Journal of Chromatography A, 1112 (2006) 121–126

An improved microwave Clevenger apparatus fordistillation of essential oils from orange peel

Mohamed A. Ferhat a, Brahim Y. Meklati a, Jacqueline Smadja b, Farid Chemat b,∗a Centre de Recherches en Analyses Physico – Chimiques CRAPC, BP 248 Alger RP 16004, Alger, Algeria

b Laboratoire de Chimie des Substances Naturelles et des Sciences des Aliments Faculte des Sciences et Technologies, Universite de La Reunion, B.P. 7151,F-97715 Saint Denis messag cedex 9, La Reunion D.O.M., France

Available online 27 December 2005

Abstract

Microwave Clevenger or microwave accelerated distillation (MAD) is a combination of microwave heating and distillation, performed atatmospheric pressure without added any solvent or water. Isolation and concentration of volatile compounds are performed by a single stage.MAD extraction of orange essential oil was studied using fresh orange peel from Valencia late cultivar oranges as the raw material. MAD has beencompared with a conventional technique, which used a Clevenger apparatus with hydro-distillation (HD). MAD and HD were compared in termof extraction time, yields, chemical composition and quality of the essential oil, efficiency and costs of the process. Extraction of essential oilsfp(©

K

1

Mdtvtwtspc

dmesm

0d

rom orange peels with MAD was better in terms of energy saving, extraction time (30 min versus 3 h), oxygenated fraction (11.7% versus 7.9%),roduct yield (0.42% versus 0.39%) and product quality. Orange peels treated by MAD and HD were observed by scanning electronic microscopySEM). Micrographs provide evidence of more rapid opening of essential oil glands treated by MAD, in contrast to conventional hydro-distillation.

2005 Elsevier B.V. All rights reserved.

eywords: Clevenger; Microwave; Extraction; Hydro-distillation; Essential oil; Orange peel

. Introduction

Oranges are grown throughout the Mitidja region of Algeria.ore than 500 Mkg were harvested in 2001 [1]. Such high pro-

uction inevitably saturates the fresh fruit market. An alternativeo consumption of orange as fresh fruit is processing it to otheralue-added products. Orange processing yields an increase inhe production of Citrus by-products, the most important ofhich is the essential oil extracted from the peel. Orange essen-

ial oil is used to add orange aroma from the orange to productsuch as carbonated drinks, ice creams, cakes, air-fresheners, anderfumes. It is also used for its germicidal, antioxidant, and anti-arcinogenic properties [2].

Citrus oil is present in oil sacs or oil glands located at differentepths in the peel of the fruit. Recovery of the oil is performedostly by cold pressing these peels. However, simultaneous

xtraction of juice and oil is considered by many fruit proces-ors to be economically attractive [3]. None of the mechanical oranual processes used to extract essential oils from fruit or peel

is able to recover the total quantity of these important deriva-tives. It is in any case essential to have a method to determine thetotal essential oil content of the fruit. Such information is funda-mental to decision on the quality of the fruit and the efficiencyof the machinery employed [4].

The Clevenger apparatus [5] based on distillation is the bestmethod to determine the essential oil content of the fruit. Ingeneral, an analytical procedure for essential oil from orangefruits or peels comprises two steps: extraction (hydro-distillationor steam distillation) and analysis (gas chromatography (GC),gas chromatography coupled to mass spectrometry (GC–MS)).Although the second step is requires only 15–30 min, extractiontakes several hours. Extraction is frequently done by a prolongedheating and stirring in boiling water.

Microwave energy, with a frequency of 2.45 GHz, is wellknown to have a significant effect on the rate of various processesin the chemical and food industry. Much attention has been givento the application of microwave dielectric heating in analyti-cal chemistry because of the reduced analysis time, simplifiedmanipulation and work-up, and higher purity of the final product[6–8]. Several classes of compounds such as essential oils, aro-

∗ Corresponding author. Tel.: +262262938182; fax: +262262938183.E-mail address: chemat@univ-reunion.fr (F. Chemat).

mas, pesticides, phenols, dioxins, and other organic compoundshave been extracted efficiently from a variety of matrices (mainly

021-9673/$ – see front matter © 2005 Elsevier B.V. All rights reserved.oi:10.1016/j.chroma.2005.12.030

122 M.A. Ferhat et al. / J. Chromatogr. A 1112 (2006) 121–126

Fig. 1. Microwave Clevenger.

soils, sediments, animal tissues, food, and plant materials). Allthe reported applications have shown that microwave-assistedsolvent-extraction (MAE) is a viable alternative to conventionaltechniques for such matrixes. The main benefits are the reductionof extraction time and solvent used.

