Integrating AI-based neural network modeling with experimental characterization for Cd(II) ion adsorption using...

Cadmium (Cd) contamination in wastewater presents serious environmental and public health challenges, requiring efficient mitigation strategies. This research focuses on assessing the biosorption capability of Sargassum fusiforme (SF) biosorbent for eliminating Cd(II) ions from water-based solutions. Based on proximate analysis, the carbohydrate content of SF was found to be 56.8 %, primarily composed of polysaccharides, which enhanced its biosorption capacity. The effective adsorption of Cd(II) ions was confirmed through characterization techniques, including SEM, EDX, FTIR, and XRD analyses. Following adsorption, there were noticeable morphological changes and the emergence of a characteristic Cd peak in the EDX spectrum. Batch experiments conducted at pH 6, using a biosorbent dose of 2 g/L, a contact time of 40 min, and an initial Cd(II) concentration of 20 mg/L, achieved a maximum removal efficiency of 99.25 %. The adsorption data aligned well with the Langmuir isotherm model, indicating monolayer adsorption with a maximum capacity of 144.87 mg/g. Kinetic studies revealed that the process followed pseudo-first-order kinetics, while thermodynamic analysis confirmed the exothermic and spontaneous nature of the adsorption. Artificial Neural Network (ANN) modelling proved to be an effective approach for predicting biosorption efficiency, demonstrating excellent prediction accuracy with R-values reaching 0.99 in training, validation, and testing datasets. Moreover, SF exhibited good regeneration potential, retaining over 60 % of its initial adsorption efficiency after eight adsorption-desorption cycles, highlighting its sustainability for practical wastewater treatment applications.