Ion exchange for solvent recovery in deacidification of vegetable oils by solvent extraction

Crude vegetable oils are extracted from oil seeds and contain approximately 98% triglycerides, lower amounts of partial glycerides (mono- and diglycerides), free fatty acids, phospholipids, carotenes, tocopherols and other impurities. Free fatty acids are more susceptible to oxidation than the other components in the oil and cause to oxidative degradation in oils. Undesirable compounds in the raw oils usually restrict their industrial application because of the adverse effect of impurities on the quality of the product and economic feasibility of the process. Thus, the crude vegetable oils are subjected to pretreatment called refining and generally refers to deacidification, bleaching and deodorization. Among these steps, the removal of fatty acids (deacidification) is the most important one because of the stated reasons above. In industrial scale, the two most widely used refining processes are physical and chemical (alkali neutralization) deacidification and they differ in the manner the free fatty acids removed. Although chemical refining process is one of the most widely used deacidification method, for high acidity oils it leads to considerable loses of neutral oil due to saponification and emulsification. There are also substantial energy requirements and generation of industrial waste in chemical refining. The physical refining, has some advantages over chemical refining such as improved product yield, elimination of soap stock as well as reduced effluent quantity, less steam, water and power consumption. However, it has also some drawbacks such as rigorous pretreatment requirement to remove phospholipids and not suitable for heat sensitive oils. Since these conventional deacidification methods are not appropriate for pretreatment of high acidity oils, other methods such as chemical reesterification, supercritical fluid extraction, membrane processing and solvent (liquid-liquid) extraction have been investigated as alternatives. Liquid-liquid extraction is based on the differences in the solubility of free fatty acids and triglycerides in an organic solvent and has many advantages. Due to the mild operational conditions (at room temperature and atmospheric pressure) the loss of neutral oil and nutraceutical compounds can be restricted by selection of an adequate solvent. As a consequence of the very high volatility difference between solvent and neutral oil, separation of the residual solvent from the refined oil and extract phases can be easily performed by distillation, stripping or evaporation. However, in terms of energy efficiency solvent recovery from the extract phases for reuse in the liquid-liquid extraction process is still a challenge. Because of the chemical stability and adsorption ability of strong anionic ion exchangers in organic solvents, ion exchange processes for separation of free fatty acids from short chain alcohols have gained wider attention as an efficient and low energy option in last decades. Also, using of ion exchange resins to remove free fatty acids from vegetable oil decreases the waste generation and minimize the loss of neutral oil. Researches using ion exchange resins for solvent recovery from the fatty acid+solvent mixtures are scarce in the literature. Moreover, most of the works to date has focused on the separation of free fatty acids from ethanol or ethanol+water mixtures. However, to the best of our knowledge, in the literature there is not any report for separation of fatty acids from metanol by ion exchange. Biodiesel production is one of the most important non-food industrial applications of vegetable oils. In the transesterification with alkali catalysis the free fatty acid content of the oil must be below 1% for acceptable quality and production efficiency of biodiesel product. For this reason, it is necessary to apply the acid removal process as a preliminary step for the high acidity oils to be used as the raw material in biodiesel production. Methanol, which is used as a reagent in biodiesel production, is also the most suitable solvent for deacidification of crude vegetable oils by liquid-liquid extraction because of its high selectivity for free fatty acid. It is known from studies in the literature that the commercial strong anion exchanger known as Amberlyst A26 OH provides high separation yields in ethanol-fatty acid solutions. However, no data has been found in the literature regarding the use of the same ion exchanger for the removal of fatty acids from methanolic fatty acid solutions. Therefore, in this study, it is aimed to investigate the usability of Amberlyst A26 OH anion exchanger for the separation of fatty acids from methanolic extract phases. For this purpose, the study was planned and carried out in two steps, namely the determination of the ion exchange equilibrium isotherms and the examination of the ion exchange kinetics. Since oleic acid is one of the major fatty acids in many vegetable oils, it was selected as representative fatty acid for experimental study. In order to obtain ion exchange equilibrium data, methanol + oleic acid mixtures having an oleic acid concentration ranging from 1 to 15% by weight were prepared to represent the methanol extract phases in solvent extraction. In all experiments the ratio of ion-exchange resin/solution and the equilibrium temperature was kept constant as 1/3 and 30 °C, respectively. In order to guarantee the swelling of the resin and the removal of the excess moisture, the resin was preconditioned before each experiment. The mixtures were allowed to stand in a shaking water bath at constant rate of 155 rpm for 5 hours until equilibrium was achieved. At the end of this period, the resin and solutions were separated by filtration. The equilibrium oleic acid concentration in the solutions (C*) was determined by titration with a standard sodium hydroxide solution in the presence of the phenolphthalein indicator. Before and after each equilibrium run the methanol content of the resins was determined by drying appropriate amounts of samples until constant weight in a vacuum oven at 30 °C. The equilibrium oleic acid content of the resin q* (mg OA/g of dry resin) was calculated from the mass balance. The experimental data were evaluated using different models known as Langmuir, Freundlich, Redlich-Peterson, and modified Langmuir isotherms to determine the model that best describes the ion exchange equilibrium for the investigated system. In kinetic experiments methanolic solutions with initial oleic acid concentrations of 2%, 8% and 15% by weight were used. A known amount of ion exchange resin was contacted with a known quantity of methanolic solution of oleic acid at 30 °C in a stirred jacketed glass reactor. The stirring rate was kept constant at 170 rpm to provide homogeneous dispersion of the ion-exchange resin in the liquid phase. The amount of oleic acid in the solution was determined in the samples taken at definite time intervals. The amount of oleic acid retained in the resin (qt mg OA/g dry resin) at the given time periods was determined from the related solution concentration by mass balance. It is generally accepted that ion exchange takes place in the following steps: • Film diffusion through the liquid layer on ion exchanger surface • Intraparticle diffusion in the pores of ion exchanger • Chemical reaction between the counter ions and functional groups of ion exchanger. The rate-determining step of ion exchange in the Amberlyst A26 OH-Oleic Acid-Methanol system was determined by evaluating experimental data with the related kinetic models. As a result of this study it was concluded that: • the separation of oleic acid with a high efficiency (99%) from the representative model solutions for methanolic extracts is technically feasible by using Amberlyst A26 OH anion exchange resin. • ion exchange equilibrium data do not comply with Freundlich, Langmuir and Redlich-Peterson isotherm models. • the modified Langmuir isotherm considering the dependency of ion-exchange capacity on the concentration of counter ions in the solution best describes the experimental data obtained. • the concentration dependence parameter of the modified Langmuir isotherm as high as x = 13 shows the ion exchange capacity of the Amberlyst A26 OH for the investigated system is highly dependent on the concentration of oleic acid in the solution. • the rate determining step in ion exchange is the diffusion of oleic acid from the liquid film layer on the surface of the resin particles • in a batch ion exchange process, removal of oleic acid from 14.5%, 7.5%, and 2.2% oleic acid solutions can be performed with efficiencies of 66%, 91%, and 99% in 5 minutes.