Characterization of the Key Aroma Volatile Compounds in Nine Different Grape Varieties Wine by Headspace Gas Chromatography–Ion Mobility Spectrometry (HS-GC-IMS), Odor Activity Values (OAV) and Sensory Analysis

During this study, the physicochemical properties, color, and volatile aroma compounds of the original wines produced from the grape varieties ‘Hassan’, ‘Zuoshaner’, ‘Beibinghong’, ‘Zuoyouhong’, ‘Beta’, ‘Shuanghong’, ‘Zijingganlu’, ‘Cabernet Sauvignon’, and ‘Syrah’ were determined and sensory evaluation was performed. Results indicated that ‘Hassan’ contained the most solids, ‘Zuoshaner’ produced the most total acid, residual sugar, total anthocyanin, and total phenol, and ‘Shuanghong’ produced the most tannin. Calculation of the chroma and hue of the wines according to the CIEL*a*b* parameters revealed that the ‘Cabernet Sauvignon’ wines were the brightest of the nine varieties and that the ‘Zuoshaner’ wines had the greatest red hue and yellow hue and the greatest saturation’. A total of 52 volatile compounds were identified and quantified in nine wine samples by HS-GC-IMS analysis, with the most significant number of species detected being 20 esters, followed by 16 alcohols, 8 aldehydes, four ketones, one terpene, and one furan, with the highest total volatile compound content being ‘Beta’. A total of 14 volatile components with OAV (odor activity value) >1 were calculated using the odor activity value (OAV) of the threshold of the aromatic compound, and the OPLS-DA analysis was performed by orthogonal partial least squares discriminant analysis (OPLS-DA) using the OAV values of the compounds with OAV values >1 as the Y variable. The VIP (Variable Importance in Projection) values of six compounds, ethyl isobutyrate, ethyl hexanoate-D, 2-methylpropanal, ethyl octanoate, ethyl butanoate-D, and Isoamyl acetate-D, were calculated to be higher than one between groups, indicating that these six compounds may influence aroma differences. It is essential to recognize that the results of this study have implications for understanding the quality differences between different varieties of wines and for developing wines that have the characteristics of those varieties.


Introduction
As the global wine market continues to grow, the question of how to enhance the flavor of the wine is increasingly becoming a hot research topic. Regular consumption of red wine in moderation has been shown to positively affect health, with wine containing phenolic compounds with antioxidant properties [1]. Wine is an alcoholic beverage product obtained by fermentation of fresh grapes or grape juice, with components derived from grape-kernels and the fermentation and aging process [2]. The organoleptic properties of wine are conferred by organoleptically active compounds, mainly polyphenols (coloring), sugars, acids, tannins (taste), and volatile flavor compounds (aroma) [3]. The blend and balance of aromas in a wine determine the quality of the wine and how well the consumer likes it, expresses the style of the wine, and is a significant indicator of the quality of the

Methodology 2.3.1. Winemaking
After harvesting, the grapes were destemmed, grape crushed by manual destemming and crushing, and fermented at room temperature (25 • C). The fermenting tank was made from 304 stainless steel thermostatic fermentation vats from Tiburth, with a volume of 12 L. Three sets of replicated winemaking experiments for each variety, and each fermenter was filled with around 10 L of crushed grapes. During fermentation, the fermenting tank was tightly closed, and an exhaust valve was used to ensure that the gas produced during fermentation was discharged smoothly. The first fermentation lasted seven days, and by testing the sugar and alcohol content during the fermentation period, the total sugar content of each variety of wine stopped decreasing at the end of the first fermentation and remained stable, and no bubbles were produced in the fermenter, while the alcohol content reached a certain concentration. The second fermentation for one month was mainly to check the difference in physical and chemical indexes of each variety at one month of aging. Fermentation temperature for the second fermentation was between 18 and 20 • C. During the second fermentation, the indexes of each variety were already stable.
The yeast used for the fermentation was CEC01 active dry wine yeast from Angel's yeast. The yeast was added at 250 mg/Kg, and the SO2 was added at 60 mg-L-1.

