In the last 10 years there has been an increased interest in using techniques involving microwave-assisted extraction and pressurised solvent extraction in analytical laboratories.
Título original
Recent Extraction Techniques for Natural Products : Microwave-assisted Extraction and Pressurised Solvent Extraction
In the last 10 years there has been an increased interest in using techniques involving microwave-assisted extraction and pressurised solvent extraction in analytical laboratories.
In the last 10 years there has been an increased interest in using techniques involving microwave-assisted extraction and pressurised solvent extraction in analytical laboratories.
Pressurised Solvent Extraction Beatrice Kaufmann and Philippe Christen* University of Geneva, School of Pharmacy, Laboratory of Pharmaceutical Analytical Chemistry, 20 bd dYvoy, CH-1211 Geneva 4, Switzerland In the last 10 years there has been an increased interest in using techniques involving microwave-assisted extraction and pressurised solvent extraction in analytical laboratories. This review gives a brief overview of both methods, and reports on their application to the extraction of natural products. The inuence of parameters such as the nature of the solvent and volume, temperature, time and particle size of the matrix is discussed. Through numerous examples, it is demonstrated that both techniques allow reduced solvent consumption and shorter extraction times, while the extraction yields of the analytes are equivalent to or even higher than those obtained with conventional methods. Copyright # 2002 John Wiley & Sons, Ltd. Keywords: Microwave-assisted extraction; solidliquid extraction; supercritical uid extraction; pressurised solvent extraction; accelerated solvent extraction; enhanced solvent extraction; natural products. INTRODUCTION A broad spectrum of solidliquid extraction (SLE) techniques is widely used for the early purication of natural products from plant material and micro-organ- isms. Classically, SLE can be divided into traditional and recent methods. Traditional methods include Soxhlet extraction, maceration, percolation, turbo-extraction (high speed mixing) and sonication. These techniques have been used for many decades; however, they are very often time-consuming and require relatively large quantities of polluting solvents. Supercritical uid extraction (SFE), microwave-assisted extraction (MAE) and pressurised solvent extraction (PSE) are fast and efcient unconventional extraction methods developed for extracting analytes from solid matrixes. Numerous review articles and textbooks have been published on SFE (Lee and Markides, 1990; Westwood, 1993; King and Bott, 1993) and a number are more particularly dedicated to SFE of natural products (Castioni et al., 1995; Smith, 1995, 1996; Jarvis and Morgan, 1997). The present review deals with the application of MAE and PSE to the extraction of secondary metabolites from plant material. MICROWAVE-ASSISTED EXTRACTION Microwaves have been used since World War II following the development of radar technology, and later the rst commercial application of microwaves con- cerned domestic ovens. The use of microwave energy as a heating source in analytical laboratories started in the late 1970s and was applied to acid digestions (Abu Samra et al., 1975). The development of microwave assisted extractions was rst reported by Ganzler and co-workers (Ganzler et al., 1986a; Ganzler and Salgo, 1987). Principle of the method and heating mechanism Microwaves are electromagnetic radiations with a frequency from 0.3 to 300 GHz (Camel, 2001). In order to avoid interferences with radio communications, domestic and industrial microwaves generally operate at 2.45 GHz (Fig. 1). Owing to their electromagnetic nature, microwaves possess electric and magnetic elds which are perpendicular to each other. The electric eld causes heating via two simultaneous mechanisms, namely, dipolar rotation and ionic conduction (Thuery, 1992; Demesmay and Olle, 1993; Sinquin et al., 1993). Dipolar rotation is due to the alignment on the electric eld of the molecules possessing a dipole moment (either permanent or induced by the electric eld) in both the solvent and the solid sample. This oscillation produces collisions with surrounding molecules and thus the liberation of thermal energy into the medium. With a frequency of 2.45 GHz, this phenomenon occurs 4.9 10 9 times per second (Ganzler et al., 1990; Sinquin et al., 1993; Barnabas et al., 1995; Onuska and Terry, 1995) and the resulting heating is very fast. Indeed, the larger the dielectric constant of the solvent (see Table 1), the more optimal the heating (Jassie et al., 1997). Consequently, unlike classical conductive heating methods, microwaves heat the whole sample simulta- neously (Fig. 2). In the case of extraction, the advantage of microwave heating is the disruption of weak hydrogen bounds promoted by the dipole rotation of the molecules. A higher viscosity of the medium lowers this mechanism by affecting molecular rotation (Sinquin et al., 1993; Camel and Bermond, 1999). Furthermore, the migration of dissolved ions increases solvent penetration into the PHYTOCHEMICAL ANALYSIS Phytochem. Anal. 13, 105113 (2002) DOI: 10.1002/pca.631 Copyright # 2002 John Wiley & Sons, Ltd. * Correspondence to: P. Christen, University of Geneva, School of Pharmacy, Laboratory of Pharmaceutical Analytical Chemistry, 20 bd dYvoy, CH-1211 Geneva 4, Switzerland. Email: philippe.christen@pharm.unige.ch Received 26 October 2001 Revised 17 December 2001 Accepted 20 December 2001 matrix and thus facilitates the solvation of the analyte (Ganzler et al., 1990). Ionic currents are also induced in the solution by the electric eld. As the medium resists these currents, frictions occur and heat is liberated by a Joule effect. This phenomenon depends on the size and charge of the ions present in the solution. The effect of microwave energy is strongly dependent on the nature of both the solvent and the