Discrimination and characterization of the volatile organic compounds of Acori tatarinowii rhizoma based on headspace-gas chromatography-ion mobility spectrometry and headspace solid phase microextraction-gas chromatography-mass spectrometry

2.1 Materials and reagents and samples

N-ketone C4-C9 standard mix (Sinopharm Chemical Reagent Beijing Co., ltd., China) was used as an external reference to calculate the retention index (RI) of the GC-IMS compound. N-alkane C8-C20 standard for GC–MS was purchased from Sigma-Aldrich Chemical Co., ltd. (Missouri, USA). Ethyl acetate (Fisher, Fair lawn, NJ, USA), Some other standard chemicals: methyl eugenol (CAS: 93–15-2, 95 %), β-asarone (CAS: 5273–86-9, 98 %), asarone (CAS: 2883–98-9, 98 %) were provided by NatureStandard (Shanghai, China), esdragole (CAS: 140–67-0, 98 %), γ-asarone (CAS: 5353–15-1, 98 %) were provided by Shanghai Yuanye Bio-technology Co., ltd. (Shanghai, China).

15 batches of ATR from 14 producing areas were collected, Anqing (ATR1, ATR4), Qianshan (ATR2), Huangshan (ATR3) and Dabieshan (ATR5) all in Anhui, Jiujiang (ATR6), Nanchang (ATR7), Shangrao (ATR8), Jingdezhen (ATR9) and Guixi (ATR10) all in Jiangxi, China; Guangyuan (ATR11), Kaijiang (ATR12), Yaan (ATR13) and Nanchong (ATR14) all in Sichuan, China; and ATR15 from Chongqing, China. Herbs purchased from Bozhou Chinese herbal market, Anhui, China. Sample details were available at Table S1. All herbs were certified by Professor Lijuan Zhang from Tianjin University of Traditional Chinese Medicine. The specimens were deposited in College of Pharmaceutical Engineering of Traditional Chinese Medicine, Tianjin University of Traditional Chinese Medicine China.

2.2 HS-GC-IMS analysis

The GC-IMS system (FlavourSpec®, Gesellschaft für Analytische Sensorsysteme mbH, Dortmund, Germany) with an autosampler (CTC Analytics AG, Zwingen, Switzerland) was used to detect the VOCs of 15 batches of ATR. Single-factor experiments were used to optimize incubation temperature (50 °C, 60 °C, 70 °C, 80 °C), and incubation time (10 min, 20 min, 30 min, 40 min). All herbs were dried and crushed through a 40-mesh sieve. Finally, 0.5 g of powder was weighed into a 20 mL headspace glass sampling bottle (Zhejiang HAMAG technology, Ningbo, China), and incubated at 60 °C for 20 min at 250 rpm. The column was FS-SE-54-CB-1 capillary column (15 m × 0.53 mm ID, 1 μm, CS-Chromatographie Service GmbH, Germany). headspace samples were automatically injected into the injector, the injection temperature was 85 ℃, and the injection volume was 500 μL. The carrier gas was N₂ (>99.999 %), the column temperature was 45 ℃. The program flow was as follows: 2 mL/min maintained for 2 min initially, increased to 10 mL/min over 10 min, then increased to 150 mL/min over 20 min, and maintained at 150 mL/min over 30 min. The drift tube has a length of 98 mm, a temperature of 45 °C, a voltage of 500 kV, and the drift gas is N₂ (>99.999 %) with a gas flow rate of 150 mL/min. Three parallel measurements per batch of samples.

