Anal. Chem. 2006, 78, 883-890
Evaluation of Leaf-Derived Extracts as an Environmentally Sustainable Source of Essential Oils by Using Gas Chromatography-Mass Spectrometry and Enantioselective Gas Chromatography-Olfactometry Barbara d’Acampora Zellner,† Maria Lo Presti,† Lauro Euclides Soares Barata,‡ Paola Dugo,§ Giovanni Dugo,† and Luigi Mondello*,†
Dipartimento Farmaco-chimico, Facolta` di Farmacia, Universita` di Messina, viale Annunziata, 98168 Messina, Italy, Instituto de Quı´mica, Universidade Estadual de Campinas, C.P. 6154 Campinas, Sa˜o Paulo, Brazil, and Dipartimento di Chimica Organica e Biologica, Facolta` di Scienze, Universita` di Messina, Contrada Papardo, 98166 Messina, Italy
In consideration of the world’s present environmental situation and the threat of species extinction, investigations concerning alternative sustainable sources of natural substances represent an extremely important issue. In this respect, the present research is focused on the analytical evaluation of Brazilian rosewood (Aniba rosaeodora Ducke) leaves, as an alternative source (with respect to wood) of rosewood essential oil and, as such, of natural linalool, which is extensively used in perfumery. Enantioselective-gas chromatography-olfactometry (EsGC-O) was used as a tool for the simultaneous stereodifferentiation and olfactive evaluation of the volatile optically active components present in the analyzed samples. In addition to Es-GC-O analyses, direct olfactive analyses were also performed, enabling the evaluation of the global aroma exerted by each sample and the influence of each linalool antipode, as also other minor compounds. The samples were also submitted to gas chromatographymass spectrometric analysis, thus establishing their chemical profiles. The assessment of enantiopure chiral compounds through Es-GC-O, along with direct olfactive analyses, confirmed that the leaves are a potential substituent for wood in the extraction of Brazilian rosewood essential oil, representing a sustainable nonwood source of natural linalool. Gas chromatography-olfactometry (GC-O) is a well-known analytical technique, which enables the assessment of odor-active components in natural and synthetic samples by associating the resolving power of GC with the selectivity and sensitivity of the human nose.1 The use of cyclodextrin stationary phases as chiral selectors, enabling a reliable separation of racemates without * To whom correspondence should be addressed. Phone: +39-090-6766536. Fax: +39-090-6766532. E-mail:
[email protected]. † Dipartimento Farmaco-chimico, Universita ` di Messina. ‡ Universidade Estadual de Campinas. § Dipartimento di Chimica Organica e Biologica, Universita ` di Messina. (1) Clery, R. In The Chemistry of Fragrances; Pybus, D. H., Sell, C. S., Eds.; Royal Society of Chemistry: Cambridge, U.K., 1999; pp 202-215. 10.1021/ac051337s CCC: $33.50 Published on Web 12/21/2005
© 2006 American Chemical Society
derivatization processes, has also been extended to GC-O. Enantioselective gas chromatography-olfactometry (Es-GC-O) is a valid method for the correct determination of sensory properties and differentiation of individual enantiomers, as demonstrated by Mosandl and his group.2 It is worthwhile to point out that the preponderance of one of the enantiomers, defined by the percentage of the enantiomeric excess, results in a characteristic aroma.3 It must be emphasized that Es-GC-O is a valuable analytical tool for the investigation of alternative sources of natural isolates. A worthwhile example is the economically relevant essential oil extracted from the trunkwood of Brazilian rosewood (Aniba rosaeodora Ducke), considered to be a valuable source of linalool (3,7-dimethyl-1,6-octadien-3-ol).4,5 