Characterization of Portuguese-Grown Geranium Oil (Pelargonium sp.)
Gomes, Paula B
Abstract
The chemical composition of the Portuguese-grown geranium essential oil (Pelargonium hybrid) cultivated in northern Portugal, which also grows wild, was examined. It was found that the geranium oil herein obtained had characteristics intermediate between Bourbon and China types, with good olfactory quality and presenting characteristics different from those previously studied. The essential oils of fresh/dried and green/yellow geranium leaves were compared. Solvent extraction methods were also employed and the composition of the respective absolutes was determined and compared with the essential oil. It was concluded that using air-dried plant and yellow leaves improved the quality of geranium oil. Moreover, the absolute obtained from a diethyl ether extract had the best organoleptic properties of the extracts studied.
Key Word Index
Geranium, Pelargonium sp., Geraniaceae, essential oil composition, extract composition, citronellyl formate, citronellol, geraniol.
Introduction
Geranium oil is widely used as a floral substitute of the rose scent, being therefore one of the most valuable natural materials for the perfumery and cosmetic industries (1). Geranium oil is obtained from a hybrid between different species of Pelargonium. Though originated from South Africa, it is reported to grow wildly in Reunion Island, Morocco and Algeria (2). This aromatic plant is cultivated extensively in Reunion Island (Bourbon-type) and Madagascar, Egypt (North African-type) and China, and more recently in India. The Bourbon-type presents the best quality, being therefore the most expensive and the most used in the composition of fine perfumes. The worldwide production is few hundred tons per year (1).
Geranium oil is employed as such or as a further refined material, which then demands advanced technical knowledge of process engineering (2). The refinement is accomplished by means of a fractionation process plant to extract fragrant isolates like commercial rhodinol (mixture of linalool, citronellol, geraniol), or to obtain a high-grade essential oil with the removal of light fractions of terpenes, known as terpeneless oils.
Organoleptic considerations: The oil is isolated from the green foliage and sometimes also from the flowers (very fragrant), by hydrodistillation or steam distillation, yielding an oil with a greenish-yellow color and a rose-mint odor. It is interesting to note that the odor of the plant changes from lemon-like to rose-like when the leaves turn from green to yellow (2).
The main constituents of the geranium oil are citronellol (sweet rose-like odor) andgeraniol (flowery-rose-like odor), which occur in different proportions according to the origin of the oil (2). Both Bourbon and North African-types contain unusual high quantities of (-)-citronellol, isomenthone and monoterpene formates. However they can be distinguished by the presence of different constituents such as guaia-6,9-diene in Bourbon oil and 10-epi-[gamma]-eudesmol in the African-type. The Chinese oil is similar to Bourbon-type, having higher content of citronellol (+40%) and lower content of Iinalool and geraniol (1,3).
State of the art: The rose geranium plant grows wildly in southern Portugal, although it is not native from Portugal. In fact, good olfactory quality geranium oil had already been produced on a small scale in Portugal and exported to England (4). Several reviews on geranium oil composition studies were published by Lawrence (5-12).
Guerere et al. (13) compared different extraction methods, namely hydrodistillation both in copper still and in laboratory glass apparatus and solvent extraction with hexane. They found that the Iinalool contents in the oil were considerable higher than in the concrete, while the geraniol concentration was lower. To justify this fact, the authors proposed a reaction mechanism of transformation of geraniol in Iinalool through water vapor action, which is catalyzed by copper ions Cu^sup 2+^.
Van der Walt and Demarne (14) examined the morphology and oil composition of two close geranium species, Pelargonium graveolens and P. radens, and their hybrids. Their composition was very similar except for minor components, such as 2-phenylethyl tiglate that occurred only in P. graveolens, while [beta]-bourbonene and guaia-6,9-diene were found just in P. radens and the hybrids contained the three components. The same authors also published a work on the comparison of Pelargonium species to find the origin of the cultivar cv. Rosé grown in Réunion Island (15).
Seabra and Carmo (4) presented a study on the cultivation of geranium in Portugal. They found that Portuguese geranium oil was intermediate between the Bourbon and African-types, containing both 10-epi-[gamma]-eudesmol (African) and guaia-6,9-diene (Bourbon), which imparted its special and appreciated olfactive characteristics. The content of citronellol was higher than commercially available geranium oils, whereas the geraniol and linalool contents were lower.
Machado et al. (16) employed headspace gas chromatography to characterize the composition of CO2 extracts and the oil obtained by steam distillation from a geranium plant also cultivated in Portugal. The authors found that the aroma of the CO2 extract more closely resembled the natural aroma and improved olfactory quality due to the higher content of geraniol and citronellol and lesser amount of linalool.
In the present paper, the composition of geranium oil obtained from Pelargonium hybrid cultivated in Portugal is reported, revealing different characteristics from those previously published by making use of several isolation methods and plant preparation.
