Essential oils and leaf hydrosols were analyzed using integrated sampling and chromatographic techniques. In particular, the volatile organic compounds (VOCs) were extracted (from EWs) or diluted (from OEs) with organic solvent and analyzed by GC-MS and the volatile profile of EWs was also analyzed by GC×GC-TOFMS after extraction of VOCs by HS-SPME. GC-MS is the well-recognized technique of choice for analysis of VOCs from plant material and plant extracts [19
] and, in this work, we applied this technique for the evaluation of the content and chemical composition of both EWs and EOs. HS-SPME is well recognized as a widespread and convenient sampling tool for VOCs and it is increasingly used coupled with GC-MS in analysis of food and more [21
]. However, for quantitation purposes, several issues [22
] (e.g., differences that arise from different absorption capacity of different fibers, changes in sorption temperature, competition between different molecules at different affinities for the absorpting material, fiber wearing) led to the need of using several devices (e.g., the use of a pool of suitable internal standards [23
]) to ensure unbiased quantification. For these reasons, we decided to apply HS-SPME-GC×GC-TOFMS analysis on EWs to better elucidate the volatile profile, thus providing a tool for the direct comparison and visualization of plant volatile components and pointing out the presence of molecules not detectable only with GC-MS. Further quantitative evaluation via HS-SPME-GC-MS analysis can be further investigated in future researches. To the authors’ knowledge, the characterization of EWs of these E
. species was not been reported in the literature, to date.
2.1. Chemical Composition of Essential Oils and Hydrolats by GC-MS
shows the chemical composition by GC-MS of the EOs and EWs summarized in Table 3 (see experimental section). Overall, 10 monoterpene hydrocarbons, 19 oxygenated monoterpenes, 2 sesquiterpene hydrocarbons, 2 aromatic monoterpenes (one of which oxygenated), 1 ester, 4 ketones, 1 aldehyde and 5 alcohols were identified. Some of the main molecules detected in the EWs and EOs are reported in Figure 1
. Relative abundance of each of the molecules identified by GC-MS was calculated as a percentage of the peak area on the total area of the identified peaks. Peak areas from the total ion current were normalized by the use of the area of internal standard (tridecane).
Monoterpene hydrocarbons (MH): these terpenes were present in the EO samples with relative abundances between 6.74% and 10.76%. Limonene was the most abundant MH, with similar percentages in all the EOs (3.00–4.65%). Different abundances of the other MHs were pointed out for different Eucalyptus
species: in Eucalyptus parvula
, similar amounts of p
-cymene and α-pinene were detected, followed by lower amounts of γ-terpinene and Z
-ocimene. In Eucalyptus cinerea
, similar amounts of α-pinene and Limonene were detected, followed by lower amounts of p
-cymene, β-myrcene, β-pinene and Z
-ocimene. Regarding the other two species (Eucalyptus pulverulenta
Sims and Eucalyptus pulverulenta
baby blue Sims), the α-pinene amount was approximately half the Limonene, with lower amounts of p
-cymene, β-myrcene, and β-pinene. Other MH (camphene, α-phellandrene, alloocimene) were in low amounts. Noteworthy, the α-phellandrene content in EOs of all these species was lower than 1%, according to the European Pharmacopoeia
specification for 1,8-cineol-rich E
. oils [16
1,8-cineole (eucalyptol) was by far the main component of the analyzed samples (see the next paragraphs); however, the differences in relative abundance of metabolites present in low amount, or even in trace, play a critical role in mediating different activities for EOs from different Eucalyptus
species with 1,8-cineole as the main component; indeed, these different activities (i.e., alleophatic [24
], protection against Parthenium hysterophorus
]) were reported as only due to differences in the relative abundance of minor components [24
], likely due to the synergistic effect of these latter compounds with other components [26
Regarding the hydrosols, no significant amounts of MH were detected, due to the hydrophobic nature of these molecules.
Oxygenated monoterpenes (OM): 1,8-cineole is the main component of EOs obtained from the leaves of Eucalyptus globulus,
the most common Eucalyptus
]. In EOs from these four species, relative abundance of 1,8-cineole ranged between 83.80% and 88.66%, higher than the 80–85% indicated by the standard ISO as the minimum amount of 1,8-cineole for EO from E. globulus.
