Differences in Volatile Organic Compounds in Rhizoma gastrodiae (Tian Ma) of Different Origins Determined by HS-GC-IMS

Headspace gas chromatography–ion mobility spectrometry (HS-GC-IMS) and principal component analysis (PCA) were used to compare the differences in volatile organic compounds (VOCs) of Rhizoma gastrodiae (Tian Ma) from six different origins in Yunnan, Sichuan, Shaanxi, Anhui, Hubei, and Guizhou. A total of 161 signal peaks were identified, and 84 compounds were characterized, including 23 aldehydes, 19 alcohols, 12 ketones, 8 heterocyclic compounds, 7 esters, 4 phenols, 4 acids, 4 ethers, 2 amines, and 1 alkane. The results of cluster analysis and fingerprint similarity analysis based on principal component analysis and Euclidean distance indicated that there were significant differences between the volatile components of Rhizoma gastrodiae from different origins. This study demonstrated that HS-GC-IMS is simple, rapid, accurate, and has a small sample size and can achieve rapid analysis of the differences in volatile compounds between samples of different origins of Rhizoma gastrodiae.


Introduction
Rhizoma gastrodiae (also known as Tian ma) is mainly produced in Asian countries such as China, Korea, Japan, and India [1]. In areas such as Guizhou, Yunnan, Hunan, Sichuan, and Shannxi in China, Rhizoma gastrodiae is used as food in stews, hot pots, and stir-fries by the inhabitants of these regions [2]. Studies have shown that there are significant differences in chemical composition and pharmacological activity between different origins of Rhizoma gastrodiae [3]. In Asia, Rhizoma gastrodiae is used both as a food and medicinal plant. In China, Gastrodia elata is recommended for treating headaches, dizziness, tremors, and epilepsy, and to help improve central nervous system disorders. Modern research has also confirmed that Rhizoma gastrodiae has shown good application in central nervous system disorders such as Alzheimer's disease, epilepsy, Parkinson's disease, cerebral ischemia/reperfusion, and cognitive dysfunction [4][5][6]. The main active ingredients of Rhizoma gastrodiae to improve central nervous system (CNS) disorders are its rich phenolic compounds, such as gastrodin, p-hydroxybenzyl alcohol, and parishin [7]. In addition, Rhizoma gastrodiae is also rich in polysaccharides, sterols, organic acids, and other components, which constitute the special smell of Rhizoma gastrodiae [7], and it is rich in volatile components [8]. In recent years, Rhizoma gastrodiae has gained much attention and popularity due to its better edible and medicinal value. It is being processed into relevant functional foods and occupys a certain proportion of consumption in the market, and volatile flavor compounds play an important role in the acceptability of Rhizoma gastrodiae products, which has undoubtedly aroused great interest in the study of Rhizoma gastrodiae flavor.
At present, there are two parts to the analysis of volatile substances in food: sensory analysis, which is a kind of sensory perception and evaluates the results with a certain subjectivity, and instrumental analysis, which is an objective analysis at the molecular level, and the combination of these two forms in order to evaluate the flavors of food more scientifically [9]. As a result, there has been an increasing amount of research in recent years on the use of instrumental analysis techniques to detect volatile flavor components in food products. The commonly used instrumental analytical techniques for volatile compounds in food are gas chromatography-mass spectrometry (GC-MS), gas chromatography-ion mobility spectrometry (GC-IMS), gas chromatography olfaction-mass spectrometry (GC-O-MS), and electronic nose (E-nose). Although E-nose has the advantage of being fast and having sensitive detection, the reproducibility of the results is somewhat lower and, therefore, electronic noses are limited in practical application [10,11]. GC-IMS has been shown to be fast, sensitive, and easy to operate [9], and compared to GC-MS, the former has a higher separation efficiency and can be used to obtain analytical results in a shorter period of time. In recent years, GC-IMS and GC-MS have been widely used for the study of volatile compounds in traditional Chinese medicine and foodstuffs [12], such as Cordyceps sinensis, [13] green tea [14], and thuja [15]. Zhang et al. [16] established a rapid and accurate method for the determination of volatile components using gas chromatographyion mobility spectrometry (GC-IMS) and gas chromatography-mass spectrometry (GC-MS). Guo et al. [17] used GC-MS and GC-IMS to analyze the aroma characteristics of oolong tea made from three tea varieties. However, few contemporary studies using HS-GC-IMS have been conducted on Rhizoma gastrodiae, and comprehensive studies on multiple origins of dried Rhizoma gastrodiae have not been reported.
Research has shown that flavors determine the organoleptic value of foods and also play an important role in identifying the nutritional value of foods. The richest source of Rhizoma gastrodiae is mainly in China, but the existence of differences in volatile flavors substances between multiple origins of dried Rhizoma gastrodiae has not been explored in current studies. Therefore, in the current study, for comparison, we compared the differences in volatile organic matter of Rhizoma gastrodiae from six origins: Yunnan, Sichuan, Shaanxi, Anhui, Hubei, and Guizhou. The volatile flavor substances identified in this study can provide some reference value for the processing and nutritional value of Rhizoma gastrodiae functional foods.