Up to now, however, there are only a few articles in the liter-ature that have reported the acceleration of essential oil extrac-tion by microwave irradiation [9–13]. The advantages of usingmicrowave energy, a non contact heat source, for the extrac-tion of essential oils from plant materials, could include: moreeffective heating, faster energy transfer, reduced thermal gradi-ents, selective heating, reduced equipment size, faster responseto process heating control, faster start-up, increased production,and elimination of process steps [14].

The aim of this work was to present an improved microwaveClevenger distillation for the extraction of essential oils fromorange peels (Citrus sinensis L. Osbeck, Rutaceae), and comparethe results with those obtained by the conventional technique, inorder to introduce this advantageous alternative in the analysisof essential oils in the agro-food industry.

2. Experimental

2.1. Plants material

Ofl4bifi

2

uT

mum delivered power of 1000 W variable in 10 W increments.Temperature was monitored by an external Infrared (IR) sensor.

Based on a relatively simple principle, this method involvesplacing orange peels in the microwave reactor, without anyadded solvent or water. The internal heating of the in situ waterwithin the orange peels distends the oil glands and sacs and leadsto rupture of the glands and oleiferous receptacles. This processthus frees essential oil, which is evaporated by the in situ waterof the plant material. A cooling system outside the microwaveoven condenses the distillate continuously. Excess water wasrefluxed to the extraction vessel in order to restore the water tothe plant material. The MAD is neither a modified microwaveassisted extraction (MAE), which use organic solvents, nor amodified hydro-distillation (HD) which use a large quantity ofwater.

In a typical MAD procedure performed at atmospheric pres-sure, 200 g of fresh orange peels were heated using a fix powerdensity of 1 W g−1 for 30 min without addition of solvents orwater. The extraction was continued at 100 ◦C until no moreessential oil was obtained. The essential oil was collected, driedover anhydrous sodium sulphate and stored at 4 ◦C until used.Extractions were performed at least three times, and the meanvalues were reported.

2.3. Conventional Clevenger or Hydro-distillationapparatus and procedure

dtfost

2

gc(ppiF

2

cppc(wfli

The oranges used in this study were Citrus sinensis L.sbeck from Valencia late cultivar. All fruits were gathered

rom the same experimental plantation Institut Technique de’Arboriculture Fruitiere (ITAF), located in the Mitidja region0 km south of Algiers (Boufarik, Algeria). Oranges were peeledy hand, separating the external part of the orange (flavedo), giv-ng a yield of 20% (w/w) of orange peel with respect to the wholeruit. Fresh plant material was employed in all extractions. Thenitial moisture content of orange peel was 90%.

.2. Microwave Clevenger apparatus and procedure

Microwave accelerated distillation MAD was performedsing the “DryDist” microwave oven illustrated in Fig. 1 [15].his is a multimode microwave reactor 2.45 GHz with a maxi-

200 g of fresh orange peels were submitted to hydro-istillation with a Clevenger-type apparatus [5,16] accordingo the European Pharmacopoeia, and extracted with 2 l of wateror 3 h (until no more essential oil was obtained). The essentialil was collected, dried under anhydrous sodium sulphate andtored at 4 ◦C until used. Extractions were performed at leasthree times, and the mean values were reported.

.4. Gas chromatography

A Hewlett-Packard 6890 GC system was used foras chromatography analysis, fitted with a fused-silica-apillary column with an apolar stationary phase HP5MSTM

30 m × 0.25 mm × 0.25 �m film thickness). The column tem-erature progress from 60 to 280 ◦C at 2 ◦C min−1. Injection waserformed at 250 ◦C in the splitless mode; 1 �L of sample wasnjected. A flow rate of 0.3 ml/min carrier gas (N2) was used.lame ionisation detection was performed at 320 ◦C.

