2.1. Composition of Lavender scCO2 Extracts
Over the last decade, the quality of lavender scent extracted by scCO
2 in the pressure ranges of 10–30 MPa and up to 40 °C were reported on in literature, utilizing either a pilot scale [
11,
13] or laboratory scale SFE units [
13,
14]. The supercritical extraction of lavender was studied by Nadalin et al. [
14]. It was shown that maintaining the pressure at 30 MPa and increasing the temperature to 59 °C resulted in the highest yield of lavender extract (around 7.5 wt%) [
14]. In accordance with this reported data, a similar extraction yield was obtained for the extraction of
Lavandula angustifolia of Polish origin (L-Pl) in this paper, resulting in 7.05 wt% as a ratio of the obtained extract mass to the feed mass for the extraction. However, the extraction temperature was significantly lower (40 °C) compared to that applied by Nadalin et al. [
14]. The Bulgarian
L. angustifolia (L-Bg) was characterized by slightly lower extraction yield (6.33 wt%) under analogous extraction conditions. The laboratory-scale investigations mostly aimed to optimize the lavender extraction process, in terms of a high total extraction yield and a relatively high amount of linalool, lavandulol, linalool acetate, and lavandulol acetate [
13,
14]. The increase of lavender flower extraction yield at a constant temperature (40 °C) was noticed when increasing the pressure from 10 to 30 MPa [
14]. Under these conditions, higher molecular weight components, such as oxygen-bearing molecules, e.g., coumarins [
13], polyphenols [
31], and non-volatile cuticular waxes [
32], may be co-extracted together with fragrance volatiles [
32,
33].
The compositions of lavender supercritical fluid extracts have been thoroughly studied in relation to extraction parameters and their impacts on the yields of advantageous flavor components [
13,
14,
34]. With the pressure increasing up to 30 Mpa at 40 °C, the CO
2 density also increases, which favors higher extraction yields and the apparent solubility of the most characteristic group of lavender odoriferous molecules—oxygenated monoterpenes, i.e., ester or alcohol components [
14]. Unavoidably, targeted fragrance molecules are co-extracted with waxy, resinous, and color matter; hence, the product is concrete-like (
Figure 1). The scCO
2 lavender extract is enriched with scent volatile compounds much more than the oil [
35] and, due to preservation of initial plant matter, it can be served as a valuable feedstock for further fractionation and valorization processes [
13].
The extract compositions from the L-Bg-E and L-Pl-E samples were determined by means of the GC–MS qualitative analysis. Results of GC–MS identification and relative amounts of components (%) are listed in
Table 1. Additionally, chromatograms corresponding to each extract are depicted in
Figure 2.
The GC–MS analysis revealed the presence of more than sixty different compounds in the extracts. (
Table 1,
Figure 2). Linalool and linalool acetate, the main representatives of lavender oxygenated monoterpenes, were the most abundant components; however, their percentages differed notably from one extract to another. The targeted compounds represented the lowest percentage composition in the case of the L-Bg-E sample with 4.56% and 12.78% for linalool and linalool acetate, respectively (
Table 1). The highest percentages of linalool; linalool acetate and lavandulol; lavandulol acetate and terpinen-4-ol were confirmed in the cases of L-Pl-E (17.02; 25.68; 1.40, 3.96 and 6.02%, respectively) (
Table 1). Lavandulol and its acetate are of a great interest in the cosmetic and perfumery industries, as it gives the oil a rosaceous, sharp floral aroma [
36,
37]; thus, their higher levels in natural oil can strongly increase its price. Together with linalool, linalyl acetate, terpinen-4-ol, and α-terpineol—the compounds were recently selected to study the influence of long-term storage on essential oil content and in the quality of two Czech
Lavandula angustifolia Mill. lavender varieties [
38]. The gradual lowering of the total content of essential oils (2.56% per year) was noticed; however, no statistically significant relationship between their initial compositions and the loss of the selected fragrance key markers in longer-term monitoring were found.
