Application and exploration of fast gas chromatography-surface acoustic wave sensor to the analysis of thymus species.

Fast gas chromatography combined with surface acoustic wave sensor (GC/SAW) has been applied for the detection of volatile aroma compounds emanated from thymus medicinal plants such as T. quinquecostotus (Jeju and Mt. Gaya in South Korea), T. quinquecostotus var. japonica (Ulreung island in South Korea), T. mongolicus (Northeastern Asia), and T. serpyllum (Europe). The GC/SAW involving the fragrance pattern analysis provides a novel analytical method with a very fast separation and characterization of aromas caused by the delicate difference of chemical composition according to botanical and geographical origin. On the comparison of experiments, the characteristic components and analytical tendency for air-dried thymus species detected by GC/SAW appear to be quite similar to those obtained by headspace solid-phase microextraction (HS-SPME)-GC-MS, but the abundance ratios between these two methods are different. In addition to that, the discrimination of various thymus species by using VaporPrint image based on GC/SAW provides a quite reliable result. On the basis of principal component analysis (PCA) results, the ability for classification among species of completely different chemotypes by HS-SPME-GC-MS is good enough, but the classification of same chemotypes species which are from different geographical origin in same country, original species and its variety, an air-drying term for 13 days and 16 months appear much lower than GC/SAW. Interestingly, the present experiment reveals that the air-drying term influences the aroma composition: the concentration of the pharmacologically active species, monoterpene phenol (thymol), reaches its highest concentrations after it was dried for 5 days or 13 days, which is much higher than in fresh or over-dried for a long times.


Data analysis for pattern recognition
Data transformation for pattern recognition was performed using MS Excel. Fifteen components were chosen based on the corresponding GC/SAW and HS-SPME-GC-MS profile. Especially, fifteen components such as -pinene, camphene, -myrcene, p-cymene, -terpinene, terpinolene, cis-sabinene hydrate, camphor, borneol, -terpineol, thymol methyl ether, thymoquinone, thymol, -caryophyllene and -bisabolene were found as characteristic components in GC/SAW and HS-SPME-GC-MS profile. And then the response of each peak was applied from the triplicate measurements. Finally, pattern recognition techniques have been used for the discrimination of the materials. Principal component analysis (PCA) is a pattern recognition technique and statistical analysis. PCA was also carried out using MVSP. 3.1 version (Kovach Computing Service, Anglesey, Wales) in order to classify thymus species.

Identification of volatile herbal aroma compounds for air-dried for 13 days of thymus species by GC/SAW
By using fast GC/SAW, volatile herbal aroma profiles for thymus species were obtained. The materials sequentially exit from the column and they land and stick on the SAW sensor. When an analyte adsorbs on the surface of the sensor, the frequency of SAW sensor is altered, which affects the detection signal in direct proportion to the amount of condensate. Fig. 1(A)-(E) shows chromatograms of volatile aroma compounds for air-dried for 13 days of thymus species. The area of each peak is correlated to its concentration and is expressed in frequency counts (Cts). The identification of each aroma compounds shown by GC/SAW was carried out by comparison with authentic standards and GC-MS analysis and their relative proportions (% total amounts) are summarized in Table 1. The herbal aroma components of thymus species consist mostly of monoterpene hydocarbons ( -pinene, camphene, -myrcene, and terpinolene), oxygenated monoterpenes (cis-sabinene hydrate, camphor, borneol, -terpineol, and thymoquinone), a monoterpene phenol (thymol), monoterpene phenol precursors (p-cymene and -terpinene), a monoterpene phenol derivative (thymol methyl ether) and sesquiterpenes ( -caryophyllene and -bisabolene). The grouping of compounds has an important meaning as responsible for the characteristic aroma of thymus. Fifteen compounds were identified, especially in T. quinquecostotus species, which are from Jeju and Mt Gaya in South Korea, the characteristic, distinctive components such as p-cymene (26.4%, 24.2%), -terpinene (10.3%, 10.5%), and active thymol (29.0%, 33.1%) were constituted 65.7%, 67.8% of the total amounts, respectively. In addition, oxygenated monoterpenes such as cis-sabinene hydrate (0.6%, 0.7%), camphor (6.6%, -), borneol (4.7%, 5.7%) and thymoquinone (11.6%, 12.5%) were constituted 23.5% and 18.9% of the total amounts as the secondly most abundance, respectively. Lesser amounts of monoterpene hydrocarbons: -pinene (0.9%, 0.8%), camphene (0.9%, -), -myrcene (5.3%, 3.3%), terpinolene (1.2%, 2.4%) were constituted 8.3% and 6.5% of the total amounts, respectively, and -caryophyllene (2.5%, 6.8%) as sesquiterpene was also found. T. quinquecostotus var. japonica species is a variety of T. quinquecostotus and its geographical origin is Ulreung island in South Korea. p-Cymene (15.8%), -terpinene (9.0%), and thymol (33.5%) were also found as characteristic components and constituted 58.3% of the total amounts. Also, oxygenated monoterpenes (32.3%), monoterpene hydrocarbons (5.5%), and sesquiterpenes (3.9%) were found.  Generally, thymol, phenolic monoterpene, defines the essential oil quality because of its active pharmacological properties. It was reported that -terpinene and p-cymene are the precursors of phenolic monoterpenes such as thymol and carvacrol in T. vulgaris by Granger and Passet [22]. It was also proposed that -terpinene assumes an important role in the aromatization whose product is p-cymene. They reported that p-cymene is the precursor of thymol by hydroxylation in T. vulgaris L. plant [23]. Moreover, the concentration of pcymene and -terpinene are found to vary in coincidence with the variation in their corresponding phenolic monoterpene products [24] and -terpinene decreases its concentration and p-cymene increases in the essential oil [25]. In T. mongolicus species (the origin of Northeastern Asia), active phenolic monoterpene (thymol) and its corresponding monoterpene hydrocarbon precursor ( -terpinene) were not found, whereas oxygenated monoterpenes, such as borneol (21.0%), -terpineol (41.8%), were identified as characteristic predominant compounds and constituted 62.8% of the total amounts. T. serpyllum species (the origin of Europe) contain almost the same chemical components with those of T. quinquecostotus. Especially, a monoterpene phenol derivative (thymol methyl ether: 12.7%) was only found as characteristic component. It was tentatively identified by comparison of its GC-MS data, because its reference standard is not commercially available.