solid matrix. Solvents generally used cover a wide range of polarities, from heptane to water. Most of the time, the chosen solvent possesses a high dielectric constant and strongly absorbs microwave energy, however, the extracting selectivity and the ability of the medium to interact with microwaves can be modulated by using mixtures of solvents (Renoe, 1994). In some cases, the matrix itself interacts with microwaves while the surrounding solvent possesses a low dielectric constant and thus remains cold (Jassie et al., 1997). This latter situation presents some obvious advantages in the case of thermosensitive compounds and has been successfully used for the extraction of essential oils (Pare, 1990, 1991; Chen and Spiro, 1994; Romele and Polesello, 1997). Indeed, microwaves interact selectively with the polar molecules present in glands, trichomes or vascular tissues. Loca- lised heating leads to the expansion and rupture of cell walls and is followed by the liberation of essential oils into the solvent (Pare et al., 1994; Jassie et al., 1997; Majors, 1998). This situation can also be obtained when a dry sample has been re-hydrated before extraction (Pare, 1990; Pare et al., 1994; Majors, 1998). In fact, moisture content is essential in MAE because water locally superheats and promotes analytes to be liberated into the surrounding medium (Onuska and Terry, 1993; Budzinski et al., 1996; Jassie et al., 1997). In addition, control of the water content of the matrix allows more reproducible results. Instrumentation Two types of instruments are commercially available and they use different approaches. The most common procedure involves extraction in a closed vessel under controlled pressure and temperature, whilst an alternative approach uses an open extracting vessel under atmos- pheric pressure. It must be strongly stressed that using a domestic microwave oven for laboratory purposes should not be considered. Application of microwave energy to highly ammable organic solvents may cause serious hazards. Furthermore, reproducibility may be poor with a domestic device because of the lack of homogeneity of the microwave eld. It is therefore seriously recom- mended that only equipment approved for MAE be used. Closed vessel systems. Such systems are generally advised for digestions or acid mineralisations or for extractions under drastic conditions, since the solvents may be heated to ca. 100C above their atmospheric boiling point (Barnabas et al., 1995; Jassie et al., 1997). Both extraction speed and efciency are enhanced in this procedure. Hazards occasioned by heating highly am- mable solvents are overcome through the use of recent security techniques such as high capacity exhaust fans, solvent vapour detectors, or pressure-burst safety mem- branes placed on each vessel (Demesmay and Olle, 1993; Jassie et al., 1997). Figure 3 shows a schematic diagram of a closed vessel system from CEM Corporation (Matthews, NC, USA). In order to overcome the non- homogeneity of the eld, the cells are placed on a rotating carousel as in a domestic oven. The solvents can be varied, and the pressure in the vessels essentially depends on the volume and boiling point of the solvents used. The temperature can be set at a xed value by adjusting the microwave power. Typically, the cells are made of Teon. In closed vessel systems, the maximal power delivered is about 6001000 W (Pare, 1990, 1991; Young, 1995), but the chosen power has to be set correctly to avoid excessive temperatures leading to possible solute degradation and overpressure problems. The vessel must be cooled to room temperature before opening: this is particularly important in the presence of volatile solutes which can partition into the head-space, but this step can considerably increase the overall Figure 1. The electromagnetic spectrum. Table 1. Dielectric constants and dipole moment values of some commonly used solvents Solvent Dielectric constant (20C) Dipole moment (25C) (Debye) Hexane 1.89 <0.1 Toluene 2.4 0.36 Dichloromethane 8.9 1.14 Acetone 20.7 2.69 Ethanol 24.3 1.69 Methanol 32.6 2.87 Water 78.5 1.87 Figure 2. Scheme of the heating principle by conduction in the classical method of extraction and by microwave irradiation in microwave-assisted extraction. 106 R. KAUFMANN AND P. CHRISTEN Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002) extraction time. Furthermore, an additional ltration or centrifugation step is necessary in order to remove the solid residue. Open cells. These cells are quartz vessels topped by a vapour condenser (Fig. 4). The system works at atmos- pheric pressure, and the maximum temperature is determined by the boiling point of the solvent used (Renoe, 1994; Letellier et al., 1999). The solvent is heated and reuxed through the sample, and in this case the microwaves are focused on the sample placed into the vessel allowing homogeneous and very efcient heating (Demesmay and Olle, 1993). The sample to be extracted can be placed into a Soxhlet-type cellulose cartridge in order to avoid ltration steps, or may be directly dipped into the solvent. Compared to closed vessel extractions, open cells offer increased safety in sample handling and, furthermore, they allow larger samples to be extracted. A typical system is the Soxwave 3.6 commercialised by Prolabo (Fontenay-sous-Bois, France) which operates at percentage power increments from 10 to 100%, corre- sponding to a maximum power of 250 W. Applications to the extraction of natural products Although many of the initial publications dealt with natural product extraction, the majority of the applica- tions concern the extraction of pollutants from environ- mental matrixes. The use of MAE for environmental analysis has recently been exhaustively reviewed (Dean et al., 1999; Camel, 2000), and, therefore, only applica- tions to natural products are reported in this review and these are summarised in Table 2. Closed vessels Pyrimidine glycosides. The rst papers to report the use of MAE for natural