Journal of the Science of Food and Agriculture
J Sci Food Agric 85:2167–2174 (2005) DOI: 10.1002/jsfa.2244
Supercritical carbon dioxide extraction of oil and α-tocopherol from almond seeds Lucia Leo,1 Leonardo Rescio,2 Loredana Ciurlia2 and Giuseppe Zacheo1∗ 1 CNR,
Istituto di Scienze delle Produzioni Alimentari, Via Lecce-Monteroni, 73100 Lecce, Italy CHIMICA srl, SS 476 Km 17, 650 Zona Industriale, 730213 Galatina (LE), Italy
2 PIERRE
Abstract: The objective of this study was to extract oil and tocopherols from almond seeds using supercritical carbon dioxide and to compare this extraction with a traditional solvent method. Oil and tocopherol extraction rates were determined as functions of the pressure (350–550 bar), temperature (35–50 ◦ C) and CO2 flow rate (10–30 kg h−1 ), using a 10-l vessel. The effects of matrix particle size on extraction yield were also studied and it was demonstrated that extraction yield is greatly influenced by particle size. Maximum recovery was obtained in the first 2–3 h of extraction at a pressure of 420 bar, a temperature of 50 ◦ C and a flow rate of 30 kg h−1 CO2 . These results suggest that the elevated initial oil and tochopherol solubility is related to the increased proportion of fatty acids in the initial extract. The results were compared with those obtained when hexane/methanol was used as a solvent. 2005 Society of Chemical Industry
Keywords: carbon dioxide; extraction; oil; solubility; tocopherol
INTRODUCTION Carbon dioxide (CO2 ) is often used in the development of supercritical fluid extraction (SFE) instead of the organic solvents normally employed in conventional extraction methodologies. The main advantages of using carbon dioxide fluid are: a reduced potential for the oxidation of extracted solutes, higher selectivity, increased sample throughput, shorter extraction time and a low critical temperature. The latter is beneficial in extracting thermally labile compounds, such as natural vegetable products. Supercritical carbon dioxide also has chemical inertness, suitable solvent strength, permits the separation of compounds of widely different polarities and molecular masses, has low cost and, what is more, can be removed from the extracted products without leaving any chemical residue. In addition, carbon dioxide is both non-toxic and non-explosive and its use can reduce the consumption of organic solvents; this is especially useful for the production of natural products used in foods and pharmaceuticals. There are several excellent articles on the use of supercritical carbon dioxide methodologies in various analytical areas, including the extraction of vegetable oil and of fat-soluble vitamins.1 – 4 Knowledge of the solubility of vegetable oil and its fat-associated products in supercritical CO2 along with pressure and temperature are important for the successful application of this technology. The extraction conditions for specific solutes or classes of solutes from the vegetable matrix can also be optimised by changing the flow rate of the supercritical fluid or
its pressure or temperature. Recently, an increasing interest in both the detection and the search for new oil and vitamin extraction techniques has been reported, and supercritical CO2 extraction methods have been successfully applied in the extraction of oils from orange peel,5 hazelnut,6 olive,7 blackcurrant and vineyard grape.8 Despite the large number of matrices processed,9 only some models of supercritical CO2 have been published. Nevertheless, to optimize the extraction conditions, a relationship between the composition of vegetable oils and their solubility in supercritical CO2 must be taken in account.8 Almond is one of the major nut tree crops of the Mediterranean region. Almond kernel, analysed by using a conventional solvent method for the isolation of oil and fat-soluble vitamins, is shown to consist mainly of (g kg−1 ): lipids (450–600), proteins (200), and carbohydrates (200), with other elements, such as tocopherols, present in much smaller quantities (tocopherols 0.4–0.8).10 – 12 The major lipid species were found to be (g kg−1 total fatty acids): oleic acid, linoleic acid (100–170), and palmitic acid (55–70). The objective of our study was to develop a supercritical CO2 extraction method for almond seed oil and subsequently to characterise the oil and tocopherols present in seed extracts. The supercritical CO2 method for extraction of the lipid components and tocopherols used a 10-l pilot plant. The effect of pressure, temperature, solvent flow and particle size on the extraction rate were analysed. The results of
∗ Correspondence to: Giuseppe Zacheo, CNR, Istituto di Scienze delle Produzioni Alimentari, Via Lecce-Monteroni, 73100 Lecce, Italy E-mail:
[email protected] (Received 1 September 2004; revised version received 27 January 2005; accepted 16 March 2005) Published online 20 June 2005
2005 Society of Chemical Industry. J Sci Food Agric 0022–5142/2005/$30.00
2167
L Leo et al
the SFE extraction were also compared with those obtained by using the conventional hexane/methanol method.