Linalool, one of the most important compounds for the perfume and flavor industries, occurs naturally in
Lavandula angustifolia Mill., as two isomers, almost exclusively predominating
®(-)-linalool (94.1%) [
39]. This enantiomer is a flowery-fresh woody odor, reminiscent of lavender, which, together with an array of its derivatives, is a common constituent of flower and honey extracts obtained from different lavender and citrus sources [
40]. Both L-Bg-E and L-Pl-E are characterized by the presence of 8-hydroxylinalool (2,6-dimethyl-2,7-octadiene-1,6-diol) resulting directly from the activity of P450 hydroxylase. Two isomers of 8-hydroxylinalool (
E-;
Z-) were found to be direct hydroxylation products of linalool, and together with α-terpineol, limonene, and nerolidol, were included in citrus honey, indicating their direct delivery from flower nectar [
41]. Other products of direct linalool hydroxylation were 8-oxolinalool and 6,7-epoxylinalool [
42]. In the lavender extracts studied,
cis- and
trans-linalool oxides (furanoid) and terpendiol isomers, i.e., 3,7-dimethyl-1,7-octadien-3,6-diol, were analyzed as products originating from further (consecutive) enzymatic reactions of 6,7-epoxylinalool (
Table 1,
Figure 3).
These linalool bioconversion products were previously detected in supercritical lavender extracts obtained under the pressure of 30 MPa and temperatures of 40–50 °C [
13]. Another component previously specified in plant and honey extracts of leatherwood (
Eucryphia lucida (Labill.)) and
Citrus spp. is 2,6-dimethyl-3,7-octadiene-2,6-diol (precursor of hotrienol) [
40], also determined in L-Bg-E and L-Pl-E at the level of 2.81 and 1.34%, respectively.
Eucalyptol (1,8-cineole) represents the oxygenated monoterpenes providing medicinal and olfactory properties of fine lavender essential oils (
Lavandulae aetheroleum) [
10]. The richest lavender oils, in terms of the content of such compounds as 1,8-cineole and camphor, are spike lavender oil and lavandin oil Grosso var. [
43]. The camphor-based products may be utilized for industrial cleaning products, detergents, or even natural and healthier alternatives to traditional solvents commonly used in fine art painting. However, in
Lavandulae aetheroleum, camphor content should not exceed 0.5% according to EU PDO specifications [
43]. High concentrations of camphor, as well as 1,8-cineole and terpinen-4-ol, can adversely affect the quality of lavender natural oils. However, the antimicrobial and antifungal bioactivities for linalool, 1,8-cineole, camphor, terpineol, and isomers of pinene, was proven [
44,
45].
Other valuable substances of industrial interest, which were found in scCO
2 lavender extracts, are sesquiterpenes, hydrocarbons, and their oxygenated forms, including alcohols. The components, i.e., β-caryophyllene, β-caryophyllene oxide, α-santalene, α-santalol, β-santalol, β-farnesene, α-bergamotene, nerolidol, α-bisabolol, tau-cadinol, ledene oxide-(II), were analyzed in both studied extracts (
Table 1). However, interpretation in the sesquiterpene region of the chromatogram may be imprecise due to structural similarities of eluted components. β-Caryophyllene oxide is a predominant sesquiterpenoid representative found in the studied extracts with percentages of 3.80% (L-Pl-E) and 4.54% (L-Bg-E). Both β-caryophyllene and its oxidation product β-caryophyllene oxide are used as cosmetics and food additives since they are approved as fragrance supplements by regulatory authorities, such as the FDA and EFSA [
46]. It was reported that SFE conditions enable high concentrations of caryophyllene oxide, linalool, and geranial at 60 °C and under pressure of 30 MPa [
47].
The studied lavender extracts contained other groups of oxygen-bearing components, i.e., lower aliphatic compounds: 1-octen-3-ol, octan-3-one, and characteristic lactones. The last group of lavender compounds includes coumarins and lavender lactone (5-ethenyldihydro-5-methyl-2(3H)-furanone), which is known for its interesting olfactory properties [
13]. Along with coumarin, its derivatives (i.e., herniarin) are recognized from their significant pharmacological properties, e.g., antitumor, anti-inflammatory, antibacterial, antifungal, diuretic, analgesic, and cardiovascular [
48]. Furthermore, natural coumarin is a desirable ingredient due to its sweet, distinguishing vanilla-like scent with grassy notes, applied either in masculine or feminine perfume compositions, and recognizable as a tonka bean fragrance [
44].
