We are unable to create an online viewer for this document. Please download the document instead.
FRYING OIL RECYCLING WITH DENSE CARBON DIOXIDE
Jesusa Rincón, Luis Rodríguez, Isaac Asencio and Virginia Ancillo
Dpto. de Ingeniería Química. Facultad de Ciencias del Medio Ambiente. UCLM.
Avda. Carlos III, s/n. 45071 Toledo. SPAIN.
e-mail: Jesusa.Rincon@uclm.es. Phone: 925268800 Fax: 925268840
Waste frying oils are a pollution source that may be eliminated by oil recycling. A
possible technology to recycle this oil is extraction with dense gases. In this work we
investigate the use of sub and supercritical carbon dioxide to recover the undegraded
triglycerides fraction from the waste oil. The experiments have been performed in a
semicontinuous extraction system. The effects of pressure (300-400 atm) and temperature
(25-80ºC) on the extraction yield have been investigated. In order to study the characteristics
of the oils recovered under different operating conditions, the concentrations of the polar
components coming from the oxidation, hydrolysis, and/or polymerisation of triglycerides
have been measured in the extracted oils.
The results obtained in the experimental range investigated showed several findings.
Extraction yields obtained with liquid CO2 slightly increased with pressure at constant
temperature. With supercritical CO2 the extraction yields increased with increasing pressure
and decreasing temperature (while keeping constant temperature and pressure, respectively);
besides, the yield increase with pressure was more significant than with liquid CO2. As
regards to the polar compounds concentrations, it has been found that the effect of pressure
and temperature on this variable followed a trend similar to that observed with the extraction
yield and that the effect of pressure was also more noticeable with the solvent at the
A large proportion of fats and oils is used in the world for the preparation of fried
foods. During the frying process many chemical complex reactions occur so the fat begins to
degrade. The principal reactions taking place include formation of conjugated dienes,
formation and decomposition of hydroperoxides, formation of low molecular weight carbonyl
compounds, hydrolysis of triglycerides, and polymerisation via complex free radical
processes at elevated temperatures .
As frying continues, the concentration of degradation products increases until the oil is
unfit for use and it must be discarded . This residue can not be reused because of its high
content of pollutants but, since it still contains a large proportion of triglycerides, not only
environmental but also economical reasons justify its regeneration.
Then, the objective of the regeneration process is the separation of the triglycerides
fraction from the polar compounds (oxidized low molecular weight material and polymers)
formed during the degradation process.
In the last decade several studies have been made to purify the used frying oil by
adsorbent and high pressure extraction treatments [3-9]. This last technology presents several
advantages over adsorption processes (such us continuous operation, low temperature
separation, good and easy adjustable selectivity) and it has been successfully applied to
extract and separate lipid components from different food products [10-12]. Therefore, it
seems interesting to analyse its use in waste frying oil recycling. As regards to the extraction
solvent to be used, CO2 should be recommended since in the food industry it has advantages
over other solvents because it is non-toxic and can be easily and completely removed from
products; furthermore, it is non-flammable, non-corrosive and readily available in large
The regeneration of used frying oil with supercritical carbon dioxide was first
proposed by Perrut and Majewski  to improve the results obtained when applying the more
classical adsorption technology. Then, several works have been published about this recycling
process [8-9], but more detailed information is needed about the operating conditions and the
characteristics of the influent and effluent process streams in order to assess its viability at the
industrial scale. Accordingly, the major aim of this paper will be to analyse the high pressure
extraction of used frying oil in order to select the best extraction conditions leading to both
high yield and quality of the oil recovered.
Carbon dioxide (purity 99%) was supplied by Praxair (Madrid, Spain). The used
frying oil was obtained by heating sunflower oil during 14 hours at 195 ºC in a frying
machine. Simulating the frying process in this way allowed to obtain a degraded oil
(Triglycerides: 70%, Polar compounds: 30%) with no food particles that may interfere in the
extraction study. Once the oil was degraded, it was stored in topaz glass bottles, in a nitrogen
atmosphere, in order to avoid further oil degradation by light, oxygen, and/or humidity.