by weighing the lipid fraction present in the extract samples. Aliquots were immediately used for further analyses or stored in the dark at −20 ◦ C. Oil content was expressed as mg g−1 of almond fresh weight.
EXPERIMENTAL Materials and sample pre-treatment Almond seeds collected in 2002 were acquired from producers and were preserved in a dry chamber at 20 ◦ C. The almond seeds were ground in a blender, using different degrees of grinding in order to obtain different particle sizes; samples were separated on the basis of their size by sieving. Particle size diameters were determined by microscopic analysis. Grinding was performed just before the extraction and within a 5-s time interval in order to prevent the heating of the almond matrix and the related degradation of oil components. Liquid carbon dioxide with a purity of 990 g kg−1 was supplied by Air Liquide srl. (Milan, Italy). Solvents (HPLC grade) were purchased from JT Baker (Milan, Italy). Standard compounds for tocopherol and lipid characterizations were obtained from SigmaAldrich (Milan, Italy).
Oil and tocopherols extraction by supercritical CO2 The schematic pilot apparatus used for CO2 supercritical extraction of almond seeds is shown in Fig 1. The system was composed of an extractor (10 l); three separators: S1 (1.5 l), S2 (1.5 l) and S3 (0.3 l); a pump; three heat exchangers; a tank for CO2 ; a filter and the instruments and all the devices necessary for the management of the automatic control of the process, including emergency jettisoning of the fluid CO2 . The matrix (3–4 kg of almond) was arranged in a small cylindrical basket (6.8 l) which was inserted inside the extractor. Pure CO2 was condensed and/or cooled to approximately −3 ◦ C (condenser E1), collected in a 10-l tank (buffer tank) and compressed to a desired pressure by means of a pump. CO2 was then heated to the predefined temperature in order to reach the supercritical state before it was passed into the extractor. From the extractor, the solution (CO2 + extracted compounds) moved into the first separator (S1) which was equipped with an internal coil that further lowers the CO2 temperature to the predefined value. The decreased temperature reduced the solubility of the less soluble components, in particular waxes and water, which then collected on the bottom of the separator. From S1 the solution passed to the S2 separator, where there was a drastic
Oil extraction by solvent Lipid fraction was extracted from 2 g of powdered almond seeds by a modified Bligh and Dyer method.13 Whole and ground almonds were stirred for 1 h with 20 ml of hexane/methanol (2:1) mixture and centrifuged for 10 min at 3000 × g. The upper layer was separated off, the residual solvent was evaporated with a rotavapor, and the oil content was determined
Figure 1. Diagram of SFE pilot plant.