Lavandula angustifolia is another source of coumarin and herniarin, extracted from flowers, with the use of supercritical CO
2. The quantitative analysis of coumarin and its 7-methoxy derivative in L-Bg-E and L-Pl-E samples, as well as distilled fractions, was performed by supercritical fluid chromatography (SFC). The less soluble and non-volatile paraffins included in the cuticle layer of flower stems are easily co-extracted (“superficial washing”); hence, finally, a higher boiling fraction dilutes the fragrance molecules in this complex matrix [
32,
33]. According to the GC–MS qualitative analysis, a similarity between the chemical composition of the paraffin fraction found in L-Bg-E and L-Pl-E was found, with predominating hentriacontane (C31) and tritriacontane (C33) in both extracts (
Table 1).
2.2. Thermal Properties of scCO2 Extracts
Thermal characteristics of lavender extracts were determined by differential scanning calorimetry (DSC). The DSC thermograms recorded during the second heating of L-Bg-E and L-Pl-E samples are depicted in
Figure 4 and compared with melting profiles of commercial waxes: carnauba and beewax. Supercritical extracts of
L. angustifolia Mill. from two European cultivars, Polish and Bulgarian, have been scarcely studied and compared according to the compositions of their waxy fractions.
The DSC profiles of L-Bg-E and L-Pl-E showed two endotherms in each analyzed sample (
Figure 4). The melting points of L-Bg-E were at 58 °C and 81 °C, while that of L-Pl-E appeared to be at 52 °C as the most intensive peak and 75 °C at a much lower intensity. Both thermograms may suggest coexistence of two different waxy structures in the lipid mixtures of lavender scCO
2 extracts and, thus, their multicomponent natures. Lower melting fractions signified by the first peaks (58 °C and 52 °C) in the heat flow curves of L-Bg-E and L-Pl-E, respectively, were also found at DSC curves of beewax (55 °C) and carnauba wax (58 °C), (
Figure 4). In the case of a beewax heat flow curve, the peaks at 40–60 °C were attributed to the heat absorption of free fatty acids (13%) and hydrocarbons (13%) [
49]. Both groups of components were qualified in lavender scCO
2 extracts with predominating oleic and linoleic fatty acids (analyzed as methyl esters), and long-chain alkanes (C31–C33),
Table 1.
The DSC scans of both lavender extracts did not give endotherms at 60 °C to 70 °C, which could approve fatty acids esters. This fraction was abundant in beewax (72%) and led to the sharp absorbing peak at 60–70 °C [
49].
Carnauba wax is noticeably different than beeswax and its chemical composition and physical properties are discussed elsewhere [
50,
51,
52]. Carnauba wax contains major proportions of esters of hydroxy acids, i.e.,
p-methoxycinnamic diesters (PCO-C), which isolates as a pure fraction, showing up in the DSC curve as an endothermic peak at 73 °C with a shoulder at 66 °C [
53]. Similar structures could be responsible for some minor features in the second heat thermal analysis of L-Pl-E found at 75 °C (
Figure 3). However, further increase of melting properties was supposed to be affected by the addition of methoxy substituent on the phenol ring, and such behavior was observed for
p-coumaric acid esters, i.e., methyl ferulate and methyl sinapate, showing fusion points of 62 and 88 °C, respectively [
54]. This might support the similarity between the highest melting fractions of L-Bg-E (81 °C) and carnauba wax (83 °C) in terms of carboxylic acid constituents (
Figure 4).
The difference between L-Bg-E and L-Pl-E reflected in DSC thermograms at the most intensive melting points may indicate that the waxy structures of both
L. angustifolia raw materials depend on its geographic source. Since Bulgarian lavender cultivars are grown on the sunny slopes in a warm and dry climate, the compositions of their cutin layers, covering flowering stems and petals of flowers, give the film excellent barrier properties against UV radiation and uncontrolled water loss. The presence of a higher melting fraction depicted as a sharp peak at 81 °C (
Figure 4) may suggest an increased proportion of phenol constituents lipophilized through its esterification. The isolation of those bioavailable natural components with antiradical properties can widen diverse uses of lavender, acknowledged in the cosmetic and pharmaceutical industries.