Apparatus and extraction procedure
A schematic diagram of the extraction system used in this study is presented in Figure
1. It consists of three main sections: the carbon dioxide supply system, the extractor assembly,
and the separator assembly. Basically the supply system was a steel cylinder (SC) that
provided liquid CO2. After cooling and filtering the CO2 was compressed by a membrane
pump. The pressure was regulated by a back pressure regulator (BPR) and checked by a
manometer (M). The compressed fluid was passed through an autoclave extraction cylinder
which had been previously filled with the used oil. As indicated in the figure, the extractor
design allowed the CO2 to enter through a 8 mm i. d. tube and, as it went upwards through
the extractor, to intimately mix with the used oil and to separate the CO2 soluble base oil
(triglycerides) from the rest of non-soluble used oil components (degradation products). A
type J thermocouple was inserted in the extractor to indicate its operation temperature. This
variable was controlled by immersing the extractor in an oil bath at the desired temperature.
The oil-laden CO2 from the extractor was passed through a heated metering valve (MV)
where the CO2 was depressurized and the separated oil was collected in a cool receiver (RE).
The CO2 flow through the extractor was measured by a turbine flow meter (FM) and totalized
by a flow computer (FC).
The separation of the polar y non polar fractions was performed by column
chromatography . The analysis of the polar fraction composition was carried out by Gel
Permeation Cromatography (GPC) according to the method specified in reference .
Figure 1. Experimental extraction system: SC, CO2 cylinder, CS: Cooling System; BPR:
Back Pressure Regulator; EX: Extractor; MV: Metering Valve; RE: Cool
Receiver; FM: Flow Meter; FC: Flow Computer.
RESULTS AND DISCUSSION
Experimental conditions of experiments performed in this work are summarised in
Table 1. The results obtained are discussed below.
Table 1. Summary of the experiments carried out with liquid and supercritical CO2.
300 300 300 350 350 350 400 400 400
25 40 80 25 40 80 25 40 80
SC SC L SC SC L SC SC
Density ( i i=1 to 9), Kg/l 0,967 0,910 0,746 0,987 0,935 0,789 1,005 0,957 0,823
1 L = Liquid. SC = Supercritical.
Figure 2 shows the effect of the solvent state, pressure and temperature on the
extraction yield. It can be observed that yields are always higher with liquid than with
supercritical CO2. On the other hand, it can also be seen that with liquid CO2, at a given
temperature, the yield slightly increase with pressure. With supercritical CO2 the yield
increases with increasing pressure (at constant temperature) and decreasing temperature (at
constant pressure), being the effect of pressure more noticeable in this case than with the
solvent in the liquid state.
These results are mainly related to the solvent density, since it is larger for liquid than
for supercritical CO2 and shows a smaller variation with pressure and temperature in the
liquid than in the supercritical region . As regards to the effects of pressure and
temperature, the variation of the solvent density with both variables is similar to that observed
with the extraction yields, indicating a direct correlation between both variables.
Nevertheless, as it can also be seen in Figure 2, the extraction yields depend also on the
solutes vapour pressures since, in some experiments performed at experimental conditions
leading to very close values of the solvent density, the variation of the yields with the variable
inverts with regard to the normal trend (increasing yield with increasing density).
Tim e (h)
Figure 2. Effect of the solvent density on the extraction yield.
Percentage of polar compounds in the recovered oil
Figure 3 shows the evolution with time of the percentage of polar compounds in the
oil recovered with CO2 at 400 atm. It can be seen that the percentage of polar compounds in
the oil decreases as the extraction progresses. In other words, both liquid and supercritical
CO2 extract selectively used frying oil polar compounds from its mixture with triglycerides.
Similar results were found at the other pressures tested.
