Reducing waste and carbon footprint in essential oil production: Artificial intelligence–optimized solvent-free microwave extraction for sustainable food and pharmaceutical applications

Essential oils are being increasingly used as flavor, fragrance, and bioactive ingredients, yet conventional hydrodistillation (HD) can require multiple hours of heating and high water use. Solvent-free microwave extraction (SFME) is a green alternative that enables rapid extraction without the addition of solvent or distillation water. In this study, we report an artificial intelligence (AI)–assisted optimization workflow that couples a central composite design (CCD) and response surface methodology (RSM) with an artificial neural network (ANN) surrogate model to identify operating conditions that maintain yield while reducing electricity demand. Bench-scale experiments were performed in a 1 L reactor and verified on a 12 L bench-scale demonstration unit (which was not a full pilot plant). To strengthen reproducibility and sustainability transparency, we (i) report the CCD structure and ANN training/validation strategy and (ii) define a gate-to-gate system boundary with auditable electricity-to-CO₂e accounting using a stated grid emission factor. Model comparison is supported by parity/residual plots and multimetric evaluation on a consistent test protocol.