Submitted to: Annual Meeting and Expo of the American Oil Chemists' Society
Publication Type: Abstract Only
Publication Acceptance Date: May 4, 2003
Publication Date: N/A
Technical Abstract: Currently, soybean oil is extracted from soybean meal using hexane as the solvent. After the extraction is completed, the hexane and soybean must be separated before the soybean oil can be used and the hexane recycled. Hexane is currently removed from the miscella (i.e., soybean oil/hexane mixture) by double-effect evaporation and steam distillation. This process, however, requires a great deal of energy and therefore is very costly. Previous research has shown that supercritical carbon dioxide (i.e., temperatures and pressures in excess of 40C and 1356 psi, respectively) could effectively be used to remove hexane from the miscella. Because supercritical CO2 extractions are generally more costly to perform than liquid CO2 extractions, this study investigated the use of liquid CO2 to remove hexane from the soybean oil/hexane mixture. CO2 at 1350 psi and 25C was passed through 50 mL of a 10% hexane in SBO solution at a rate of ca. 3-4 L/min (expanded gas). Four different volumes of CO2 (i.e., 100, 300, 500, and 1000 L) were compared. After passing through the SBO/hexane mixture, the CO2 was passed through a chilled collection vial to capture the hexane as well as any triglycerides which were extracted. After the extraction, the 10% hexane in SBO solution was removed from the extraction cell and analyzed for residual hexane using ISO Method 9832:2002. This method involves a GC analysis of the headspace over the SBO with an added internal standard and comparison to a standard curve. The amount of SBO in the collection vial increased as the amount of CO2 used increased, although it was still generally very low (i.e., less than 1.5g). Although it was expected the amount of hexane would also increase as the amount of CO2 used increased, there was not a clear relationship between the two. It is hypothesized that the collection efficiency was low and the hexane tended to be entrained with the CO2 and lost rather than collected. As expected, the ppm residual hexane decreased as the amount of CO2 used increased and even with the lowest amount of CO2 used (i.e., 100 L), the residual hexane was less than 2 ppm. This level of residual hexane is well below the 1000 ppm water and volatile matter allowed.