2.3 HS-SPME-GC–MS qualitative analysis

SPME fiber (supelco, Bellefonte, Penn.) was installed on a MultiPurpose sampler (Gerstel, GER) and combined with a 7890B-7000D GC–MS (Agilent Technologies, Palo Alto, CA, USA) to detect VOCs of ATR. All herbs were dried and crushed through a 40-mesh sieve. 0.1 g of ATR was weighed into a 20 mL headspace glass sampling bottle. Single-factor optimization of SPME conditions: Four fiber coatings were tested, polydimethylsiloxane 100 µm phase thickness (PDMS), polyacrylate 85 µm phase thickness (PA), polydimethylsiloxane/divinylbenzene 65 µm phase thickness (PDMS/DVB) and polydimet-hylsiloxane/Carboxen/divinylbenzene 50/30 µm phase thickness (PDMS/CAR/DVB). The incubation temperatures were 50 °C, 60 °C, 70 °C and 80 °C, respectively. And the extraction times were 10 min, 20 min, 30 min and 40 min. The final, extraction conditions were the following: The PDMS/DVB coating was chosen and incubated for 5 min at 60 °C, followed by 30 min in an extract. vial penetration was 21.00 mm, the injection penetration was 54.00 mm, the desorption time was 5 min, and the desorption temperature was 250 °C.

The chromatographic column was HP-5MS phenyl methyl siloxane (30 m × 0.25 mm × 0.25 μm, 19091S-433, J&W Scientific, Folsom, CA, USA) elastic quartz capillary column. the initial temperature was maintain a temperature of 40 ℃ for 2 min; increased to 100 ℃ at a rate of 20 ℃/min; increase to 130 ℃ at a rate of 7 ℃/min; then, increased to 136 ℃ at a rate of 2 ℃/min, and maintain for 6 min; increased to 139 ℃ at a rate of 2 ℃/min, and maintained for 2 min; increased to 146 ℃ at a rate of 3 ℃/min; finally increased to 250℃ at a rate of 30 ℃/min and maintained for 5 min. the carrier gas was Helium (>99.999 %) with a flow rate of 1 mL/min; the split ratio was 1:1; the injection port temperature is 250 ℃. The conditions of MS were set as follow: the ion source temperature was 230 °C, electron impact (EI) with ionization energy was 70 eV. MS1 full scan mode was adopted with a scan range of m/z 30–650. Three parallel measurements per batch of samples.

2.4 Volatile oil extraction

All herbs were dried and crushed through a 40-mesh sieve. 50 g of ATR powder and 500 mL of pure water were placed in a 2 L round bottom flask, then added a small amount of zeolite. Heated to a slight boil and started timing, distilled for 5 h until the volatile oil no longer increased in the volatile oil tester, read the volatile oil volume. Added the volatile oil to ethyl acetate to make a solution.

2.5 GC–MS quantitative analysis

Liquid injection volatile oil solution 1 μL. The collision energies of 10 eV, 15 eV, 20 eV and 25 eV were optimized in product ion scan. Based on the optimized energy for multi reaction monitoring (MRM). In MRM scan, the precursor ions of estragole, methyleugenol, γ-asarone, β-asarone and asarone were m/z 147.1, m/z 179, m/z 208.1, m/z 208.1 and m/z 207.9. Then, determined their respective CE: 15 eV, 10 eV, 10 eV, 10 eV, 10 eV; the product ions were m/z 105.2 and m/z 91; m/z 164.1 and m/z 148.1; m/z 193.1 and m/z 165; m/z 193.1 and m/z 165; m/z 193.1 and m/z 165. Retention time range correspond to 3.5–10 min, 10–17 min, 17–20 min, 20–22 min, 22–32.59 min. Other conditions are consistent with the qualitative method.

2.6 Data analysis

Laboratory Analytical Viewer (LAV) and the plug-in (Reporter, Gallery Plot) were used to analyze the VOCs detected by HS-GC-IMS. The two-dimensional spectrograms of the samples were compared by Reporter, the fingerprint was established by Gallery Plot. The components were then retrieved from the GC-IMS database according to RI and drift time (Dt). The VOCs detected by HS-SPME-GC–MS were identified based on NIST17 library (matching degree > 80, RI) and reference compounds. Histogram drawn by origin 2021(Origin Lab Corporation, Northampton, MA, USA), PCA analyzed by SIMCA 14.1 (Umetrics, Sweden), heatmap and PatternHunter were performed using the MetaboAnalyst 5.0.