2168
J Sci Food Agric 85:2167–2174 (2005)
Supercritical extraction of oil and α-tocopherol from almond seeds
reduction of pressure and, therefore, of the solubility of the majority of the oil components. The reduction of solubility in S2 could be controlled by the valve PV1 situated at the entrance to the separator. The solution then left S2 and flowed into S3 where, by means of the pressure-reducing valve PV2, there was a further fall in pressure and the separation of the solvent and the solute. The CO2 from S3 flowed through a filter, was newly cooled and recycled in the service tank. The extracts were manually recovered by opening the valves at the bottom of S1, S2 and S3 at certain time intervals. The extracted samples were collected, weighed and then stored at −20 ◦ C for the further analyses. Oil characterisation The almond oil obtained by conventional and CO2 extraction was separated into polar and neutral fractions by using the standardised IUPAC method.14 Total lipids (1 g) were separated on a silica gel column by eluting sequentially the neutral lipids (NL) with chloroform and then the polar lipids (PL) with a methanol/chloroform (6:1 v/v) mixture. The solvents were evaporated and the remaining lipid fractions were weighed and used for further analysis. NL were further separated into their subclasses by TLC on 0.25 mm silica gel plates (Merck, Darmstadt, Germany) using a hexane/diethyl ether/formic acid (80:20:1 v/v) mixture. The plates were air dried and analytes detected by exposure to iodine vapour. Individual bands were identified with the aid of the standard Sigma (Milan, Italy) procedure and fractions corresponding to each lipid type were extracted from the plates with 100 ml l−1 methanol in diethyl ether and weighed to calculate chemical yelds.8 Fatty acid methyl esters (FAMEs) were prepared by adding 5 ml of sodium hydroxide (0.5 M) in methanol at 100 ◦ C to 0.5 g of oil in a water bath for 10 min. The solution was cooled to room temperature, added to 5 ml of 150 mg g−1 boron trifluoride in methanol and heated at 100 ◦ C for a further 10 min. After cooling, 5 ml of hexane and 10 ml of aqueous saturated solution of sodium chloride were added and the mixture was vigorously shaken. The organic layer was separated and concentrated under vacuum. FAMEs were analysed by a Shimadzu (Kyoto, Japan) GC-17A using a gas chromatograph equipped with a flame ionisation detector (FID) and a 60 m Stabilwax (Restek Corp, Bellafonte, PA, USA) capillary column (0.25 mm ID, 0.25 µm film thickness). The oven temperature was programmed as follows: 100 ◦ C for 1 min, then to 240 ◦ C at 8 ◦ C min−1 and maintained at this temperature for 30 min. The injector and detector temperatures were 250 and 280 ◦ C, respectively. Helium was used as carrier gas at a flow rate of 10 ml min−1 . Tocopherol analysis Collected oil samples were analysed by HPLC using a Beckman-Coulter (Milan, Italy) chromatograph J Sci Food Agric 85:2167–2174 (2005)
equipped with a diode array (System Gold 168; 190–600 nm) detector, a Shimadzu spectrofluorimetric detector (RF-10AXL) and a Marathon autosampler. Tocopherols were separated by isocratic elution in a normal phase column (Ultrasphere Silica, 4.6 mm × 2.5 cm, Beckman-Coulter, Fullerton, CA, USA.). The mobile phase was hexane/2-propanol (99.5:0.5 v/v) flowing at 0.5 ml min−1 . The elute was monitored with fluorescence detector at an excitation wavelength of 289 nm and an emission wavelength of 330 nm and with UV detector at 294 nm. Quantification was performed by comparing the sample peak areas to those of known amounts of standard compounds. Tocopherol content was expressed as µg g−1 of almond dry weight.
RESULTS AND DISCUSSION Oil extraction The almond oil obtained from the various fractions, extracted with either hexane/methanol or supercritical CO2 , was intensely yellow in colour having a variable content of tocopherols. After oil extraction, the residual defatted almond was a white flour with variable grain size, more or less fine, depending on the degree of grinding of the fresh matrix and on the effect of the operating conditions (pressure, flow rate, temperature) used to remove the oil. Influence of particle size on conventional oil extraction Mature dry almond seeds with water content estimated at 50–60 g kg−1 dry weight were homogenized and sieved to obtain four representative sub-samples: whole, broken (4–8 mm), milled (0.5–3 mm) and powered samples. The influence of particle size on the conventional extraction process of almond seeds is reported in Table 1. The oil value obtained from the powdered almond seeds was assumed to be the maximum possible extractable oil and was used as an absolute value (100%) to calculate the proportion of extracted oil. Influence of particle size on CO2 supercritical oil extraction Preliminary tests allowed us to determine suitable operating conditions (temperature, pressure and flow rate). At the same time the ability of the process to concentrate the almond oil components selectively in the separator vessel was also established. From these tests it was evident that the yield of extraction not only depended on the utilised pressure, temperature and flow rate of the solvent through the extraction bed, but also on the chemical and physical characteristics of the matrix such as oil composition, moisture content and particle size. The almond matrix was blended at various particle sizes in order to obtain data on the extraction rate and the total yield of oil. The yield, ie the proportion of oil extracted by our pilot apparatus in relation 2169
L Leo et al Table 1. Influence of particle size on conventional almond oil and tocopherol extractions
Oil
Almond Powdered Milled (0.5–3 mm) Broken (4–8 mm) Whole
As mg g−1 of fresh weighta
As mg g−1 of extractable oil
As µg g−1 of dry weighta
As mg g−1 of extractable tocopherol
545 ± 23 298 ± 12 151 ± 13 55 ± 4
1000 ± 40 550 ± 40 280 ± 80 100 ± 7
398 ± 25 106 ± 11 44 ± 1 3 ± 0.2
1000 ± 60 270 ± 10 110 ± 20 0
Mean of four values ± standard deviations.
