Characterisation of lavender essential oils by using gas chromatography–mass spectrometry with correlation of linear retention indices and comparison with comprehensive two-dimensional gas chromatography
Introduction
Lavandula essential oils are obtained from the flowering tips of the plants Lavandula angustifolia (lavender), Lavandula hybridia (lavandin) and Lavandula latifolia (spike lavender). These oils have a popular and easily recognisable fragrance. Pure L. angustifolia essential oils are used in aromatherapy, and are thought to have calmative, anti-flatulence, and anti-colic properties [1]. The primary use of lavender oils however is as raw ingredients in industrial perfume and fragrance materials, with the bulk of this market filled by lavandin oils [2]. Lavender essential oil is characterised by high levels of linalool, and linalyl acetate, moderate levels of lavandulyl acetate, terpinen-4-ol and lavandulol. The amount of 1,8-cineole and camphor often varies between very low to moderate [2]. This is a somewhat simplified description; indeed lavender oil typically contains many more than 100 individual components (many minor ones often unidentified and/or not quantitated), each contributing to the chemical and sensory properties of the oils.
For many years, gas chromatography–mass spectrometry (GC–MS) has been the benchmark technique for the qualitative analysis of flavour and fragrance volatiles, and commercially available MS libraries contain many hundreds of mass spectra of such compounds. MS libraries allow tentative identification of essential oil components. Thus GC and GC–MS have been used extensively for the characterisation of lavender essential oils (as reviewed by Boelens [3]). The analysis of living lavender flowers using solid-phase microextraction and GC–MS has also been described [4]. Identification of individual components of essential oils however is not always possible using MS data alone. Differences in mass spectra may be observed if the spectra were obtained using a quadrupole MS, as opposed to using an ion trap MS [5]. Often different spectra are reported in a library for a single compound, with different common names, or systematic name, corresponding to an individual component sometimes apparent. The spectral similarity of a great number of essential oil components causes difficulty in obtaining positive identification of individual components; mass spectra for sesquiterpenes are often identical or nearly identical [6]. More than 230 naturally occurring sesquiterpenes have a molecular mass of 204 [7].
Chromatographic retention data can support MS data, providing an independent parameter on which to base compound identity. The reproducibility and reliability of retention indices allows assignment of identity to unknown components with greater confidence. Both retention indices and MS data of essential oil components are reported in compilations such as Adams [8], Jennings and Shibamoto [9], and Davies [10], and in a number of more recent publications [11], [12], [13]. An MS library incorporating the use of linear retention indices (LRI) as part of an interactive library search has also been described [14]. Called FFC (flavour and fragrance compounds), the library was used to characterise citrus essential oils, and more recently used to characterise several varieties of Australian tea tree oil [15].
The GC–MS analyses and the GC–flame ionisation detection (FID) quantitative analyses in the present investigation were all performed using two independent columns. The sample complexity demanded two independent analyses, on dissimilar stationary phase columns. Adequate resolution of many individual components was not possible in a single analysis. Comprehensive two-dimensional gas chromatography (GC×GC) provides a substantial increase in peak capacity by serially coupling two capillary columns. The principles, modulation processes and applications of GC×GC have been reviewed recently [16], [17]. Cryogenic modulation was used in the present investigation to achieve the GC×GC result, allowing effective modulation of effluent from the first column [18]. Separation of many unresolved components from the first column is achieved in the second column. The application of GC×GC to the analysis of essential oils has been reported [19], [20], [21], [22], [23], with two studies focussing on lavender essential oil [19], [20]. By using a low-polarity column–polar column combination, and by using suitable operating conditions, the two-dimensional separation achieves an increase in peak capacity of the order of 7–12 times.
The present study describes the use of high-resolution GC–MS with LRI to characterise a range of lavender essential oils obtained by steam distillation from a number of different lavender cultivars used for the production of essential oils in Australia. Results are also compared to those obtained from GC×GC analysis of a similar sample. Future opportunities of this technique are discussed.
Section snippets
GC–MS analysis
For all analyses, a Shimadzu QP5050A GC–MS, fitted with a Shimadzu AOC-20i auto sampler, and Shimadzu Class-5000 Chromatography Workstation software (Shimadzu, Italy) was used.
All analyses were carried out by using two different stationary phase columns. Column 1 was an Rtx-5MS (0.25 μm film thickness) fused-silica capillary column. The column dimensions were 30 m×0.25 mm. Column 2 was an Rtx-WAX (0.25 μm film thickness) fused-silica capillary column. The column dimensions were 30 m×0.25 mm.
Results and discussion
Qualitative characterisation of nine different lavender essential oil samples was performed using GC–MS. Some of the difficulties faced whilst characterising the individual components are demonstrated in Table 1 which is the library report for an authentic α-terpinene standard using the conditions described in the experimental section above. Thus comparison of library spectra alone as the method of identification of unknown compounds is not reliable. The first nine library matches to the
Conclusion
The essential oils industry depends on reliable techniques for characterising essential oils to ensure product quality. For essential oils to be used as pharmaceutical products or as food products, reliable characterisation techniques are especially important. By incorporating LRI data into the MS library and using these two independent parameters to make assignments, identification of the components in the samples investigated was greatly simplified. Automation of library search algorithms to
Acknowledgements
The authors would like to acknowledge the assistance provided by Shimadzu Italy to conduct this investigation. R.S. acknowledges support to conduct this project under the UROP program at RMIT. We thank Australian Botanical Products for their assistance in sourcing samples for analysis. The authors wish to express their thanks to Mr Peter Dawes, SGE International for continuing to support our research work on GC×GC. Agilent Technologies is acknowledged for the support of facilities development
References (28)
- et al.
J. Chromatogr. A
(2001) J. Chromatogr. A
(2001)J. Chromatogr.
(1990)- et al.
J. Chromatogr.
(1963) - et al.
Phytother. Res.
(1999) - et al.
Lavender. A growers guide for commercial production
(1999) Perfum. Flavorist
(1995)- et al.
J. Ess. Oil Res.
(1992) - et al.
Flavour Fragr. J.
(1999) Identification of Essential Oil Components By Gas Chromatography/Mass Spectrocsopy
(1995)