SciELO - Scientific Electronic Library Online

vol.17 issue3Pigments and their solubility in and extractability by supercritical CO2 - I: the case of curcuminComparision of conventional and supercritical CO2-extracted rosehip oil author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Brazilian Journal of Chemical Engineering

Print version ISSN 0104-6632On-line version ISSN 1678-4383

Braz. J. Chem. Eng. vol.17 n.3 São Paulo Sept. 2000 



R. Eggers, A. Ambrogi and J. von Schnitzler
Technical University of Hamburg-Harburg, Germany
Phone: +49-40-42878-3561, Fax: +49-40-42878-2859,


(Received: February 10, 2000 ; Accepted: May 9, 2000)



Abstract - Supercritical CO2 extraction has shown great potential in separating vegetable oils as well as removing undesirable oil residuals from natural products. The influence of process parameters, such as pressure, temperature, mass flow and particle size, on the mass transfer kinetics of different natural products has been studied by many authors. However, few publications have focused on specific features of the raw material (moisture, mechanical pretreatment, bed compressibility, etc.), which could play an important role, particularly in the scale-up of extraction processes. A review of the influence of both process parameters and specific features of the material on oilseed extraction is given in Eggers (1996). Mechanical pretreatment has been commonly used in order to facilitate mass transfer from the material into the supercritical fluid. However, small particle sizes, especially when combined with high moisture contents, may lead to inefficient extraction results. This paper focuses on the problems that appear during scale-up in processes on a lab to pilot or industrial plant scale related to the pretreatment of material, the control of initial water content and vessel shape. Two applications were studied: deoiling of wheat gluten with supercritical carbon dioxide to produce a totally oil-free (< 0.1 % oil) powder (wheat gluten) and the extraction of oil from rose hip seeds. Different ways of pretreating the feed material were successfully tested in order to develop an industrial-scale gluten deoiling process. The influence of shape and size of the fixed bed on the extraction results was also studied. In the case of rose hip seeds, the present work discusses the influence of pretreatment of the seeds prior to the extraction process on extraction kinetics.
Keywords: separating vegetable oils, SCF Solid extraction.




Figure 1 shows a flow sheet of the high-pressure extraction plant at the Technical University of Hamburg, which was used to perform the experiments.





Wheat Gluten

Wheat gluten is used in the food industry for purposes such as the conditioning of meal qualities and as a basic substance for different products. Deoiling of this product using supercritical CO2 in a lab-scale plant shows good extraction kinetics. Wheat gluten is a side product of the milling process of wheat. It is a fine powder, with 98 % of its particles less than 0.2 mm in diameter, and is rich in protein (~80 %) with a water content of around 8 % and an oil content of 1.4-1.8 %. These properties combined with high-pressure extraction conditions are responsible for the compacting of the fixed bed. Scaling up while still using a relatively small extraction vessel (0.6 l inner volume) showed quite unsatisfactory results. Extraction kinetics are slow and the extracted material was not homogeneous, and on an industrial scale (300 l extractor) it was impossible to carry out an extraction due to the total compacting of the bed. Table 1 shows the results of the experiments which were performed to study the influence of different kinds of extraction conditions on the final product (oil content and bed compaction). The first experiments were carried out as follows: to remove the air, the extractor was filled with CO2, until extraction pressure was reached. The automatic valve was then opened so that the CO2 could circulate in a closed cycle.



The first two lines in Table 1 represent a series of experiments which show that at higher pressure and CO2 mass flow the bed compacts during extraction. Reduction of CO2 mass flow leads to longer extraction times, and especially on an industrial scale, to the poor economic performance of the process. A way to reduce the CO2 circulation velocity throughout the bed is to increase the volume of the extractor, which also implies an elevated cost for the plant. Line 4 shows that CO2 circulation interrupted during the extraction process and low fluid flows improve the quality of the final product, but not the deoiling of the material.

Finally an inner basket consisting of a series of three extraction units with different height/diameter (H/D) ratios was used. The appearance of the final product after extraction is shown in the seventh column in Table 1. The best results were obtained at lower H/D ratios.

As mentioned above, any of the solutions proposed to reduce the final oil content and to avoid compaction of the gluten, imply higher costs and construction complications. Hence, the reasons for the high degree of adhesion between the fine particles needs to be studied in more detail. A combination of different factors, such as elevated pressure and the presence of water, most probably results in such problems. During extraction water is co-extracted and partially separated at the bottom of the separator. Under these conditions (6 MPa and 40 °C) the solubility of water in CO2 is around 0.2 %; thus, the remaining water circulates in the closed cycle at a given concentration in the bed.