Discrimination of thymus species using VaporPrint image based on GC/SAW
By comparing the relative contents, chemical composition for characteristic components of air-dried for 13 days of T. quinquecostotus species which are originated from Jeju and Mt. Gaya in South Korea were almost same: the active monoterpene phenol (thymol: 29.0%, 33.1%), its corresponding precursors (p-cymene: 26.4%, 24.2%, -terpinene: 10.3%, 10.5%) and oxygenated monoterpenes (borneol: 4.7%, 5.7%, thymoquinone: 11.6%, 12.5%). It is interesting to note that the components found in the same species of different geographical origin in same country are almost same. Also, the proportion of such components is same. The GC/SAW provides a visually recognizable fragrance pattern (VaporPrint image) derived from the frequency of SAW sensor. This image is created by transforming the time variable to a radial angle with the beginning and end of the analysis. This image transfers the olfactory response to a visual response [18]. These fragrance images are a useful for comparing delicate differences of various thymus species for species identification. The fragrance patterns for air-dried for 13 days of thymus species are shown in Fig. 1(A')-(E'). As shown in Fig. 1(A') and (B'), these same species present almost same fragrance patterns. In its variety species (the origin of Ulreung island in South Korea), the most components are similar in chemical compositions between original and its variety, specific components which are p-cymene (original species: Jeju 26.4%; Mt. Gaya 24.2%, its variety species: 15.8%), borneol (original species: Jeju 4.7%; Mt. Gaya 5.7%, its variety species: 18.4%) were shown a different compositions. Also, -bisabolene (2.4%) was found additionally in variety species. Therefore, such differences seem to result in a small differences in fragrance pattern ( , respectively. Therefore, distinctive differences seem to result in a substantial differences in fragrance pattern ( Fig. 1(E')). As a result, it turned out that each species has own characteristic fragrance pattern owing to its own chemical compositions and its own characteristic fragrance patterns are conducive to species identification.

Comparison of GC/SAW and HS-SPME-GC-MS method for air-dried for 13 days of thymus species
The compositions of aroma compounds of thymus species extracted by HS-SPME using CAR/DVB/PDMS fiber and then analyzed by GC-MS are presented in Table 2. Their GC-MS total ion chromatograms are shown in Fig. 2. The alphabetic numbers of peaks shown in Fig. 2 correspond to the numbers indicated in GC/SAW chromatograms (Fig. 1). These results are comparable to those given by GC/SAW, including the characteristic components and chemical composition. Fifty-four compounds were detected by HS-SPME-GC-MS. As shown in Table 3, characteristic components and analytical tendency for air-dried for 13 days of thymus species detected by GC/SAW and HS-SPME-GC-MS are similar, but the abundance ratios between these two methods are different. In T. quinquecostotus species, monoterpene phenols (thymol, carvacrol) and monoterpene phenols precursors (p-cymene, -terpinene) were also dominant (76.0%) like GC/SAW (66.8%). But, the amounts of total monoterpene phenols are 11.7% by HS-SPME-GC-MS, 31.1% by GC/SAW and monoterpene phenols precursors are 64.3% by HS-SPME-GC-MS, 35.7% by GC/SAW. In comparison of T. quinquecostotus var. japonica species and its original species, these results show similar analytical tendency with GC/SAW. For instance, p-cymene decreases from 46.3% to 22.2% by HS-SPME-GC-MS and from 25.3% to 15.8% by GC/SAW (an average value is indicated in bold) ( Table 1 and 2). While oxygenated monoterpene increases from 11.3% to 17.6% by HS-SPME-GC-MS and from 21.2% to 32.3% by GC/SAW. In T. mongolicus species, the characteristic and dominant components are borneol and -terpineol in similar to the results by GC/SAW. But, the amounts of borneol and -terpineol are 8.0% and 31.0% by HS-SPME-GC-MS; 21.0% and 41.8% by GC/SAW, respectively (Table 1 and 2). Interestingly, the significant amount of thymol methyl ether and carvacrol methyl ether which are not almost found in other species were found in T. serpyllum species. The amounts are 20.5% by HS-SPME-GC-MS and 12.7% by GC/SAW.

