Synthesized CaO-Based Nanocatalysts for Sustainable Biodiesel Production: Artificial Intelligence (AI) Optimization of a Green Nanotechnological Strategy toward Reactor Design

This work introduces a sustainable approach for biodiesel production through the synthesis and catalytic evaluation of three heterogeneous nanocatalysts, CaO, BaOCaO, and BaOTiOCaO, derived from waste eggshells. The CaO catalyst was obtained by calcination, while Ba and Ti were incorporated to enhance surface basicity and catalytic efficiency. Characterization using SEM/EDX, TEM, FTIR, and XRD confirmed the formation of mixed-oxide structures with uniform Ba and Ti dispersion and a surface area of 154.6 m²·g^–1. Catalytic testing in the transesterification of waste cooking oil demonstrated the activity trend CaO < BaOCaO < BaOTiOCaO, with maximum biodiesel yields of approximately 80%, 90%, and 96%, respectively. The superior performance and reusability of BaOTiOCaO were attributed to synergistic interactions among the metal oxides. To complement experimental optimization and support scale-up decision-making, machine learning models, Random Forest, Gradient Boosting, and XGBoost, were trained on 189 experimental datasets to predict biodiesel yield based on key process parameters. The Random Forest model showed the highest predictive accuracy (R² = 0.91, RMSE = 5.01), identifying methanol volume, temperature, and nanocatalyst composition as the dominant factors. The optimized conditions predicted a 95.31% maximum yield at 54 °C, 169 min, and a 4:1 methanol-to-oil mole ratio. The combined catalytic and ML insights provide actionable guidance for reactor operation, including temperature windows, residence time estimation, and catalyst loading strategies suitable for batch and continuous transesterification systems. From an environmental perspective, the waste valorization strategy, high catalytic efficiency, and reusability collectively contribute to reduced energy demand, lower waste disposal, and improved life-cycle sustainability. Overall, this work advances green nanocatalyst design and AI-assisted process optimization toward scalable, low-carbon biodiesel production.