1000
1000
900
900
Oil yield (as g kg-1 extractable oil)
800
Milled almond Broken almond
700
Whole almond 600 500 400 300
800 Oil yield (as g kg-1 extractable oil)
a
Tocopherol
700 600 500 400 300 P=550 bar
200
200
P=450 bar
100
P=350 bar
100 0 10
20
30
40
50
60
CO2 kg kg-1 matrix Figure 2. Comparison of oil extraction curves for various particle sizes of almond seeds. Experiments carried out at 50 ± 2 ◦ C, 420 ± 20 bar, 20 ± 5 kg h−1 (mean ± SD, n = 4).
to the amount extracted by solvent in powdered seeds, is given as a function of the ratio of the CO2 mass used and the initial seed mass loaded into the extractor (Fig 2). As expected, it was observed that extraction yield strongly increased in broken and milled almonds. During the first part of the extraction process the curves were characterised by a linear increase, subsequently they approached a constant value. The latter was caused by a reduction in oil concentration in the matrix as well being more transfer-controlled in the latter stages of extraction. Nevertheless, under identical operating conditions and using the same quantity of CO2 (300 g kg−1 of CO2 per kg of matrix) broken almonds yielded 300 g kg−1 oil as whole almonds and milled almonds as much as 700 g kg−1 (Fig 2). This means that, at predetermined operating conditions, the reduction of particle size resulted in an increase in oil yield. This increase was explained by the increase of ‘contact surface’ in the milled matrix compared with that available in the whole almonds. 2170
0 10
20
30
40
50
60
CO2 kg kg-1 matrix Figure 3. Effect of pressure on SFE oil extraction yield at a temperature of 50 ± 2 ◦ C and a flow rate of 20 ± 5 kg h−1 (mean ± SD, n = 4).