products were published by Ganzler and co-workers (Ganzler et al., 1986a, b; Ganzler and Salgo`, 1987) and concerned the extraction of vicine and convicine from faba beans: these toxic pyrimidine glycosides preclude the use of faba beans as a source of nutritional proteins. Ground beans were suspended in a methanol:water mixture (1:1, v/v), and the suspension was subjected to two successive microwave irradiations (30 s each) with a cooling step in between. No degradation could be observed under these conditions, but further irradiation was found to decrease the yield of vicine and convicine. The yield obtained was 20% higher than with the conventional Soxhlet extraction method. Gossypol. As for faba beans, the presence of gossypol in cotton seeds limits their use for human consumption or animal feeding (Ganzler et al., 1986a; Ganzler and Salgo`, 1987). Extraction yields of gossypol were much higher with three cycles of 30 s each of MAE than with the traditional 4 h Soxhlet extraction, probably because the compound is sensitive to high temperatures. This phenomenon was conrmed by further MAE irradiations which also lead to the degradation of gossypol. Alkaloids. Sparteine, a lupine alkaloid, was extracted from lupine seeds (Lupinus mutabilis) with methanol:a- cetic acid (99:1, v/v) in a domestic microwave oven (see caution above!), and the treatment (one to ve cycles of 30 s with a cooling step in between) gave 20% more sparteine than was obtained with a shaken-ask extrac- tion using the same solvent mixture for 20 min (Ganzler et al., 1990). Pyrrolizidine alkaloids were extracted from dried plant material (Senecio sp.) by Bicchi et al. (1992) using closed vessel technology with temperature feed-back control (65100C). Qualitative and quantitative chromato- graphic results were identical to those obtained with traditional techniques, with a signicant reduction in extraction time and solvent consumption with good reproducibility. Terpenes. Five terpenic compounds (linalool, a-terpi- neol, citronellol, nerol and geraniol) associated with grape (Vitis vinifera) aroma were extracted from must samples by MAE (Carro et al., 1997). Four variables (extracting solvent volume, extraction temperature, amount of sample and extraction time) were optimised by means of two- and three-level factorial designs. Several conditions were xed, such as the extraction time (10 min) and the applied power (475 W). The solvent volume appeared to be the only statistically signicant factor, but was limited to 15 mL by the cell size. The highest extraction yield was obtained with both the solvent volume and the temperature at their maximum tested values. In contrast, the sample amount had to be minimised in order to obtain the best recoveries. The nal Figure 3. Schematic diagram of a closed-vessel microwave system for extraction. Figure 4. Schematic diagram of an open focused-microwave system for extraction. MICROWAVE-ASSISTED EXTRACTION 107 Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002) optimised extraction conditions were as follows: 5 mL sample amounts extracted with 10 mL of dichloro- methane at a temperature of 90C for 10 min with the microwave power set at 50% (475 W). Essential oils. MAE of essential oils has been patented by Pare (1991). The samples were suspended in hexane and the microwaves reached the inner glandular and vascular systems of the plant material. Owing to the high moisture content of these structures they were heated almost specically and this promoted disruption of cell membranes releasing the analytes into the solvent. The same author also patented the extraction of volatiles from biological material (Pare, 1990); the method involved the exposure of oil-containing cellular matter or glandular systems to microwaves and the dissolution of essential oil in a suitable organic solvent. Another patent has been deposited by Pare (1994) concerning the extraction of volatile oils from plants and biological material. This procedure was applied by Chen and Spiro (1994) for the extraction of essential oils from rosemary and peppermint leaves suspended in hexane, ethanol or mixtures of the two solvents. Scanning electron micrographs of the leaves showed structural changes in the oil-containing glands after microwave extraction: some of them were found to be collapsed and others completely disintegrated. Two distinct extraction mechanisms are plausible, one involving diffusion of the essential oil across the gland wall and the other involving rupture of the gland and liberation of the constituents into the solvent. Either of the two phenomena may be prominent according to the different maturation stages of the glands. MAE has also been compared to classical hydrodis- tillation for the extraction of essential oils from 10 different plant species using a domestic microwave oven (Collin et al., 1991). The yields were generally similar, but the chromatographic proles varied dramatically, especially with respect to the ratios between the different substances. The microwave technique was used to promote desorption of essential oil from Lippia sidoides with a modied commercial microwave oven (Craveiro et al., 1989). Fresh leaves were irradiated in the oven and the Table 2. Application of MAE to natural product extraction Compounds Matrix System Extraction conditions Reference Vicine, convicine (pyrimidine glycosides) Faba beans (Vicia faba) Domestic oven Methanol:water (1:1); two successive irradiations (30 s) with an intermediate cooling step Ganzler et al. (1986a, 1986b); Ganzler and Salgo (1987) Gossypol Cotton seeds Domestic oven Three cycles of irradiation (30 s) with cooling steps in between Ganzler et al. (1986a); Ganzler and Salgo (1987) Sparteine (alkaloid) Lupine seeds Domestic oven Four cycles (30 s) with cooling steps in between Ganzler et al. (1986b, 1990) Terpenes (linalool, terpineol, citronellol, nerol and geraniol) Must (Vitis vinifera) Closed vessels 10 mL dichloromethane; 475 W; 10 min; 90C Carro et al. (1997) Essential