.5. Gas chromatography–mass spectrometry identification

The essential oils were analysed by gas chromatographyoupled to mass spectrometry (GC–MS) (Hewlett-Packard com-uterized system comprising a 6890 gas chromatograph cou-led to a 5973A mass spectrometer) using a fused-silica-apillary column with an apolar stationary phase HP5MSTM

30 m × 0.25 mm × 0.25 �m film thickness). GC–MS spectraere obtained using the following conditions: carrier gas He;ow rate 0.7 mL min−1; split 1:20; injection volume 0.1 �L;

njection temperature 250 ◦C; oven temperature progress from

M.A. Ferhat et al. / J. Chromatogr. A 1112 (2006) 121–126 123

60 to 280 ◦C at 2 ◦C min−1; the ionisation mode used was elec-tronic impact at 70 eV.

2.6. Qualitative and quantitative analyses

Most constituents were tentatively identified by comparisonof their GC Kovats retention indices (RI), determined with refer-ence to a homologous series of C5–C28 n-alkanes and with thoseof authentic standards available in the authors’ laboratory. Iden-tification was confirmed when possible by comparison of theirmass spectral fragmentation patterns with those stored in the MSdatabase (National Institute of Standards and Technology andWiley libraries) and with mass spectra literature data [17,18].Component relative concentrations were obtained directly fromGC peak areas obtained with GC-FID.

2.7. Olfactory evaluation

The two essential oils were evaluated olfactorically by JeanClaude ELLENA (Perfurmer from Symrise, France).

2.8. Physical constants

Orange essential oils have been analysed according to thestandard method AFNOR. The usual physical constants definingtr[

2

htvs

3

3

tFtstoef

w1mta

Fig. 2. Temperature profiles (� MAD � HD) and yields (© MAD � HD) as afunction of time for the MAD and HD extractions of essential oil from orangepeels.

essential oil per 100 g of orange fruit. These results mean a sub-stantial saving of time, energy and plant material.

3.2. Composition of essential oil

The comparison of yields, extraction times, oxygenated frac-tion, composition of chemical families, and detailed compositionfor each extract is reported in Table 1. Limonene, �-myrcene,linalool, �-sisensal and decanal were the main components in theessential oil extracted from orange peels but the relative amountsdiffered for the two extraction methods. Limonene, a monoter-pene hydrocarbon, is the most abundant component present at76.7% and 78.5%, respectively for MAD and HD. Linalool, anoxygenated monoterpene, is present at 3.1% and 2.0%, respec-tively for MAD and HD.

Substantially higher amounts of oxygenated compounds andlower amounts of monoterpenes hydrocarbons are present inthe essential oil of orange peels extracted by MAD in compari-son with HD. The oxygenated fraction in essential oil samplesfrom MAD (11.7%) was higher than HD (7.9%). Monoterpeneshydrocarbons are less valuable than oxygenated compounds interms of their contribution to the fragrance of the essential oil.Conversely, the oxygenated compounds are highly odoriferousand, hence, the most valuable. The greater proportion of oxy-genated compounds in the MAD essential oils is probably dueto the diminution of thermal and hydrolytic effects, comparedwiaa

3

sssi

he essential oil have been determined at 20 ◦C: specific gravity,efractive index, optical rotation, and solubility in 95% ethanol19].

.9. Scanning electron micrographs (SEM)

The specimens were freeze-dried, fixed on the specimenolder with aluminium tape and then sputtered with gold. Allhe specimens were examined by a TOPCON ABT60, underacuum condition and accelerating voltage of 15 kV, with a spotize 5 and a working distance of 15 mm.

. Results and discussion

.1. Extraction yield and time

MAD is clearly quicker than conventional HD. The extrac-ion takes 30 min, whilst 3 h were required by hydro-distillation.or HD or MAD, the extraction temperature is equal to boiling

emperature of water at atmospheric pressure (100 ◦C). Fig. 2hows the temperature profiles during MAD and HD extrac-ions. To reach this extraction temperature (100 ◦C) and thusbtain the distillation of the first essential oil droplet, it is nec-ssary to heat for only 3 min with MAD compared with 30 minor HD.