Influence of pressure, temperature and flow rate on CO2 supercritical oil extraction The functional relationships between the various physical parameters such as pressure, temperature and flow rate were investigated. The influence of pressure on the CO2 supercritical oil extraction from almond seeds was determined at pressures of 350, 450 and 550 bar, respectively, and at a constant temperature (50 ± 2 ◦ C) and a constant flow rate (20 ± 5 kg h−1 ). These results are reported in Fig 3 where the oil yield is plotted versus kg of CO2 per kg of matrix. The extraction curves obtained at 450 and 550 bar were characterised by an initial period in which the oil yield rose more steeply as the pressure was increased and a second period characterised by a higher value of the plateau as the pressure increased. The extraction curve showed a linear period without a flex point at low pressure (350 bar). In view of these results it can be stated that at the beginning of the extraction process, when the physical phenomenon J Sci Food Agric 85:2167–2174 (2005)
Supercritical extraction of oil and α-tocopherol from almond seeds
of quantitative extraction mass depends mainly on the available amount of oil extracted into the CO2 , an increment of the extraction pressure determines a yield increment of oil. Figure 3 shows that at a flow rate of 30 kg CO2 kg−1 an increase in pressure of 28% from 350 to 450 bar resulted in a 100% increase in yield, while an increase of 57% from 350 to 550 bar in the extraction pressure led to a further increase in yield. As expected, using the same quantity (40 kg of CO2 per kg of matrix), as pressure increased the solubility of the almond oil in CO2 increased, as found in other studies on oil extraction from seeds.15 – 17 The results of the pressure effect on the oil yield suggest that the increased yield resulted from the increasing density of the solvent and consequently from increased solubility of almond oil in the CO2 .18 Temperature is also a critical parameter which affects the oil yield in CO2 extraction. In Fig 4 the effect of temperatures of 35, 40 and 50 ◦ C on the extraction yield is reported. The results obtained clearly indicate that the extraction yield was promoted by an increase in temperature. At fixed conditions of pressure and flow rates, an increase in temperature of 15 ◦ C produced an almost fourfold yield increment (Fig 4). An explanation of this phenomenon is that the oil solubility is enhanced in the CO2 . However, the temperature effect on CO2 extraction is more difficult to assess than the other parameters, pressure and solvent flow, because it influences both the solvent diffusion in the matrix and the oil dissolution in the solvent.19 As the temperature rises there is a lower
solvent strength of the fluid due to the decrease in fluid density. In contrast, an increase in temperature can improve the extraction efficiency despite the decrease in fluid density, since the vapour pressure of the oil is increased.9,19 To determine the effect of flow rate during almond oil extraction a constant pressure of 420 bar and temperature of 50 ◦ C were used. In our pilot plant, the optimum extraction yield was obtained at a flow rate between 20 and 30 kg h−1 of CO2 . From the results reported in Fig 5 it is evident that oil yield, at the initial stage of extraction, increased with increasing CO2 flow rate from 10 to 30 kg h−1 . It seemed that flow rate of supercritical fluid directly affects the extraction rate. Flow rates below 10 kg h−1 resulted in a relatively flat yield curve and, consequently, in an increase in the extraction time, while flow rates beyond 30 kg h−1 resulted in only a slight increase of oil yield (data not shown). The latter observation indicated that for higher flow rates the carbon dioxide leaving the extractor was less saturated and the effect of increasing flow rate on extraction process was very small. In conclusion, these experiments demonstrated that when extracting a biological matrix at the described conditions (see Materials and Methods), the oil yield increased with increasing pressure, temperature and flow rate. Tocopherol extraction and lipid characterisation Lipids from oil extracted by hexane/methanol and CO2 fraction extracted oil were analysed in order to find the relationship between changes in oil solubility and
1000 1000 900 900 T=50 °C
30 kg h-1 CO2 800
T=40 °C
700
Oil yield (as g kg-1 extractable oil)
Oil yield (as g kg-1 extractable oil)
800
T=35 °C 600 500 400 300
20 kg h-1 CO2 10 kg h-1 CO2
700 600 500 400 300
200
200
100
100 0
0 10
20
30 CO2 kg
40 kg-1
50
60
matrix
Figure 4. Effect of temperature on SFE oil extraction yield at a pressure of 420 ± 20 bar and a flow rate of 20 ± 5 kg h−1 (mean ± SD, n = 4).
J Sci Food Agric 85:2167–2174 (2005)
10
20
30 CO2 kg
40 kg-1
50
60
matrix
Figure 5. Effect of CO2 flow rate on SFE oil extraction yield at a pressure of 420 ± 20 bar and a temperature of 50 ± 2 ◦ C (mean ± SD, n = 4).
2171
L Leo et al
Pressure (bar) 370 400 420 430 450 480 500 a
Tocopherol extraction rate (mg min−1 kg−1 CO2 )a
Oil extraction rate (g min−1 kg−1 CO2 )a
0.047 ± 0 0.100 ± 0 0.108 ± 0 0.043 ± 0 0.029 ± 0 0.028 ± 0 0.028 ± 0
0.10 ± 0 0.21 ± 0.02 0.22 ± 0.01 0.22 ± 0.02 0.27 ± 0.01 0.26 ± 0.01 0.27 ± 0.02
Mean of four values ± standard deviations.
