SUPERCRITICAL CO2 SEPARATION OF BERGAMOT ESSENTIAL OIL 429 FLAVOUR AND FRAGRANCE JOURNAL Flavour Fragr. J. 2003; 18: 429–435 Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ffj.1245
Supercritical carbon dioxide separation of bergamot essential oil by a countercurrent process Marco Poiana,1* Antonio Mincione,1 Francesco Gionfriddo2 and Domenico Castaldo2 1
2
Dipartimento di Biotecnologie per il Monitoraggio Agroalimentare e Ambientale, University of Reggio Calabria, Piazza S. Francesco 4, I-89061 Gallina, Reggio Calabria, Italy Stazione Sperimentale per le Industrie delle Essenze e dei Derivati Dagli Agrumi, Reggio Calabria, Italy
Received 15 March 2002 Revised 25 July 2002 Accepted 28 August 2002
ABSTRACT: The efficiency of separation of bergamot essential oil, performed by a countercurrent column filled with Raschig rings and using supercritical carbon dioxide as partition solvent, is affected by various parameters. In the experiments explained in this work, the direct effect of CO2 density was shown and the ratio between the amount of oil loaded to on the column and the amount of CO2 used were discussed. The conditions that produced extracts with a similar volatile fraction composition of starting material and with a high yield (more than 80% of recovery) were those with a low feed:solvent ratio; the lowest bergaptene content was obtained at low CO2 density or at high feed:solvent ratio. A good result was observed at a CO2 density of 206 g/dm3 (8 MPa of pressure and a temperature gradient of 46–50–54 °C) and a feed:solvent ratio of 9.4–9.6; in this separation, a yield of 74–77% and a bergaptene content lower than 0.01% was measured. Copyright © 2003 John Wiley & Sons, Ltd. KEY WORDS:
bergamot essential oil; supercritical carbon dioxide; bergaptene; furocoumarines; volatiles
Introduction Bergamot essential oil, obtained from the peels of the fruit, is composed of a volatile part and a non-volatile fraction. In the volatile fraction, the oxygenated compounds are included in a superior amount to that contained in other citrus essential oils extracted from peels. This oxygenated terpene fraction provides much of the characteristic flavour of bergamot oil and its high amount makes this citrus essential oil unique by its fragrance and aroma. The principal ingredients of the non-volatile fraction are components with a heterocyclic nucleus containing oxygen, which belong to the coumarine and psoralene families, the main components being bergamotene, 5geranyloxy-7-methoxycoumarin, citropten and bergapten.1 Some of these non-volatile compounds, particularly bergapten, have a well-ascertained phototoxic activity. The bergamot oil is also used in the food industry as a flavouring in ‘Earl Grey’ -type tea and as an ingredient in citrus soft drink flavourings and some natural fruit flavourings. It is a common industrial practice to remove highboiling compounds from cold-pressed oils. The oils are distilled, or chilled to induce the precipitation of high
* Correspondence to: P. Marco, Dipartimento di Biotecnologie per il Monitoraggio Agroalimentare e Ambientale, University of Reggio Calabria, Piazza S. Francesco 4, I-89061 Gallina, Reggio Calabria, Italy. E-mail:
[email protected]
Copyright © 2003 John Wiley & Sons, Ltd.
molecular weight compounds, or alkali-treated to remove furocoumarines. Supercritical carbon dioxide (SC-CO2) extraction could be an alternative method for elimination of the non-volatile residue from bergamot peel oil. SCCO2 offers unusual possibilities for selective extraction and fractionation due to control of solubility via manipulation of its temperature and density. The non-toxicity and volatility of CO2 are particularly important for its application in food industry. So far, few studies have been conducted on the SC-CO2 extraction and fractionation of citrus peel oils. Temelli et al. discussed the application of SC-CO2 extraction to remove terpene hydrocarbons from cold-pressed orange oil and reported phase equilibrium data for the system under different conditions.2,3 The SC-CO2 extraction of bergamot flavedo was studied using CO2 as extraction solvent: a six-fold reduction of bergaptene content in the extract was obtained compared with cold-pressed bergamot essential oil.4,5 The deterpenation and elimination of coumarins and psoralens from some citrus peel oils by SC-CO2 desorption from silica gel has been studied.6–9 Fractionation of citrus oil for the removal of hydrocarbon monoterpenes has been studied.10 The authors used a rectification column in a continuous mode and a model mixture of citrus essential oil was employed to study the separation of limonene from linalool. Other researchers have developed an industrial process for deterpenating citrus essential oils by countercurrent SC-CO2 in a packed column with a work capacity of 250 l feed