However, the destructive harvesting of wild trees is a threat not only for the survival of this species but also for the maintenance of biodiversity of the region. In the 1960s, the production and export of this essential oil, which is steam distilled in yields averaging 1%, led to an annual consumption of ∼50 000 tons of trunkwood.6 However, with the introduction of synthetic linalool in the market and the employment of alternative natural linalool sources, the Brazilian rosewood oil industry declined significantly.4 In the mid-1980s, the annual demand for this essential oil decreased to ∼100 tons6 and reaching in the years 2000-2003, an average of 26 ton/year.4,7 Nevertheless, nearly 2600 tons/year of trunkwood are required to produce such an amount. The monoterpene alcohol, linalool, is not only an aroma chemical in its own right but also a precursor for other fragrance compounds, such as linalyl acetate, frequently used to impart a sweet freshness to top notes.8 Linalool represents an excellent (2) Lehmann, D.; Dietrich, A.; Hener, U.; Mosandl, A. Phytochem. Anal. 1995, 6, 255-257. (3) Boelens, M. H.; Boelens, H. Perfum. Flavor. 1993, 18, 2-16. (4) Barata, L. E. S.; May P. Econ. Bot. 2004, 58, 257-265. (5) Lawrence, B. M. Perfum. Flavor. 1984, 9, 87-95. (6) Osashi, S. T.; Rosa, L. dos S.; Santana, J. A. Perfum. Flavor. 1997, 22, 1-5. (7) Maia, J. G. S.; Zoghbi, M. G.; Andrade, E. H. Aromatic Plants in the Amazon and their Essential Oils; PR/MCT-Museu Paraense Emilio Goeldi, Colec¸ ˜ao Adolfo Ducke: Bele´m, 2001.
Analytical Chemistry, Vol. 78, No. 3, February 1, 2006 883
example of antipodes generating distinct odor properties and threshold values and can be found as two optical forms, licareol, the (R)-(-)-enantiomer, and coriandrol, the (S)-(+)-enantiomer. Brazilian rosewood oil contains both forms in variable proportions6,9-11 according to the cultivation region of the species.12 It must be added that the majority of studies concerning Brazilian rosewood are based on the essential oil obtained from the wood.6,7,12,13 The aim of this research was to provide an accurate evaluation on rosewood leaf essential oil, considering a possible use in perfumery. This was achieved by comparing the chemical (GC/ MS and chiral GC analyses) and olfactive (Es-GC-O and direct olfactive analyses) profiles of wood and leaf-derived essential oils, as well as the distilled linalool-rich fractions of both oils (rectified essential oil). Although the enantiomeric distribution of linalool present in Brazilian rosewood oil has been previously described in the literature,6,8,10 to the authors’ knowledge, this is the first time that Es-GC-O has been applied to this matrix and, most importantly, with such a purpose. EXPERIMENTAL SECTION Samples. The crude essential oil of rosewood (A. rosaeodora Ducke) wood and leaves, as well as their linalool-rich distilled fractions, were supplied by Prof. Dr. Lauro Barata (Universidade Estadual de Campinas, Sa˜o Paulo, Brazil). All the samples were stored at 4 °C. The following standards compounds were purchased from Sigma-Aldrich (Bellefonte, PA): (+)-R-pinene, (-)-R-pinene, (+)β-pinene, (-)-β-pinene, (-)-limonene, (+)-limonene, (()-linalool, (+)-terpinen-4-ol, (()-terpinen-4-ol, (-)-R-terpineol, and (+)-Rterpineol. Sample Preparation. Each crude essential oil, distilled fractions, and standard components were diluted 1:10 v/v in n-hexane. GC/MS Analyses. GC/MS analyses were carried out on a Shimadzu GCMS-QP2010 gas chromatograph-mass spectrometer equipped with autoinjector AOC-20i and autosampler AOC-20s (Shimadzu, Kyoto, Japan) and the Flavour and Fragrance Natural and Synthetic Compounds (FFNSC) ver. 1.2 MS Library (Shimadzu). Data were collected