Spectral Analysis of Rhizoma gastrodiae Samples from Different Origins
HS-GC-IMS was used to analyze the VOCs in six Rhizoma gastrodiae samples. A twodimensional top view of the VOCs in the six Rhizoma gastrodiae samples was plotted using the Reporter plug-in (Figure 1), enabling a detailed comparison of the differences in VOCs between the different Rhizoma gastrodiae samples. The vertical coordinates indicate the retention time of the GC, and the horizontal coordinates indicate the drift time and the reactive ion peak (RIP). The different colored dots to the right of the RIP represent the different VOCs detected. The difference in color reflects the signal intensity of the different volatiles detected in each Rhizoma gastrodiae sample, with the red signal indicating a higher concentration of the detected volatiles and, the darker the color, the greater the intensity, indicating a greater concentration of that volatile substance and vice versa. As can be seen in Figure 1A, most of the signal occurs at retention times of 80-970 s and drift times of 1.0-1.9 ms. A differential contrast model was used to compare the differences between samples, using the YN sample as the reference contrast and the remaining sample minus the reference. If the two VOCs agree, the background after subtraction is white; if it is red, it means that the substance concentration is higher than the reference, and blue means that it is lower than the reference. In the differential contrast model plot ( Figure 1B), the concentrations of the volatile substances can be clearly seen. Comparing samples YN, SC, and SX, it can be seen that in the range of retention times of 190-270 s and drift times of 1.4-1.8 ms, the concentrations of petanal, hexanal, pentan-1-ol, methyl isobutyl ketone, 2-hexanone, and butanoic acid in SC and SX methyl esters and other VOC concentrations were significantly higher than YN. Comparing samples YN, SX, and AH, it can be found that the VOC concentrations of heptanal, 2-heptanone, n-hexanol, styrene, 1,2-dimethylbenzene, and pentanoic acid in SX and AH were significantly higher and were in the retention time range of 350-400 s and 1.0-2.0 ms. Comparing samples YN and HB, it can be found that in the drift time range of 1.0-1.7 ms, the concentrations of 5-methylfurfural, butanoic acid methyl eater, 3-hydroxybutan-2-one, alpha-pinene, 1-heptanol, and ethyl hexanoate in HB were significantly higher than YN. Heptanol, ethyl hexanoate, 3-octanol, 2-methylbutan-1-ol, N, N-diethylethanamine, isopropyl acetate, octan-1-ol, diethylene glycol dimethyl ether, alpha-phellandrene, 2,6-dimethylphenol, 2-penthlfuran, and other VOCs were significantly higher than in YN. However, comparing samples YN and GZ, it can be seen that the concentration of volatile components in the GZ sample is significantly lower than YN. These differences in data indicate that the differences in climate, soil, environment, and seed quality among different regions have led to differences in volatile components in Rhizoma gastrodiae from each production area. This also indirectly indicates that under the current experimental conditions, compared with other regions, the Rhizoma gastrodiae from Hubei contains more flavorful substances, which explains why it has a more obvious aroma in all the samples.
In this study, the differences in volatile organic compounds in Rhizoma gastrodiae from 6 origins in Yunnan, Sichuan, Shaanxi, Anhui, Hubei, and Guizhou were analyzed using HS-GC-IMS, and a total of 161 signal peaks were identified. These compounds were characterized by comparing their IMS drift times and retention indices with authentic reference compounds. A total of 84 typical target compounds were identified from the topography by the GC × IMS library. As shown in Figure 2, the horizontal coordinates indicate the drift time, the vertical coordinates indicate the retention time, and the white numbers correspond to the compounds in Table 1. A total of 84 compounds were characterized, including 23 aldehydes, 19 alcohols, 12 ketones, 8 heterocyclic compounds, 7 esters, 4 phenols, 4 acids, 4 ethers, 2 amines, and 1 alkane compound. Of these, there were nonanal, (E)-hept-2-enal, octanal, 3-methylbutanal, hexanal, butanal, heptanal, (E)-2-pentenal, 1-octen-3-one, 3-octanol, isopropyl alcohol, alpha-pinene, and ortho-guaiacol. These 13 compounds were present as monomers and dimers. The method appears to be able to detect a wider range of organic compounds with richer results than the conventional GC-IMS analysis technique. Previously, Sun et al., analyzed the differences in volatile compounds in fresh Rhizoma gastrodiae of three varieties using GC-IMS, and a total of 75 volatile compounds were detected, including 45 identified substances such as aldehydes, esters, alcohols, ketones, and acids [19]. Our study also found that aldehydes were the highest substance in the volatile components of Rhizoma gastrodiae, and our method using HS-GC-IMS could detect more abundant volatile components. We speculated that this may be due to the more abundant volatile components contributed by Hubei Rhizoma gastrodiae.
The above results indicate that the volatile components contained in Rhizoma gastrodiae from different origins vary more significantly and can be distinguished from those of different origins by HS-GC-IMS.