3.1 Qualitative analysis of the VOCs by HS-GC-IMS

Separation of components in two dimensions by HS-GC-IMS and a three-dimensional (retention time, migration time, and peak intensity) map was provided. It provided a visual method for the identification of VOCs. The X-axis was the normalized migration time, the Y-axis was the retention time. It could be seen that most of the ion peaks appear at the retention time of 100–1600 s and the migration time of 1.0–2.5. The red line at a migration time of 1.0 is the reaction ion peak (RIP). The color represented the peak intensity. The peak intensity decreased from red to white and then to blue. The dimers or polymers usually appear in the high concentration components, and the appearance of them will make the result identification more accurate (Fig. 1A). However, the differences among the 15 batches of ATR could not be quickly identified by the Topographic plot alone (Fig. S1). Therefore, the fingerprint was established (Fig. 1B).

Close

Fig. 1

(A) Topographic plot; (B) Fingerprints of 15 batches of ATR deteced by HS-GC-IMS. Each row represents a selected signal peak in a sample and each column represents a signal peak of a compound in a different sample. Compounds correspond to Table 1.

Export to PPT

A total of 33 VOCs were identified corresponding to 1–50, there are 10 alcohols, 4 ketones, 5 terpenes, 5 aldehydes, 4 esters and 5 other compounds (Table 1). These components correspond to the five regions a, b, c, d, e and f (Fig. 1B). Compounds 51–109 were unidentified compounds (Fig. 1B-g). Alcohols were less abundant in ATR1, ATR4 and ATR9; linalool was higher in ATR7, ATR10 and ATR13; (-)-myrtenol, nerol and 2-methylbutanol were more abundant in ATR2 and ATR10, nerol has some hepatoprotective effects (Islam et al., 2021); ATR3 and ATR8 contain more 1-hexanol; more 2-butoxyethanol in ATR6 (Fig. 1B-a). ATR4 had more 2-heptanone; 3-octanone was high level in ATR9, ATR13 and ATR15; the ketone component was also relatively low in ATR1 (Fig. 1B-b). Terpenoids were relatively few in ATR1 and ATR4; β-ocimene was most abundant in ATR13; α-phellandrene was more abundant in ATR3 and ATR9, it has anti-inflammatory and antioxidant effects (de Christo Scherer et al., 2019); α-terpinene was more in ATR2 and ATR9, it has a diastolic effect on tracheal smooth muscle(Pereira-de-Morais et al., 2020); Fig. 1B-c summarized the content of terpenoids. ATR4 and ATR14 contained more pentanal; and furfural was also higher in ATR4 (Fig. 1B-d). The content of isoamyl acetate in ATR2, ATR9, ATR10 and ATR13 was more; the content of methyl 2-methylbutanoate was higher in ATR2, ATR5 and ATR10, highest in ATR2; ATR1, ATR9 and ATR10 contained more ethyl acetate (Fig. 1B-e). Decalin had a high content in ATR2, ATR9, ATR10 and ATR13; ATR1 and ATR5 contained more acrylonitrile and 2-propanethiol; ATR11 contained high levels of isovaleric acid; 2-pentylfuran was higher in ATR4 (Fig. 1B-f). Among the unidentified components, there are some differences in ATR of different batches, especially ATR11, components 77–86 are higher (Fig. 1B-g).

3.2 Qualitative analysis of the VOCs by HS-SPME-GC–MS

PDMS is a non-polar fiber coating, PA is a polar fiber coating, PDMS/DVB and PDMS/CAR/DVB are bipolar fiber coating (Shirey 2012). Optimization of SPME conditions. After filtering the miscellaneous peaks, compared the number of peaks under different conditions. Firstly, the different SPME fiber coatings were optimized, PDMS has the least number of peaks and PDMS/DVB has the most. Secondly, optimized incubation temperature, the number of peaks at 50 °C was the highest, and the number of peaks at other temperatures was similar. Finally, optimized the extraction time, the higher the temperature the higher the number of peaks (Fig. 2A). Most of the major peaks were present under different conditions, The peak areas were further analyzed. The peak area of eight major components of in ATR were used as criteria. Compare different fiber coatings, the peak areas of both PA and PDMS were low in general. The PDMS/CAR/DVB had the highest peak area only for estragole and asarone. most of the peaks had the highest area in PDMS/DVB. So PDMS/DVB was selected as the extraction fiber (Fig. 2B-a). At different incubation temperatures, the peak areas of components with large retention times were almost positively correlated with the incubation temperature. But the peak areas of components with small retention times were generally negatively correlated with incubation temperature. When the incubation temperature was 60 °C, the peak areas of most components were moderate at this temperature (Fig. 2B-b). At different incubation times, it found that the extraction time increased, the peak areas increased, except estragole. And when the extraction time reached 40 min, methyleugenol, cis-methyl isoeugenol and shyobunone were also declined. A high peak area was achieved for all components at an extraction time of 30 min (Fig. 2B-c). The final SPME conditions were determined.