Other papers in the literature reported the characterization of EOs from Eucalyptus cinereal
] and Eucalyptus pulverulenta
], while, to the authors’ knowledge, no reports on EO from Eucalyptus parvula
have been published, to date. In such papers, the relative abundance of 1,8-cineole showed great variability ranging usually from 58.0 to 69.0% and sometimes reaching 87.8% in E. cinerea
EOs and being approx. 75% in E. pulverulenta
EOs. Consequently, also the relative abundances of the other minor terpenes showed a great variability. This variability might be due to the effect of climatic and geographical factors and harvesting season.
In our study, Eucalyptus parvula and Eucalyptus pulverulenta Sims were the species with the highest amount of 1,8-cineole. Since EOs are totally composed by the volatile fraction, the relative abundance of each compound can be assumed as the amount of this molecule in the oil expressed as g/100g.
Regarding the analyzed hydrosols (EWs), 1,8-cineole was in the range 88.40–90.78%. In order to better characterize these hydrosols, 1,8-cineole was quantified using an external calibration curve, as reported in the experimental section. Table 2
shows that the absolute concentration of 1,8-cineole in the EWs extracts varied in the range 0.74–1.58 g/L, highlighting that this molecule was also recovered in water samples.
In EWs, OMs constituted 98–99% of the total VOCs, according to their water solubility, higher than MHs. Regarding OMs other than 1,8-cineole, in our samples α-terpineol was the most abundant one (4.19–6.24%), followed by lower amounts of terpinen-4-ol, linalool oxides (furanoid, cis and trans), terpineol isomer, and other minor OMs (<1.5%).
OMs constituted about 90% of the EOs. In these samples, α-terpineol was in the range 2.00–3.11%. Noteworthy, in the EO from Eucalyptus parvula, the highest amount of α-terpineol and no presence of its ester, namely α-terpinyl acetate, were detected. In the other three species, α-terpinyl acetate was detected and the sum of the percentages of α-terpineol and α-terpinyl acetate was similar to that of α-terpineol of EO of the Eucalyptus parvula. The other OMs didn’t exceed 0.72%.
Other terpenes: no sesquiterpene hydrocarbons were identified in EWs, in agreement with their insolubility in water. In EOs, very low percentages of β-caryophyllene (≤0.07%) in all samples, and slightly higher amounts of alloaromandrene in Eucalyptus pulverulenta Sims (0.44%) and Eucalyptus pulverulenta baby blue Sims (0.60%) were detected.
One aromatic monoterpene, namely p-cymenene, was identified in very low amounts (≤0.04%) only in EOs samples, while one oxygenated aromatic monoterpene (p-cymen-8-ol) was identified in low amounts (≤0.11%) only in EWs, according to their different water solubility.
Other compounds: no significant amounts of esters and aldehydes were identified in our samples, the only exceptions being traces of nonanal in one EW sample and very low amounts of isoamyl acetate in EO from Eucalyptus parvula. Ketones (the three linear isomers of heptanone and lower percentages of 6-methylhept-5-en-2-one) were identified in low amounts (0.20–0.27% in both EWs and EOs). Finally, alcohols were identified in very low amounts in EOs (heptan-2-ol and heptan-3-ol for a total amount up to 0.16%), while in EWs they were present in percentages up to 2.01%, with 3-methylbutanol as the main molecule, followed by 2-phenylethanol and Z-hex-3-en-1-ol.
2.2. Fingerprint Analysis by HS-SPME-GC×GC-TOFMS
Solid-phase microextraction (SPME) is a rapid and simple procedure for extraction of volatile fraction from aromatic and medicinal plants [28
]. As reported, the divinilbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS) fiber is the most effective SPME fiber able to isolate the volatile fraction from commercial hydrosols of several plants [12
]. HS-SPME and GC×GC-TOFMS fingerprint analysis are ideal tools to analyze complex volatile fraction from several matrices, and to provide a sensitive method for the direct comparison and visualization of plant volatile components. As previously reported, 1,8-cineole was the major component in the EOs and EWs, but differences in the other metabolites present in low amounts are very important. Utilization of comprehensive two-dimensional GC (GC×GC) increases separation power with respect to that of the one-dimensional GC in complex matrices where the presence of low abundant components is critical, such as Eucalyptus
]. To our knowledge, there has been no study reporting the volatile profile of EWs from these Eucalyptus
species; therefore, hydrosols from the four Eucalyptus
species were analyzed by HS-SPME-GC×GC-TOFMS to better elucidate the volatile profile of these by-products, also pointing out the presence of molecules not detectable with only GC-MS. HS-SPME-GC×GC-TOFMS analyses of the complex volatile fraction of EWs were submitted to advanced fingerprinting analysis of 2D chromatographic data.
