Meng-en LiuLing LiuHasan Sh. MajdiIbrahim AlbaijanMeldi SuhatrilDalia H. ElkamchouchiJose Escorcia-GutierrezRiadh MarzoukiASSILZADEH, HAMIDHAMIDASSILZADEH2025-11-062025-11-062025-12https://doi.org/10.1016/j.seppur.2025.133899This study introduces FeZr-MOF/Graphene Oxide (GO) composite membranes as an innovative application of inorganic membrane technology for advanced separation and purification. These membranes were engineered to efficiently remove industrial dyes such as Congo red and methylene blue from aqueous solutions by leveraging the high surface area, chemical stability, and selective adsorption capabilities of FeZr-Metal-Organic Framework (FeZr-MOF). Based on laboratory evaluations, the composite membranes exhibited strong adsorption performance, achieving 87.5 mg/g capacities for Congo red and 80.2 mg/g for methylene blue. Integrating GO into the membrane matrix enhanced hydrophilicity, water permeability, and anti-fouling behavior, as evidenced by over 90 % permeability recovery after multiple cleaning cycles. This research reflects the latest advancements in inorganic membrane technology and bridge material innovation with industrial water purification needs. Machine Learning (ML), implemented via the Random Forest (RF) algorithm, was used to predict the impact of key operational parameters—pH, dye concentration, and contact time—on removal performance, and the model's predictions aligned closely with experimental trends. These findings, supported by experimental data and predictive modeling, highlight the potential of FeZr-MOF/GO membranes as a scalable and robust material for sustainable water purification applications.enAnti-Fouling PropertiesComposite MembranesFeZrInorganic Membrane TechnologyMachine Learning OptimizationMetal-Organic FrameworkWater Purificationtext::journal::journal article