oils Monarda stulosa, Allium sp. Modied domestic oven Hexane Pare (1990) Volatile oils Mentha piperita, Thuja occidentalis Modied domestic oven Hexane, alkanes (transparent solvents) Pare (1994) Essential oils Rosemary and peppermint leaves Domestic oven Hexane, carbon tetrachloride, toluene; 750 W; <60 s Chen and Spiro (1994) Essential oils Plant leaves Domestic oven Hexane; <60 s Collin et al. (1991) Essential oils Fresh leaves of Lippia sidoides Domestic oven High cooking level; 5 min Craveiro et al. (1989) Carotenoids Paprika powder Closed vessels 50 W; 120 s; <60C Csiktunadi Kiss et al. (2000) Taxanes (paclitaxel) Needles of Taxus sp. Closed vessels 5 g fresh needles pre- soaked with 5 mL water prior to extraction with 10 mL of 95% ethanol; 100% power; 54 s; 85C Incorvia Mattina et al. (1997) Ergosterol Fungal contaminations Domestic oven 375 W; 35 s Young (1995) Withanolides Iochroma gesnerioides leaves Open cell, focused 100 mg material pre- soaked with 0.6 mL water prior to extraction with 5 mL methanol; 25 W; 40 s Kaufmann et al. (2001a) Cocaine and benzoylecgonine Erythroxylum coca leaves Open cell, focused Methanol; 125 W; 30 s Brachet et al. (2002) Alkaloids Senecio sp. Closed vessels 65100C Bicchi et al. (1992) 108 R. KAUFMANN AND P. CHRISTEN Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002) volatiles were extracted by an air current passing through the system. The yields obtained after 5 min were comparable to those obtained after 6090 min of steam distillation, and both qualitative and quantitative perfor- mances were found similar to the conventional method. Carotenoids. The MAE of carotenoids from paprika has been optimised by Csiktusnadi Kiss et al. (2000). Thirty different water:organic solvent mixtures were evaluated, and both extraction efciency and selectivity were signicantly dependent on the dielectric constant of the extracting solvent mixture. The extractions had to be performed at temperatures lower than 60C because of the possible rearrangement of molecules at higher temperatures. Steroids. Recently, a system has been developed simultaneously to saponify and extract ergosterol by MAE (Young, 1995). The determination of this com- pound in ltered air or contaminated corn or dust can be used as an indicator of fungal contamination. The samples were placed in culture tubes containing 2 mL methanol and 0.5 mL 2 M sodium hydroxide. Microwave irradiation was applied at 375 W for 35 s and the samples were cooled for 15 min before neutralisation with 1 M hydrochloric acid followed by pentane extraction. It was demonstrated that only 3040 s were sufcient to extract ergosterol quantitatively and that the yield was similar to or even higher than that obtained with the traditional methanolic extraction followed by alkaline saponication and pentane extraction. Taxanes. The application of microwave energy to the extraction of taxanes from Taxus biomass was reported by Incorvia Mattina et al. (1997). Various parameters, including temperature, extraction time, solvent choice and water content were investigated in order to optimise extraction efciency. Recoveries of taxane reached 100% of the conventional method when the biomass was freeze-dried to less than 10% moisture and pre-soaked with water prior to extraction using 95% ethanol. As some degradation of paclitaxel occurred at temperatures above 115C, the temperature was set at 85C and MAE was performed for 54 s. The extracts were quantitatively and qualitatively equivalent to those obtained with the conventional extraction method, but with considerable reduction of both extraction time and solvent consump- tion. Open vessels Very few papers have been published concerning the focused microwave extraction of natural products. Alkaloids. Cocaine and benzoylecgonine were extracted from coca leaves (100 mg) by focused MAE (Brachet et al., 2002). The extraction was optimised by taking into account several parameters such as the nature of the extracting solvent, particle size distribution, sample moisture, applied microwave power and radiation time. MAE was found to generate similar extracts to those obtained by conventional SLE but in a more efcient manner. Indeed, 30 s were sufcient to extract cocaine quantitatively from leaves, using methanol as solvent and a microwave power of 125 W. Steroids. Focused MAE was applied to the extraction of withaferin A, iochromolide and withacnistin from the leaves of Iochroma gesnerioides (Kaufmann et al., 2001a). Withanolides are steroidal lactones derived from an ergostane-type skeleton. Six extraction variables, i.e. the nature and volume of the extracting solvent, sample moisture, extraction time, power of irradiation and particle size of the matrix were investigated. The most favourable conditions were obtained using powdered plant material (<220 mm), previously moistened with water for 15 min, and extracted with methanol for 40 s at 25 W. The extraction yield obtained with the optimised method was comparable to that achieved with a Soxhlet extraction. PRESSURISED SOLVENT EXTRACTION Pressurised solvent extraction (PSE) is a SLE technique which has been developed as an alternative to current extraction methods such as Soxhlet, maceration, percola- tion or reux, offering advantages with respect to extraction time, solvent consumption, extraction yields and reproducibility. PSE uses organic solvents at elevated pressure and temperature in order to increase the efciency of the extraction process. Increased tempera- ture accelerates the extraction kinetics and elevated pressure keeps the solvent in the liquid state, thus enabling safe and rapid extractions. Furthermore, high pressure forces the solvent into the matrix pores and hence should facilitate extraction of analytes. High temperatures decrease the viscosity of the liquid solvent, allowing a better penetration of the matrix and weakened solutematrix interactions. In addition, elevated tempera- tures enhance diffusivity of the solvent resulting in increased