2172
Tocopherol concentration (as mg g-1 oil)
0.6
0.5
0.4
0.3
0.2
0.1
0 60
120
180
240
300
360
Time (min) Figure 6. Concentration of tocopherols in SFE extracted oil as function of time at a temperature of 50 ± 2 ◦ C, a pressure of 420 ± 20 bar and a flow rate of 20 ± 5 kg h−1 (mean ± SD, n = 4).
0.35
0.2 Tocopherol 0.18
Oil
0.3
0.16 0.25
0.14 0.12
0.2
0.1 0.15
0.08 0.06
0.1
0.04
Oil extraction rate (g min-1 kg-1 CO2)
Table 2. Tocopherol and oil extraction rate at different pressure values during 2 h of extraction time
and decreased slowly in the remaining extraction time. The extraction rate of tocopherol was highest (0.13 mg min−1 kg−1 CO2 ) in the initial phase of the process and was characterised by a constant decrease over the remaining time. The high tocopherol concentration during the initial stage could be due to a high proportion of more soluble lipids such as free fatty acids, in the oil extracted (Fig 8). This, in a way, is
Tocopherol extraction rate (mg min-1 kg-1 CO2)
its components. Tocopherol extraction from almond seeds by the conventional solvent method increased as particle size decreased (Table 1). In powdered almond a mean of 0.4 mg g−1 dry weight (dw) was obtained and this recovery was comparable with results obtained by Zacheo et al.12 The value was assumed to be the maximum possible extractable almond tocopherols and it was assumed as the reference value for calculating the proportion of tocopherol extracted by the SFE process (Table 1). In all the analyses of solvent and CO2 extracts, α-tochopherol was the main isomer; β, γ , δ- tochopherols were not detected. In order to assess the ability of the process to obtain oil fractions rich in tocopherol various supercritical extraction conditions were employed. The optimum co-extraction conditions for oil and tocopherols were a temperature of 50 ± 2 ◦ C, a pressure of 420 ± 20 bar and a flow rate of 25 ± 5 kg h−1 using powdered almond seeds. The oil fractions obtained at the above conditions during the first 2 h of the extraction process contained the highest tocopherol concentration, which decreased in the remaining extraction time. This tendency can be observed in Fig 6, where the proportion of tochopherol obtained is plotted against extraction time. Experiments conducted by varying extraction pressure led to changes of the oil tocopherol concentration during the first 2 h. It can be seen that increasing the pressure from 370 to 400 and 420 bar, resulted in an increase in the tocopherol extraction rate (expressed as mg min−1 kg−1 CO2 ) in the fraction collected during the first 2 h of extracting time, but a further increase in pressure induced a decrease in the extraction rate (Table 2). This result can be explained by considering that the mass transfer of tocopherol is related to its solubility in the oil, and both its initial amount present in the matrix and the subsequent depletion of the solid phase (Fig 6). The tendency of extraction curves can be seen in Fig 7 where the oil and tocopherol SFE extraction rate (expressed as g min−1 kg−1 CO2 and mg min−1 kg−1 CO2 respectively) is reported as function of time. The figure clearly indicates that the oil extraction rate during the first stage was very high and remained almost constant at a mean value of 0.25–0.30 g min−1 kg−1 CO2 for a 3 h period
0.05 0.02 0
0 60
120
180
240
300
360
Time (min) Figure 7. Comparison of oil and tocopherol extraction rate as function of time at a temperature of 50 ± 2 ◦ C, a pressure of 420 ± 20 bar and a flow rate of 20 ± 5 kg h−1 (mean ± SD, n = 4).