by the GCMS Solution software (Shimadzu). The column set consisted of a fused-silica capillary column of 5% diphenyl-95% dimethylpolysiloxane (MDN-5S) phase (0.25-µm film thickness, df) with dimensions 30 m × 0.25 mm i.d. (Supelco, Bellefonte, PA). The oven temperature was kept at 50 °C for 2 min, then increased to 250 °C at 3.0 °C/min, and kept for 10 min. The GCMS-QP2010 was equipped with a split/ splitless injector (250 °C); injection volume, 1.0 µL, in split mode (50:1); inlet pressure, 37.1 kPa; carrier gas, He; constant linear velocity, 32.4 cm/s; interface temperature, 250 °C; MS ionization (8) Bauer, K.: Garbe, D.; Surburg, H. Common Fragrance and Flavor Materials; VCH Verlagsgesellschaft: Weinheim, 1990. (9) Guenther, E. The Essential Oils; Van Nostrand Co.: New York; 1950; Vol. IV. (10) Siani, A. C.; Tappin, M. R. R.; Ramos, M. F. S.; Mazzei, J. L.; Ramos, M. C. K. V.; de Aquino Neto, F. R.; Frighetto, N. J. Agric. Food Chem. 2002, 50, 3518-3521. (11) Casabianca, H.; Graff, J. B.; Faugier, V.; Fleig, F.; Grenier, C. J. High Resolut. Chromatogr. 1998, 21, 107-112. (12) Mors, W. B.; Gottlieb, O. R.; Djerassi, C. J. Am. Chem. Soc. 1957, 79, 45074511. (13) Buccellatto, F. Perfum. Flavor. 1988, 13, 35-36.
884
Analytical Chemistry, Vol. 78, No. 3, February 1, 2006
mode, electron ionization; detector voltage, 0.9 kV; acquisition mass range, 40-400 uma; scan speed, 10 000 uma/s; acquisition mode, full scan; scan interval, 0.50 s (2 Hz); solvent delay, 5 min. The analyses were executed in triplicate. GC Analyses. Conventional and chiral GC analyses were carried out on a Shimadzu GC-2010 equipped with identical autoinjector, autosampler, and split/splitless injector, such as the GCMS-QP2010. Data were collected by the GC Solution software (Shimadzu). The GC analyses, as also the chiral analyses, were executed in triplicate. Conventional GC analyses were carried out by applying the temperature program and column as previously reported for the GC/MS analyses: injection volume, 1.0 µL, in the split mode (100: 1); inlet pressure, 100 kPa; carrier gas, He; constant linear velocity (uˆ ), 30.0 cm/s; detector, flame ionization detector (FID) (275 °C); H2 flow, 50.0 mL/min; air flow, 400.0 mL/min; makeup (He), 50.0 mL/min; sampling rate, 20 ms. Chiral GC analyses were carried using a column consisting of diethyl-tert-butyl-silyl-β-cyclodextrin (DETTBSBETA) phase (0.25µm film thickness, df) with dimensions 25 m × 0.25 mm i.d. (Mega, Legnano, Italy). The oven temperature was kept at 45 °C for 6 min, increased to 200 °C at 2.0 °C/min, and kept for 5 min: injection volume, 1.0 µL, in the split mode (30:1); inlet pressure, 96.0 kPa; carrier gas, H2, delivered with constant pressure: 37.1 kPa; linear velocity (uj ), 35.1 cm/s; detector, FID (250 °C), under the above-described conditions; sampling rate, 80 ms. All four samples have been analyzed in triplicate, and the average accuracy was established through the calculation of the mean of the obtained values. Furthermore, for individual variations from the average an interval higher or lower than the standard deviation (s ) 0.045) has been accepted. Es-GC-O Analyses. GC-O chiral analyses were carried out on a Shimadzu GC-2010 (Shimadzu), also equipped with the aforementioned autoinjector and autosampler, and hyphenated to the Phaser Sniffing Port OP275 (ATAS GL International B.V., Veldhoven, The Netherlands). A splitter stand with four ports was located in the GC oven; two ports were connected to the analytical column outlet and to an auxiliary gas outlet; retention gaps connected the two remaining ports to the detector (FID) and to the transfer line (sniffing port). The former ends up on an ergonomic glass nose cone. Sniffing port working