Winemaking
After harvesting, the grapes were destemmed, grape crushed by manual destem and crushing, and fermented at room temperature (25 °C). The fermenting tank was m from 304 stainless steel thermostatic fermentation vats from Tiburth, with a volume L. Three sets of replicated winemaking experiments for each variety, and each ferm was filled with around 10 L of crushed grapes. During fermentation, the fermenting was tightly closed, and an exhaust valve was used to ensure that the gas produced du fermentation was discharged smoothly. The first fermentation lasted seven days, an testing the sugar and alcohol content during the fermentation period, the total sugar tent of each variety of wine stopped decreasing at the end of the first fermentation remained stable, and no bubbles were produced in the fermenter, while the alcohol tent reached a certain concentration. The second fermentation for one month was m to check the difference in physical and chemical indexes of each variety at one mon aging. Fermentation temperature for the second fermentation was between 18 and 2 During the second fermentation, the indexes of each variety were already stable.
The yeast used for the fermentation was CEC01 active dry wine yeast from An yeast. The yeast was added at 250 mg/Kg, and the SO2 was added at 60 mg-L-1.

Testing the Basic Physical and Chemical Properties of Raw Wine Grapes
The soluble solids must be determined by handheld refractometer, and the titra acid content of wine was determined by the indicator method according to GB/ T 15 2006 General Analysis Method of Wine and Fruit Wine. The alcohol content was d mined by the alcohol meter method according to GB/T 15,038-2006 General Ana Method of Wine and Fruit Wine. Anthrone and sulfuric acid colorimetry were use determine the total sugar content in grapes wine, and the standard curve was prep with standard glucose solution; the Folin-Denis reagent method was used to deter tannin content in grapes Juice, and the standard curve was created with different ta concentrations. It was reacted with phosphomolybdic acid in sodium carbonate solu to form the blue compound after being soaked in water at 85 °C for three hours. Th sorbance value was measured at 740 nm.; total anthocyanin content in grape juice determined by the pH difference method by reacting anthocyanins with potassium ride buffer (0.025 M, pH = 1) and acetic acid buffer (0.4 M, pH = 4.5), then calcul differences at 520 nm and 700 nm. Total phenol content: Folin-Ciocalteu colorim

Testing the Basic Physical and Chemical Properties of Raw Wine Grapes
The soluble solids must be determined by handheld refractometer, and the titratable acid content of wine was determined by the indicator method according to GB/T 15,038-2006 General Analysis Method of Wine and Fruit Wine. The alcohol content was determined by the alcohol meter method according to GB/T 15,038-2006 General Analysis Method of Wine and Fruit Wine. Anthrone and sulfuric acid colorimetry were used to determine the total sugar content in grapes wine, and the standard curve was prepared with standard glucose solution; the Folin-Denis reagent method was used to determine tannin content in grapes Juice, and the standard curve was created with different tannin concentrations. It was reacted with phosphomolybdic acid in sodium carbonate solution to form the blue compound after being soaked in water at 85 • C for three hours. The absorbance value was measured at 740 nm.; total anthocyanin content in grape juice was determined by the pH difference method by reacting anthocyanins with potassium chloride buffer (0.025 M, pH = 1) and acetic acid buffer (0.4 M, pH = 4.5), then calculating differences at 520 nm and 700 nm. Total phenol content: Folin-Ciocalteu colorimetric method [42]. Dry extraction content: refer to the dry extraction test method of the national standard (GB/T 15,038-2006).