Flavour Fragr. J. 2003; 18: 429–435
430
M. POIANA ET AL.
oil/day.11 They obtained a concentrated oil with five or 10 times as much aroma by single passage through the column, the concentration factor being a function of the flow ratio CO2/feed. We were interested in applying SC-CO2 to the fractionation of citrus oils. Scale-up of the process to industrial scale requires a continuous process, which could be carried out using a countercurrent column. In this preliminary study we used the supercritical carbon dioxide fractionation process on bergamot essential oil with the aim of reducing high molecular weight compounds such as psoralenes.
Experimental Materials The cold-pressed bergamot essential oil was purchased from Gatto (Messina, Italy). Distilled bergamot essential oil was manufactured by the same company, using a patented process with a 3 m × 20 cm i.d. column filled with Raschig rings. The distillation pressure was 2.67– 4.00 kPa and the temperature of the distillate was 40– 50 °C. Alkali-defurocumarinized bergamot essential oil was obtained by a treatment of cold pressed oil with an equal volume of a 6% (w/v) sodium hydroxide solution for 12 h under stirring, then the two phases were separated and the oil was washed twice with pure water. Finally the essential oil was centrifuged to separate the trace of water.
Supercritical Extraction Equipment The supercritical pilot plant built by Mueller GmbH (Coburg, Germany) was modified to improve its working and is constituted of the following parts. The extraction column is made of stainless steel, 3 m high and 30 mm i.d. It is divided into three sections, each having a thermostated water jacket. The column is packed with Raschig rings and has three insertion points of feed material at 1, 2, and 3 m high. A high-pressure membrane pump (Lewa, Germany) transfers the supercritical fluid from the reservoir vessel to the plant at the desired pressure. The SC-CO2 enters the column through the bottom and the flow is therefore directed to the top. The SC-CO2 flow can be regulated by varying the piston pump run. A manually adjustable valve controls the chosen extraction pressure. A separator vessel immediately after the expansion valve permits the expansion of the fluid and collects the solutes. A volumetric pump (Lewa, Germany) feeds the cold-pressed oil in the column. A flowmeter measures the CO2 quantity passed through the column.
Copyright © 2003 John Wiley & Sons, Ltd.
Supercritical Separation Conditions The CO2 compressed by the high-pressure pump passes through the pressure expansion valve and reaches the separator autoclave, which is maintained at a pressure of 1.5–2.0 MPa and a temperature of 15–20 °C. The expanded gaseous CO2 finally passes through the flowmeter to measure the quantity used. The essential oil charged into the column by the volumetric pump comes into contact with SC-CO2 and the soluble fraction is carried out the tower from the top. The top product (extracted by supercritical fluid) precipitates into the separator, while the insoluble fraction is collected from the bottom of the column. The column is run with CO2 for 2 h before starting to introduce the bergamot essential oil. The cold-pressed essential oil was charged at the top of column (3 m high). After the experiment was started, the quantity of essential oil in the extracted fraction was measured every 20 min. Steady state was usually reached after 5 h.
Gas Chromatography–Mass Spectrometry A Hewlett-Packard 5973 mass selective detector connected with a 6890 Hewlett-Packard gas chromatograph was used. The separation was achieved by a HP-5-MS fused-silica capillary column (30 m × 0.2 mm i.d., film thickness 0.25 µm). The column temperature was programmed from 50 °C (5 min) to 280 °C (10 min) at 4 °C/min. The flow rate of helium as a carrier gas was 1 ml/min, kept constant for the whole gas chromatographic run. 0.1 µl neat sample was injected in split mode with a split ratio of 1:50 and at a temperature of 250 °C. The ionization voltage was 70 eV. The sample components were identified by matching their mass spectra with those of the Wiley library by comparison with pure standard components and confirmed by their GC retention times and with data in literature.12–15 The quantitative results were expressed as total ion chromatography area normalization. Every sample was injected in triplicate.
