Available online at www.sciencedirect.

com

Ultrasonics Sonochemistry 15 (2008) 898902 www.elsevier.com/locate/ultsonch

Improved extraction of vegetable oils under high-intensity ultrasound and/or microwaves

Giancarlo Cravotto a,*, Luisa Boa a, Stefano Mantegna a, Patrizia Perego b, Milvio Avogadro b, Pedro Cintas c,*
` di Torino, Via P. Giuria 9, 10125 Torino, Italy Dipartimento di Scienza e Tecnologia del Farmaco, Universita ` di Genova, Via Pia 15, I-16145 Genova, Italy Dipartimento di Ingegneria Chimica e di Processo, Universita c mica Orga nica e Inorga nica, Universidad de Extremadura, Facultad de Ciencias-UEX, E-06071 Badajoz, Spain Departamento de Qu
b a

Received 19 October 2007; accepted 30 October 2007 Available online 9 November 2007

Abstract Ultrasound-assisted extraction (UAE) and microwave-assisted extraction (MAE) techniques have been employed as complementary techniques to extract oils from vegetable sources, viz, soybean germ and a cultivated marine microalga rich in docosahexaenoic acid (DHA). Ultrasound (US) devices developed by ourselves, working at several frequencies (19, 25, 40 and 300 kHz), were used for USbased protocols, while a multimode microwave (MW) oven (operating with both open and closed vessels) was used for MAE. Combined treatments were also studied, such as simultaneous double sonication (at 19 and 25 kHz) and simultaneous US/MW irradiation, achieved by inserting a non-metallic horn in a MW oven. Extraction times and yields were compared with those resulting from conventional procedures. With soybean germ the best yield was obtained with a cavitating tube prototype (19 kHz, 80 W), featuring a thin titanium cylinder instead of a conventional horn. Double sonication, carried out by inserting an immersion horn (25 kHz) in the same tube, improved the yield only slightly but halved the extraction time. Almost comparable yields were achieved by closed-vessel MAE and simultaneous US/MW irradiation. Compared with conventional methods, extraction times were reduced by up to 10-fold and yields increased by 50 500%. In the case of marine microalgae, UAE worked best, as the disruption by US of the tough algal cell wall considerably improved the extraction yield from 4.8% in soxhlet to 25.9%. Our results indicate that US and MW, either alone or combined, can greatly improve the extraction of bioactive substances, achieving higher eciency and shorter reaction times at low or moderate costs, with minimal added toxicity. 2007 Elsevier B.V. All rights reserved.
Keywords: Ultrasound-assisted extraction; Microwave-assisted extraction; Soybean germ; Seaweed; Docosahexaenoic acid

1. Introduction and background Extraction techniques are widely employed for the isolation of bioactive substances from natural sources [1]. However, they are usually time-consuming and, unless carefully controlled, are liable to cause degradation or unwanted chemical changes in the products. Current techniques of solidliquid extraction are essentially based on diusion

Corresponding authors. Tel./fax: +39 0 11 6707684. E-mail address: giancarlo.cravotto@unito.it (G. Cravotto).

and osmosis. To shorten extraction times and improve yields treatments may be carried out at a higher temperature, and/or be repeated several times with fresh solvent. The simplest and cheapest technique is maceration; at any rate stirring is indispensable if local saturation at the matrix surface is to be avoided, so that diusion may work more eciently. In the extraction of essential oils and, generally, of volatile fractions, steam distillation often proves very useful, although exposure to heat may cause degradation of thermally labile compounds. The same drawback aects Soxhlet extraction, the usual preliminary step in the processing of solid samples according to many ocial

1350-4177/$ - see front matter 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.ultsonch.2007.10.009