Colorimetry
The colorimetric analysis was based on the CIEL*a*b* colorimetric standard, and the color characteristics of the wine samples were measured spatially using a Lambda 365 UV-Vis spectrophotometer with continuous scanning (400-700 nm) and distilled water as a blank control group. L*, a*, b*, Cab*, hab*, and ∆Eab* were calculated based on the four absorbance values, L value indicates brightness, a* = red-green deviation, b* = blue-yellow deviation, hab* indicates hue angle, Cab* indicates red grape color index, and ∆Eab* indicates the total color difference (∆E ab *) 2 = (L* − L 0 *) 2 + (a* − a 0 *) 2 + (b* − b 0 *) 2

Determination of Organic Acid Content
The organic acids were detected using high-performance liquid chromatography (HPLC), following the previously published literature as a reference, under the following conditions: aqueous phosphoric acid solution at pH = 2.3, methanol as mobile phase, and the test conditions were: a C18-XT column (4.6 × 250 × 5) at a column temperature of 25 • C and a set flow rate of 0.4 mL/min [43]. The standard curves for the six organic acids tested were as Table 1. Headspace-gas chromatography-ion mobility spectrometry (HS-GC-IMS) was used for the determination of volatiles in wine. The instrument used in the experiment was a G.A.S. FlavourSpec ® flavor analyzer. Briefly, 1 mL of the sample was taken in a 20 mL headspace vial, 10 µL of 20 ppm 4-methyl-2-pentanol was added, incubated at 60 • C for 15 min, and then injected into the sample. Chromatographic conditions: The column was a WAX column (15 m × 0.53 mm, 1 µm), column temperature 60 • C, carrier gas N2, and IMS temperature 45 • C. The conditions for the automatic headspace injection were as follows: injection volume 100 µL, incubation time 10 min, incubation temperature 60 • C, injection needle temperature 65 • C, and incubation speed 500 rpm; The analysis was carried out using 4-methyl-2-pentanol as an internal standard at a concentration of 198 ppb. The volume of the signal peak was 493.34 and the intensity of each peak was approximately 0.401 ppb. And Table 2 shows the gas chromatography conditions.
Ci is the calculated mass concentration of any component in µg/L, Cis is the mass concentration of the internal standard used in µg/L, and Ai/AIS is the volume ratio of any signal peak to the signal peak of the internal standard. The NIST database and IMS database are built into the software for the qualitative analysis of the substances. The contribution of the overall aroma of wine is evaluated using the odor activity value (OAV). The OAV is calculated by dividing the concentration of volatile compounds by the odor threshold (OT). Volatile compounds with an OAV > 1 are considered to be aromatically active and play an essential role in developing the wine's aromatic profile.

Sensory Evaluation
A quantitative descriptive analysis (QDA) of the wines was carried out by a trained sensory panel of 13 sommeliers (8 women and 5 men, aged from 22 to 52 years, average 34 years). The experts were recruited based on their motivation and availability and had been trained according to national standards ISO 6658 and ISO 8586 prior to the sensory evaluation. The experts discussed the aromatic composition of the wines in depth over three preliminary meetings (2 h each) until they all agreed on the degree of aromaticity. Wine odors were quantified using six sensory descriptors (floral, fruity, botanical and herbal, fermented, tarry, and confectionery) for sensory characteristics descriptors, based on definitions in the published literature, according to the national standard GB 15,038-2006 and in conjunction with references [44,45], taking into account the results of the discussions. Samples were labeled with three numbers and presented to the tasters in random order. Panel members were asked to rate the intensity of each attribute on a 10-point scale, where a score of 10 indicated the highest intensity and a score of 0 indicated absence. Each sample was assessed in triplicate, and the mean of each sample was expressed by the average of the three scores based on a ten-point scale.

Data Processing
The trial data were statistically collated using Excel 2010, and an analysis of variance (ANOVA) was performed by SPSS (version 22.0, IBM, Armonk, NY, USA). Statistical analysis was performed to check for significant differences in individual results for the trial data, and all data were expressed as mean ± standard deviation. Differences between the two groups were considered significant at p < 0.05. Simca software was used for OPLS-DA and VIP value analysis; the GC-IMS assay was smoothed and denoised with Savitzky Golay. The migration time was normalized by setting the RIP position to 1, i.e., dividing the actual migration time by the RIP peak exit time to obtain the approximate migration time. Direct comparison of spectral differences between samples was performed using the Reporter plug-in and comparison of fingerprint profiles using the Gallery Plot plug-in for visual and quantitative comparison of volatile organic compound differences between different samples. Heat map analysis and correlation analysis were performed using the OmicShare tools, a free online platform for data analysis (https://www.omicshare.com/tools (accessed on 27 August 2022)).