Results The main process parameters and the balance matter are shown in Table 1. The density value of CO2 was calculated considering the state parameters (temperature and pressure) of the fluid at the top of the column (when fluid and solutes exit from the tower). Figure 1 shows the three-dimensional (3D) relationship between the solubility of essential oil components (expressed as ml essential oil dissolved into 1 kg CO2 passed through the column), the SC-CO2 density and
Flavour Fragr. J. 2003; 18: 429–435
SUPERCRITICAL CO2 SEPARATION OF BERGAMOT ESSENTIAL OIL
431
Table 1. SC-CO2 process parameters, yield and solubility of the bergamot essential oil extracts Test
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Pressure (MPa)
Column temperature gradient, bottom–top (°C)
CO2 density (g/cm3)
CO2 flow (Kg/h)
Feed (ml/h)
Feed/CO2 ratio (ml essential oil/kg CO2)
Yield (%)
Solubility (ml essential oil/kg of CO2)
8 8 8 8 8 8 8 8 8 8 8 8 9 9 9
52–56–60 52–56–60 52–56–60 52–56–60 52–56–60 46–50–54 46–50–54 52–56–60 52–56–60 42–46–50 42–46–50 42–46–50 52–56–60 52–56–60 52–56–60
191.5 191.5 191.5 191.5 191.5 205.6 205.6 213.6 213.6 219.6 219.6 219.6 235.7 235.7 235.7
4.01 3.90 5.52 3.63 3.85 5.01 4.90 3.90 4.20 3.48 3.68 3.52 3.73 4.09 3.35
23 24 60 90 95 48 46 24 95 80 80 155 38 41 150
5.7 6.2 10.9 24.8 24.7 9.6 9.4 6.2 22.6 23.0 21.7 44.0 10.2 10.0 44.8
81.94 82.87 62.56 31.06 32.42 74.11 76.70 91.00 35.37 35.67 37.26 27.25 89.32 89.78 33.28
4.7 5.1 6.8 7.7 8.0 7.1 7.2 5.6 8.0 8.2 8.1 12.0 9.1 9.0 14.9
Figure 1. Relation between solubility of bergamot essential oil, carbon dioxide density and feed:solvent ratio
the feed:solvent ratio (as ml essential oil charged on the column/kg CO2). The solubility ranged from 4.7 ml of oil dissolved in 1 kg CO2 at a density of 191.5 g/dm3 and a feed:CO2 ratio of 5.7, to 14.9 ml soluble in 1 kg SC-CO2 at 235.7 g/dm3 and a feed:solvent ratio of 44.8. As expected, the amount of soluble oil increases with an increase of the fluid density. At a similar oil loaded:CO2 ratio of 10–11, the soluble amount was 6.8 ml/kg solvent at CO2 density of 191.5 g/dm3 (test 3), 7.1–7.2 and 9.0–9.1 ml/kg at respectively 205.6 g/dm3 and 235.7 g/dm3 of fluid density (tests 6, 7, 13 and 14). This trend was not so clear at an oil-loaded:solvent ratio of 23–25. Analogously to the density effect, a similar increasing trend was observed for the solubility value influenced by the feed:solvent ratio: at constant density an increase of the dissolved amount was measured for feed:solvent
Copyright © 2003 John Wiley & Sons, Ltd.
Figure 2. Relation between recovery yield of extracted bergamot essential oil, carbon dioxide density and feed:solvent ratio
ratio increases. At a CO2 density of 191.5 g/dm3 and a feed:solvent ratio of about 5–6 ml/kg (tests 1 and 2) a solubility of 4.7–5.1 ml separated material/kg CO2 was observed. This value increased to nearly 8 ml/kg at a feed:solvent ratio of 25 ml oil charged per kg CO2 (tests 4 and 5). At a fluid density of 213.6 g/dm3 and a feed:solvent ratio of 6 (test 8), the solubility measured was about 5.6 ml/kg; at a high ratio (22.6), as applied in separation test 9, the soluble amount increased to 8 ml/kg. The increasing trend, due to density and feed:CO2 ratio, influences the real yield of separation. Figure 2 shows the 3D relationship between the recovery yield of essential oil (as amount of the oil recovered from the separator compared with the material charged on the column), CO2 density and feed:CO2 ratio. At a CO2 density