conditions, such as temperature and auxiliary gas flow, are set on a control station, independent from the Shimadzu GC-2010. The gas chromatographic effluent of the representative isolate of volatile compounds is sniffed through the sniffing port, and the odor activity is described by trained evaluators. The sniffing port is equipped with a device that minimizes the discomfort from sniffing hot dry effluent gases through the combination of the hot GC effluent with humidified air, reducing nasal dehydration. The sniffing procedure was divided in 20-min sessions with a 15-min interval in order to avoid lassitude. The applied temperature program and column set were identical to the chiral GC analyses: split/splitless injector (250 °C); injection volume, 0.5 µL, in the split mode (30: 1); carrier gas, He, delivered with constant pressure, 117.4 kPa; linear velocity (uˆ ), 35.1 cm/s; FID (250 °C); under the aforementioned conditions; sampling rate, 80 ms. (a) Panel. The panel used in the Es-GC-O analyses was composed of eight evaluators: three working in our laboratory
Table 1. GC/MS Result for the Essential Oils of Leaves and Wood of Brazilian Rosewood and Their Respective Distilled Fractiona relative % peak areas
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
compound
LRIb
3-hexen-1-ol R-pinene camphene Bois de rose oxide β-pinene 6-methyl-5-hepten-2-one myrcene p-cymene limonene 1,8-cineole 2 (3H)-furanone (E)-β-ocimene trans-linalool oxide (furanoid) cis-linalool oxide (furanoid) linalool unknown oxygenated monoterpene unknown oxygenated monoterpene cis-linalyl oxide (pyranoid) trans-linalool oxide (pyranoid) terpinen-4-ol p-cymen-8-ol unknown oxygenated monoterpene R-terpineol unknown oxygenated monoterpene nerol geraniol geranial unknown oxygenated monoterpene unknown oxygenated monoterpene R-cubebene R-copaene β-elemene R-gurjunene E-caryophyllene R-guaiene R-humulene alloaromadendrene unknown sesquiterpene β-selinene R-selinene R-muurolene γ-cadinene δ-cadinene trans-calamenene unknown oxygenated sesquiterpene E-nerolidol spathulenol caryophyllene oxyde guaiol unknown oxygenated sesquiterpene unknown oxygenated sesquiterpene selin-11-en-4-R-ol unknown oxygenated sesquiterpene unknown oxygenated sesquiterpene unknown oxygenated sesquiterpene unknown oxygenated sesquiterpene benzyl benzoate monoterpenes oxygenated monoterpenes sesquiterpenes oxygenated sesquiterpenes other
853 931 947 968 976 985 989 1023 1029 1031 1036 1045 1070 1087 1109 1129 1146 1170 1175 1180 1187 1190 1196 1223 1225 1253 1268 1273 1342 1346 1374 1388 1405 1417 1434 1455 1458 1469 1488 1494 1500 1514 1517 1520 1549 1561 1575 1581 1595 1607 1615 1657 1667 1707 1723 1724 1766
crude wood oil 0.14 trc 0.16 0.07 trc 0.02 0.04 0.46 0.25 0.05
distilled wood fraction
0.05 0.02 0.04 0.28 0.21 0.10
2.43 2.21 85.00
3.14 2.84 85.89 0.07
0.14 0.19 0.06 0.03 0.44 1.48
0.14 0.15 0.05 0.02 1.06 1.09 0.03 0.03 0.06 0.04 0.31 0.11
0.03 0.19 0.05 0.13 0.09 1.19 0.33 0.31 0.04 0.15 0.71 0.35 0.04 0.08 0.03 0.02 0.02 0.11 0.05 0.13 0.04 0.08 0.03 0.11 0.09 trc 0.09 0.06 0.11 0.74 92.88 3.25 0.82 2.31
0.41 0.16
trc 0.18 0.08
crude leaf oil
distilled leaf fraction
0.10 0.97 0.04
trc
0.70
0.03
0.03 trc 0.18 0.15
trc 0.19 0.04
0.03 1.19 0.88 81.45 0.06 0.09 0.04
1.00 0.95 94.87 0.21 0.08 0.04 0.06 trc
0.04 1.21
0.03 0.72
0.13 0.35
0.07 0.20 0.14 0.05
0.14 1.12 0.25 0.11 0.20 0.05 0.07 0.04 0.29 1.65 1.39
0.49 0.23
0.07 0.04
0.07 0.29 0.06 0.17 1.27 0.33 0.15 0.12
0.34 95.29 0.84 3.53
0.36 0.35 0.29 0.32 0.57 0.25 1.93 85.62 5.73 3.93 2.79
0.23 98.47 0.83 0.47
a Relative percent peak areas were determined through GC analysis. b Linear retention index calculated on the MDN-5S column. c Trace compound (% peak area < 0.01).