Basic Physico-Chemical Indicators of Different Grape Varieties of Raw Wine
As a whole, there were significant differences between the fundamental physicochemical indicators of the wine samples of each variety (Table 3). The residual sugar and solids content of the wines determine the type of wine, which can be classified as dry, semi-dry, semi-sweet, or sweet according to the sugar content [46]. The phenolics are factors that produce bitterness and astringency [47]. The main coloring compound in red wines is anthocyanin [48,49], whose composition and content influence the color characteristics of the wine. As can be seen in the table above, 'Hassan' has the highest solid content at 9.4 g/L, which differs significantly (p < 0.05) from the other eight varieties of wine samples; 'Zuoshaner' has the highest total acid, residual sugar, total anthocyanin, and total phenol content at 16.25 g/L, 5.7 g/L, 1477.85 mg/L, and 3.56 g/L, respectively, which differs significantly (p < 0.05) from the other varieties. Tannin is the source of bitterness and astringency in fruit wines. It is also an essential component of the backbone of fruit wines and has a very positive effect on color stabilization, prevention of oxidation, and removal of off-flavors [50]. 'Shuanghong' had the highest tannin content of 3.64 g/L, which was significantly different from the other varieties (p < 0.05), while the lowest tannin content was 1.3 g/L in 'Syrah'. The results of the data show that the total anthocyanins, total phenols, and tannins of 'Hassan', 'Zuoshaner', 'Beibinghong', 'Zuoyouhong', 'Shuanghong', and 'Zijngganlu' are significantly higher than those of the American variety of grape 'Beta' and the Eurasian varieties of 'Cabernet Sauvignon' and 'Syrah'. This is also in line with the results of previous studies [51,52]. Because of their small size and thick skin, Vitis amurensis is very rich in anthocyanins, tannins, and total phenols and produce wines with the right taste. It is also because of the richness of the anthocyanin content of the mountain grape variety that the original wine is darker than the Eurasian and American varieties. The dry extract content is an essential indicator of wine quality, mainly determined by the variety and the age of the wine [53], and the analysis of the dry leachate content in wine can tell whether the wine is adulterated with water, alcohol, etc. According to China's national standards, the dry leachate of red wine should not be less than 18.0 g/L, the dry extract content of the nine tested varieties met the winemaking standards; the alcohol content of the nine raw grape wine samples ranged from 11 • to 13 • . The absorbance of the nine wine samples increased continuously in the 400-520 nm band (Figure 2), reaching a maximum of around 520 nm, and decreased significantly in the 520-700 nm band, gradually converging to zero as the wavelength approached 700 nm.

Sample CIELab Parameters
The chromaticity of wine is an essential criterion for evaluating the quality of pearance, and the degree of oxidation and quality of a bottle of wine can be judged chromaticity and hue. New red wines have a purplish-red hue due to anthocyan they mature, the blue-violet hue disappears as the anthocyanins combine with oth