Compositions of volatile herbal aroma compounds of thymus species of elapsed for 16 months by GC/SAW
The compositions of volatile herbal aroma compounds of thymus species of elapsed for 16 months extracted by GC/SAW are presented in Table 1. As a result, according to elapse for 16 months, there are slightly differences in the relative quantities of their characteristic constituents than those of 13 days: in T. quinquecostotus species and its variety, active thymol precursor p-cymene increases from 25.3% to 29.6% and from 15.8% to 25.1%, respectively, whereas active thymol decreases from 31.1% to 26.7%, 33.5% to 24.1%, respectively (an average value is indicated in bold). However, for T. serpyllum species, there appears distinctive differences compared to those dried for 13 days: p-cymene also increases from 15.8% to 30.4% and thymol substantially decreases from 34.1% to 1.1%, while thymol precursor -terpinene is not found and sesquiterpene -bisabolene increases 2 times. In T. mongolicus species, -terpineol as the characteristic and dominant component also slightly decreases from 41.8% to 35.9%. Fig. 3(A) shows principal component analysis (PCA) for GC/SAW responses of air-dried for 13 days and 16 months of thymus species. PCA were carried out using MVSP. 3.1 version in order to classify the thymus species. As a result, a good classification among four different species: T. quinquecostotus, T. quinquecostotus var. japonica, T. mongolicus, and T. serpyllum was obtained. Additionally, the classification of same species which are from different geographical origin in same country, classification of original species and its variety for T. quinquecostotus, classification as an air-drying ter m f o r 1 3 d a y s a n d 1 6 m o n t h s f o r T. quinquecostotus showed good results. In these results, it turned out that discrimination of various thymus species by using VaporPrint image based on GC/SAW are very clear, which was reproducible data points shown in Fig. 3 Table 3. Comparison of composition of aroma compounds for air-dried for 13 days of thymus species by HS-SPME-GC-MS and GC/SAW Fig. 3(B), (B') shows PCA for HS-SPME-GC-MS responses of air-dried for 13 days and 16 months of thymus species. As a result, a good classification among species of completely different chemotypes was obtained. But, the classification of T. quinquecostotus which are same species of different geographical origin in same country (from Jeju and Mt. Gaya in South Korea), classification of original species and its variety for T. quinquecostotus, grouped as an air-drying term 13 days and 16 months for T. quinquecostotus appeared to be quite lower than those achieved by GC/SAW.  Fig. 4 shows chromatograms of volatile aroma compounds according to elapse for T. quinquecostotus var. japonica by GC/SAW. The changes of composition of their volatile aroma compounds are summarized in Table 4.  In the raw sample which are prepared as soon as possible after it was picked, monoterpene phenol (thymol: 20.4%) is lower than its corresponding precursor (p-cymene, -terpinene: 30.7%). Whereas, in the case of two samples which are elapsed for 5 days at 5 °C and airdried for 13 days at room temperature, active thymol (33.5%~37.4%) is higher than its corresponding precursor (23.8%~24.8%). Additionally, thymoquinone (13.5%), borneol (18.4%) in 13 days sample increased than those of 5 days sample. After elasping for 16 months, the concentration of active thymol (24.1%) and borneol (9.6%) considerably decreased, whereas thymoquinone (18.4%) increased in high intensity. As these results, they also show some interesting characteristic feature for the influence of air-drying on the volatile aroma compositions. In the aspect of pharmacological effects, it was found that active monoterpene phenol (thymol) reaches its highest concentrations after it was dried for 5 days or 13 days, which is much higher than in fresh or over-dried for a long times.

Changes of composition of aroma compounds according to elapse for T. quinquecostotus var. japonica by GC/SAW
quantitatively. It also enables to discriminate the botanical and geographical origin of thymus species by recognizable fragrance pattern analysis. As a result, a new methodology by GC/SAW can serve as an alternative analytical technique for the analysis of discrimination of thymus species that provides second unit analysis, simple, highly sensitive analytical method, and fragrance pattern recognition compared to the conventional HS-SPME-GC-MS technique. In addition, it would be a first report to deal with the volatile herbal actual compositions and discrimination of thymus species by GC/SAW and will be applied to a variety of applications.