Flavour Fragr. J. 2003; 18: 429–435
432
M. POIANA ET AL.
Figure 3. Relation between bergaptene content of the bergamot essential oil extracts, carbon dioxide density and feed:solvent ratio
of 191.5 g/dm3 the best yield (82–83%) was obtained at the lowest feed:CO2 ratio (5–6 ml essential oil charged/kg of CO2) applied in tests 1 and 2. The ratio of about 10 ml of essential oil charged on the column per kg of CO2 showed a yield of 67%; a further rise of the ratio to about 25 gave a separation yield of 31–32%. In the same manner, at a fluid density of 236 g/dm3 and a feed:solvent ratio of 10 (tests 13 and 14), a yield of nearly 90% was attained, but at the higher ratio of 45 the yield decreased to 33%. Table 2 shows the components and their percentage compositions measured as peak area ratios of the total ion chromatogram, detected in different bergamot extracts obtained under different separation conditions, in the cold-pressed essential oil used as starting material, in the distilled essential oil and in the alkali-treated oil. The sequence of the compounds is according to their retention times obtained with a HP5-MS capillary column. Figure 3 shows the 3D relationship between the bergaptene content of the extracts, CO2 density and feed:CO2 ratio. From this picture is possible to observe the solvent density and the feed:solvent effect on bergaptene solubility and consequently on its content in the separated fractions. At low density (192 g/dm3) the bergaptene content varied from a undetectable amount to 0.015% (expressed as peak area ratio). In separation tests 1–5 the influence of the ratio between oil loaded and CO2 used was not clear. The highest bergaptene amount was measured in the extract of the test 14, with a CO2 density, calculated at the exit stage, of 236 g/dm3 (CO2 pressure of 9 MPa and a gradient temperature of 52–56–60 °C) and a feed:solvent ratio of 10. In these process conditions the bergaptene content was at 0.038%. The extract of test
Copyright © 2003 John Wiley & Sons, Ltd.
Figure 4. Relation between linalyl acetate content of bergamot essential oil extracts, carbon dioxide density and feed:solvent ratio
number 15 was obtained with the same fluid density as separations 13 and 14 (236 g/dm3) but with a higher feed:solvent ratio (44.8). In this extract the bergaptene was measured at 0.007%. A similar decreasing trend was observed for separations performed at CO2 density of 214 and 220 g/dm3. From Table 2 an effect of SC-CO2 fractionation is observed. The starting material, distilled and alkali-treated oil, contained limonene (the most important monoterpene hydrocarbon) at 32.1–33.1%, while in SC-CO2 fractions the percentage total content of this compound was 31.4– 41.8%. The content of linalyl acetate, the characteristic ester of bergamot essential oil, was 29.7–31.3% in coldpressed, distilled and alkali-defurocumarinized bergamot oil. The same compound comprised 15.0–30.1% in the SC-CO2 separated bergamot oils. The effect of separation parameters on the extract composition is shown in Figure 4. In this picture the linalyl acetate content was considered as a marker of good quality extract. The highest amount of linalyl acetate (about 30%) was obtained in separations 1 and 2, in which the lowest CO2 density was applied (191 g/dm3), and in test 14, where the highest density was used.
Discussion Sato et al. reported that a temperature gradient along the column improves the separation conditions in a SC-CO2 fractionation process.10 In the same way, a temperature gradient was applied in the separation tests performed in this work. As previously reported by other researchers, the main components of essential oils are highly soluble
