Biodiesel production by iron oxide supported on light weight expanded clay aggregate (Fe2O3-f-LECA) Catalyst: Box-Behnken design-based optimization and ANFIS modeling with GA and PSO

This study presents the synthesis of an iron oxide-supported functionalized lightweight expanded clay aggregate (f-LECA) catalyst via the microwave solution combustion method for biodiesel production. Characterization techniques, including XRD, FESEM, TEM, BET-BJH, and TPD-NH₃, confirmed the uniform dispersion of iron oxide on f-LECA, leading to enhanced catalytic performance. BET-BJH analysis revealed a specific surface area of 8.35 m²/g, an average pore width of 10.47 nm, and a well-developed pore structure conducive to catalytic reactions. The catalyst exhibited strong acidity, as evidenced by ammonia desorption peaks at 346 °C and 390 °C, indicating the presence of medium-strength acidic sites essential for transesterification. A maximum biodiesel conversion of 96.47 % was achieved under optimized conditions: 100 °C, a methanol-to-oil molar ratio of 20, 3 wt% catalyst loading, and 1-h reaction time. Optimization using Response Surface Methodology (RSM) and Adaptive Neuro-Fuzzy Inference System (ANFIS) modeling demonstrated a strong correlation (R² > 0.96) between predicted and experimental results. Further refinement using the Particle Swarm Optimization (PSO) algorithm enhanced model accuracy, yielding minimal error values based on MAE, MSE, RMSE, and R² metrics. The excellent catalytic activity, economic feasibility, and environmental sustainability of f-LECA-supported iron oxide make it a promising candidate for large-scale biodiesel production from oleic acid.