Melamine-Derived MgO/AC Nanocomposite for Efficient Dye Removal: Integrating RSM and Machine Learning Modeling.

In this study, a melamine-derived MgO/activated carbon (MgO/AC) nanocomposite was synthesized and evaluated for methylene blue (MB) removal from aqueous solutions. The incorporation of MgO nanoparticles and nitrogen functionalities originating from melamine generates chemically active adsorption sites that enhance dye-adsorbent interactions beyond surface area effects. Despite a relatively low BET surface area (15.19 m2/g), the MgO/AC composite achieved a high MB removal efficiency of 94.72% under optimized conditions using a low adsorbent dosage. Response surface methodology was employed to statistically optimize the operating parameters, while machine learning models were applied to capture nonlinear adsorption behavior. Among the evaluated models, ANFIS exhibited superior predictive performance (R2 = 0.981), indicating its suitability for modeling complex adsorption systems. This integrated experimental-modeling approach highlights the potential of melamine-derived MgO/AC as an efficiency-oriented adsorbent for dye-contaminated wastewater treatment.