Mass transfer study in soybean oil extraction using ethanol and mixtures with p-cymene as co-solvent: Early uses of pore volume and surface diffusion model

Abstract

Green solvents are widely explored for extracting oil from raw vegetable materials. As an alternative to hexane, ethanol is used in oil extraction but exhibits partial miscibility, requiring a co-solubilizer. p-Cymene, selected based on Hansen solubility parameter ($\mathrm{HSP}$), enhances miscibility due to the reduced distance between solute and solvent, as indicated by the decrease in relative energy difference ($\mathrm{RED}$). This study evaluated the pore volume and surface diffusion model (PVSDM) for vegetable oil extraction, with statistical results supporting its suitability. The best fits obtained were 0.990, 0.076, and 0.004 for $R^2$, $\mathrm{SAE}$, and $\mathrm{RMSE}$, respectively. The experimental data fit better to the PVSDM than the power-law models. While the So and MacDonald model fit more accurately, inconsistencies arose due to the non-stationary profile of the experimental data. Mechanistically, the PVSDM provided richer insights into mass transfer compared to the mass transfer kinetic model (MTKM). Extraction at 25 °C was influenced by the co-solvent addition, which may aid energy savings. ${\chi }^2$ and Fisher tests confirmed the model’s predictive accuracy. These findings are valuable for further exploration of soybean oil as a feedstock for food and biodiesel production.