G. Cravotto et al. / Ultrasonics Sonochemistry 15 (2008) 898902

899

methods of analysis. Industrial plants, that must produce large amounts of extract in a short time, resort to percolation: the solvent drips through large cylinders packed with up to several cubic meters of solid matrix. As a single pass would yield but a limited amount of extract, the euent is enriched by returning it several times to the top of the column. Because eciency limits are ultimately set by diusion rates that increase with temperature, the abovementioned strictures also apply to the possibility of working with hot solvents. Ultrasound-assisted extraction (UAE) and microwaveassisted extraction (MAE) are now recognized as ecient extraction techniques that dramatically cut down working times, increasing yields and often the quality of the extract. Although chiey exploited on the laboratory scale, both have already found industrial applications in this eld, particularly the latter [2]. This employ of ultrasound (US) to enhance extraction yields began in the fties [3]; more recently we showed that it dramatically improves the extraction of oils from plant materials [4], mainly through the phenomenon of cavitation. The mechanical eect of US promotes the release of soluble compounds from the plant body by disrupting cell walls, enhancing mass transfer and facilitating solvent access to the cell content. This eect is much stronger at low frequencies (1840 kHz) and practically negligible at 400800 kHz. If the matrix has been previously dried, US accelerates its rehydration and swelling. Vinatoru et al. [5] showed that it greatly enhanced the hydration process which takes place simultaneously with matrix fragmentation, without any appreciable chemical degradation. UAE has proved to be a versatile technique that can be scaled up to the benet of industrial production. The use of microwave (MW) dielectric heating in analytical laboratories began in the late 1970s and was rst seized upon by the food industry. Dielectric heating depends on the ability of materials to absorb MW energy and convert it to heat. Unlike conductive heating, MW heats the whole sample volume simultaneously. It disrupts weak hydrogen bonds by promoting the rotation of molecular dipoles, an eect that is opposed by the viscosity of the medium. Furthermore, the movements of dissolved ions increase solvent penetration into the matrix and thus facilitate analyte solvation. The eect is strongly dependent on the nature of both solvent and matrix. Sometimes MW aects mainly the latter, while the surrounding liquid, having a low dielectric constant, remains relatively cold. This set-up, that presents obvious advantages when dealing with thermosensitive compounds, has been successfully used for the extraction of natural oils [6]. MW heating of fresh plant material is also a convenient way to achieve direct distillation of essential oils [7]. Liquid-phase extraction (often used to isolate essential oils from plants) is based on the dierent absorbance for MW of materials having dierent dielectric constants. The matrix to be extracted (usually water-rich) is mixed with a solvent having a low dielectric constant, so that most of the heating eect will be concentrated on the plant material. Ganzler and associates [8]

COOH - linolenic acid (18:3 n-3)

COOH linoleic acid (18:2 n-6)

COOH DHA (22:6 n-3)

Fig. 1. Chemical structures of main fatty acids from soybean-germ oil and docosahexaenoic acid (DHA) from seaweed.

were the rst to report the advantages of this method over conventional ones for the extraction of natural compounds. MAE requires less solvent and takes less time, aording better products at lower costs. Some biologically active compounds have been so extracted, e.g. taxanes from Taxus brevifolia [9], glycyrrhizic acid from Glycyrrhiza glabra [10], and artemisinin from Artemisia annua [11]. In a previous paper, we compared yields and times of rice bran extractions using both UAE and conventional methods [4]. These results prompted us to extend our study to related techniques, introducing novel apparatuses as well. After widely experimenting with the design, construction and testing of new reactors using US [12] and MW alone or combined, we report here the rst instances of extraction employing simultaneous US and MW irradiation [13] as well as double sonication combining two transducers working at dierent frequencies (19 and 25 kHz). We applied these environmentally benign methods to: (1) soybean germ, an important source of tocopherols, carotenoids, and polyunsaturated fatty acids (PUFAs); (2) a cultivated seaweed that is rich in docosahexaenoic acid (DHA) (Fig. 1). Gas-chromatographic analyses of methyl ester derivatives, revealed little or no oxidative degradation of PUFAs, which was not the case for parallel extraction runs carried out with conventional methods. In this context, it is worth mentioning a recent study in which Zhao and coworkers investigated the eects of MW and US on the stability of a keto-carotenoid, (all-E)-astaxanthin, that is present in marine animals and algae; they found that MW induced isomerization to its Z forms, while US promoted degradation to unidentied compounds [14].

