Experimental Optimization and Modeling for Total Removal of Basic Red Cl18 Dye by Walnut Shell: An Integrated ANN-PSO

Abstract

In this study, purified walnut shell powder was utilized as a natural adsorbent for the removal of the cationic textile dye Basic Red Cl18 (BR-Cl18) from aqueous solution. Comprehensive characterization of the walnut shell powder, using scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy, BET analysis, X-ray diffraction, and Fourier transform Infrared spectroscopy, revealed its mesoporous structure, crystalline cellulose composition and respective functional groups, as well as its important specific surface area. An optimization methodology combining artificial neural networks and particle swarm optimization was employed to identify the optimal conditions for achieving complete BR-Cl18 removal. The optimal conditions to achieve 100% removal efficiency were computed, including an initial dye concentration of 94.72 mg/L, a contact time of 53.41 min, a temperature of 27.12 °C, a pH of 5.30, an adsorbent dosage of 4.88 g/L, and a stirring speed of 173.74 rpm. A Multi-Layer Perceptron model with Adam optimizer and ReLU activation (128 nodes) predicted removal with high accuracy (R² = 0.9623, MSE = 0.00096). The adsorption isotherms and kinetics at various dye concentrations indicated that the adsorption process was best described by the Freundlich isotherm (R² > 0.96) and the pseudo-second-order kinetic model (R² > 0.98), as indicated by their high correlation coefficients. Furthermore, the thermodynamic study demonstrated that the adsorption was spontaneous, exothermic, and characterized by chemisorption. The findings highlight the significant potential of walnut shell powder as an efficient sorbent for dye removal from wastewater and confirm the effectiveness of the integrated ANN-PSO approach in optimizing operational conditions.

Graphic Abstract