2.3. Application of Molecular Distillation in Lavender scCO2 Extract Fractionation
Molecular distillation is a type of vacuum distillation that separates components of a mixture by their difference in volatility. A distillation unit designed for short-path operations (molecular distillation) is made up of evaporator and internal spiral condensers placed in its center. The distance from the evaporation surface to the condensation surface does not exceed the mean free path of molecules, which means that, once evaporated, they reach the condensing surface without delay. Additionally, due to specific working conditions provided in the system by a vacuum pump set (0.1–100 Pa), the relative volatility of compounds increases, which allows separation of mixture components at lower temperature. For the optimal product distribution, the feed product is pumped on top of a rotating wiper basket plate, before it is mixed to a thin film by a wiper system. The most important for the system is that it ensures a short residence time of complex mixtures in the distillation unit and largely prevents thermal decomposition of their components, divided during the process into distillate and/or residue streams.
The objective of the present study was to evaluate the potential of a selected approach, to incorporate fractionation under vacuum into the refinement of scCO
2 produced lavender extracts. Molecular distillation as a gentle technique of physical refining was applied to separate essential oil components from heavy components co-extracted with scCO
2 on a pilot scale. Hence, it was feasible to recover higher yields of precious oxygenated monoterpenes from scCO
2 extracts enriched under the selected conditions (30 MPa, 40 °C). The SFE-MD strategy was already found to be more efficient in a separation of essential oils from artemisia argyi Lévl. Et Vant, compared to hydrodistillation described as time- and energy-consuming and destructive for thermolabile compounds (high water abundance and high process temperatures leading to hydrolysis) [
28]. What is more important, the latter mentioned disadvantage is crucial to overcome especially in relation to recovery of linalyl and lavandulyl acetates. Caryophyllene oxide is one of the highest boiling constituents identified in the lavender essential oil analyzed with GC–MS [
55]. However, the boiling point of lavender essential oil (204 °C) as a mixture is a function of the vapor pressures of its major and minor volatile constituents. Coumarin and herniarin are representatives of lactones, which, similar to monoterpenoids, are abundant components of scCO
2 extracts obtained under increased extractant density [
13]. Thus, the highest boiling components from the studied group of volatiles can also feed the distillate stream, depending on the applied molecular distillation conditions.
The essential oils included in the scCO
2 lavender extracts (
Figure 1) are diluted with higher boiling components with a waxy character, causing congelation at room temperature. The studied feedstocks differed slightly in densities between L-Bg-E and L-Pl-E (0.98 g/mL and 0.93 g/mL, respectively) and in composition/proportion to a higher melting fraction, as confirmed by the DSC curve of L-Bg-E (81 °C). Those features of lavender extracts may further impact the performance of thin film evaporator. The boiling points of the studied volatiles are listed in
Table 2.
For the purpose of the experiment, fractionation was performed under the pressure down to 1 Pa and the temperature below 100 °C, to assure mild processing conditions for major constituents of lavender aroma. The rest of the crucial parameters were checked and adjusted individually to each extract. The experimental conditions of molecular distillation processes performed on L-Bg-E and L-Pl-E feedstocks are listed in
Table 3.
Accordingly, as can be seen in
Table 3, the temperatures of the feed tank (FT) were different in order to keep L-Bg-E and L-Pl-E in a liquid form at 50 °C and 45 °C, respectively. The values of the evaporator temperatures (EVT) and condenser temperature (CTs) were chosen individually to each of the studied lavender scCO
2 extracts, based on the first distillate drops falling down the condenser. Since there was no oil distilled out at 50 °C and below that temperature, 55 °C was chosen as the initial EVT for both extracts, and five experiments with evaporator temperatures ranging from 55 °C to 95 °C, with an interval of 10 °C, were performed with a constant feed flow (FF) rate of 0.833 mL/min on L-Bg-E and L-Pl-E (
Table 3). The CT temperature was kept constant at 10 °C during the entire distillation experiment of the L-Pl-E extract, while it was varied from −5 °C to 6 °C in distillations D1–D5 in the case of L-Bg-E fractionation. Additionally, the residue discharge temperature (RdT) was raised in a stepwise fashion from 55 °C to 70 °C in D1–D5 distillations of L-Bg-E to keep fluidity of the residue stream.
2.4. The Effect of EVT on Distillate Enrichment with Key Lavender Fragrance Molecules
The lavender extracts were submitted to MD experiments, which were performed with evaporator temperatures (EVT) ranging from 55 °C to 95 °C (five experiments per extract), at a constant pressure (1 Pa) and under a constant feed flow rate (FF) of 0.833 mL/min (
Table 4). The results were analyzed in terms of contents of targeted oxygenates (mg/g) in the obtained distillates and the ratio of distillate stream mass to residue stream mass (D/R) obtained at every single step of the MD processes (D1–D5).