Sample CIELab Parameters
The chromaticity of wine is an essential criterion for evaluating the quality of its appearance, and the degree of oxidation and quality of a bottle of wine can be judged by its chromaticity and hue. New red wines have a purplish-red hue due to anthocyanins. As they mature, the blue-violet hue disappears as the anthocyanins combine with other substances. The wine gradually gains a yellowish hue due to polymeric tannins, with the hue gradually changing from initially purplish-red to tile-red or brick-red. The color tone of the wine is also a critical factor in the marketing of wine [54]. As can be seen from Table 4, the L* values of all nine varieties of raw grape wine samples were high, ranging from 41.65 to 84.65, indicating that all nine varieties of wine have a good luster. The largest L* was 'Cabernet Sauvignon' at 84.65, with the brightest color of the wine, followed by 'Syrah' and 'Beta', and the smallest L* was 'Zuoshaner' at 41.65, with the darkest color. Means with different letters in the same column express significant differences (Duncan's test p < 0.05).
Chroma a* values indicate the red hue of the wines, with the largest a* value being 167.48 for 'Zuoshaner', followed by 'Shuanghong' and 'Zuoyouhong', and the smallest being 17.5 for 'Cabernet Sauvignon'; chroma b* values indicate the yellow hue of the wines, with nine samples ranging from 1.58 to 21.76, the largest being 'Zuoshaner' and the smallest being 'Syrah'; saturation cab* is a combination of a* and b*, indicating the color of the wines The greatest saturation was for 'Zuoshaner', followed by 'Zouyouhong' and 'Shuanghong', the smallest was for 'Cabernet Sauvignon' at 17.93. The saturation and chroma values for the nine varieties were close to each other, which is characteristic of young wines; the hue angle of the nine samples ranged from 4.02 to 23.52, all close to zero, i.e., all close to the purple-red hue. The largest is 'Zijingganlu', which is less red than the other eight varieties, and the smallest is 'Syrah', which is the closest to purplish red.
The nine wine samples were analyzed for the color difference using Cabernet Sauvignon, which had the highest L* value, as the base value. They were calculated according to the range of color difference units (National Bureau of Standards Unit (NBS)) given by the CIE1976Lab color space system to describe the degree of color difference between the wine samples [55]. As seen from the values in the table, there is a significant color difference between Cabernet Sauvignon and the other eight varieties, with substantial differences.

Comparison of Organic Acids in Wine Samples of Different Varieties
The type of organic acid affects the acidity and, therefore, the taste of the wine. The organic acid content varies from variety to variety (Table 5). Organic acids are an essential part of the structure of the wine, with tartaric acid, the characteristic organic acid of wine with a sour and astringent taste [56], being the most abundant of the six organic acids. 'Zuoshaner' had the highest tartaric acid content at 7.99 g/L, significantly higher than the other eight varieties, followed by 'Zuoyouhong', with the lowest tartaric acid content being 'Cabernet Sauvignon' at 3.16 g/L. Malic acid is present in high levels in grapes and at the end of alcoholic fermentation. It gradually decreases after MLF, a crucial secondary fermentation in most wine production, usually carried out by lactic acid bacteria after the completion of alcoholic fermentation, converting sharp L-malic acid into soft L-lactic acid, which improves microbial stability through residual lactic acid bacteria nutrients and can promote sensory regulation of wine in the secondary metabolism of lactic acid bacteria [57]. The highest malic acid content was in 'Beta' at 9.51 g/L, followed by 'Zuoshaner' and 'Hassan', while the lowest malic acid content was in 'Syrah' at 2.03 g/L. The highest levels of lactic acid were found in 'Hassan' and 'Zijingganlu' at 0.18 g/L, followed by 'Beibinghong', and the lowest level of lactic acid was in 'Syrah'; the levels of acetic acid detected were all low, with the highest level of acetic acid in 'Zuoshaner', followed by 'Beibinghong' and 'Zijingganlu', and the lowest level in 'Syrah'. The highest citric acid content was 0.89 g/L for 'Zuoyouhong', significantly higher than the other eight varieties, followed by 'Syrah', and the lowest citric acid content was 0.31 g/L for 'Hassan'; the highest succinic acid content was 0.85 g/L for 'Zuoshaner', followed by 'Beibinghong' and 'Zijingganlu', and the lowest succinic acid content was 0.39 g/L for 'Hassan'. Table 5. Comparison of Organic Acids in Different Variety Wine Samples.