2. Experimental (materials and methods) 2.1. Chemicals All reagents and solvents were purchased from Carlo Erba Reagenti and Acros Organics. The extraction course was followed by thin layer chromatography (TLC) on Fluka F254 (0.25 mm) plates, which were visualized by

900

G. Cravotto et al. / Ultrasonics Sonochemistry 15 (2008) 898902

heating after a molybdic acid spray. Milled pure soybean germ (150 mesh) and milled seaweed (150 mesh) were kindly provided by BF Pharma Spa Italia (Fossano, Italy). Both were extracted with hexane (Carlo-Erba, Italy): ltered extracts were evaporated under vacuum and weighed. 2.2. Microwave and ultrasonic equipment All MW runs were carried out with a professional multimode oven operating at 2.45 GHz (Microsynth, Milestone, Italy). US treatments were conducted with the following devices developed in our laboratories in collaboration with Danacamerini sas (Torino, Italy), working at dierent frequencies (19, 25, 40, and 300 kHz): (1) high-power probe systems with immersion titanium horns, working at 25 and 40 kHz, (2) a titanium cup-horn working at 300 kHz. The transducer can work without overheating at high power (up to 500 W) thanks to an ecient cooling system that employs a chiller and an oil circulation circuit; (3) the high-power cavitating tube (Fig. 2), an innovative cuphorn-like system, consisting of a thin hollow titanium cylinder whose base is xed on a booster. Measurements with a small hydrophone showed that acoustic energy was eciently transferred to the liquid both from its bottom and wall. A thermostatting uid was circulated through the outer polytetrauoroethylene (PTFE) jacket. Inserting an immersion horn in the cavitating tube made it possible to combine two transducers working at dierent frequencies (Fig. 2); (4) simultaneous US/MW irradiation in a single reaction vessel was achieved by inserting in a MW oven a horn made of PEEK (polyether ether ketone). Fig. 3 shows a professional oven currently employed in our laboratory for such combined US/MW irradiations. 2.3. Chromatographic analyses A Shimadzu GC-14 B (FID detector) with a Shimadzu C-R6A Chromatopac integrator was used for analyses by

Fig. 3. Apparatus for simultaneous US/MW irradiation.

gas chromatography (GC). All data presented below are mean values from three extractions. The prole of fatty acids (derivatized as methyl esters) was determined according to the Ocial Method of Analysis AOAC [15] by gas chromatography using an OV1 apolar column (length 8 m; i.d. 0.25 mm; lm thickness 0.25 lm). GC conditions were: injection split 1:20, injector temperature 250 C, detector temperature 300 C; temperature program: from 50 C (2 min) to 100 C at 3 C/min, then to 280 C (5 min) at 5 C/min, then the temperature was kept constant at 280 C for 20 min; hydrogen as carrier gas at 25 kPa. 3. Results and discussion The great potential of US- and MW-assisted extractions has not been adequately exploited as yet; the combined use of both energy sources has hardly been investigated at all [16]. High-intensity US generate cavitational bubbles and streaming. Bubble collapse in the vicinity of plant teguments may cause microfractures in these tissues [17]. MAE generally works better than conventional methods, as it saves both time and solvent volume, has a high extraction eciency and a low environmental impact [18]. While the Soxhlet method usually takes hours (up to 15 or more), MAE only takes a few minutes, using (at most) 10 times less solvent. In the present MAE treatments samples were suspended in an apolar solvent such as n-hexane, so that MW heated the fresh plant material almost exclusively owing to its high moisture content. The resulting disruption of cell membranes released the oil to the solvent. As previously mentioned, we tested several US devices working at dierent frequencies and dierent MW programs, as well as US/MW combinations. Table 1 compares results obtained for hexane extraction of soybean germ under dierent conditions, including conventional techniques (static extraction at room temperature and reuxing

Fig. 2. Immersion horn (25 kHz) inserted in the cavitating tube (19 kHz); the inset shows the latter as seen from the top.