According to Tovar et al. [
20], two parameters were important for effective separation of citral from lemongrass essential oil, i.e., EVT and FF. The last parameter, FF, was kept the same during both MD processes, while EVT was increased stepwise from 55 to 95 °C (
Table 3). However, processing of L-Bg-E appeared to be more difficult in terms of a mass and heat transfer and required adjusting more parameters at the same time, compared to MD of L-Pl-E (
Table 3). Different physicochemical properties of L-Bg-E, and its “congelation” over the distillation time, required maintaining a higher wiper basket speed of 350 rpm. Constant mixing of the falling film caused by the roller wiper action provided the heat transfer into deeper layers of the extracts more abundant with volatile molecules, and favored their concentrations in the vapor phase. The faster wiper basket movement (more rotations per minute) and increased evaporation efficiency resulted in intensified condensation. Hence, in order to ensure concentrations of the desired lavender fragrance components in the distillate stream, and to minimize material loss in the vapor phase and condense in the cold trap, the CT temperature was adjusted (changed from −5 °C to 6 °C in distillations D1–D5) at each stage of the L-Bg-E processing (
Table 3). Low amounts of material condensed in the trap (lower than 8%) of the vacuum system previously found an indicator of the effective separation [
20]. The yields of water and light oil (%wt
CT), collected after each stage of MD of both lavender extracts did not exceed, in total, 8% (
Table 4). However, the loss of light oil increased with the increase of the EVT, to the maximum level at 95 °C.
Additionally, with the highest applied EVT favoring highest recovery of the lower boiling fraction, the viscosity of a residual part increased and, consequently, the residue discharge temperature was increased up to 70 °C (
Table 3). Further processing of L-Bg-E under selected conditions with higher EVT exceeding 95 °C can be unworkable because of the residue viscosity.
Results show that, with increasing EVTs, the percentage weight of the distillate stream (%wt
D) also increased (
Table 4). The highest values of %wt
D and, thus, the highest split ratio (D/R), were obtained for experiments D4 and D5, with EVTs of 85 °C and 95 °C, respectively. The highest was also the percentage weight of the cold trap fraction (%wt
CT), referring to the content of water and light oil collected in the direct cooled cold trap (−80 °C). However, removal of the light oil fraction from L-Pl-E proceeding with increasing EVTs (55–95 °C) did not affect the viscosity of residuals as much as in the case of L-Bg-E. Hence, the RdT was quite low and remained unchanged over processing of L-Pl-E up to an EVT of 95 °C. A further slight increase in EVT might cause a further increase in %wt
D and enhance recovery of oxygenated compounds in the stream of distillates. In the example of citral, Tovar et al. [
20] confirmed that with the highest applied EVT (60–120 °C) and the highest feed flow rate (1.5–4.5 mL/min), the concentration of this compound in the distillate stream doubled to 40.963 mg/mL compared to the initial concentration, proving the high product quality. However, according to Li et al. [
27], the EVT exceeding 120 °C was found deleterious for scCO
2 extract of
Artemisia annua separated with MD under a similar vacuum (1.67 Pa). The combination of SFE and MD purification methods below 120 °C allowed producing high-quality essential oils, mainly composed of limonene, (
1S,5S)-α-pinene, β-pinene, β-farnesene, α-caryophyllene, and γ-elemene, exhibiting antimicrobial and antioxidant activities [
27].
The contents of target oxygenates quantified in lavender distillates obtained in the applied EVT range (55–95 °C) are listed in
Table 5. The increase of EVT was crucial for the increase in contents of oxygenated monoterpenes and caryophyllene oxide in the distillate streams of both processed lavender extracts. Once the EVT increased up to 85 °C, the contents of 1,8-cineole, linalool, linalyl acetate, terpinen-4-ol, lavandulyl acetate, lavandulol, and caryophyllene oxide increased 2.0–2.4 times (L-Pl-D4) and 2.0–2.2 times (L-Bg-D4) in relation to the crude extracts (
Table 5).