In Figure 2
, “contour plots” from HS-SPME-GC×GC-TOFMS analyses of the four Eucalyptus
species are reported: each 2D-peak corresponds to a single volatile compound. In this case, SPME and comprehensive comparative analysis of 2D chromatographic data showed visual differences among EW samples. 1-16-EW and 2-16-EW showed a larger number and a higher intensity of peaks, with respect to 3-16-EW and 4-16-EW. The most intense peak corresponded to 1,8-cineole.
For example, a total of about 400 compounds was detected by GC×GC analysis in 1-16-EW (estimated from the number of peak contours in 2D plots) and, after subtracting baseline peaks, corresponding to fiber blending or background interferences, 137 peaks/compounds were identified. These results were in agreement with Wong et al. [20
], where the 2D rational separation pattern aids the identification of ca. 400 metabolites in Eucalyptus
spp. leaf oils, 183 of which were identified or tentatively identified and represented percentages between 50.8–90.0% of the total ion count, comprising various chemical families.
HS-SPME-GC×GC-TOFMS provided a high metabolic coverage of VOCs: monoterpenes, oxygenated monoterpenes (the main class), oxygenated monoterpenes acetate, and others (ketones, aldehyde, alcohols).
GC×GC is currently adopted as separation technique not only because of its high separation power and sensitivity, but also for its ability to produce more widely distributed and rationalized peak patterns [30
] for chemically correlated group of analytes. Terpenic compounds of Eucalyptus
hydrosols were organized mainly in three clusters in 2D separation space: monoterpenic hydrocarbons, oxygenated monoterpenes and monoterpenes acetate, except for the 1,8-cineol that wrapped around, resulting in monoterpenes zone (Figure 2
A). 1,8-cineol showed high secondary retention and fall outside the range of secondary retention time (wrapped around). As previously reported for volatile oil from leaves of Eucalyptus dunnii
], one molecule that wraps around, does not affect the separation and identification of the compounds, since the more strongly retained components (those that wrap-around) did not overlap peaks that were weakly retained in the subsequent modulation.
Up to 31 peaks/compounds belonging to the class of oxygenated monoterpenes were distributed in a defined part of the contour plot for 1-16-EW (Figure 2
A, braced region “b”). The number of oxygenated monoterpenes were 33 for 2-16-EW, 23 for 3-16EW and 24 for 4-16-EW.
An advanced approach known as comprehensive template matching fingerprinting [32
] was adopted (Figure 2
B). This method considers, as a comparative feature, each individual 2D peak together with its time coordinates, detector response and MS fragmentation pattern, and includes them in a sample template that is created by the analyst and can be used to compare plots from different samples directly and comprehensively. A template could be used to correctly interpret visual differences in further analyses. To create the template, the peak identification was performed by matching the experimental mass spectra against spectra databases combined with GC-MS data.
The main differences that emerged between the four varieties could be summarized as follows: 1-16-EW showed the presence of 6-methylhept-5-en-2-one that was not present in the other species (see also Table 1
) and the presence of exo
-2-hydroxycineole acetate isomers. 1-16-EW did not show terpinyl acetate and α-terpinene, which instead were found in the other three species. 2-16-EW was the only species that showed the presence of β
-phellandrene, piperitone and citral. 1-16-EW and 2-16-EW showed the presence of cis
-jasmone and carvone, while 3-16-EW and 4-16-EW didn’t show the presence of these molecules. Volatile profiles presented in 2D contour plots allow visual discrimination of the metabolic composition among interspecies of Eucalyptus
aromatic waters, as reported for leaf oils of other different Eucalyptus