G. Cravotto et al. / Ultrasonics Sonochemistry 15 (2008) 898902 Table 1 Hexane extraction of soybean germ and seaweed: optimal extraction times and oil yields as % of matrix weight Entry 1 2 3 4 5 6 7 8 9 10
a b c

901

Extraction method Separatory funnel (static) Soxhlet Immersion horn 25.0 kHz, 80 W Immersion horn 40.0 kHz, 80 W Cup horn 300 kHz, 70 W Cavitating tube 19.0 kHz, 80 W Cavitating tube 19.0 kHz, 65 W and immersion horn 25.0 kHz, 60 W Comb. US/MW (60/100 W)b MW, open vesselc MW under pressurec

Matrice weight (g) 15 15 8 8 8 8 8 8 3 2

Solvent volume (ml) 50 100 50 50 40 50 50 40 40 35

Temp. (C) r.t. rfx 45 45 45 45 45 45 60 120

Time (h) 8 4 1 1 1 1 0.5 1 1 0.5

Yield SGa (%) 3.5 8.6 12.2 8.9 6.4 17.7 17.9 14.1 10.0 16.5

Yield SWa (%) 2.0 4.8 18.7 13.6 12.0 24.7 25.9 24.0 12.5 17.8

SG, soybean germ; SW, seaweed. US horn made of PEEK. Magnetic stirrer (P/N 86116) and weon button for apolar solvents (P/N WO1703).

in a Soxhlet). As expected, signicant dierences are seen in terms of extraction times and yields. The best result was achieved by our novel US device we call a cavitating tube, working at 19 kHz. Inserting in the cavitating tube a titanium immersion horn (25 kHz and same US power) improved the oil yield only slightly but halved the extraction time. To avoid solvent splashing, the horn and the external wall of the tube were joined by an elastomeric sleeve to seal the system. Good results were likewise obtained (Table 1, last entry) with closed-vessel MW, although this required a much higher temperature. Simultaneous US/MW irradiation cannot be performed by introducing conventional metallic horns inside the MW chamber, as this would cause arcing and explosions. We resorted to a PEEK horn (21 kHz, 50 W) inserted in the MW oven (100 W) (Fig. 3); this setup aorded satisfactory extraction yields at moderate temperatures. Although MAE and UAE can be conducted at lower temperatures than those attained in Soxhlet extraction, nevertheless superheating and pyrolytic degradation can occur; therefore much care must be exerted in applying these techniques to the extraction of bioactive compounds. Soybean-germ oil is rich in thermolabile unsaturated compounds [19,20] that include PUFAs (both omega-3 and omega-6 acids), carotenoids, tocopherols, phytosterols and coenzyme Q10. UAE of the whole soybean has been reported [21] as well as its application to separating isoavones from freeze-dried ground soybean [22]. Gas chromatography (GC) analyses generally showed no dierences between the methyl esters proles of soybean germ oils extracted under conventional and ultrasonic conditions. Owing to the very brief sonication times and the low moisture content (4.5%) of the matrix, we observed but a very small decrease in the relative percentage of unsaturated fatty acids. Data for oils extracted by Soxhlet and simultaneous double sonication are shown in Table 2. Results indicate that only a very slight oxidation occurs, irrespective of the degree of unsaturation, upon application of high-intensity US using metallic horns.

These positive, reproducible results encouraged us to extend our study to other matrices, in particular the axeniTable 2 Methyl ester proles of soybean-germ oil from UAE and Soxhlet extraction Fatty acid compositiona C18:2 C18:3 C16:0 C18:1 C18:0 C20:0
a b

Soxhlet (%) 56.8 15.7 12.7 10.6 3.6 0.4

Double sonicationb (%) 56.3 15.4 12.7 10.6 3.7 0.4

linoleic linolenic palmitic oleic stearic arachidic

GC analyses of fatty acid methyl esters were replicated twice. Cavitating tube 19.0 kHz, 65 W + immersion horn 25.0 kHz, 60 W.