As is shown in Table 1 and Fig. 2, an extraction time of 30 minith MAD provides yields comparable to those obtained after80 min by means of HD, which is the one of the referenceethods in essential oil extraction. The ultimate yield of essen-

ial oil obtained from orange peels was 0.42 ± 0.02% by MADnd 0.39 ± 0.02% by HD. Yields are expressed as in grams of

ith hydro-distillation, which uses a large quantity of water ands time and energy consuming. Water is a polar solvent, whichccelerates many reactions, especially reactions via carbocations intermediates.

.3. Physical constants and olfactory evaluation

Orange essential oils have been analysed according to thetandard method AFNOR to determine the usual physical con-tants defining the essential oil extracted either by MAD or HD:pecific gravity, refractive index, optical rotation, and solubilityn 95% ethanol at 20 ◦C (Table 2). There is no significant differ-

124 M.A. Ferhat et al. / J. Chromatogr. A 1112 (2006) 121–126

Table 1Chemical compositions, grouped compounds, oxygenated fractions, and yields of essential oils obtained by MAD and HD extractions from fresh orange peels

No. Compounds RI MAD (%) HD (%) Method of identification

Monoterpene hydrocarbons 85.7 89.81 �-Pinene 918 0.5 1.6 GC, GC–MS2 Sabinene 961 1.2 1.2 GC, GC–MS3 �-Pinene 983 2.4 2.7 GC, GC–MS4 �-Myrcene 1003 4.3 5.3 GC, GC–MS5 Limonene 1075 76.7 78.5 GC, GC–MS6 �3-Carene 1098 0.2 0.1 GC, GC–MS7 �-Terpinene 1099 0.3 0.2 GC, GC–MS8 Terpinolene 1110 0.2 0.2 GC, GC–MS

Oxygenated monoterpenes 7.0 4.89 Linalool 1132 3.1 2.0 GC, GC–MS10 trans-Limonene oxide 1144 0.3 0.1 GC, GC–MS11 Citronellal 1159 0.4 0.3 GC, GC–MS12 Terpin-4-ol 1178 0.3 0.5 GC, GC–MS13 �-Terpineol 1192 0.8 0.4 GC, GC–MS14 trans-Carveol 1217 0.2 0.3 GC, GC–MS15 Nerol 1228 0.6 0.3 GC, GC–MS16 Neral 1239 0.4 0.3 GC, GC–MS17 Geraniol 1254 0.3 0.1 GC, GC–MS18 Geranial 1268 0.5 0.5 GC, GC–MS19 Perilla Alcohol 1284 0.1 0.0 GC, GC–MS

Sesquiterpene hydrocarbons 1.0 1.320 �-Copaene 1357 0.1 0.1 GC, GC–MS21 �-Cubebene 1371 0.1 0.1 GC, GC–MS22 �-Elemene 1373 0.1 0.1 GC, GC–MS23 (E)-Caryophyllene 1391 0.2 0.2 GC, GC–MS24 �-Gurjunene 1406 0.1 0.1 GC, GC–MS25 (E)-�-Farnesene 1442 0.1 0.1 GC, GC–MS26 Germacrene D 1465 0.1 0.1 GC, GC–MS27 Valencene 1479 0.2 0.4 GC, GC–MS28 �-Cadinene 1513 0.1 0.2 GC, GC–MS

Oxygenated sesquiterpenes 0.9 0.329 �-Sinensal 1647 0.2 0.2 GC, GC–MS30 �-Sinensal 1675 0.7 0.1 GC, GC–MS

Other oxygenated compounds 4.0 2.731 1-Hexanol 852 0.2 0.0 GC, GC–MS32 1-Octanol 1102 1.1 0.5 GC, GC–MS33 n-Nonanal 1118 0.2 0.1 GC, GC–MS34 Decanal 1210 1.9 1.7 GC, GC–MS35 Octanol Acetate 1212 0.1 0.1 GC, GC–MS36 1-Decanol 1266 0.3 0.0 GC, GC–MS37 Undecanal 1299 0.1 0.1 GC, GC–MS38 Dodecanal 1391 0.2 0.2 GC, GC–MS

Extraction time (min) 30 180Yield (%) 0.42 ± 0.02 0.39 ± 0.02Oxygenated fraction 11.7 7.9

ence between the physical constants of essential oils obtainedby MAD or HD.

