANN model was developed by using a three-layer, feed-forward network with an optimum architecture of 4:10:1. Sigmoid functions were employed in both inputs and hidden layers, as also those hidden in the output layers. The results indicated a high correlation value ( R = 0.976) to predict the entire experimental dataset, which indicated the applicability of the ANN tool, to describe the MG adsorption data in a highly accurate manner. The important conclusion of this study, after comparison with other similar adsorbents used in the adsorption process of dye wastewater, revealed that almond peel waste is a cheap, recyclable, and effective adsorptive agent, thus a good alternative to remove dyes from aqueous solutions. Graphical abstract