extraction speed. Instrumentation ASE 2 (accelerated solvent extraction) is a form of PSE and the rst instrument was commercialised by the Dionex Corporation (Sunnyvale, CA, USA) in 1994 (Richter et al., 1996; Reighard and Olesik, 1996). More recently, ESE (enhanced solvent extraction) or PSE instruments have appeared (Bautz et al., 1998). Accelerated solvent extraction. Figure 5 shows a Figure 5. Scheme of an accelerated solvent extraction (ASE) system (from Richter et al., 1996). MICROWAVE-ASSISTED EXTRACTION 109 Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002) scheme of an ASE apparatus. A solid or semi-solid sample is placed into a stainless steel extraction cell which is lled with solvent and heated (50200C) in an oven (Hoer et al., 1995a; Richter et al., 1996). The heating process generates solvent expansion and thus pressure in the extraction cell, typically in the region of 5003000 psi. In order to prevent over-pressurisation of the cell, a static valve pulses open and closed auto- matically when the cell pressure exceeds the set point. The solvent that escapes during this venting is collected in a vial. A static extraction stage of about 510 min is followed by pumping fresh solvent through the system to rinse the sample and the tubing. All the solvent present in the system is then purged with a compressed gas, generally nitrogen (Richter et al., 1995, 1996; Ezzel et al., 1995; Levy, 1998), and the total solvent volume (extractant and rinsings) is collected in the vial. The Dionex apparatus offers the possibility of automation, and up to 24 samples can be sequentially extracted. Cells can be of different sizes: 1, 5, 11, 22 and 33 mL (Richter et al., 1995; Morf et al., 1998; Richter, 1999). Enhanced solvent extraction. An alternative approach to the above is to perform accelerated solvent extractions with a supercritical uid extraction (SFE) apparatus (see Fig. 6; Li et al., 1998) in which the pressure is controlled by the restriction device. The extraction can be static or dynamic. In the dynamic mode, fresh solvent is continuously pumped through the sample at elevated temperature and pressure (Bautz et al., 1998). According to Ficks rst law of diffusion, the transfer rates are accelerated, leading to improved extraction efciencies and reduced extraction times. Carbon dioxide can be used to ush the remaining solvent out of the extraction chamber (Ashraf-Khorassani et al., 1999). In this case, a wide range of cell sizes is available, from 0.2 to a few hundreds of millilitres. Furthermore, the possibility to perform either PSE or SFE using the same apparatus has obvious economical advantages. Environmental applications To date most of the published applications of ASE have been in the area of environmental research. The tech- nique was applied successfully to the extraction of different pollutants from various environmental matrixes (Hoer et al., 1995b). After comparison with the corresponding reference methods, recoveries and preci- sion were generally found equivalent, or even better. Most papers stress the shorter extraction times and lower solvent consumption of ASE in comparison to other conventional preparation techniques of environmental samples with the exception of SFE (Richter et al., 1995; Jensen et al., 1996; Kreisselmeier and Durbeck, 1997; Fisher et al., 1997). ASE and SFE can be considered as complementary, the rst allowing the extraction of polar compounds, while the second is more selective for compounds of low or medium polarity. The ASE methodology is accepted in the US EPA SW-846 Method 3545A for selected priority pollutants from environmen- tal matrixes (Richter et al., 1996; Levy, 1998). Applications to the extraction of natural products Very few applications of ASE have been published in the eld of natural product research and the major contribu- tions are summarised in Table 3. An evaluation of ASE has been made for the extraction of various metabolites covering a large range of structures and polarities (curcuminoids, saponins, avo- nolignans, terpenes) present in different vegetal matrixes such as leaves, roots, fruits, herbs and rhizomes (Benthin et al., 1999). Performances were compared to corre- sponding European Pharmacopoeia methods. Yields were found to be equivalent or even higher with ASE, with a reduction in the extraction time (especially when consecutive extractions with solvents of increasing polarity were made) and solvent consumption (by a factor from two to ve), and with good reproducibility probably occasioned by the minimal sample handling required during the extraction procedure. Curcuminoids. Three 6 min ASE cycles with methanol at 80C were carried out on turmeric rhizomes (Curcuma xanthorrhiza). Almost all of the curcuminoids present in the sample were extracted during the rst cycle, and the second and third extracts afforded only minimal additional curcuminoids. As the amount determined by the corresponding Pharmacopoeia method was much lower, the authors assumed that the latter method was not exhaustive. Saponins. Horse chestnut (Aesculus hippocastanum) contains a saponin mixture called aescin. The Pharma- copoeia method for the extraction of this material involves 3 h defatting followed by 3.5 h of maceration in 65% methanol at room temperature and a 30 min reux. A preliminary ASE defatting step (two cycles of 5 min each with dichloromethane at 100C) was followed by extraction of saponins with 65% methanol at 100C in two 6 min cycles. As previously mentioned, ASE recovery was higher than the corresponding Pharmaco- poeia method. Flavonolignans. Silybin is a avonolignan contained in milk thistle fruit (Silybum marianum) and it possesses hepatoprotective properties. A single ASE treatment with 20 mL hexane (for defatting) followed by 5 min in methanol lead to the quantitative extraction of the avonolignans. The amounts extracted were similar to those obtained using the traditional method which Figure 6. Scheme of a supercritical uid extraction (SFE) system slightly modied for application in pressurised solvent extraction (PSE). 