Linalool, lavandulyl acetate, and linalyl acetate, the most abundant oxygenated ingredients of L-Bg-E and L-Pl-E, differed in contents between both feedstocks. They were quantified at almost twice lower levels in L-Bg-E than in L-Pl-E (
Table 5). The same applied to terpinen-4-ol, lavandulol, and caryophyllene oxide. In the case of L-Pl-E, the content of these three oxygenated compounds quantified by GC–FID were 59.16, 43.61, and 89.53 mg/g, respectively. The contents of linalool, lavandulyl acetate, and linalyl acetate gradually increased in the distillates across stages with a maximum in L-Pl-D4 (131.79, 92.53 and 185.73 mg/g, respectively) (
Table 5). However, the increase of EVT from 85 °C to 95 °C caused a slight decrease in contents of those monoterpenoids (except from coumarins and caryophyllene oxide) compared to distillates L-Pl D4. The contents of less abundant monoterpenoids in L-Pl-E: 1,8-cineole (1.65 mg/g), terpinen-4-ol (8.96 mg/g), lavandulol (14.40 mg/g), and caryophyllene oxide (15.88 mg/g) also increased twice after MD at an EVT of 85 °C. A further increase in EVT by 10 °C caused a slight increase in those contents.
The content of 1,8-cineole was similar in both lavender extracts and was concentrated to the same extent (2.3-times for L-Bg-E and two-times in L-Pl-E) upon increasing the EVT to 85 °C. According to
Table 5, the molecular distillation of L-Bg-E under the highest applied EVT of 95 °C generally caused a greater decrease in contents of monoterpenoids and caryophyllene oxide in the distillate compared to L-Pl-E. This might be an effect of dilution since the second heating DSC graph of the L-Bg-D5 distillate revealed a solid–liquid melting transition at 115 °C (
Figure 5). The co-distilled higher-boiling components caused a slight cloudiness of L-Bg-D5 chilled to −4 °C.
The molecular distillation of L-Pl-E under the highest applied EVT temperature of 95 °C neither caused a significant change in D/R nor in the contents of the oxygenated monoterpenes compared to the process performed at 85 °C (
Table 5). At the same time, it was noticed that the amount of the cold trap fractions (%wt
CT) collected in a glass cylinder covering a steely cool finger (−80 °C) increased remarkably after the distillation process at 95 °C (experiment D5) of L-Pl-E, yielding, in total, 7.95%. The essential oil condensed with water and then separated as a lighter layer was the predominating cold trap material (4.53%). Thus, the use of EVT = 95 °C only slightly influenced the fraction L-Pl-D5, compared with L-Pl-D4. At the same time, lighter fragrance components were intensified in the vapor phase as the loss and were collected in the cold trap.
Coumarin and herniarin were the less volatile components amongst the analyzed oxygenated molecules (
Table 3) co-distilled in some parts with oxygenated monoterpenes. According to the SFC analysis, it was confirmed that the yield of coumarin was higher than herniarin in both studied extracts (L-Pl-E and L-Bg-E). However, similar to the content of monoterpenoids, the content of both coumarins was almost twice higher in L-Pl-E compared to L-Bg-E (25.96 vs. 12.30 mg/g). Although both components were detected in distillates D1–D5 obtained from both feedstocks, their contents under processing conditions were lower than in the initial extract samples.
The other co-distilled components affecting the quality of oxygenated monoterpenes were pigments, which influenced the yellowish color of the distillates.
Figure 6 depicts distillates (L-Pl-D4 and L-Bg-D4) obtained at EVT of 85 °C, light oil fractions collected after the process (L-Pl-O and L-Bg-O), and yellow–green residue fractions (L-Pl-R4 and L-Bg-R4, EVT = 85 °C) obtained in the process based on parameters listed in
Table 4. The distillates and colorless light oil fractions were used in the following tests of quality and antimicrobial activities.
2.5. Quality Evaluation of Distillates and Light Oil Fractions
A non-targeted fingerprint HATR-FTIR analysis was performed to evaluate aromatic lavender concentrated fractions. Recently, vibrational spectroscopy methods, including mid- (MIR) and near-infrared (NIR), combined with chemometric data analyses, were used to confirm the identity and quality of lavender essential oils for commercial purposes [
56,
57]. Mid-infrared spectroscopy, primarily used qualitatively, provides structural characterization according to the functional group vibration and fingerprint region, which are crucial for molecular identification. The GC–MS is another technique, an alternative one, used for the quality assessment of concentrated fractions, i.e., distillates and light oils (cold trap fractions). The HATR-FTIR mid-infrared spectra of lavender distillates (L-Pl-D4 and L-Bg-D4) and light oil fractions (L-Pl-O and L-Bg-O) obtained from studied lavender extracts; L-Pl-E and L-Bg-E are depicted in
Figure 7 and
Figure 8, respectively.