and five from local flavor and fragrance industries with previous experience in GC-O. Due to the subjectivity of odor perception,
all evaluators were submitted to a daily training, for two weeks, by sniffing different dilutions of standard compounds in ethanol. Analytical Chemistry, Vol. 78, No. 3, February 1, 2006
885
Figure 1. Es-GC chromatograms of Brazilian rosewood (A) crude wood oil and (B) crude leaf oil. The numbered peaks correspond to the enantiomers presented in Table 2.
With the aim to normalize the language between panelists, the olfactive quality description was based on a glossary of olfactive descriptors. The panel was also trained for intensity estimation by using a five-point intensity interval scale (1, extremely weak; 2, weak; 3, moderate; 4, strong; 5, extremely strong) in the evaluation of the different dilutions of standard compounds. The following analyses were carried out in duplicate. (b) Olfactometry Global Analysis (Frequency Response). Es-GC-O frequency detection was performed using an adaptation 886
Analytical Chemistry, Vol. 78, No. 3, February 1, 2006
of the method presented by Pollien et al.14 The adaptation consists not only of the indication of each assessor of the perception of an odor but in the simultaneous description of the odor quality according to the glossary of olfactive descriptors adopted by the group. (c) Posterior Intensity Method. The estimations of the perceived intensity were registered after the elution of the (14) Pollien, P.; Ott, A.; Montigon, F.; Baumgartner, M.; Munoz-Box, R.; Chaintreau, A. J. Agric. Food Chem. 1997, 45, 2630-2637.
Figure 2. Es-GC chromatograms of Brazilian rosewood distilled fractions obtained from (A) wood oil and (B) leaf oil. The numbered peaks correspond to the enantiomers presented in Table 2.
chromatographic peak.15 The perceptions were rated on the aforementioned intensity scale. Linear Retention Index (LRI). The LRI determination was carried out in triplicate by injecting an homologous series of n-alkanes containing 24 n-hydrocarbons (C7-C30) purchased from Supelco (Bellefonte, PA), each at 1000 ppm in hexane. The indices
were calculated according to the equation proposed by Kova´ts in 1958.16 Sensorial Analyses. The samples were submitted to direct olfactive analyses by a panel of 12 judges, the 8 aforementioned evaluators of the Es-GC-O analyses and another 4 evaluators working in different laboratories at the University of Messina with previous experience in descriptive sensory evaluations. Prior to
(15) van Ruth, S. M.; O’Connor, C. H. Food Chem. 2001, 74, 341-347.
(16) Kova´ts, E. S. Z. Helv. Chim. Acta 1958, 41, 1915-1932.
Analytical Chemistry, Vol. 78, No. 3, February 1, 2006
887
Table 2. Distribution and Odor Description of the Enantiomers Present in the Essential Oils of Leaves and Wood of Brazilian Rosewood and Their Respective Distilled Fractions enantiomer distribution (%) enantiomers 1 2 3 4 5 6 7 8 9 10 a
(+)-β-pinene (-)-β-pinene (-)-limonene (+)-limonene (-)-linalool (+)-linalool (+)-terpinen-4-ol (-)-terpinen-4-ol (-)-R-terpineol (+)-R-terpineol
crude wood oil
distilled wood fraction
8.0 92.0 5.4 94.6 38.3 61.7 68.9 31.1 51.5 48.5
tra tra 5.6 94.4 38.1 61.9 tra tra 52.2 47.8
crude leaf oil
distilled leaf fraction
2.8 97.2 0.4 99.6 29.3 70.7 63.2 36.8 48.6 51.4
tra
odor description camphoraceous, fresh, pungent camphoraceous, fresh, pungent harsh, turpentine-like citric, fresh woody, floral, lavender, fresh sweet, citric, herbaceous musty, dusty musty green, coniferous floral
tra 0.7 99.3 26.8 73.2 tra tra 54.7 45.3
Trace compound (% peak area < 0.01).