110 R. KAUFMANN AND P. CHRISTEN Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002) consists of a 4 h defatting step (100 mL petroleum ether) followed by Soxhlet extraction with 100 mL methanol. With ASE, the solvent consumption was reduced by a factor of 5. Terpenes. Essential oil of thyme (Thymus vulgaris) was obtained by hydrodistillation and, simultaneously, the plant material was extracted by ASE. Thymol (the major component) was quantied by GC-FID and comparable results were demonstrated for both extraction methods. The ASE conditions were one 5 min cycle with hexane at 50C, followed by one 5 min cycle with dichloromethane at the same temperature. Exhaustive extraction of the essential oil by hydrodistillation required 2 h. Taxanes. ASE with methanol was used for the extraction of taxanes from the bark of Taxus cuspidata (Kawamura et al., 1999) and permitted the extraction of paclitaxel, baccatin III and 10-deacetylbaccatin III (10-DAB) in higher yields than could be obtained using classical SLE (0.10 MPa; 2 24 h; 22C). For ASE various tempera- tures, pressures, extraction times and solvent composition were evaluated. The recoveries of taxanes were highest in the temperature range 100150C for paclitaxel and 120 150C for 10-DAB. Degradation problems occurred at temperatures greater than 160C. The yield hardly changed within the pressure range 1.0120.27 MPa, when extraction was performed at 150C for 15 min. Particular attention has been given to ASE performed with water. Paclitaxel, although reported to be poorly soluble in this solvent, showed high recovery some 50 times higher than that obtained with cold water extraction (22C; 48 h) and ve times higher than with hot water extraction (100C; 1 h). Extraction at 10.13 MPa for 15 min at a temperature within the range 130140C for paclitaxel, and 120130C for 10-DAB, gave the best recoveries (Kawamura et al., 1999). Various methanol: water ratios were also tested and it was shown that the higher the proportion of water the greater the recovery of overall extract from the bark: the highest yields of taxanes were obtained with a methanol:water ratio of 90:10. Naphthodianthrones. ASE was applied to the recovery of naphthodianthrones (e.g. hypericin) from extracts and pharmaceuticals containing St Johns wort (Hypericum perforatum; Morf et al., 1998; Benthin et al., 1999). These compounds are known to show poor solubility in most organic solvents. The inuence of different par- ameters on extraction efciency, such as solvent, time, pressure, temperature and presence of lling material, was studied. The results obtained were compared with those of conventional methods (i.e. sonication and Soxhlet extraction) which are time consuming and require a preliminary pre-extraction step with dichloro- methane followed by different treatments before a colorimetric measurement. A rst ASE cycle with dichloromethane at 100C was performed in order to Table 3. Application of ASE or PSE to natural product extraction Compounds Matrix System Extraction conditions Reference Taxanes (paclitaxel, baccatin III and 10- deacetylbaccatin III) Taxus cuspidata bark ASE Methanol:water (90:10); 10.13 MPa; 15 min; 150C Kawamura et al. (1999) Naphthodianthrones (hypericin) St John's wort herb, extracts, pharmaceuticals ASE Methanol; 100 bar; two cycles of 5 min with Extrelut in the cell; 40C Morf et al. (1998) Naphthodianthrones St John's wort herb (Hypericum perforatum) ASE Defatting with dichloromethane; 5 min; 100C; followed by methanol extraction; three cycles of 5 min; 50100C Benthin et al. (1999) Saponins Horse chestnut seeds (Aesculus hippocastanum) ASE Defatting with dichloromethane; two cycles of 5 min; 100C; followed by extraction with 65% methanol; two cycles of 6 min; 100C Benthin et al. (1999) Terpenes Thyme (Thymus vulgaris) ASE Hexane; one cycle of 5 min; 50C; followed by dichloromethane; one cycle of 5 min; 50C Benthin et al. (1999) Flavonolignans Milk thistle fruit (Silybum marianum) ASE Defatting with hexane (20 mL); one cycle; 100C; followed by extraction with methanol (20 mL); one cycle of 5 min; 100C Benthin et al. (1999) Curcuminoids Turmeric rhizome (Curcuma xanthorrhiza) ASE Methanol; one cycle of 6 min; 80C Benthin et al. (1999) Xanthones and avanones Osage orange tree (bark) (Maclura pomifera) ASE Dichloromethane; 13.8 MPa; three cycles of 5 min and 90 s purge; 40, 80 or 100C Da Costa et al. (1999) Cocaine and benzoylecgonine Erythroxylum coca leaves PSE Methanol; 20 MPa; 1 mL/min; 10 min; 80C Brachet et al. (2001) Withanolides Iochroma gesnerioides leaves PSE Methanol:water (1 :1); 25 MPa; 0.5 mL/min; 10 min; 100C Kaufmann et al. (2001b) MICROWAVE-ASSISTED EXTRACTION 111 Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002) defat the plant material, and this was followed by three cycles (5 min each) with methanol at a temperature ranging from 50 to 100C. The total amounts extracted by ASE were greater than those obtained using classical methods, and the precision was around 1%. Xanthones and avanones. SLE, SFE and ASE have been compared with respect to the recovery of several phenolic compounds from bark of the Osage orange tree (Maclura pomifera; Da Costa et al., 1999). The ASE involved three cycles of 5 min static extraction at 13.8 MPa with 5 min equilibration time in between and a 90 s purge. Three temperatures (40, 80 and 100C) were tested, and about 15 mL dichloromethane was used for each extraction. The ASE required 35 min compared with 45 min for SFE (using carbon dioxide modied with 20% methanol) or 48 h for solvent extraction. Both SFE and ASE extracted xanthones and avanones from the plant material at similar or higher yields than those obtained with SLE. Alkaloids. PSE was applied to the rapid extraction of cocaine and benzoylecgonine from coca leaves (Brachet et al., 2001). Several parameters, including the nature of the extracting solvent, the pressure, temperature, extrac- tion time, addition