The organoleptic properties of essential oils extracted byMAD and HD are shown in Table 3. According to Mr. J.C.Ellena, a “Nose” or perfumer, the MAD method offers the pos-sibility for a better reproduction of natural aroma of the orangeessential oil than the hydro-distilled essential oil.

3.4. Structural changes after extraction

The various extraction methods produced distinguishablephysical changes in the orange peels. Fig. 3a is a micrograph ofthe untreated peels, which can be compared with structures of the

treated orange peels in Fig. 3b (MAD) and Fig. 3c (HD). Fig. 3bshows the typical structure after MAD extraction; cells are emptybut still intact. In the case of HD extraction, we observed signif-icant damage on the external surface of the orange peel togetherwith some dispersed cellular material. This indicates that themechanical strain induced by the rapid decompression and theviolent vaporization of water have two main effects: the dehy-drating effect due to vaporization and a subsequent change inthe surface tension of the glandular wall, causing it to crumbleor rupture more readily (Fig. 3).

Similar effects were pointed out by Pare and Belanger [20],and Chen and Spiro [21] for the microwave extraction of rose-mary leaves in hexane. When the glands were subjected to more

M.A. Ferhat et al. / J. Chromatogr. A 1112 (2006) 121–126 125

Table 2Physical properties of essential oils obtained by MAD and HD extraction fromfresh orange peels

Physical properties MAD HD

Specific gravity d2020 0.86 0.86

Refractive index n20D +1.475 +1.477

Optical rotation in degree [α]20D 38 39

Solubility (v/v) in 95% ethanol 0.4 0.4

Table 3Organoleptic properties of essential oils obtained by MAD and HD extractionfrom fresh orange peels

Organolepticproperties

MAD HD

Colour Colourless Pale yellowOdour Fresh, light and

sweet citrusyodour

Fresh, pungent butdifferent from freshfruit

Aspect Mobile liquid Mobile liquid

severe thermal stresses and localized high pressures, as in thecase of microwave heating, the pressure build-up within theglands could have exceeded their capacity for expansion, andcaused their rupture more rapidly than in conventional extrac-tion.

3.5. Cost, cleanliness and up-scaling

The reduced cost of extraction is clearly advantageous forthe proposed MAD method in terms of time and energy. Hydro-distillation required an extraction time of 30 min for heating 2 lof water and 200 g of orange peels to the extraction temperature,

Table 4Energy consumption of MAD and HD methods

MAD HD

Extraction time (min) 30 180Electric consumption (kWh) 0.25 4.33CO2 rejected (g) 200 3464

followed by evaporation of water and essential oil for 150 min.The MASD method required heating for 3 min only and evap-oration for 27 min of the in situ water and essential oil of theorange peels. The energy required to perform the two extractionmethods are respectively 4.33 kWh for HD, and 0.25 kWh forMAD (Table 4). The power consumption has been determinedwith a Wattmeter at the microwave generator entrance and theelectrical heater power supply.

Regarding environmental impact, the calculated quantity ofcarbon dioxide rejected in the atmosphere is higher in the case ofHD (3464 g CO2/g of essential oil) than for MASD (200 g CO2/gof essential oil). These calculations have been made accordingto literature: to obtain 1 kWh from coal or fuel, 800 g of CO2will be rejected in the atmosphere during combustion of fossilfuel [22].

In this study, we present MAD as an “environmentallyfriendly” extraction method suitable for sample preparation priorto essential oil analysis. MAD is a very clean method, whichavoids the use of large quantity of water and voluminous extrac-tion vessels (HD). MAD could also be used to produce largerquantities of essential oils by using existing big scale microwaveextraction reactors [23]. These microwave reactors are suitablefor the extraction of 10, 20 or 100 kg of fresh plant materialper time. Theses reactors could be easily modified and used forMAD extractions.

ter MA

Fig. 3. Electron micrograph of orange peels: untreated (a); af D extraction (30 min) (b); and after HD extraction (3 h) (c).

