USING LIQUID CARBON DIOXIDE TO SEPARATE HEXANE FROM SOYBEAN OIL/HEXANE MIXTURES

Authors: Eller, Fred; Taylor, Scott

Submitted to: Intl Symposium on Supercritical Fluid Chromatography and Extraction
Publication Type: Abstract Only
Publication Acceptance Date: 7/28/2004
Publication Date: 7/28/2004
Citation: Eller, F.J., Taylor, S.L. 2004. Using liquid carbon dioxide to separate hexane from soybean oil/hexane mixtures. Intl Symposium on Supercritical Fluid Chromatography and Extraction. Abstract Book. p. 272.

Interpretive Summary:

Technical Abstract: The use of liquid carbon dioxide (L-CO2) was investigated as a means to separate hexane from the mixture of soybean oil (SBO) and hexane resulting from the hexane extraction of soybeans. Using a 2.5 meter tall fractionation tower, five different volumes of CO2 (i.e., 100, 200, 300, 500 and 1000 liters expanded gas) were passed through hexane/SBO mixtures to determine the amount of L-CO2 required to effectively remove the hexane from the mixture. Two concentrations of n-hexane (i.e., 25% and 10% w/w) as well as a 25% mixture of isohexane in SBO were investigated. The 25% and 10% hexane/SBO mixtures represent the concentrations of hexane in SBO after the first and second stage evaporators, respectively. L-CO2 at 25 deg C and 1350 psi was passed through 50 mL of the hexane/SBO mixture at a rate of 3-4 L/min (expanded gas). After passing through the hexane/SBO mixture, the expanded CO2 was passed through a chilled round bottom flask to capture extracted hexane and SBO. The raffinate SBO was removed from the column and analyzed for residual hexane using ISO Method 9832:2002. There was no clear relationship between the amount of hexane collected and liters of CO2 and it is hypothesized that this was a result of low collection efficiency for the hexane using the round bottom flask. The amount of SBO extracted with the L-CO2 increased linearly with liters of CO2. This is likely due to the low solubility of SBO in L-CO2. Significantly more SBO was collected from the 25% hexane/SBO mixtures than from the 10% hexane/SBO mixture. In this case, the hexane may be acting as a co-solvent to increase the solubility of the SBO in the L-CO2. Residual hexane was inversely proportional to the volume of L-CO2 used and was generally less than 20 ppm after 200 liters CO2 (Table 1). The original concentration of hexane (i.e., 25% or 10%) in SBO did not have a signidicant effect on the residual hexane level. Fatty acid analysis of the raffinate as well as the SBO carried over with the hexane revealed that the L-CO2 selectively carried over triglycerides with lower molecular weights leaving higher molecular weight trigylcerides behind in the raffinate. This research demonstrates the utility of using L-CO2 to remove hexane from mixtures of hexane and SBO at both low pressures and temperatures to reduce energy costs associated with conventional distillation.