Experimental
Plant material: The samples consisted of foliage of rose-geranium plant Pelargonium hybrid, using fresh/green, fresh/yellow, air-dried/green and oven-dried/green leaves. The samples were collected in October-November, after the flowering season, from plants cultivated in northern Portugal (humid climate). These plants were propagated by stem cuttings from a wild plant found in southern Portugal (dry climate). The crops were two years old.
Oil and extract isolation: The geranium oils were obtained by hydrodistillation of plant material (25-30 g) in a Clevenger-type glass apparatus for 2 h. Three kinds of extracts from solvent extraction with hexane, ethanol and diethyl ether were also obtained. The fresh plant material (30 g) was initially extracted with each solvent (250 mL) during 1 h. The extract was then filtered and concentrated by a rotary evaporator under reduced pressure at 50°C, 65°C and 20°C, respectively. The solid residue (the concrete) was extracted with ethanol and again filtered under reduced pressure, resulting finally in the absolute. Each isolation procedure was performed twice. The oils and absolutes were stored in glass vials and kept in a freezer prior to their analysis.
GC/MS: Gas chromatography (GC/FID) and GC/MS were carried out in parallel, using a Varian CP-3800 instrument equipped with two split/splitless injectors, two CP-Wax 52 CB bonded fused silica polar columns (50 m x 0.25 mm, 0.2 µm film thickness), a Varian FID detector and a Varian Saturn 2000 MS ion-trap mass spectrometer, both controlled by Saturn 2000 WS software. The oven temperature was programmed isothermal at 50°C for 5 min, then increased from 50°-200°C at 2°C/min and held isothermal for 40 min. The injectors were set at 240°C, with a split ratio of 1/50 for FID and 1/200 for MS. The FID detector was maintained at 250°C. The sample volume injected was 1 µL. The carrier gas was He N60, at a constant flow rate of 1 mL/min. There were three injections per sample and the sample GC vial was kept in the refrigerator between injections.
Component identification and quantification: The essential oil composition was expressed in percentage values calculated directly from GC peak areas and in a base without solvent. The areas were used without applying correction factors. The linear retention indices (LRI) were determined relative to the retention times obtained for a standard mixture containing a series of n-alkanes C8-C20 (Fluka, Spain).
The component identification was made by NIST-98 spectral library search, comparison with MS spectra of authentic reference compounds (own laboratory library) and with literature data. The identification was also confirmed by comparison of retention times with authentic reference compounds run under the same analytical conditions. The composition of each sample was determined as the average of the three GC injections.
Results and Discussion
The main components identified in the Portuguese-grown geranium oil (which constituted 76% of the total composition) are shown in Table I. Germacrene D was only tentatively identified based on NIST library search and on the fact that it has been already reported in geranium oil (5,14,15). The identification of guaia-6,9-diene was made by comparison with published MS spectra (17), presented in Figure 1.
Comparing the composition of the obtained geranium oil with the data published by Teisseire (3) for the three kinds of commercial geranium oils, it was concluded that the composition was intermediate between the Bourbon and Chinese oils, with the same content of guaia-6,9-diene and the absence of 10-epi-[gamma]-eudesmol, as can be seen in Figure 2. The citronellol/geraniol ratio was 3.3, which was a value closer to the Bourbon [1.2] than to China-type [8.3]. The main composition of our geranium oil was also very similar with the oil obtained from geranium grown in Réunion Island (cv. Ros ) described by Demarne and van der Walt (15), except for the lower concentration of geraniol.
When comparing our analysis with other Portuguese characterization studies, the composition approached that obtained by Seabra and Carmo (4) except for the absence of 10-epi-[gamma]-eudesmol characteristic of the African-type (Figure 3). Therefore, the present work reports the characterization of a new geranium oil.
The oil yield calculated from fresh plant material was 0.1-0.2%. The influence of preparation and condition of the plant on the composition of the oil is shown in Table II. The geranium oil obtained from yellow leaves presented a content of rose oxide and isomenthone (more volatile) higher than the oil obtained from green leaves, increasing their concentration 6 and 1.5 times, respectively. On the other hand, the content of geraniol and its esters and of linalool decreased 40-70%, while citronellol and its esters maintained their levels. The concentration of guaia-6,9-diene decreased significantly. The intensification of the rose-like odor that is reported when using yellow leaves is therefore due to the high content of rose oxide. Drying the leaves in an oven prior to extraction resulted in an oil that was poorer in all the considered compounds except for geranyl and 2-phenylethyl tiglates, the less volatile compounds. In fact, the contents of isomenthone decreased more than 80%, while the concentration of the tiglates doubled. On the other hand, the air-dried geranium oil was found to contain the highest content of Iinalool and geraniol, having though approximately the same amount of citronellol and esters as using fresh geranium leaves.
In conclusion, the quality of our geranium oil improved with air-drying because the ratio citronellol/geraniol decreased from 3.3 to 2.3, being closer to the Bourbon-type. When yellow leaves are included, the olfactory quality of the oil should also improve due to the higher content of rose oxide.
