Mechanistically Guided Machine Learning Modeling of Azithromycin Extraction Using TiO 2 Pickering‐Stabilized Emulsion Nanofluid Membranes

ABSTRACT Azithromycin is a widely prescribed antibiotic frequently detected in aquatic environments due to its extensive consumption and incomplete removal by conventional wastewater treatment processes. In this study, a novel TiO 2 Pickering‐stabilized emulsion nanofluid membrane (ENM) system was developed for the selective extraction of azithromycin from aqueous solutions. The membrane phase consisted of n‐hexane as the diluent, tri‐octylamine (TOA) as the carrier/extractant, Span‐80 as the surfactant, and TiO 2 nanoparticles as stabilizing agents. The incorporation of TiO 2 nanoparticles enhanced emulsion stability, interfacial rigidity, and mass‐transfer characteristics through Pickering stabilization. The effects of nanoparticle concentration, ultrasonication time, and extractant concentration on extraction efficiency were optimized using response surface methodology (RSM). Under the optimized conditions, the ENM system achieved a maximum azithromycin extraction efficiency of 99.5% within only 3 min of contact time. Artificial neural network (ANN) modeling was employed to validate the experimental results and demonstrated excellent predictive capability, with correlation coefficients exceeding 0.99 for the training, testing, and validation datasets. The developed TiO 2 Pickering‐stabilized ENM system offers a rapid, stable, and sustainable approach for pharmaceutical wastewater treatment and selective antibiotic recovery from aqueous environments.