The application of HATR-FTIR enabled the analysis of selected fractions in the form of ultra-thin films with no additional pre-treatment. HATR-FTIR absorption spectra showed characteristic bands identified previously in lavender essential oils [
58,
59]. Since major components of lavender oils are linalyl acetate and linalool, FTIR spectra are dominated by vibrational modes from those monoterpenoids. The carbonyl groups (C=O) present in linalyl acetate and lavandulyl acetate were characterized by peaks at ca. 1735 cm
−1. The corresponding band in L-Pl-D4 had a small shoulder with a maximum at 1680 cm
−1, indicating the formation of a hydrogen bond between C=O and –OH groups [
57]. The area between 1100 and 1300 cm
−1 included absorptions, representative for C-O stretching vibrations; those were documented at around 1239 cm
−1, 1171 cm
−1, 1112 cm
−1, and 1110 cm
−1, with some shifts in the 1176–1109 cm
−1 region (
Figure 7 and
Figure 8). The other characteristic vibrational frequencies for linalool and linalyl acetate were associated with the vinyl group vibration (−C=CH
2); however, the intensity of a related peak around 1646 cm
−1 is very weak. Additionally, the −C=CH
2 in-plane deformation vibration can be found at 1416 cm
−1 as a weak band (
Figure 7 and
Figure 8). The feature characteristic of the O-H stretching vibration of alcohol functional groups present in linalool, its derivatives: 8-hydroxylinalool, furanoid linalool oxides, 3,7-dimethyl-1,5-octadiene-3,7-diol (terpendiol I), and lavandulol, could be found as a broadband, in the region of 3400–3500 cm
−1. Those bands were found to be some of the strongest affecting the principal components in the
Lamiaceae family group essential oils on the basis of ATR-FTIR and PCA analysis [
57]. In the spectra of lavender distillates, L-Pl-D4 and L-Bg-D4 (
Figure 7 and
Figure 8), the signal of the O-H stretching mode was broad and more distinct compared to the spectra of corresponding light oils, which indicates hydrogen-bonding interactions between molecules. The differences in band positions, shapes, and intensities were also found in the area between 920 and 1240 cm
−1 when comparing the spectra of lavender distillates and corresponding light oil fractions. The bonds, which have absorptions in the mentioned part of the fingerprint region, are those assigned to the C-O stretching vibrations, O-C-O from primary alcohols (i.e., lavandulol) and =C-H below 1000 cm
−1. The bands at approximately 2800–3200 cm
−1 may be related to C-H stretching and C=C-C ring vibrations, both documented as absorbing around 2950 cm
−1 [
56]. Since the last mentioned vibrations are attributed to molecular fragments of linalyl acetate, linalool, and its derivatives (
Figure 3), which are major constituents of the obtained lavender fractions, they would not impart significant changes in the FTIR spectra.