Table 3 Methyl ester proles of seaweed oil from UAE and Soxhlet extraction Fatty acids compositiona C12:0 lauric C14:0 myristic C15:0 pentadecanoic C18:1 oleic C18:0 stearic C20:0 arachidic C16:0 palmitic C16:1 palmitoleic C17:0 margaric C18:2 linoleic C18:3 c linolenic C18:4 stearidonic C22:0 behenic C20:3 eicosatrienoic n-6 C20:4 arachidonic C20:5 eicosapentaenoic C22:5 docosapentaenoic n-6 C22:5 docosapentaenoic n-3 C22:6 docosahexaenoic C24:0 lignoceric Others
a b

Soxhlet (%) 0.2 3.8 2.0 0.1 1.5 0.1 38.0 0.1 0.5 0.1 0.1 0.3 1.3 0.2 0.1 0.4 8.8 0.4 39.5 0.1 2.4

Double sonicationb (%) 0.2 3.9 2.0 0.1 1.6 0.1 37.9 0.1 0.6 0.1 0.1 0.2 1.2 0.2 0.1 0.3 8.6 0.4 39.3 0.1 2.9

GC analyses of the fatty acid methyl esters were replicated twice. Cavitating tube 19.0 kHz, 65 W + immersion horn 25.0 kHz, 60 W.

902

G. Cravotto et al. / Ultrasonics Sonochemistry 15 (2008) 898902

cally cultivated marine microalga Crypthecodinium cohnii [23]. Some varieties of seaweed, long known as traditional foodstu, possess high added value due to their content in biologically active substances [24]. Of particular importance is DHA [25], an omega-3 fatty acid, present in seaweed; this is in fact the source from which shes get their DHA, also serving as a natural antifreeze. Health benets of DHA and other PUFAs contained in marine plants have been reported in the literature; they include prevention of cardiovascular disease, obesity, besides their well-known roles in inammation and the immune response [2630]. Results for MAE and UAE of the microalga Crypthecodinium cohnii are shown in Table 1. As was found with soybean germ oil, the methyl ester proles of seaweed oil obtained under conventional and ultrasonic conditions were comparable, showing that no signicative PUFAs degradation had occurred (Table 3). 4. Conclusions UAE of soybean-germ oil using several types of US apparatuses (cup horn, immersion horn, cavitating tube) working at dierent frequencies (19, 25, 40 and 300 kHz) has been evaluated. In a newly developed apparatus simultaneous US/MW irradiation was achieved by inserting a PEEK horn in a multimode oven. Optimum extraction times were determined and yields were compared with those obtained by MAE (open- and closed-vessel) and conventional methods. The best oil yield was obtained with the cavitating tube (19 kHz) and double sonication employing an additional immersion horn (25 kHz). Compared with conventional methods much higher yields were also achieved with closed-vessel MW irradiation at 120 C and simultaneous US/MW irradiation. Results were even more striking in the case of seaweed extraction, as the cell wall of the microalga is very tough. Extraction times were reduced up to 10-fold and yields increased by 50500% in comparison with conventional methods. GC analyses showed only slight or negligible dierences in methyl esters proles of oils extracted under high-intensity US and in Soxhlet. It should also be pointed out that techniques using US or combined US/MW irradiation should be well suited for other processes, such as two-step extraction and transesterication for the production of biofuels. Acknowledgements The present work was supported by Regione Piemonte CIPE bando 2004, BF Pharma (Fossano, Italy), and the