J Sci Food Agric 85:2167–2174 (2005)
Supercritical extraction of oil and α-tocopherol from almond seeds Table 3. Lipid composition (as mg g−1 weighta ) of almond oil extracted by conventional and CO2 methods
SFE extractionc Lipid classesb TAGs FFA DAGs MAGs Other
Hexane/methanol extraction
Fract I
Fract II
Fract III
Fract IV
Fract V
Fract VI
750 ± 20 150 ± 10 34 ± 1 21 ± 1 15 ± 3
630 ± 30 310 ± 20 29 ± 2 25 ± 2 11 ± 1
640 ± 20 290 ± 20 25 ± 1 22 ± 2 11 ± 5
710 ± 40 200 ± 10 23 ± 6 30 ± 2 12 ± 1
790 ± 20 120 ± 4 32 ± 1 32 ± 2 9±0
860 ± 30 40 ± 3 38 ± 2 33 ± 1 5±0
840 ± 30 12 ± 1 38 ± 2 35 ± 3 2±0
Mean of four values ± standard deviations. Lipid classes: TAGs (triglycerides), FFA (free fatty acids), DAGs (diglycerides), MAGs (monoglycerides). c The fractions (Fract I–VI) were collected at intervals of 1h. a
b
Table 4. Fatty acid composition (as mg g−1 of total FAMEa ) in hexane and SFE extracted almond oil
SFE extractionb Fatty acid
Hexane extraction
Fract I
Fract II
Fract III
Fract IV
Fract V
Fract VI
Palmitic (C16:0) Palmitoleic (C16:1) Stearic (C18:0) Oleic (C18:1) Linoleic (C18:2) Other MUFA PUFA
70 ± 6 4±0 22 ± 0 720 ± 20 177 ± 9 7±0 72 ± 2 17.7 ± 0.9
68 ± 4 6±0 22 ± 0 710 ± 20 188 ± 9 4±0 72 ± 2 18.8 ± 0.9
66 ± 2 5±0 21 ± 1 730 ± 20 180 ± 10 1±0 73 ± 2 18 ± 1
67 ± 4 5±0 17 ± 2 730 ± 30 180 ± 10 1±0 74 ± 3 18 ± 1
67 ± 4 .5 ± 0 19 ± 0 710 ± 30 200 ± 10 1±0 72 ± 3 20 ± 1
69 ± 2 5±0 22 ± 1 690 ± 20 210 ± 10 8±0 69 ± 2 21 ± 1
71 ± 3 .5 ± 0 17 ± 0 680 ± 20 220 ± 20 2±0 69 ± 2 22 ± 2
b
Mean of four values ± standard deviations. The SFE samples (Fract I–VI) were collected at intervals of 1 h.
0.7
400 Tocopherol Free Fatty Acids
0.6
350
Tocopherol (mg g-1 of oil)
300 0.5 250 0.4 200 0.3 150 0.2 100 0.1
Free Fatty Acids (mg g-1 of oil)
a
50 0
0 30
60
90
120 180 240 300 360 Time (min)
Figure 8. Comparison of tocopherol and free fatty acid content in the extracted oil as function of time at a temperature of 50 ± 2 ◦ C, a pressure of 420 ± 20 bar and a flow rate of 20 ± 5 kg h−1 (mean ± SD, n = 4).
