Supercritical carbon dioxide extraction of carrot fruit oil - Comparison with hydrodistillation and modelling

Abstract

In this study the compositions of Carrot fruit (Daucus carrota, cultivar Chanteney) essential oil obtained by hydrodistillation and its extract obtained by supercritical carbon dioxide extraction at 313 K and 10 MPa were examined. The qualitative and quantitative analyses of the essential oil and supercritical extract were performed using GC and GC/MS methods. Mathematical modeling of experimental results obtained by supercritical extraction at 313 K and 10 MPa with different particle size, as well as at 323 K and 9 MPa was also performed. GC/MS analysis showed that in the case of both, hydrodistillation and supercritical extraction, the main component of carrot oil was carotol. However, the composition, as well as the yield of obtained oil was different depending on applied technique of extraction/distillation. The yield of essential oil of 0.69 wt% was obtained by hydrodistillation, while extraction with CO2 under supercritical conditions (at 313 K and 10 MPa) gave yield of 1.17 wt%. The main reason for this difference was that heavier compounds like olerosines and waxes were also extracted by supercritical CO2 and not only the essential oil. Moreover, the lighter compounds like pinenes were not found in the supercritical extract. Obtained yields by supercritical fluid extraction were higher at higher pressures and lower temperatures. Results of supercritical extractions showed that particle size had no influence on the extraction yield. It seems that plant structure which contains secretory ducts as oil reservoirs, proved by SEM (Scanning Electron Microscopy) is responsible for such influence of particle size on the extraction yield. Mathematical modeling of supercritical extraction results included the application of a new model based on the secretory structure (MSS) of carrot fruit. Model showed good agreement with experimental results, and confirmed that particle size had no influence on the yield. This model had been previously used to predict similar behavior in the case of other plant materials with same structural characteristics.