Dynamic PCA of Samples
Principal component analysis (PCA) is a multivariate statistical method used to examine correlations among multiple variables [20]. PCA serves as a powerful visualization tool that provides researchers with a way to reduce the dimensionality of data, thereby eliminating non-essential information [21,22]. In this study, a PCA analysis was performed on Rhizoma gastrodiae samples, and the results are shown in Figure 4, where different colors represent different samples of Rhizoma gastrodiae, the distance between individual points represents the level of similarity, and the dispersion of the same points represents the homogeneity of the same sample. The PCA (Figure 4) illustrates the differences in the contribution of different volatile flavor substances to different samples. When samples are closely located, it can be understood that the differences in flavor compounds between samples are relatively small. Conversely, it indicates a significant difference in volatile components between the two. According to Figure 4, Dim1 accounts for 47.5% and Dim2 accounts for 17.5%, with a total cumulative contribution of 65% by the two principal components. Figure 4 also illustrates significant differences in aroma components of Rhizoma gastrodiae from different production areas. Rhizoma gastrodiae from Anhui, Sichuan, and Yunnan are grouped together, while Shaanxi and Hubei are significantly different than other production areas, and there are also significant differences between Rhizoma gastrodiae samples from the three production areas. This indicates that HS-GC-IMS data contains effective information that can distinguish Rhizoma gastrodiae samples from different production origins.

Fingerprint Similarity Analysis Based on Euclidean Distance
Euclidean distance, similar to PCA analysis, is a method of cluster analysis in which similarity is determined by a distance coefficient; if the coefficient is large, the difference between the two is also large and shows a positive correlation. Conversely, the smaller the coefficient, the smaller and more similar the difference between the two [23]. By applying the Euclidean distance similarity algorithm, it is possible to evaluate the quality of two samples, and it has been found that the algorithm can accurately and reliably evaluate the quality of samples. [23]. Figure 5 shows the fingerprint similarity based on Euclidean distance and Table S1 represents the Euclidean distance values between the samples of different origins of Rhizoma gastrodiae. The results of the Euclidean distance analysis show that the distances between the Rhizoma gastrodiae samples of different origins can be clearly distinguished. Among them, the average Euclidean distance between SX and AH was 6,257,676, the average Euclidean distance between AH and SC was 5,688,077, the average Euclidean distance between SC and YN was 6,284,151.444, the average Euclidean distance between YN and GZ was 8,169,782, the average Euclidean distance between GZ and HB was 23,577,777.778, and the average Euclidean distance between HB and SX was 34,733,333.333. So, HB and SX are the furthest apart, and the difference between them is considered to be the most significant.

Hierarchical Cluster Analysis Heatmap
To further analyze the differences in VOCs between different Rhizoma gastrodiae samples, a hierarchical cluster analysis (HCA) heatmap was generated, which can be used to distinguish between two main clusters [24] and is an important method of cluster analysis [25]. It has been widely used to analyze the degree of variation in food composition [26,27]. Figure 6 shows the HCA of volatiles in different Rhizoma gastrodiae samples. The outer circle represents the volatiles detected, and the contents of the column at the opening of the circle indicate the name of each Rhizoma gastrodiae sample. Purple indicates low relative intensity, while brown indicates high relative intensity, and the darker the color, the greater the intensity, and vice versa. It is evident in Figure 6 that the relative content of volatile substances varied between the samples, with the HB (Hubei Rhizoma gastrodiae) sample containing a higher and more diverse range of volatile substances, which was significantly different from the other Tianma samples. The other samples also showed some differences in the volatile substances of Rhizoma gastrodiae. Previous studies have shown that the conventional GC-MS technique is usually used to analyze VOCs and thus distinguish fresh Rhizoma gastrodiae from different origins and varieties [19]. In contrast, this study used HS-GC-IMS to analyze VOCs in Rhizoma gastrodiae samples from different origins, which had many advantages over the traditional GC-MS method such as more convenient operation, faster response, higher sensitivity, and lower cost.