Close

Fig. 2

(A) Effect of different SPME conditions on the number of peaks. (B) Effect of different SPME conditions on the peak areas of eight major compounds in ATR. (a) SPME fiber coating; (b) incubation temperature; (c) extraction time.

Export to PPT

The total ion chromatograms (TIC) of ATR are shown in Supplementary Information (Fig. S2). The content of individual components in different batches varies significant. A total of 95 VOCs were identified by HS-SPME-GC–MS, there were 24 alcohols, 9 ethers, 2 aldehydes, 2 acids, 39 terpenes, 1 epoxyalkane, 13 ketones, 1 alkane and 4 esters correspond to the a, b, c, d, e, f, g and h regions of heatmap (Fig. 3). The peak area was taken as log₁₀ for the heatmap, the color changed from brownish red to white and then to dark blue, indicating a sequential decrease. Components methyleugenol, cis-methyl isoeugenol, estragole, γ-asarone, β-asarone, asarone, β-elemen (anti-tumor activity), β-caryophyllen (lilac scent), aromandendrene, viridiflorene, eremophilene, δ-guaiene, β-cadinene, shyobunone, isoshyobunone, eremophila ketone, β-oplopenone, cyperenone (anti-dysmenorrhea effect) were more abundant in the ATR.

Close

Fig. 3

Heatmap of 15 batches of ATR detected by HS-SPME-GC–MS. Each column represents a different component in a batch of ATR, each row represents the same component in different batches. Compounds correspond to Supplementary Information (Table S2).

Export to PPT

β-Asarone and asarone are its main medicinal components of ATR. β-Asarone has central nervous system depressant effect, sedative, hypothermic and tranquilizing, and anticonvulsant properties; asarone has the effect of calming asthma, expectorant, cough, antispasmodic and choleretic (Park et al., 2015, Chellian et al., 2017, Chellian and Pandy 2018, Saki et al., 2020). But the two ingredients have some toxicity (Chellian et al., 2017). Methyleugenol has a sweet clove-anise pungent aroma, the aroma is more translucent and long-lasting, with a mild tea pungent aroma, can be used in perfumes and fragrances. It also has the effect of suppressing anorexia (Zhu et al., 2018) and analgesic effect (Rocha-Gonzalez et al., 2020). However, genotoxicity and carcinogenicity should not be ignored (Suparmi et al., 2019). Estragole has an aniseed aroma, mainly used in spices, has antibacterial activity (Moghaddam et al., 2014), insecticidal effect (Kumari et al., 2017). γ-Asarone has a strong correlation with C6 glioma cells in terms of antioxidant, anti-inflammatory and neurotrophic (Yan et al., 2020). β-Caryophyllen has antibacterial, anti-inflammatory, anti-cancer and analgesic properties (Fidyt et al., 2016). Shyobunone interact with AChE with a kind of non-competitive inhibition (Martins et al., 2021). These components with pharmacodynamics and high content can be used as quality markers detected by ATR in HS-SPME-GC–MS.