This experimental finding, the selectivity of a non-polar solvent as CO2 for the polar
fraction of the used oil, can be explained only if the molecular weights of the polar
compounds extracted are smaller than those of the triglycerides, non-polar compounds that, in
principle, would dissolve more easily in a non-polar solvent than the polar matter. Therefore,
this result could suggest that the molecular weights of the polar compounds in the used frying
oil are smaller than those of the triglyceride fraction. It could also indicate that both liquid and
supercritical CO2 may dissolve only the polar compounds of low molecular weight and that,
as these components are exhausted, the extraction of polar components decreases due to that
also decreases the capacity of the supercritical solvent to dissolve the remaining polar
compounds (those with larger molecular weight). This last hypothesis was confirmed by
analysing by Gel Permeation Cromatography (GPC) the evolution with time of the polar
fraction composition of the recovered oil .
As regards to the effect of pressure and temperature on the percentage of polar
compounds in the oil recovered, it was observed that it followed a trend similar to that of
yields and, as in such case, although it mainly depended on the solvent density, the used oil
components vapour pressure also affected to the variable .
Figure 3. Effect of the solvent density on the percentage of polar components in the
The high pressure extraction of used frying oil seems very promising. In the
experimental range analysed it has been observed that with liquid CO2 yields slightly increase
with pressure (at constant temperature) and with supercritical CO2 they increase with
increasing pressure and decreasing temperature (while keeping constant temperature and
pressure, respectively); besides, the yield increase with pressure was much more important
than with liquid CO2. As regards to the polar compounds concentrations, it has been found
that the effect of pressure and temperature on this variable followed a trend similar to that
observed with the extraction yield and that the effect of pressure was also more noticeable
with the solvent at the supercritical state. These results may be explained by considering both
the solvent density and the solutes vapour pressure.
In order to assess the viability of the process at the industrial scale an effort has been
made to select the best extraction conditions. It has not been a simple task since the
percentage of undesirable polar compounds in the recovered oil was larger at those conditions
leading to higher yield. Nevertheless, operating at 300 atm and 40 ºC the extraction yield was
50% and the product recovered contained only about 10% of polar matter.
This research has been financially supported by the Junta de Comunidades de Castilla-
La Mancha (Spain) through Contract PBI-05-048.
 QUILES, J. L, RAMÍREZ-TORTOSA, M. C., GÓMEZ, J. A., HUERTAS, J. R.,
MATAIX, J., Food Chem., Vol. 76, 2002, p. 461.
 European Parliament, Recycled Cooking Oils, PE n. 289.889, February 2001.
 YATES, R. A., CALDWELL, J. D., JAOCS, Vol. 69, 1992, p. 894.
 YATES, R. A., CALDWELL, J. D., JAOCS, Vol. 70, 1993, p. 507.
 JUNG, M.Y., RHEE, K.C., Foods and Biotechnol., Vol. 3, 1994, p. 65.
 LIN, S., AKOH, C. C., REYNOLDS, A. E., JAOCS, Vol. 76, 1999, p. 739.
2000, French Patent PCT/FR00/01669.
 YOON, J., HAN, B.S., KANG, Y.C., KIM, K.H., JUNG, M.Y., KWON, Y.A., Food
Chem., Vol. 71, 2000, p. 275.
 SESTI-OSSÉO, L., CAPUTO, G., GRACIA, I., REVERCHON, E., Proc. of the 6th
International Symposium on Supercritical Fluids, Tome I, 2003, p. 397.
 STAHL, E., QUIRIN, K.W., GERARD, D., “Dense Gases for Extraction and Refining”.
Ed. Springer-Verlag Publishing, Berlin, 1988.
 KING, J. W. , LIST, G. R., “Supercritical Fluid Technology in Oil and Lipid
Chemistry”. Ed. AOCS Press, 1996.
 LUCAS, A., MARTÍNEZ DE LA OSSA, E., RINCÓN, J., BLANCO, M. A., GRACIA,
I., J. Sup. Fluids. Vol. 22, 2002, p. 221.
 Spanish Official Bulletin (BOE) No. 26/1989, January 31 of 1989, p. 2665.
 DOBARGANES, M.C., PÉREZ-CAMINO, M.C., MÁRQUEZ, G., Fat Sci. Technol.
Vol. 90, 1988, p. 308.
 ANCILLO, V., DEA Thesis, 2005, University of Castilla-La Mancha, Spain.