Soft Computing and Density Functional Theory-Based Forecasting of Fatty Acid Extraction from Fat, Oil, and Grease (FOG) Deposits Using <scp>d</scp>-Limonene

Fat, oil, and grease (FOG) deposits in wastewater systems present persistent operational and environmental challenges due to their high fatty acid (FA) content and resistance to degradation. Conventional FA extraction methods frequently employ petroleum-based solvents such as n-hexane, which, despite their efficiency, pose significant health and environmental risks. This study investigated d-limonene, a citrus-derived green solvent, as an alternative for FA extraction from FOG matrices. It was supported by a thermostatistical assessment incorporating soft computing models COSMO-RS and density functional theory (DFT). A double-extraction approach was applied under varying temperatures, times, and solid-to-liquid (S/L) ratios to optimize the process. Experimental data were collected using a factorial design and modeled with an artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS), as conventional empirical models failed to accurately predict the results. The findings indicate that higher extraction yields are obtained at elevated temperatures (55 °C) and higher solid-to-liquid ratios (1:55 g/mL). Notably, d-limonene enables FA extraction at significantly lower temperatures (25 °C) while maintaining yields comparable to those achieved with conventional solvents. The pseudo-second order model (R2 = 0.9988) best described the extraction kinetics, yielding a rate constant of 1.903 × 10–4 ± 2.09 × 10–5 min–1. Small positive induced charges in d-limonene molecules may enhance its FA extraction potential, with entropy values comparable to those of conventional solvents, which revealed more favorable solute–solvent interactions for d-limonene compared to n-hexane. This study highlights d-limonene as a promising alternative to petroleum-based solvents, contributing to the development of more sustainable extraction processes.