Spanish Ministry of Education and Science (CTQ200507676). The authors thank Danacamerini sas for designing and assembling the US apparatuses. References
[1] R. Self, Extraction of Organic Analytes from Foods. A Manual of Methods, Royal Society of Chemistry, Cambridge, 2005. [2] G. Cravotto, P. Cintas, Extraction of avourings from natural sources, in: A. Taylor, J. Hort (Eds.), Modifying Flavour in Food, Woodhead Publishing Ltd., Cambridge, 2007, pp. 4163 (Chapter 3). [3] A. Schmall, Schweiz. Brau.-Rundschau (1953); A. Schmall, Chem. Abstr. 47 (1954) 2932. [4] G. Cravotto, A. Binello, G. Merizzi, M. Avogadro, Eur. J. Lipid Sci. Technol. 106 (2004) 147151. [5] M. Toma, M. Vinatoru, L. Paniwnyk, T.J. Mason, Ultrason. Sonochem. 8 (2001) 137142. [6] B. Kaufmann, P. Christen, Phytochem. Anal. 13 (2002) 105113. [7] M.E. Lucchesi, F. Chemat, J. Smadja, J. Chromatogr. A 1043 (2004) 323327. [8] K. Ganzler, A. Salgo, K. Valko, J. Chromatogr. 371 (1986) 299306. [9] M.J.I. Mattina, W.A.I. Berger, C.L. Denson, J. Agric. Food Chem. 45 (1997) 46914696. [10] X. Pan, H. Liu, G. Jia, Y.Y. Shu, Biochem. Eng. J. 5 (2000) 173 177. [11] J. Hao, W. Han, S. Huang, B. Xue, X. Deng, Sep. Purif. Technol. 28 (2002) 191196. [12] G. Cravotto, G. Omiccioli, L. Stevanato, Ultrason. Sonochem. 12 (2005) 213217. [13] G. Cravotto, P. Cintas, Chem. Eur. J. 13 (2007) 19021909. [14] L. Zhao, G. Zhao, F. Chen, Z. Wang, J. Wu, X. Hu, J. Agric. Food Chem. 54 (2006) 83468351. [15] AOAC (Association of Ocial Analytical Chemists), Ocial Method of Analysis 2 (1990) 963964. ve que, G. Cravotto, Chimia 60 (2006) 18. [16] J.M. Le [17] M. Vinatoru, M. Toma, O. Radu, P.I. Filip, D. Lazurca, T.J. Mason, Ultrason. Sonochem. 4 (1997) 135139. [18] A. Sides, K. Robards, S. Helliwell, Trends Anal. Chem. 19 (2000) 322329. [19] F. Bonina, C. Puglia, M. Avogadro, E. Baranelli, G. Cravotto, Arch. Pharm. Life Sci. 338 (2005) 598601. [20] G. Cravotto, B. Robaldo, A. Binello, F. Viola, S. Oliaro-Bosso, M. Avogadro, Eur. J. Lipid Sci. Technol. 107 (2005) 701705. [21] H. Li, L. Pordesimo, J. Weiss, Food Res. Intern. 37 (2004) 731738. [22] M.A. Rostagno, M. Palma, C.G. Barroso, J. Chromatogr. A 1012 (2003) 119128. [23] T. Iizuka, K. Uehara, T. Takeuchi, Patent 1997, JP 09191890, CAN 127:134791. [24] A. Sukenik, H. Takahashi, S. Mokady, Annals Nutrit. Metab. 38 (1994) 8596. [25] W. Yongmanitchai, O.P. Ward, Phytochemistry 30 (1991) 2963 2967. [26] L.A. Horrocks, Y.K. Yeo, Pharmacol. Res. 40 (1999) 211225. [27] A.P. Simopoulos, J. Am. Coll. Nutr. 21 (2002) 495505. [28] W.E. Hardman, J. Nutr. 132 (2002) 35083512. [29] K.W. Lee, G.Y.H. Lip, Q. J. Med. 96 (2003) 465480. [30] J.L. Breslow, Am. J. Clin. Nutr. 83 (2006) 14771482.