The composition of the hexane extract herein obtained was in agreement with the results reported by Guerere et al. (13), in particular regarding the trace amount of Iinalool and the higher contents of geraniol found in the extract.
The composition of the absolutes obtained from hexane and diethyl ether extractions were very similar, having linalool only in trace amounts, isomenthone and citronellyl formate in half of the quantity, 20% less of citronellol and a higher content of geraniol and tiglates when compared to the oil composition. They differed in the higher content of rose oxide in the absolute from diethyl ether extract, which supports the better odor quality imparted by this absolute. Ethanol extraction resulted in an opaque absolute, with a high level of non-volatiles. The higher percentage of linalool observed in this ethanol extract, associated to a decrease in geraniol content, resulted from the conversion of geraniol to linalool due to the use of a higher temperature and longer time to remove the solvent, during the extraction procedure, as explained by Guerere et al. (13).
From the organoleptic point of view, the geranium oil produced was transparent light-green colored, with an initial green odor that evolved to a sweet-floral aroma, slightly fresh. The absolutes were green-brownish, dark green and olive-green colored from hexane, ethanol and diethyl ether extracts, respectively. The odor was intensely green at first, evolving afterwards to a sweet-floral aroma. In the case of ethanol extract, the odor changed from an unpleasant to a somewhat balsamic odor. All the absolutes stained the smelling strip, which means that they contained non-volatiles. The ‘tea’ odor was noticed in the hydrodistilled oils.
This work on the characterization of the Portuguese-grown geranium, the volatiles of which isolated by the traditional methods of hydrodistillation and organic solvent extraction, will be the basis for the future work on this aromatic plant, namely the study of the supercritical fluid extraction.
Acknowledgments
We gratefully acknowledge the financial support of FCT Fundação para a Ciência e Tecnologia, project grant POCTI/EQU/ 39990/2001.
References
1. K. Bauer, D. Garbe and H. Surburg, Common Fragrance and Flavor Materials: Preparation, Properties and Uses. Wiley-VCH, 3rd ed., Weinheim (1997).
2. B.C. Gulati, S. P.S. Duhanand R.K. Thappa, Geranium. In: Cultivation and Utilisation of Medicinal and Aromatic Plants. Edits., C.K. Atal and B.M. Kapur, Regional Research Laboratory Jammu-Tawi, New Delhi (1977).
3. P. Teisseire, Industrial Quality Control of Essential oils by Capillary GC. In: Capillary Gas Chromatography in Essential Oil Analysis. Edits., C.B. Pat Sandra, Huethig, Heidelberg (1987).
4. R.M.C. Seabra and M.M. Carmo, Estudo da Cultura de Pelargonium graveolens L’Heritier ex Alton. 1as. Jornadas lb ricas de Plantas Medicinales y de Aceites Essenciales, Madrid, Spain (1989).
5. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 9(5), 87-95 (1984).
6. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 13(5), 61-71 (1988).
7. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 17(2), 39-52 (1992).
8. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 17(6), 51-60(1992).
9. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 19(1), 51-44 (1994).
10. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 21(2), 55-63 (1996).
11. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 24(1), 53-63(1999).
12. B.M. Lawrence, Progress in Essential Oils. Perfum. Flavor., 28(1), 60-81 (2003).
13. M. Guerere, F. Demarne, J.M. Mondon and A. Pajaniaye, Etuded’Huiles Essentielles de Géranium Bourbon Obtenues de Trois Façons Différentes. Ann. Fals. Exp. Chim., 78, 131-136 (1985).
14. J.J.A. van der Walt and F. Demarne, Pelargonium graveolens and P. radens: A comparison of their morphology and essential Oils. S. Af r. J. Bot., 54, 617-622 (1988).
15. F. Demarne and J.J.A. van der Walt, Origin of the rose-scented Pelargonium cultiver grown in Reunion Island. S. Afr. J. Bot., 55, 184-191 (1989).
16. A.S.R. Machado, E.G. Azevedo, R.M.A. Sardinha and M.N.d. Ponte, High pressure CO2 Extraction from Geranium Plants. J. Essent. Oil Res., 5, 185-189 (1993).
17. C.L. Steele, J. Crock, J. Bohlmann and R. Croteau, Sesquiterpene Syntheses from Grand Fir (Abies grandis). J. Biol. Chem., 273, 2078-2089 (1998).
Paula B. Gomes, Vera G. Mata* and A.E. Rodrigues
Laboratory of Separation and Reaction Engineering (LSRE), Department of Chemical Engineering, Faculty of Engineering of University of Porto, Rua Dr Roberta Frias, 4200-465 Porto, Portugal
* Address for correspondence
Copyright Allured Publishing Corporation Sep/Oct 2004
Provided by ProQuest Information and Learning Company. All rights Reserved
