Multi-walled carbon nanotubes green-functionalized with iron nanoparticles for continuous removal of pharmaceutical pollutants through fixed-bed adsorption: Integrated experimental and machine learning approaches.

Pharmaceutical residues, including losartan and diclofenac, are insufficiently removed by conventional wastewater treatment plants, leading to persistent environmental contamination and potential public health risks. This study addresses this issue by investigating the continuous adsorption of these pharmaceuticals in a fixed-bed column utilizing green-functionalized carbon nanotubes as a sustainable and efficient adsorbent. The adsorbent material was underwent to comprehensive characterization through particle size analysis, zeta potential measurement, CHNS elemental analysis, and X-ray fluorescence, confirming its physicochemical suitability and successful functionalization. Experimental adsorption tests indicated that flow rate significantly influences removal efficiency, with lower flow rates (0.2 mL/min) enhancing retention and extending the mass transfer zone, particularly for losartan. Additionally, higher initial concentrations resulted in earlier breakthrough and saturation, but increased adsorptive capacity. For mass transfer modeling, the modified dose-response (MDR) and dual-site diffusion (DualSD) models provided the best fit to the experimental data. Furthermore, an artificial neural network model demonstrated high predictive accuracy (R = 0.9772; MSE = 0.0033), reinforcing the robustness of the system. Among the approaches tested, the DualSD model exhibited the most reliable performance based on parametric statistics (Radjust and AICc). These findings demonstrate the potential of this green adsorbent for scalable application in the treatment of pharmaceutical-contaminated effluents under continuous flow conditions.