This study presents liquid−liquid equilibrium data for systems of interest in biodiesel production and the refining of edible oils by liquid−liquid extraction. These systems are composed of vegetable oils + anhydrous ethanol + water, at temperatures ranging from (298.15 to 333.15) K. The following vegetable oils were investigated: refined canola oil, refined corn oil, semiprocessed Jatropha curcas oil, and semiprocessed macauba (Acrocomia aculeata) pulp oil. According to the results, the enhancement of temperature within the study ranges increased oil−solvent miscibility in both phases. The liquid−liquid equilibrium experimental data were correlated using the non-random two liquid (NRTL) model. For all of these systems, the global deviations between calculated and experimental data were lower than 0.7 %, showing the good descriptive quality and applicability of the NRTL model in liquid−liquid analysis and computational simulation processes.

The aim of this work was to determine liquid−liquid equilibrium data for systems of interest in ethyl ester (biodiesel) production. The following systems were investigated: refined soybean oil + anhydrous ethanol + hexane at (313.15 ± 0.1) K and (328.25 ± 0.1) K, refined canola oil + anhydrous ethanol + hexane at (313.25 ± 0.1) K and (328.15 ± 0.1) K, and refined palm oil + anhydrous ethanol + hexane at (318.15 ± 0.1) K and (328.25 ± 0.1) K. The instrumental technique of near-infrared spectroscopy allied to multivariate calibration (in this work, this union was denominated “NIR method”) was used for quantification of the phase compositions. The NIR method presented an appropriate experimental accuracy for quantifying equilibrium data at higher temperatures. Good reliability of the phase compositions was confirmed from values of global mass balance deviations ranging from (0.06 to 0.16) %. The experimental data were correlated using the NRTL model with an average global deviation of 0.83 %.

Phase diagrams of aqueous two-phase systems (ATPS) composed of 400 g • mol -1 poly(ethylene glycol) (PEG) + sodium sulfate or magnesium sulfate were determined at (298.15, 308.15, and 318.15) K. The temperature effect on the biphasic area was not relevant, indicating a small enthalpic contribution associated with the phase separation. The ability of the salts in inducing the formation of the ATPS with PEG 400 followed the order sodium sulfate < magnesium sulfate. The liquid-liquid equilibrium experimental data were correlated using the NRTL (nonrandom two-liquid) activity coefficient model, and new interaction parameters were estimated. The mean deviations between calculated and experimental compositions were lower than 0.99 %, showing the good descriptive quality and applicability of the NRTL model.

This work presents experimental data of systems composed of grape seed oil + oleic acid + ethanol + water from (283.2 to 298.2) K. The solvent selectivity and oleic acid distribution coefficient were more affected by water content in the alcoholic solvent than by temperature. The enhancement of the water content in the solvent and the reduction of temperature within the studied ranges led to greater values of solvent selectivity. Nevertheless, greater oleic acid distribution coefficients were found as a result of the reductions in temperature and water content in the solvent.

This work presents liquid-liquid equilibrium data for systems of interest in biodiesel production, composed of Jatropha curcas oil (1) + oleic acid (2) + ethanol (3) + water (4), at temperatures ranging from (288.15 to 318.15) K and with water mass fractions in the solvent (ethanol + water) of (0.00, 2.00, and 4.00) %. The experimental data were correlated using the nonrandom two-liquid (NRTL) model, and for all systems, the global deviations between calculated and experimental data were lower than 0.96 %, showing the good descriptive quality of the NRTL model. To verify the effect of temperature (T), the water mass fraction of the solvent (w 4s ), and the fatty acid mass fraction in the oil phase (w 2 OP ) on the distribution coefficient of the oleic acid (k 2 ), a complete second-order model was fitted with a determination coefficient value higher than 0.9. The results showed that the oleic acid distribution coefficient was more affected by water content in the solvent than by temperature. Moreover, as a result of the reductions in temperature and water mass fraction in the alcoholic solvent, greater oleic acid distribution coefficients were found.

This work presents densities, viscosities, interfacial tensions, and liquid−liquid equilibrium data for systems of interest in vegetable oil refining and biodiesel production, composed of soybean oil + commercial cis,cis-octadeca-9,12-dienoic acid (linoleic acid) + ethanol + water at 298.2 K with water mass fractions in the alcohol solvent (ethanol + water) of 0.05, 0.08, 0.11, and 0.14. Experimental phase compositions at equilibrium were correlated using the nonrandom two-liquid (NRTL) model. For all systems, the global deviations between calculated and experimental phase compositions were no greater than 0.0005. All physical properties were evaluated by empirical correlations as functions of phase compositions. The absolute relative deviations between calculated and experimental data were no greater than 6.36 %. Water content in the solvent presented a considerable effect on the physical properties of the systems and phase compositions at equilibrium, even more so on its interfacial tensions.

Dissertação submetida à comissão examinadora como parte dos requisitos para obtenção do título