the analyses, the evaluators were trained to recognize linalool. The odor attributes were generated according to the abovementioned glossary of olfactive descriptors. The overall odor intensity of the olfactive impact of the samples was also assessed; the perceived intensity was rated using the five-point intensity interval scale applied in the posterior intensity method. The analyses were made in duplicate, at room temperature, through the application of each sample, previously diluted 1:10 v/v in n-hexane, to appropriate smelling strips. Statistical Evaluation. The data obtained from the Es-GC-O and sensory analyses were subjected to statistical treatment using one-way ANOVA analysis. A significance level of 5% was used throughout the study. The evaluation was executed using the SigmaStat, Statistical Software (Jandel Corp., San Rafael, CA). RESULTS AND DISCUSSION GC/MS and GC analyses. The essential oils obtained from the wood and the leaves of Brazilian rosewood (both in yields ranging 1.00%) and their respective distilled fractions were initially investigated with regard to the identity of the compounds present in these matrixes. GC/MS analyses enabled the identification of the compounds through the comparison of their calculated LRIs and obtained mass spectra with those present in the recently developed FFNSC ver. 1.2 MS Library. The latter has been created registering spectra obtained from pure standard compounds, essential oils, and perfumes. The LRI, a GC qualitative parameter independent of column type, flow, and dimension, is used as a filter, shortening so the matching results, and enhancing the credibility of the identification through MS. The detected compounds and their relative percentage peak areas (determined through GC), are presented in Table 1. A total of 48 compounds were detected in the essential oil obtained from trunkwood; among these, 37 were positively identified. In the oil obtained from the leaves, 37 constituents were identified out of the 44 detected. When considering the distilled fraction of each of the oils, 27 compounds were detected, and among these 21 identified in the fraction obtained from the wood, while 18 out of 23 constituents were identified in the rectified leaf oil. As expected, linalool was identified as the main compound in all four samples, and with regard to the crude oils, a similar percentage of this compound was present in the oil obtained from the leaves (81.45%) when compared to that of the wood (85.00%). 888 Analytical Chemistry, Vol. 78, No. 3, February 1, 2006
Table 3. Gas Chromatographic Data of the Enantiomers Present in the Leaf- and Wood-Derived Essential Oils of Brazilian Rosewood and Each Respective Distilled Fraction (Es-GC Analyses) resolution Rs b
1 2 3 4 5
enantiomers
LRIa
(+)-β-pinene (-)-β-pinene (-)-limonene (+)-limonene (-)-linalool (+)-linalool (+)-terpinen-4-ol (-)-terpinen-4-ol (-)-R-terpineol (+)-R-terpineol
933 945 1050 1057 1168 1181 1249 1251 1295 1307
crude wood oil
distilled wood fraction
crude leaf oil
distilled leaf fraction
4.4
ndc
3.0
ndc
5.7
2.3
2.4
3.0
1.9
2.3
1.5
1.9
1.8
ndc
2.3
ndc
2.0
3.1
3.7
3.6
a Linear retention index measured in triplicate on DETTBSBETA column (25 m × 0.25 mm i.d. × 0.25 µm film thickness) programmed from 45 (6 min) to 200 °C (5 min) at 2.0 °C/min. Samples (1 µL) were injected in the split mode (30:1). b Resolution Rs was calculated according to 2(tR2 - tR1)/(wb1 + wb2), where tR2 and tR1 are retention times and (wb1 + wb2), are peak widths at base. c The resolution factor could not be determined due to the presence of these enantiomers in trace concentrations.