Sample Preparation
The samples of the six species are detailed in Figure 7; each sample was crushed using a pulverizer (a high-speed crusher (model: FW80, produced by Tianjin, Tianjin, China, Tianjin Tasty Instruments Co., Ltd.)) and was further passed through a 24-mesh sieve for HS-GC-IMS analysis, and the processing information for each sample is given in Table 2.

HS-GC-IMS System
We used the Gas-phase Ion Mobility Spectrum Flavour Spec ® (G.A.S. department of Shandong Province, Qingdao, China, Shandong Hai Neng Science Instrument Co., Ltd.) to analyze 6 different regions of Rhizoma gastrodiae powder prepared previously. We placed 2 g of Rhizoma gastrodiae powder in a 20 mL top-empty bottle and incubator for 20 min at 70 • C and 500 rpm under gas phase temperature. Next, set the temperature of the injection needle to 85 • C and inject 300 microliters of the sample.
Then, perform gas chromatography separation using an MXT-5 chromatography column (15 m × 0.53 mm × 1 µm) at a column temperature of 60 • C. Set the IMS temperature to 45 • C and use N 2 (purity ≥99.999 %) as the carrier/drift gas with a flow rate of 2 mL/min (0-2 min), 10 mL/min (2-10 min), 100 mL/min (10-20 min), 150 mL/min (20-30 min), and stop the analysis after 30 min. The drift tube is maintained at 45 • C under the N 2 drift gas with a flow rate of 150 mL/min. Three samples are measured for each sample.

Data Analysis
GC-IMS library Search software (Version number: 1.0.3) and the Laboratory Analytical Viewer (LAV) are data analysis software (Version number: 2.2.1) that allow different perspectives to be examined. The LAV includes VOCal and three plug-ins, and VOCal is used to view analytical spectra and qualitative and quantitative analysis data. Volatile organic compounds are represented by each point on the graph. With the software's built-in database, qualitative analyses of substances can be performed. A Reporter plug-in can be used to compare the spectral differences between different products, such as 2D top views and sample difference spectra. Using the library plot plug-in, differences in VOCs between the samples were visually compared using inter-sample fingerprinting; in order to facilitate rapid identification of unknown sample types, the dynamic PCA plug-in was used for cluster analysis of the samples. A principal component analysis was used to investigate the relationship between different samples and VOCs; using the clustering heat map tool, heat maps were created.

Conclusions
In this study, headspace gas chromatography-ion mobility spectrometry (GC-IMS) and principal component analysis (PCA) were used to compare the differences in volatile organic compounds in Rhizoma gastrodiae from six origins in Yunnan, Sichuan, Shaanxi, Anhui, Hubei, and Guizhou. A total of 161 signal peaks were identified, and 84 compounds were characterized, including 23 aldehydes, 19 alcohols, 12 ketones, 8 heterocyclic compounds, 7 esters, 4 phenols, 4 acids, 4 ethers, 2 amines, and 1 alkane compound. Due to the limitations of the HS-GS-IMS assay and the fact that 77 signal peaks have not yet been identified, further qualitative analysis can be carried out at a later stage by other analytical techniques, such as HPLSMS (High-Performance Liquid Chromatography-Mass Spectrometry).
The results of the cluster analysis and fingerprint similarity analysis based on the principal component analysis, as well as Euclidean distance, showed that Rhizoma gastrodiae mainly contained 2-pentanone, nonana, benzaldehyde, tert-butylmethylether, 3-methylbutanal, hexanal, 1,2-dimethoxyethane, 2-ethyl furan, butanal, isopropyl alcohol, 6-methyl-5-hepten-2-one, 2-methyl-1-propanol, and 2,3-butanedione, with some variation between different origins of Rhizoma gastrodiae. The higher content and variety of volatile substances were contained in the HB (Hubei Rhizoma gastrodiae) samples. This indicates that the differences between different origins influenced the results of the detection of volatile substances in Rhizoma gastrodiae. It also shows that under the present experimental conditions, the quality of Hubei Rhizoma gastrodiae is better compared to other origins and varieties of Rhizoma gastrodiae.
In summary, we used HS-GC-IMS to perform three parallel tests on Rhizoma gastrodiae samples from each region. The operation is simple, fast, accurate, and requires a small sample size. Our study demonstrates that through the multivariate data analysis method of HS-GC-IMS, it is possible to analyze and distinguish Rhizoma gastrodiae from different geographic sources.