126 M.A. Ferhat et al. / J. Chromatogr. A 1112 (2006) 121–126

3.6. Safety considerations

Microwave extraction process is simple and can be readilyunderstood in terms of the operating steps to be performed.However, the application of microwave energy can pose seri-ous hazards in inexperienced hands. A high level of safety andattention to details when planning and performing experimentsmust be used by all the persons dealing with microwaves. Theyhave to ensure that they seek proper information from knowl-edgeable sources and that they do not attempt to implement thistechnique unless proper guidance is provided. Only approvedequipment and scientifically sound procedures should be used.

4. Conclusion

Microwave Clevenger or microwave accelerated distillation(MAD) technique has been compared with the conventionalhydro-distillation method, for the extraction of essential oilfrom fresh orange peels. This microwave extraction methodoffers important advantages over traditional hydro-distillation,namely; water and solvent free process, shorter extraction times(30 min against 3 h for hydro-distillation); better yields (0.42%against 0.39% for HD); higher oxygenated compounds; environ-mentally friendly; lower production of by-products (as no wateror solvent is used); lower cost; and the possibility for a betterreproduction of natural aroma of the orange essential oil thantubci

A

tS

References

[1] A. Ferhat, L’agrumiculture Algerienne, Internal Report, CRAPC,Algiers, Algeria, 2004.

[2] E. Guenther, The Essential Oils, 3, R.E. Kreiger Publishing, New York,1974.

[3] M. Mukhopadhyay, Natural Extracts Using Supercritical Carbon Diox-ide, CRC Press, New York, 2000.

[4] G. Dugo, A. Di Giacomo, The Genus Citrus, Taylor & Francis Publish-ing, London, 2002.

[5] J.F. Clevenger, Am. Perfumer Essential Oil Rev. (1928) 467.[6] H.M. Kingston, L.B. Jassie, Introduction to Microwave Sample Prepa-

ration, American Chemical Society, Washington, 1988.[7] H.M. Kingston, S.J. Haswell, Microwave Enhanced Chemistry: Fun-

damentals, Sample Preparation and Applications, American ChemicalSociety, Washington, 1997.

[8] A. Zlotorzynski, Crit. Rev. Anal. Chem. 25 (1995) 43.[9] M.D. Luque de Castro, M. Jiminez Carmona, V. Fernandez-Prez, Trends

Anal. Chem. 18 (1999) 708.[10] A.A. Craveiro, F.J.A. Matos, J.W. Alencar, Flavour Frag. J. 4 (1989) 43.[11] J.R.J. Pare, European Patent, EP 0,485,668, 1992.[12] P. Mengal, B. Mompon, European Patent, EP 2,705,035, 1993.[13] F. Chemat, J. Smadja, M.E. Lucchesi, United Stated Patent, US

0,187,340, 2004.[14] A.C. Metaxas, R.J. Meredith (Eds.), Industrial Microwave Heating, Pere-

grinus Ltd., London, 1993.[15] Milestone, Bergamo, Italy, www.miestonesrl.com.[16] Conseil de l’Europe, Pharmacopee Europeenne 1, Maisonneuve S.A.

Editions, Sainte Ruffine, 1996.[17] R.P. Adams, Identification of Essential Oil Components By Gaz Chro-

[

[

[

[[[

he hydro-distilled essential oil. SEM images of orange peelsntreated or subjected to MAD or HD emphasize the differenceetween the two extraction methods used. Microwaves seem toause the rupture of the cells and the glands more rapidly thann conventional hydro-distillation.

cknowledgements

The authors gratefully acknowledge Mme Esme Spicer fromhe University of Stellenbosch for SEM micrographs, and Prteven Bradshaw for his valuable comments.

matography/Mass Spectroscopy, Allured Publishing Corporation, CarolStream, 1995.

18] S. Arctander, Perfume and Flavor Chemicals, Allured Publishing Cor-poration, Carol Stream, 1994.

19] Recueil des normes francaises, Huiles essentielles, Editions AFNOR,Paris, 2000.

20] J.R.J. Pare, J.M.R. Belanger, Instrumental Methods in Food Analysis,Elsevier, Amsterdam, 1997.

21] S.S. Chen, M. Spiro, Flavour Frag. J. 10 (1995) 101.22] J. Bernard, Sciences et vie 214 (2001) 68.23] Archimex, Quimper, France, www.archimex.com.