These analyses confirmed the similarities between the chemical profile of the two crude essential oils. It must be emphasized that the oil extracted from the leaves is characterized by a higher concentration of oxygenated sesquiterpenes in comparison to the wood oil, while the latter is composed of a slightly higher amount of oxygenated monoterpenes (Table 1). Further, a difference was observed in the concentration of oxygenated derivates of linalool, such as linalool and linalyl oxides, with the leaf oil presenting lower amounts. Considering that the most expressive terpenes used in perfumery are the oxygenated monoterpenes,17 the essential oil obtained from the wood proved to be, indeed, richer. On the other hand, the oil extracted from the leaves presented a higher concentration of oxygenated, and also nonoxygenated sesquiterpenes. These compounds are less volatile and, hence, are being used as fixatives in perfumes.18 (17) Sell, C. In The Chemistry of Fragrances; Pybus, D. H., Sell, C. S., Eds.; Royal Society of Chemistry: Cambridge, U.K., 1999; pp 24-50.
Figure 3. Odor descriptor graphs of the Brazilian rosewood samples: (A) wood essential oil; (B) leaf essential oil; (C) distilled fraction of the wood essential oil; (D) distilled fraction of the leaf essential oil.
Es-GC analyses. Es-GC analyses enabled the determination of the racemic distribution of linalool in each sample, revealing a ratio of 38.3% (-)-linalool to 61.7% (+)-linalool in the crude wood oil against 29.3 to 70.7% in the leaf oil, respectively. The chromatograms of Brazilian rosewood obtained from wood and leaf sources are presented in Figure 1 (A and B, respectively), and of each respective distilled fraction in Figure 2 (A and B, respectively). Other enantiomers were also identified in the essential oils, presenting, as linalool, differences in the racemic distribution, as reported in Table 2. In general, these differences were slight between the crude essential oil and its respective distilled fraction, except for the enantiomers of β-pinene and terpinen-4-ol. These were present in trace concentrations due to the fractionated distillation process, which aim is to modify the content of certain components present in an essential oil, in this specific case to obtain a sample containing mainly linalool. The LRI of each enantiomer and the resolution (Rs) between the enantiomeric pairs are presented in Table 3. Es-GC-O and Direct Olfactive Analyses. It is well known that although chemical detectors used in GC provide relevant information on the composition of essential oils, most of them are not as sensitive as the human nose in the detection of odoractive compounds. Prior to Es-GC-O analyses, the four samples were submitted to direct olfactive analyses in order to confirm the impact of linalool on the fragrance of Brazilian rosewood oil. As expected for such complex matrixes, the olfactive profile confirmed to be very rich. (18) Curtis, T.; Williams, D. G. Introduction to Perfumery; Micelle Press: New York, 2001.
The impressions given by the assessors are reported in the odor descriptor graphs (Figure 3). The notes floral, sweet, and perfumed were used to describe all four samples, although sweet was considered less relevant for the wood-derived oil. In both crude oils, a linalool-like characteristic note was perceived, reinforcing the strength of this monoterpene alcohol on the impact odor profile of each oil. The notes woody, citric, and green represented the main differences between the oils of wood and leaves, the first being more representative for the wood-derived, with the others for the leaf-derived oil. This was confirmed in the analyses of the distilled fractions. Although subtle, a powdery note was also perceived in these latter mentioned samples. In general, when applied directly as a fragrant ingredient, Brazilian rosewood oil is defined as a subnote of the woody family and is considered a high-impact middle note, presenting a balanced bouquet of floral, sweet, woody, and citric odor.18 The two different antipodes of linalool present in all four samples were investigated through olfactometry applying the previously described methods. The detection frequency method enabled the evaluation of the linalool enantiomers present in each sample, verifying not only the presence of the target compound but also the respective retention time. All assessors detected both antipodes of linalool, giving simultaneously the odor quality description of each. According to the assessors, the (-)-enantiomer was described to present woody, flowery, lavender-like, fresh notes, whereas the (+)-enantiomer elicited sweet, citric, and herbaceous impressions. The perceived intensities (Table 4), evaluated through the posterior intensity method, enabled the magnitude estimation of each antipode. Comparing the enantioAnalytical Chemistry, Vol. 78, No. 3, February 1, 2006
889
Table 4. Average Intensities and Resolution of the Linalool Enantiomers Present in the Essential Oils of Leaves and Wood of Brazilian Rosewood and Their Respective Distilled Fractions (Es-GC-O Analyses) crude wood oil
dist wood fraction
crude leaf oil
dist leaf fraction
enantiomers
Ia
Rsb
Ia
Rsb
Ia
Rsb
Ia
Rsb
(-)-linalool (+)-linalool
4 4
4.2
4 4
4.0
4 4
4.2
4 5
4.1
a Intensity (I) according to a five-point intensity interval scale (1extremely weak, 2-weak, 3-moderate, 4-strong, 5-extremely strong). b Resolution R was calculated according to 2 (t s R2 - tR1)/(wb1 + wb2), where tR2 and tR1 are retention times and (wb1 + wb2), are peak widths at base.