supported from the data reported in Table 3 where in fraction I (1 h after starting the CO2 extraction process) 310 mg free fatty acids (FFA) g−1 were observed. During the rest of the extraction process the free fatty acid component gradually decreased from 310 J Sci Food Agric 85:2167–2174 (2005)
to 10 mg g−1 while an increase in triglycerides (TAG) fractions was observed. Similar data were reported by Sovov´a et al 8 during the extraction of vegetable oils where a close relationship between the oil solubility and the amount of free fatty acids was observed. The methyl ester fatty acid analysis in both the hexane/methanol and the CO2 extracted oil is presented in Table 4. No differences in the fatty acid compositions of the various extracts are apparent. From the analysis, the main components of almond oil were as follows (g kg−1 ): oleic acid (700), linoleic acid (200) and palmitic acid (70). Mono- and polyunsaturated fatty acids accounted for about 700 and 200 g kg−1 respectively of the total fatty acid composition. The use of SFE, provides the advantage of fractioning the components of almond oil and obtaining oil enriched in tocopherol oil. Similar results for the extraction and enriched tocopherol fractions have been obtained using supercritical CO2 by several authors, as reviewed by Turner et al.9
CONCLUSION In conclusion, the use of supercritical CO2 in almond oil extraction, under the specific conditions reported above, resulted in tocopherol-rich oil fractions in the first part of the extraction process. The results clearly demonstrate that the oil extraction rate differed for large and small almond pieces, indicating that particle size is a limiting parameter to the extraction rate. 2173
L Leo et al
REFERENCES 1 Chester TL, Pinkston JD and Raynie DE, Supercritical fluid chromatography and extraction. Anal Chem 66:106–130 (1994). 2 Chester TL, Pinkston JD and Raynie DE, Supercritical fluid chromatography and extraction. Anal Chem 68:487–514 (1996). 3 Chester TL, Pinkston JD and Raynie DE, Supercritical fluid chromatography and extraction. Anal Chem 70:301–320 (1998). 4 Smith RM, Supercritical fluids in separation science-the dreams, the realty and the future. J Chromatogr A 856:83–115 (1999). 5 King JW, Favati F and Taylor SL, Production of tocopherol by supercritical fluid extraction and chromatography. Sep Sci Tech 31:843–857 (1996). 6 Bernardo-Gil MG, Grenha J, Santos J and Cardoso P, Supercritical fluid extraction and characterisation of oil from hazelnut. Eur J Lipid Sci Technol 104:402–409 (2002). 7 De Lucas A, Martinez de la Ossa E, Rincon J, Blanco MA and Gracia I, Supercritical fluid extraction of tocopherol concentrates from olive tree leaves. J Supercrit Fluids 22:221–228 (2002). ´ 8 Sovov´a H, Zarevucka M, Vacek M and Str´ansk´y K, Solubility of two vegetable oils in supercritical CO2 . J Supercrit Fluids 20:15–28 (2001). 9 Turner C, King JW and Mathiasson L, Supercritical fluid extraction and chromatography for fat-soluble vitamin analysis. J Chromatogr A 936:215–37 (2001). 10 Kader AA, In-plant storage, in Almond production Manual. Publication N◦ 3364, Micke CW, University of California, Division of Agriculture and Natural Resources, 274 pp (1996).
2174
11 Zacheo G, Cappello AR, Perrone LM and Gnoni GV, Analysis of factors influencing lipid oxidation of almond seeds during accelerated ageing. Lebensm Wiss Technol 31:6–9 (1998). 12 Zacheo G, Cappello MS, Gallo A, Santino A and Cappello AR, Changes associated with post-harvest ageing in almond seeds. Lebensm Wiss Technol 33:415–23 (2000). 13 Bligh EG and Dyer WS, A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37:911–917 (1959). 14 Ruiz-Gutierreez V and Perez-Camino MC, Update on solidphase extraction for the analysis of lipid classes and related compounds, J Chromatogr A 885:321–341 (2000). 15 Reverchon E, Supercritical fluid extraction and fractionation of essential oils and related products, J Supercrit Fluids 10:1–37 (1997). 16 Eggers R, Supercritical fluid extraction of oil seeds, in Supercritical Fluid Technology in Oil and Lipid Chemistry, ed by King JW and List GR, AOCS Press, Champaign, IL, pp 35–43 (1996). 17 King JW, Sub- and supercritical fluid processing of agrimaterials: extraction, fractionation and reaction models, in Supercritical Fluids Fundamentals and Applications, ed by Kiran E, Debenedetti PG and Peters CJ, NATO Science Series, Series E, Applied Sciences, Vol 366, Kluwer Academic, Dordrecht, pp 451–462 (2000). 18 Del Valle JM and Aguilera JM, An improved equation for predicting the solubility of vegetable oils in supercritical CO2 . Ind Eng Chem Res 27:1551–1553 (1988). 19 Luque de Castro MD, Valcarcel M and Tena MT, Analytical supercritical fluid extraction, Springer-Verlag, New York, 319 pp (1994).
J Sci Food Agric 85:2167–2174 (2005)