The GC–MS semi-quantitative characterization results of distillates L-Pl-D4 and L-Bg-D4, and the corresponding light oil fractions L-PL-O and L-Bg-O, which were collected in the cold trap, are reported in
Table 6. These multicomponent mixtures are characterized by a few major compounds in higher contents (20–70%). The major monoterpenoids of lavender fractions were linalool and linalyl acetate; however, there was a significant difference between the contents of each component when comparing fractions obtained from molecular distillations of L-Pl-E and L-Bg-E. The lowest linalool percentages were confirmed in L-Bg-D4 (7.58%) and L-Bg-O (17.44%), and, at the same time, the furanoid linalool oxides appeared in the highest content. The other late-eluting linalool derivatives/bioconversion products: 2,6-dimethyl-3,7-octadiene-2,6-diol, 2,6-dimethyl-1,7-octadiene-3,6-diol, 8-hydroxylinalool, 6,7-epoxylinalool, 3,7-dimethyl-1,7-octadiene-3,6-diol (terpendiol II) were found in higher amounts in L-Bg-D4 than in L-Pl-D4 (
Figure 3,
Table 6). Linalyl acetate was the most abundant component of the studied extracts, distillates, and light oil fractions; however, its percentage in distillate L-Bg-D4 was notably lower compared to L-Pl-D4 (
Table 6). Since, in both distillates, physical refining caused a removal of heavy compounds, i.e., fatty acids and waxes, which diluted essential oil components in feedstocks, the late-eluting components remaining in L-Bg-D4 and L-Pl-D4 were coumarins, sesquiterpenes, and their derivatives (
Figure 9). The identification of individual components using GC–MS and library spectra might not be reliable, particularly in the sesquiterpene region of the chromatogram where there were eluted, structurally-related components [
60]. Some of the components eluted between 30 and 55 min (
Figure 9) remained uncharacterized (
Table 6). Nevertheless, according to a quantitative analysis by GC–FID, employing an external standard technique, caryophyllene oxide was quantified in both distillates, L-Pl-D4 and L-Bg-D4, at different levels, 34.75 and 13.80 mg/g (
Table 6), respectively. Caryophyllene (and its oxidation product) was one of the few sesquiterpenes assigned in the light oil fractions (
Table 6) using GC–MS.
2.6. Antimicrobial Activity
This is the first report to study MD fractions in terms of antimicrobial activities. As presented in
Table 7, the distillates (L-Pl-D4, L-Bg-D4) and the related cold trap light oil fractions (L-Pl-O and L-Bg-O) isolated from two lavender scCO
2 extracts showed antibacterial activity. Gram-positive bacteria were more susceptible to both distillates and light oils according to the minimum inhibitory concentration (MIC
) (MIC = 0.5–4 mg/mL) when compared with
E. coli, being Gram-negative bacterium (MIC = 8 mg/mL). The yeast strains were more susceptible to the distillates and light oils (MIC = 0.5–1 mg/mL) than the most bacterial strains. There were some differences between the antimicrobial activities of distillates (L-Pl-D4 and L-Bg-D4) and the corresponding light oil fractions. Predoi et al. [
61] reported MIC and MBC values of <0.1 for essential oil hydrodistilled from
L. angustifolia. However, the studied
L. angustifolia hydro-distillated from southern Romania was characterized by a higher content of linalool (47.75%) when compared with L-Pl-D4 (19.79%) and L-Bg-D4 (7.58%).
Subsequently, the activity of the scCO2 crude extracts from lavender (L-Pl-E and L-Bg-E) against bacteria and yeasts was also examined. Gram-positive bacteria were found to be more-or-less susceptible to both extracts as compared to the distillates and light oils, depending on the bacterial species. The highest activity of both extracts was observed against Bacillus subtilis ATCC 6633 with MIC = 0.25 mg/mL (L-Pl-E) and MIC = 0.06 mg/mL (L-Bg-E). Candida spp. showed lower susceptibility to both extracts (MIC = 2–4 mg/mL) than to distillates and light oils.
It was found that all studied fractions (distillates and light oils) possessed bactericidal and fungicidal effects, confirmed by MBC minimum bactericidal concentration/MIC = 1–4 and minimum fungicidal concentration (MFC)/MIC = 1–2. It is generally accepted that antimicrobials are usually regarded as bactericidal or fungicidal if the MBC/MIC or MFC/MIC ratio is ≤4 [
62]. However, it should be noted that the crude lavender extracts had bactericidal (MBC/MIC = 1–4) or bacteriostatic effects (MBC/MIC = 4–16), depending on the bacteria species, while both extracts exerted fungicidal effects (MFC/MIC = 1–2).
The reference substances, such as coumarin, herniarin (7-methoxycoumarin), linalool, linalyl acetate, caryophyllene oxide, lavandulol, and lavandulyl acetate were also used for antibacterial and antifungal activity tests (
Table 8). Generally, lavandulol and lavandulyl acetate were mostly active against Gram-positive bacteria, with yeasts MIC ≤ 0.5 mg/mL showing the widest spectrum of activity.
B. subtilis ATCC 6633 and
M. luteus ATCC 10240 were found susceptible to all compounds included. They had bactericidal (MBC/MIC = 1–4) or bacteriostatic effects (MBC/MIC = 8–128), depending on the bacteria species, while all exerted fungicidal effects (MFC/MIC = 1–4).