Flavour Fragr. J. 2003; 18: 429–435
Copyright © 2003 John Wiley & Sons, Ltd.
tr, trace.
α-Thujene α-Pinene Camphene Sabinene β-Pinene Myrcene Octanal α-Phellandrene α-Terpinene p-Cymene Limonene (Z)-β -Ocimene (E)-β -Ocimene δ -Terpinene cis-Sabinene hydrate Octanol Terpinolene Linalool Terpinen-4-ol α-Terpineol Decanal Octyl acetate Nerol Neral cis-Sabinene hydrate acetate Linalyl acetate Geranial Lynalyl propionate α-Terpinyl acetate Neryl acetate Geranyl acetate β-Caryophyllene trans-α-Bergamotene β-Bisabolene Nootkatone Bergaptene
Compound
0.4 1.5 tr 0.9 7.0 1.0 tr tr 0.2 0.7 32.1 tr 0.3 8.5 0.1 tr 0.4 12.1 tr 0.1 0.1 0.1 0.1 0.2 0.1 29.7 0.3 0.1 0.3 0.5 0.4 0.6 0.6 0.8 0.1 0.5
Coldpressed
0.4 1.4 tr 1.1 6.7 0.9 tr tr 0.3 0.5 33.1 tr 0.3 8.7 tr tr 0.5 11.9 0.1 0.1 0.1 0.1 0.1 0.2 0.1 30.6 0.3 0.1 0.2 0.4 0.3 0.5 0.5 0.5 tr tr
Distilled
0.4 1.4 tr 0.9 6.7 0.9 tr tr 0.1 0.6 32.5 tr 0.2 8.5 tr tr 0.4 11.8 tr 0.1 0.1 0.1 0.1 0.2 0.1 31.3 0.3 0.1 0.3 0.4 0.4 0.6 0.5 0.7 0.1 tr
Alkalitreated
0.4 1.4 tr 1.2 6.6 1.0 tr tr 0.2 0.7 32.0 0.1 0.3 8.8 0.1 tr 0.5 12.6 tr 0.1 0.1 0.1 0.1 0.2 0.1 29.8 0.3 0.1 0.3 0.5 0.4 0.6 0.5 0.7 tr tr
1 0.4 1.5 tr 1.2 6.8 1.1 tr tr 0.2 0.7 31.8 0.1 0.3 8.9 0.1 tr 0.5 12.7 tr 0.1 0.1 0.1 0.1 0.2 0.1 29.6 0.3 0.1 0.3 0.5 0.4 0.5 0.5 0.6 tr tr
2 0.6 1.9 0.1 1.4 8.5 1.3 0.1 0.1 0.4 0.7 40.7 0.1 0.4 11.3 0.1 tr 0.7 11.4 tr 0.1 tr 0.1 tr 0.1 0.1 19.1 0.1 tr 0.1 0.1 0.1 0.1 0.1 0.1 tr tr
3 1.0 3.2 0.1 1.9 11.0 1.6 0.1 0.1 0.4 0.8 41.8 0.1 0.4 10.5 0.1 tr 0.6 9.3 tr 0.1 tr 0.1 tr 0.1 0.1 16.1 0.1 tr 0.1 0.1 0.1 0.1 0.1 0.1 tr tr
4 1.2 3.5 0.1 2.0 11.2 1.8 0.1 0.1 0.5 0.8 41.2 0.1 0.4 10.6 0.1 0.1 0.6 9.6 tr 0.1 tr 0.1 tr 0.1 0.1 15.0 0.1 tr 0.1 0.1 0.1 0.1 0.1 0.1 tr tr
5 0.4 1.3 tr 1.1 6.7 1.4 tr 0.1 0.2 0.7 35.3 0.3 0.6 8.7 0.1 tr 0.5 11.4 tr 0.1 0.1 0.1 tr 0.1 0.1 28.4 0.2 tr 0.2 0.5 0.6 0.3 0.2 0.2 tr tr
6 0.5 1.8 0.1 1.2 7.1 1.2 tr 0.1 0.3 0.7 31.7 0.1 0.4 9.4 0.1 0.1 0.6 13.9 0.1 0.2 0.1 0.2 0.1 0.2 0.1 27.9 0.3 0.1 0.2 0.3 0.2 0.4 0.3 0.2 tr tr
7 0.5 1.8 0.1 1.2 7.1 1.2 tr 0.1 0.3 0.6 31.6 0.1 0.4 9.0 0.1 0.1 0.6 13.1 0.1 0.2 0.1 0.2 0.1 0.2 0.1 27.4 0.3 0.1 0.3 0.5 0.5 0.6 0.6 0.8 0.1 tr
8 1.0 3.3 0.1 1.5 11.4 1.7 0.1 0.1 0.4 0.7 41.3 0.1 0.4 10.6 0.1 0.1 0.6 9.5 tr 0.1 tr 0.1 tr 0.1 0.1 15.9 0.1 tr 0.1 0.1 0.1 0.1 0.1 0.1 tr tr
9 0.9 3.1 0.1 1.9 10.9 1.5 tr 0.1 0.4 0.7 41.4 0.1 0.4 10.5 0.1 tr 0.5 9.4 tr 0.1 tr 0.1 tr 0.1 0.1 16.8 0.1 tr 0.1 0.1 0.1 0.1 0.1 0.1 tr tr
10 0.9 3.1 0.1 1.9 10.7 1.6 0.1 0.1 0.4 0.7 41.4 0.1 0.4 10.6 0.1 tr 0.6 9.6 tr 0.1 tr 0.1 tr 0.1 0.1 16.6 0.1 tr 0.1 0.2 0.1 0.1 0.1 0.1 tr tr
11 0.9 3.0 0.1 1.9 10.6 1.5 0.1 0.1 0.4 0.7 41.1 0.1 0.4 10.4 0.1 tr 0.5 9.6 tr 0.1 tr 0.1 tr 0.1 0.1 17.3 0.1 tr 0.1 0.2 0.1 0.1 0.1 0.1 tr tr
12
0.5 1.8 0.1 1.2 7.0 1.2 tr 0.1 0.3 0.6 31.7 0.1 0.4 8.9 0.1 tr 0.6 12.8 0.1 0.2 0.1 0.1 0.1 0.2 0.1 27.7 0.3 0.1 0.3 0.6 0.5 0.7 0.6 1.0 0.1 tr
13
SC-CO2 extracts (number of test; see Table 1 for separation conditions)
0.4 1.5 tr 1.1 6.6 1.1 tr tr 0.3 0.6 31.4 0.1 0.3 8.6 0.1 tr 0.5 12.4 tr 0.1 0.1 0.1 0.1 0.2 0.1 30.1 0.3 0.1 0.3 0.5 0.5 0.6 0.6 0.8 0.1 tr
14
0.9 3.0 0.1 1.8 10.4 1.5 tr 0.1 0.4 0.8 40.6 0.1 0.3 10.1 0.1 tr 0.5 9.7 tr 0.1 tr 0.1 tr 0.1 0.1 18.4 0.1 tr 0.1 0.2 0.1 0.2 0.1 0.1 tr tr
15
Table 2. Composition of bergamot treated and fractionated essential oil. Explanations of different treatments are reported in Experimental, and the conditions of SC-CO2 extractions are reported in Table 1
SUPERCRITICAL CO2 SEPARATION OF BERGAMOT ESSENTIAL OIL 433
Flavour Fragr. J. 2003; 18: 429–435
434
M. POIANA ET AL.