of alkaline substances and the granular nature of the sample, were investigated in order to nd the best extraction conditions. It was demonstrated that 10 min were sufcient to extract cocaine quantitatively at 80C and 20 MPa. Furthermore, addition of an alkaline substance did not improve cocaine recovery, since degradation occurred during the extraction process. Steroids. A PSE method was developed for the recovery of three withanolides (withaferin A, iochromolide, and withacnistin) from the leaves of Iochroma gesnerioides in the dynamic mode (Kaufmann et al., 2001b) and the performance of the method was compared to traditional Soxhlet extraction. The inuence of the nature of the solvent, the ow rate, the pressure and the temperature of the extracting solvent, as well as the particle size of the plant material, on the recovery of analytes was investigated. A 1:1 mixture of methanol:water at a ow rate of 0.5 mL/min was found to be the most efcient extracting solvent and allowed a quantitative extraction in 10 min. PSE produced similar results to Soxhlet in terms of recovery, repeatability and selectivity; however, both total handling time and solvent consumption were dramatically reduced with PSE. CONCLUSIONS MAE and PSE are emerging as attractive alternatives to conventional extraction methods such as Soxhlet, percolation, digestion, extraction under reux, sonication and in some cases steam distillation. Initially employed as a digestion method for different sample types such as environmental, biological and geological matrixes, MAE is now widely accepted in analytical laboratories. The main advantage of MAE resides in the performance of the heating source. The high temperatures reached by microwave heating reduces dramatically both the extrac- tion time and the volume of solvent required. PSE works according to the principle of SLE with elevated temperature and high pressures in order to keep the solvent in a liquid state. Enhanced diffusivity of the solvents leads to an increase in extraction speed and efciency. With both MAE and PSE, recoveries of analytes and reproducibilities are improved and, there- fore, both methods should be considered as interesting alternatives with the limitation that experimental condi- tions must be chosen in order to avoid possible thermal degradation. Finally the costs of the specialised equip- ment (especially PSE) may also inuence the choice of the extraction technique (Camel, 2001). REFERENCES Abu-Samra A, Morris JS and Koirtyohann SR. 1975. Wet ashing of some biological samples in a microwave oven. Anal Chem 47: 14751477. Ashraf-Khorassani M, Hinmann S and Taylor LT. 1999. A unied sample preparation system for the extraction of various analyte/matrix combinations. J High Resol Chro- matogr 22: 271275. Barnabas IJ, Dean JR, Fowlis IA and Owen SP. 1995. Extraction of polycyclic aromatic hydrocarbons from highly contaminated soils using microwave energy. Analyst 120: 18971904. Bautz H, Polzer J and Stieglitz L. 1998. Comparison of pressurised liquid extraction with Soxhlet extraction for the analysis of polychlorinated dibenzo-p-dioxins and dibenzofurans from y ash and environmental matrixes. J Chromatogr A 815: 231241. Benthin B, Danz Hand Hamburger M. 1999. Pressurized liquid extraction of medicinal plants. J Chromatogr A 837: 211 219. Bicchi C, Bellardo F and Rubbiolo P. 1992. Estrazione di alcaloidi da specie di Senecio. Laboratorio 2000 6: 3638. Brachet A, Christen P and Veuthey J-L. 2002. Focused microwave-assisted extraction of cocaine and benzoy- lecgonine from coca leaves. Phytochem Anal (in press). Brachet A, Rudaz S, Mateus L, Christen P and Veuthey J-L. 2001. Optimisation of accelerated solvent extraction of cocaine and benzoylecgonine from coca leaves. J Separation Sci 24: 865873. Budzinski H, Baumard P, Papineau A, Wise S and Garrigues P. 1996. Focused microwave assisted extraction of poly- cyclic aromatic compounds from standard reference materials, sediments and biological tissues. Polycyclic Arom Comp 9: 225232. Camel V. 2000. Microwave-assisted solvent extraction of environmental samples. Trends Anal Chem 19: 229248. Camel V. 2001. Recent extraction techniques for solid matrices-supercritical uid extraction, pressurized uid extraction and microwave-assisted extraction: their po- tential and pitfalls. Analyst 126: 11821193. Camel V and Bermond A. 1999. Un gain de temps notable pour l'analyse chimique: l'utilisation des micro-ondes dans le traitement de l'e chantillon. Ann Fals Exp Chim 92: 117135. Carro N, Garca CM and Cela R. 1997. Microwave-assisted extraction of monoterpenols in must samples. Analyst 122: 325329. Castioni P, Christen P and Veuthey J-L. 1995. L'extraction en phase supercritique des substances d'origine ve ge tale. Analusis 23: 95106. Chen SS and Spiro M. 1994. Study of microwave extraction of essential oil constituents from plant materials. J Microw Electromagn Energy 29: 231241. 112 R. KAUFMANN AND P. CHRISTEN Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002) Collin GJ, Lord D, Allaire J and Gagnon D. 1991. Huiles essentielles et extraits micro-ondes. Parfums Cosme t Aro mes 97: 105112. Craveiro AA, Matos FJA, Alencar JW and Plumel MM. 1989. Microwave oven extraction of an essential oil. Flavour Fragr J 4: 4344. Csiktusnadi Kiss GA, Forgacs E, Cserhati T, Mota T, Morais H and Ramos A. 2000. Optimisation of the microwave- assisted extraction of pigments from paprika (Capsicum annuum L.) powders. J Chromatogr A 889: 4149. Da Costa CT, Margolis SA, Benner BAJ and Horton D. 1999. Comparison of methods for extraction of avanones and xanthones from the root bark of the Osage orange tree using liquid chromatography. J Chromatogr A 831: 167 178. Dean JR, Fitzpatrick L and Heslop C. 1999. Microwave- assisted solvent extraction in organic analysis. Extr Meth Org Anal 166173. Demesmay C and Olle M. 1993. Utilisation des micro-ondes dans les laboratoires d'analyse. Spectra Analyse 175: 27 32. Ezzel JL, Richter BE, Felix WD, Black SR and Meikle JE. 