Integrating Artificial Intelligence with Graphene Nanosheet-Modified Polyamide Membranes for Wastewater Treatment: A Dual Environmental Solution

Global water scarcity and increasing industrial effluent discharge have intensified the need for advanced wastewater treatment technologies. With decreasing freshwater resources, advanced and optimized polyamide membrane technologies are urgently needed to enable sustainable purification, pollutant removal, and enhanced water recycling. In this study, polyamide/support membranes modified with amine-functionalized graphene oxide (GO) nanosheets were synthesized and evaluated for their ability to simultaneously reject multiple pollutants, salts, heavy metals, and hydrocarbons under different operating pressures. The GO-modified membrane exhibited enhanced hydrophilicity, a pure water flux of 200 L m –2 h –1, and sulfate rejection of 78% at approximately 500 kPa. To support process optimization, ensemble machine learning (ML) algorithms, including Random Forest, Gradient Boosting, and eXtreme Gradient Boosting, were applied to predict the performance of membranes in wastewater treatment. Model performance was evaluated using independent 80–20 train–test splitting and 5-fold cross-validation. The models achieved high predictive accuracy, with R 2 values of 0.99 for sulfate, 0.89 for metals, and 0.91 for hydrocarbons. Rather than relying solely on labor-intensive empirical trials, ML served as a sustainability-driven optimization tool, reducing experimental effort by predicting optimal operating conditions and membrane compositions. These results demonstrate the potential of combining GO-based membranes with ML-assisted operational optimization for wastewater treatment. This study provides a sustainable framework for integrating functional nanomaterials and data-driven modeling in membrane-based water purification systems.