Benzyl alcohol was higher in ATR14; elemol was high level only in ATR5, it has immunosuppressive effects and has therapeutic potential in the treatment of atopic dermatitis (Yang et al., 2015); longiborneol was more abundant in ATR3, ATR8, ATR10 and ATR14; guaiol (anti-proliferative, autophagy-promoting, anthelmintic and insecticidal bioactivities) was present in ATR8, ATR9 and ATR14, it significantly inhibite cell growth of non-small cell lung cancer cells in vitro and in vivo (Yang et al., 2016); α-eudesmol was more abundant in ATR9 and T-muurolol was only less abundant in ATR14 (Fig. 3a). cis-Anethole, anethole (fennel odor, leukocyte-raising effect) were higher levels in ATR2, ATR4, ATR5 and ATR10, anethole can be used as a potential therapeutic ingredient for non-alcoholic steatohepatitis (Zhang et al., 2021) and has osteoprotective effects (Qu et al., 2021); more allyltetramethoxybenzene in ATR1 and ATR3 (Fig. 3b). Cinnamal (cinnamon aromatic scent) was higher in ATR2 and ATR10. The content of acetosyringone (raw materials for the preparation of asarone) in ATR1, ATR4 and ATR14 was higher (Fig. 3c). The acid content was low (Fig. 3d). Low content of α-longipinene, longicyclene, α-guaiene and more content of γ-maaliene in ATR9; cyclosativene was relatively more in ATR3 and ATR14, it has antioxidant ability, can reduce oxidative damage in the field of neurodegenerative diseases, anti-cancer (Turkez et al., 2015). The content of α-guaiene in ATR14 and ATR15 was low. γ-Himachalene was less abundant in ATR3, ATR6 and ATR8. The content of δ-guaiene and 3,7(11)-selinadiene were higher in ATR14. The contentα-calacorene in ATR10 was high level (Fig. 3e). The content of d-camphor was abundant in ATR14 it has sedative effect (Oshima and Ito 2021), relief of inflammation (Dos Santos et al., 2021). ATR14 had more dehydrofukinone, ATR9 and ATR14 contained more zierone (Fig. 3f). It showed less in both epoxides and esters (Fig. 3g and h) different components in a batch of ATR, each row represents the same component in different batches. Compounds correspond to Supplementary Information (Table S2).

3.3 Quantification of 5 major components by GC–MS

Estragole, methyleugenol, γ-asarone, β-asarone and asarone were high in ATR and have some biological activity (Moghaddam et al., 2014, Park et al., 2015, Chellian et al., 2017, Zhu et al., 2018, Yan et al., 2020), so the five components were quantified. The content of five components in the powder was calculated after the quantitative research of the volatile oil by GC–MS. In Table 2, the standard curve showed good linearity in the corresponding range (r2 > 0.999). The concentrations at signal-to-noise ratios of 3 and 10 were used as limits of detection (LOD) and limits of quantification (LOQ). Methodological investigations were shown in Supplementary Information (Table S3). Took the mixed standard solution and calculated the intraday precision by continuously measuring the concentration of 6 times of the standard solution (<1.87 %). The stability was calculated by injecting the one sample at 6 times points: 0 h, 2 h, 4 h, 6 h, 12 h, 24 h (<2.14 %). The 1 batch of samples was weighed precisely in 6 portions to prepare the test solution, each solution was measurement once, then calculated the reproducibility (<2.45 %). 9 samples of the 1 sample were divided into 3 groups, each group was blended with 50 %, 100 % and 150 % of the target standard compound. 3 injection per solution, then calculated the recovery (93.32–105.68).

ATR1-ATR15 were extracted volatile oils: 0.59 mL, 1.47 mL, 0.70 mL, 0.68 mL, 0.69 mL, 0.88 mL, 0.86 mL, 1.08 mL, 0.59 mL, 1.03 mL, 1.37 mL, 0.87 mL, 0.69 mL, 1.18 mL, 0.98 mL. The content of the five components in the volatile oil and powder were shown in Supplementary Information (Table S4). The content of esdragole were between 0.05 and 21.78 mg/g, methyl eugenol were between 0.04 and 5.16 mg/g, γ-asarone were between 0.22 and 2.97 mg/g, β-asarone were between 3.06 and 21.61 mg/g and asarone were between 0.15 and 4.55 mg/g in powder of 15 batches of ATR. The content of esdragole was particularly pronounced only in ATR2 and ATR10; methyl eugenol was more abundant in ATR12, ATR3, ATR6, ATR8, most in ATR8; the content of β-asarone was high in all batches, with ATR4, ATR11, ATR14 and ATR15 being particularly high; asarone is present in high levels in ATR1; ATR5, ATR9 and ATR12 have more γ-asarone (Fig. 4). It could be seen that there were differences in active ingredients among the different batches of ATR.