in dense carbon dioxide at already relatively low gas density; in the same manner, bergamot essential oil components are easily dissolved in SC-CO2.4 For this reason, and for a good separation, CO2 densities lower than 240 g/dm3 were applied. The densities tested were 191.5– 235.7 g/dm3, obtained in a range of CO2 pressure and temperature already reported (Table 1). Some parameters affect the solubility of a component in a countercurrent SC-CO2 separation process system. The most important is the density of the solvent, moreover the temperature influences the solvent density value and the vapour pressure of the solute increases its solubility with increase of temperature. Also the feed:solvent ratio plays an important role in a continuous solubilization process. With high ratios the solvent phase could not completely dissolve the fraction soluble in it, but at low ratios a long process time occurs. The solubility range data obtained in the separation tests is in agreement with those reported for citrus essential oil and for single constituents representing a model system of essential oil.2,16 A direct relation between solubility with fluid density and with oil feed:CO2 ratio was observed. This is due to the increase of solvent power and a more saturated solvent, which permits better solubilization of the oil. The recovery yield is influenced by density increase, which causes a yield increase, but a feed:CO2 ratio increase causes a decrease of the yield. The essential oil feed:CO2 ratio affects the transport matter and influences the saturation of the solvent phase, particularly the contact time between the solute (essential oil) and the solvent (SC-CO2). A saturation state of the solvent causes a decrease of the yield. The best yield (91%) was obtained at a CO2 condition of test 8 (density of 214 g/dm3); similar results were observed in tests 13 and 14 at a density of 236 g/dm3. Among the extracts obtained under different process conditions, some differences in composition could occur. Vapour pressure and polarity have a special influence on the solubility behaviour in compressed CO2. The compounds are better extractable the lesser polar they are, and the smaller their molecular weight and the larger their vapour pressure. The essential oils are mixtures of a number of terpene derivative molecules, which have different solubility properties. All components of bergamot essential oil are very soluble in dense CO2 at already low gas density. Furanocumarines, such as bergaptene, also showed a high solubility at low CO2 density. Bergaptene was extracted from bergamot peels at a CO2 density of about 280 g/dm.3–5 There is a decrease in selectivity when CO2 density is increased. This is due to the increase in solubility of furocumarine compounds in CO2 as its density is increased. In order to maximize the selectivity, low CO2 density should be preferred. In this study, we tested solubilities smaller than 240 g/dm3 to separate bergamot
Copyright © 2003 John Wiley & Sons, Ltd.
essential oil from bergaptene and other high-boiling substances. A solubilization effect due to the feed:solvent ratio is apparent. At low ratios the solvent phase could well dissolve the components and so also the less soluble, such as bergaptene, while high ratios caused an overload and the solvent preferentially dissolves the more soluble compounds. Bergaptene in the starting material essential oil was determined to be 0.53% of the total quantified by GC–MS; in the distilled oil it was present as a trace and in the alkali-treated oil at 0.01%. Compared with the starting material, all the separation tests showed a lesser amount of this psoralene. A bergaptene content of