A series of experiments, which are summarized in Table 2, were performed to analyze the influence of humidity on the quality of the extracted material. An inner basket consisting of two vessels was used. Feed material was placed in the upper vessel, and the other contained silica gel to retain the humidity of the circulating CO2.



The best results were obtained using dry-frozen gluten: not only a powdery product was obtained, but also the oil content after extraction was lower than 0.1 %. The first three experiments showed that the presence of humidity in the raw material, although at different levels, produces inefficient extraction and extracted material of a poor quality.

Alternative drying methods, such as predrying in normal air dryers or predrying using carbon dioxide, must be better studied in order to find a less expensive method to eliminate humidity prior to extraction.

Rose Hip Seeds

The second example of application is the extraction of oil from rose hip seeds (Rosa Mosqueta; Latin: Rosa Aff. Rubiginosa). Rosa Mosqueta, a wild rose hip which grows in the southern and central part of Chile, is well known for the oil extracted from its seeds. It has a high percentage of polyunsaturated fatty acids (44% linoleic 36 % alphalinolenic). The difficulties arise from the low percentage of oil in the seeds and the thick shell. The efficiency of the extraction process is directly related to the pretreatment of the feed material. A few studies on extraction of rose hip seeds have been published. Mabe et al. (1998) compared extraction by supercritical CO2 to the use of hexane but at pressures near the critical point (7 MPa). The influence of pressure and CO2 mass flow during extraction were studied by Schalcke (1998) but no mention was made of pretreatment conditions.

In this work rose hip seeds were processed prior to extraction in three different ways:

1. By milling in a blade grinder, obtaining the following particle distribution:

1-1.5 mm: 54 %

0.5-1 mm: 36.2 %

0.1-0.25 mm: 9.7 %.

2. By flaking with a roller mill with a gap of 1 mm.

3. By pressing with a cage screw press.

Figure 2 shows the residual oil content in the extracted materials as a function of CO2 consumption, all related to the initial mass of the rose hip seeds. In contrast to the extraction of oil from sunflowers (Ambrogi et al., 1996) and other oilseeds, CO2 requirements were very similar for milled and flaked material (from the roller mill). In the case of rose hip seeds, a considerable amount of energy is required to destroy the thick shell of the seed without significantly affecting cell structure. However, prepressing of seeds leads to shorter extraction times and consequently, smaller extractor volumes.



Figures 3 and 4 show the influence of temperature and pressure on mass transfer kinetics. At 50 and 70 MPa all the extraction curves are close together, suggesting that above 50 MPa the influence of pressure diminishes. Increasing the mass flow of CO2 produces opposite effects: while hold-up is reduced, mass transfer is enhanced without any noticeable dependence of extraction kinetics on mass flow. Furthermore, at 50 MPa dependence on temperature in the range studied is negligible. The so-called cross-over point (equal solubility in CO2 at different temperatures) is probably situated nearby.





The saturated fatty acid composition of rose hip oil at different extraction conditions is shown in Table 4. No significant differences were observed when comparing CO2 - extracted oil from pressed cakes (lines 1 to 3 of Table 4) to values for mechanically pressed oil. However, at the beginning of CO2 extraction (line 5) a lower content of linolenic acid was determined. A plausible explanation is that free fatty acids together with triglycerides containing saturated fatty acids were more easily extracted at the beginning of the process.




Egerrs, R., Supercritical Fluid Extraction of Oilseeds/Lipids in Natural Products in Supercritical Fluid Technology in Oil and Lipid Chemistry. AOCS Press, Illinois. 1996.        [ Links ]

Mabe, G.D.; Foco, G.M.; Brignole, E.A. and Bottini, S.B., Extraction of Rosa Mosqueta Oil with Dense Fluids. ISHS, International Society for Horticulturae Science, Acta Horticulturae vol 4, l998.        [ Links ]

Schalke, R., The Supercritical CO2 Extraction of Oil from Rose Hip Seeds. 8th Symposium onSupercritical Fluid Chromatography and Extraction. St Louis. 1998.        [ Links ]

Ambrogi, A; Mattea, M. and Eggers, R., Extraction of Sunflower Oil Using Supercritical Carbon Dioxide. VI Latinamerican Congress on Heat and Mass Transfer. Floreanopolis – 1996.        [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License