Close

Fig. 4

The content of the 5 components in ATR powder.

Export to PPT

3.4 Comprehensive analysis

The two instruments have different detection ranges. 1,8-Cineole, linalool, d-camphor and borneol are the compositions detected by both approaches. The results measured by both approaches were analyzed based on PCA. PCA is a multivariate statistical analysis technique based on unsupervised analysis. By reducing the dimension of data, an unsupervised data driven approach is provided. Then, the regularity and difference among samples are evaluated according to the contribution rate of principal component factors in different samples (Farrugia et al., 2021). The peak area is taken as log₁₀, then subjected to UV-scaling. Since the magnitude-level difference between the data detected by the two instruments, UV-scaling was used to increase the importance of low-response variables (Cheng et al., 2017). PCA of HS-GC-IMS and HS-SPME-GC–MS were established with named as model A and model B, respectively (Fig. 5). The sample differences in the graph are proportional to their distances, large distance results in large difference, small distance results in small difference. The ellipse in the figure is defined as the Hotelling's T2 (95 %) critical limit. Exceeding the critical limit is considered as an outlier. Most of the ATR are in the critical limit, except for ATR6 of model B, which is at the edge of the critical limit. In model A, Q2 was 0.757, indicating that the model predicts well. In model A, ATR1, ATR4 and ATR14 deviated significantly from the other batches (Fig. 5A-a), 2-Pentylfuran, 2-Heptanone and other components were positively correlated with them (Fig. 5A-b). The others were not more obviously separated (Fig. 5A). In model B, Q2 was 0.623, indicating that the model predicted well. Model A and model B had similar distribution trends. However, the distribution of PCA for Model A and Model B were not entirely consistent. Because there are some differences in the detection methods (Fig. 5B-a). Few components were positively correlated with ATR1 and ATR4, asarone, β-asarone, and so on were positively correlated components; γ-elemene, 3,7(11)-selinadiene, epi-γ-eudesmol and others in ATR14 are positively correlated components; acorenone B and other components were positively correlated with ATR6 (Fig. 5B-b).

Close

Fig. 5

PCA score plot of 15 batches of ATR. (A) HS-GC-IMS, compounds correspond to Table 1; (B) HS-SPME-GC–MS, compounds correspond to Supplementary Information (Table S2). (a and b are score scatter plot and biplot; in biplot, the yellow hexagons represent ATR, the green circles represent components.).

Export to PPT

PatternHunter was performed based on the percent peak area content (Fig. 6). Blue indicates negative correlation and pink indicates positive correlation. cis-Methyl isoeuge and some components are negatively correlated with asarone, benzyl alcohol and others were positively correlated with asarone (Fig. 6A). Meanwhile, β-selinene and some components were also negatively correlated with β-asarone, asarone and others were positively correlated with β-asarone (Fig. 6B). The components detected by HS-GC-IMS showed more positive correlations with asarone and β-asarone compared to HS-SPME-GC–MS. Benzaldehyde and some components were positively correlated with asarone, while borneol and other components were negatively correlated with asarone (Fig. 6C). 2-Pentylfuran and some components showed a positively correlation with β-asarone, β-pinene and others showed a negative correlation with β-asarone (Fig. 6D). β-Asarone and asarone are the main medicinal components of ATR. To some extent, they determine the quality of ATR. When more of the VOCs have a positive correlation with the target components and less of a negative correlation, it means that the target components are high content. This is helpful to quickly distinguish the content of asarone and β-asarone in ATR by HS-GC-IMS. Furthermore, the quality of ATR can also be quickly identified.

Close

Fig. 6

PatternHunter of ATR. (A) The top 25 components detected by HS-SPME-GC–MS showed the highest correlation with β-asarone; (B) The top 25 components detected by HS-SPME-GC–MS showed the highest correlation with asarone; (C) The top 25 components detected by HS-GC-IMS showed the highest correlation with β-asarone; (D) The top 25 components detected by HS-GC-IMS showed the highest correlation with asarone.

Export to PPT