Machine learning and mechanistic studies on p-nitrophenol remediation using sustainable activated carbon

Abstract

Abstract This study presents a sustainable approach for p -nitrophenol ( p NP) removal by synthesizing activated carbon with a large surface area from waste Pistacia vera shells using H 3 PO 4 activation. Comprehensive characterization confirmed a high specific surface area (670.25 m 2 g − 1 ) and a heterogeneous structure rich in functional groups, which are beneficial for adsorption. Adsorption followed pseudo-second-order kinetics, and the equilibrium dataset followed the Langmuir isotherm, with a maximum adsorption capacity of 142.93 mg g − 1 . The thermodynamic results showed that the adsorption was a spontaneous and exothermic (− 13.43 kJ mol − 1 ) process. ANN and ANFIS models were developed to predict the adsorption behavior. The ANFIS model exhibited the best predictive performance, with an R 2 of 0.9935. ANFIS sensitivity analysis identified that contact time and initial p NP concentration were the key factors. Regeneration tests demonstrated that the adsorbent could be reused for five cycles, supporting its practical applicability. Real water matrix studies have demonstrated robust p NP removal in real-world water scenarios, highlighting its environmental relevance. Thus, waste-derived adsorbents present a low-cost and sustainable option for p NP removal.