Artificial Neural Networks-Adaptive Neuro-Fuzzy Inference Systems Modelling of Heavy Metal Adsorption from Steel Mill Wastewater Using Periwinkle Shell Hydroxyapatite

Abstract This study explores the removal of lead ions (Pb²⁺) and cadmium ions (Cd²⁺) from steel rolling mill wastewater using hydroxyapatite synthesized from periwinkle shells (PSHAP) as a new adsorbent. Characterization of PSHAP through SEM, EDX, FTIR, and XRD confirmed its porous structure, elemental makeup, and crystalline features suitable for heavy metal adsorption. Batch experiments assessed the effects of pH, adsorbent amount, contact time, metal ion concentration, and temperature. Results showed maximum removal efficiencies of 94.6% for Pb²⁺ and 94.9% for Cd²⁺ under optimal conditions. Adsorption followed the Langmuir isotherm model (R² >0.99), indicating monolayer adsorption, and pseudo-second-order kinetics, implying chemisorption. Thermodynamic analysis indicated the process is spontaneous and exothermic. RSM and ANFIS were employed for modelling and optimization. ANFIS models exhibited high predictive accuracy (R² = 0.9965 for Pb²⁺, 0.9366 for Cd²⁺) with Gaussian and linear membership functions. The study confirms PSHAP's potential as an effective, sustainable adsorbent for removing heavy metals from industrial wastewater