1995. A comparison of accelerated solvent extraction with conventional solvent extraction for organophosphorus pesticides and herbicides. LC-GC 13: 390398. Fisher JA, Scarlett MJ and Stott AD. 1997. Accelerated solvent extraction: An evaluation for screening of soils for selected US EPA semi-volatile organic priority pollutants. Environ Sci Technol 31: 11201127. Ganzler K and Salgo A. 1987. Microwave-extractiona new method superseding traditional Soxhlet extraction. Z Lebensm Unters Forsch 184: 274276. Ganzler K, Salgo Aand Valko K. 1986a. Microwave extraction. Anovel sample preparation method for chromatography. J Chromatogr 371: 299306. Ganzler K, Ba ti J and Valko K. 1986b. A new method for the extraction and high-performance liquid chromatographic determination of vicine and convicine in faba beans. Chromatography 84: 435442. Ganzler K, Szinai I and Salgo A. 1990. Effective sample preparation method for extracting biologically active compounds from different matrixes by a microwave technique. J Chromatogr 520: 257262. Ho er F, Ezzell J and Richter B. 1995a. Beschleunigte Lo semittel-extraktion (ASE). Labor Praxis 19: 6267. Ho er F, Ezzell J and Richter B. 1995b. Anwendungen der ASE in der Umweltanalytik. Labor Praxis 19: 5862. Incorvia Mattina MJ, Ianucci Berger WA and Denson CL. 1997. Microwave-assisted extraction of taxanes from Taxus biomass. J Agric Food Chem 45: 46914696. Jarvis AP and Morgan ED. 1997. Isolation of plant products by supercritical-uid extraction. PhytochemAnal 8: 217222. Jassie L, Revesz R, Kierstead T, Hasty E and Matz S. 1997. Microwave-assisted solvent extraction. In Microwave- Enhanced Chemistry. Fundamentals, Sample Prepara- tion, and Applications, Kingston HMS, Haswell SJ. (eds). American Chemical Society: Washington, DC; 569609. Jensen D, Ho er F, Ezzell J and Richter B. 1996. Rapid preparation of environmental samples by accelerated solvent extraction (ASE). Polycyclic Arom Comp 9: 233 240. Kaufmann B, Christen P and Veuthey J-L. 2001a. Parameters affecting microwave- assisted extraction of withanolides. Phytochem Anal 12: 327331. Kaufmann B, Christen P and Veuthey J-L. 2001b. Study of factors inuencing pressurised solvent extraction of polar steroids from plant material. Chromatographia 54: 394398. Kawamura F, Kikuchi Y, Ohira T and Yatagai M. 1999. Accelerated solvent extraction of paclitaxel and related compounds from the bark of Taxus cuspidata. J Nat Prod 62: 244247. King MB and Bott TR. 1993. Extraction of Natural Products using Near-Critical Solvents. Blackie: London. Kreisselmeier A and Du rbeck HW. 1997. Determination of alkylphenols, alkylphenolethoxylates and linear alkylben- zenesulfonates in sediments by accelerated solvent extraction and supercritical uid extraction. J Chroma- togr A 775: 187196. Lee ML and Markides KE. 1990. Analytical Supercritical Fluid Chromatography and Extraction. Chromatography Con- ferences: Provo. Letellier M, Budzinski H and Garrigues P. 1999. Focused microwave-assisted extraction of polycyclic aromatic hydrocarbons. LC-GC Int 12: 222225. Levy JM. 1998. Sample preparation options: Supercritical uid extraction and accelerated solvent extraction. Am Lab 30: 38F. Li K, Landriault M, Fingas M and Llompart MP. 1998. Pressurised solvent extraction of environmental organic compounds in soils using a supercritical uid extractor. Analusis 26: 365369. Majors RE. 1998. Sample preparation perspectives. New approaches to sample preparation. LC-GC Int 8: 128133. Morf S, Debrunner B, Meier B and Kurth H. 1998. Auto- matische Probenvorbereitung von Panzlichen Arznei- mittel. Labor Praxis 22: 5662. Onuska FI and Terry KA. 1993. Extraction of pesticides from sediments using a microwave technique. Chromatogra- phia 36: 191194. Onuska FI and Terry KA. 1995. Microwave extraction in analytical chemistry of pollutants: polychlorinated biphe- nyls. J High-Resol Chromatogr 18: 417421. Pare JRJ. 1990. Microwave extraction of volatile oils and apparatus therefore. Patent no. 90250286.3 (0 485 668 A1). Pare JRJ. 1991. Microwave-assisted natural products extrac- tion. Patent no. 519,588 (5,002,784). Pare JRJ. 1994. Microwave extraction of volatile oils. Patent no. 29,358 (5,338,557). Pare JRJ, Be langer JMR and Stafford SS. 1994. Microwave- assisted process (MAP); a new tool for the analytical laboratory. Trends Anal Chem 13: 176184. Reighard TS and Olesik SV. 1996. Bridging the gap between supercritical uid extraction and liquid extraction tech- niques: alternative approaches to the extraction of solid and liquid environmental matrixes. Crit Rev Anal Chem 26: 6199. Renoe BW. 1994. Microwave assisted extraction. Am Lab 26: 3440. Richter BE. 1999. The extraction of analytes from solid samples using accelerated solvent extraction. LC-GC 17: S22S28. Richter BE, Ezzel JL, Felix D, Roberts KA and Later DW. 1995. An accelerated solvent extraction system for the rapid preparation of environmental organic compounds in soil. Am Lab 27: 2428. Richter BE, Jones BA, Ezzel JL, Porter NL, Avdalovic N and Pohl C. 1996. Accelerated solvent extraction: a technique for sample preparation. Anal Chem 68: 10331039. Romele L and Polesello S. 1997. How to minimize the use of solvents. Laboratorio 2000 11: 102110. Sinquin A, Go rner T and Dellacherie E. 1993. L'utilisation des micro-ondes en chimie analytique. Analusis 21: 110. Smith RM. 1995. Supercritical uid extraction of natural products. LC-GC 13: 930939. Smith RM. 1996. Supercritical uid extraction of natural products. LC-GC Int 9: 815. Thue ry J. 1992. Microwaves and matter. In Microwaves: Industrial, Scientic and Medical Applications, Part 1, Grant EH (ed.). Artech House: London; 83125. Westwood SA. 1993. Supercritical Fluid Extraction and its Use in Chromatographic Sample Preparation. Blackie: London. Young JC. 1995. Microwave-assisted extraction of the fungal metabolite ergosterol and total fatty acids. J Agric Food Chem 43: 29042910. MICROWAVE-ASSISTED EXTRACTION 113 Copyright # 2002 John Wiley & Sons, Ltd. Phytochem. Anal. 13: 105113 (2002)