meric ratios of all four samples and the obtained average intensity values, (+)-linalool was perceived as very strong when present in enantiomeric ratios of 70.70 and 73.20%, while in lower ratios the intensity was classified as strong. However, the (-)-antipode was perceived as strong in all samples. Although the odor threshold determination has not been carried out in the present research, the strong perceived intensity of (-)-linalool could be attributed to previously reported studies3,19 on the odor threshold of linalool enantiomers, where the (-)-antipode presents a significantly lower threshold value in air than the (+)-antipode. It must be emphasized that a good correlation has been achieved between the percentage amount of each linalool antipode and its perceived intensity through Es-GC-O. In agreement with the enantiomeric ratios obtained through Es-GC, the oil extracted from the leaves showed, indeed, a strongly positive optical rotation value, presenting as such, according to olfactometry, a slightly sweeter, petitgrain note (citric, light floral, and fresh). Since other compounds were also identified as enantiomers in all four samples, and the olfactive impression elicited by them could be determined through Es-GC-O, their odor quality was attributed and is reported in Table 2. However, according to the (19) Padrayuttawat, A.; Yoshizawa, T.; Tamura, H.; Tokunaga, T. Food Sci. Technol. Int. 1997, 3, 402-408.
890 Analytical Chemistry, Vol. 78, No. 3, February 1, 2006
average of the perceived intensities, the odor influence of the enantiomers, besides linalool, was not of great relevance to the global odor profile (according to the five-point intensity interval scale the reported values were from 1, extremely weak, to 3, moderate). Another substantial influence is the presence of unidentified oxygenated sesquiterpenes in the leaf essential oil; when investigated through olfactometry, they elicited predominantly green and balsamic olfactive impressions of relevant intensity, enabling the differentiation between the two crude essential oils. CONCLUSIONS Even if the enantiomeric distribution of linalool present in Brazilian rosewood oil has been previously described in the literature, the investigation through Es-GC-O has been revealed to be a new and effective approach in the ongoing research of this genuine essential oil. Emphasis must also be placed on the importance of the evaluation of the odor properties of enantiomers, not only when these are incorporated in an essential oil but especially after the essential oil was submitted to fractionated distillation. In this case, in fact, a more precise estimation of the olfactive activity of each antipode is attained. The chemical profile determination and sensorial evaluation, through GC/MS and Es-GC-O, of Brazilian rosewood essential oil derived from wood and leaf sources, and each respective distilled fraction, confirm that the use of sustainable harvested rosewood leaves, rather than wood, in the extraction of natural linalool may represent a reasonable guarantee of long-term raw material source. Moreover, the oil content of the leaves might be improved by standardizing cultivation parameters The nondestructive harvesting of aerial parts, as leaves, and also terminal branches, would enable a predictable large raw material supply. ACKNOWLEDGMENT The authors gratefully acknowledge Shimadzu Corp. and ATAS GL International B.V. for the continuous support. Received for review July 27, 2005. Accepted November 19, 2005. AC051337S