Varieties
Tartaric Acid/g/L Malic Acid/g/L Lactic Acid/g/L Acetic Acid/g/L Citric Acid/g/L Succinic 3. 16  Means with different letters in the same column express significant differences (Duncan's test p < 0.05).
The cluster analysis results can better reflect the characteristics of the organic acid substances in the different wine samples ( Figure 3). According to the cluster analysis of the organic acids of each variety, it can be seen that when the cross-cutting line takes values between 5 and 6, the nine varieties of wine samples can be divided into three categories: the first category is 'Beta' and 'Hassan', the second category is 'Cabernet Sauvignon', and 'Syrah', and the third category is 'Zuoyouhong', 'Zijingganlu', 'Shuanghong', 'Zuoshaner', and 'Beibinghong', indicating that the samples contained in each category have similarity in organic acids when the cross-cutting line takes values between 5 and 6. The result is also better in bringing together different types of grapes.

HS-GC-IMS Analysis of Wine Samples of Different Varieties
The aroma description of a wine is one of the keys to its quality [58]. The type and content of volatile compounds and their interactions are the main factors influencing the quality of grapes and wines and also determine the wine's uniqueness [59]. Gas chromatographymass spectrometry (HS-GC-MS) is a commonly used method for separating and quantifying volatile compounds in foodstuffs [60].

Fingerprinting of Volatile Components of Wine Samples from Different Varieties
The fingerprints of the volatile flavor compounds of different wine varieties were constructed based on all the peaks in the HS-GC-IMS two-dimensional profile ( Figure 4). Each sample was measured in parallel, with darker colors indicating higher peak intensities and higher contents. The fingerprints revealed the composition of and differences in the volatile flavor compounds of the wine samples of different varieties. As the graph shows, 'Hassan' has a high content of ethyl acetate, 1-propanol, pentanal, 2-pentanone, and 4-methyl-2-pentanone; 'Zuoshaner' has a high content of 1-penten-3-ol, 2,5-dimethylfuran, ethyl isobutyrate, and methyl acetate. The content of substances such as 1-butanol is higher in 'Zuoyouhong'; 1-pentanol, 1-hydroxy-2-propanone, 2-Methylpropanal, and isobutyl acetate are higher in Beta; acetone, 1-hexanol, 2-butanol, Ethyl hexanoate, hexyl acetate, Ethyl octanoate, isoamyl acetate, Ethyl butanoate, Ethyl propanoate, and isobutyl butyrate, in 'Zijingganlu'. Acetaldehyde is present in higher amounts; ethyl lactate and acetic acid are present in higher amounts in 'Syrah'. the organic acids of each variety, it can be seen that when the cross-cutting line takes values between 5 and 6, the nine varieties of wine samples can be divided into three categories: the first category is 'Beta' and 'Hassan', the second category is 'Cabernet Sauvignon', and 'Syrah', and the third category is 'Zuoyouhong', 'Zijingganlu', 'Shuanghong', 'Zuoshaner', and 'Beibinghong', indicating that the samples contained in each category have similarity in organic acids when the cross-cutting line takes values between 5 and 6. The result is also better in bringing together different types of grapes.

HS-GC-IMS Analysis of Wine Samples of Different Varieties
The aroma description of a wine is one of the keys to its quality [58]. The type and content of volatile compounds and their interactions are the main factors influencing the quality of grapes and wines and also determine the wine's uniqueness [59]. Gas chromatography-mass spectrometry (HS-GC-MS) is a commonly used method for separating and quantifying volatile compounds in foodstuffs [60].

Fingerprinting of Volatile Components of Wine Samples from Different Varieties
The fingerprints of the volatile flavor compounds of different wine varieties were constructed based on all the peaks in the HS-GC-IMS two-dimensional profile ( Figure 4). Each sample was measured in parallel, with darker colors indicating higher peak intensities and higher contents. The fingerprints revealed the composition of and differences in the volatile flavor compounds of the wine samples of different varieties. As the graph shows, 'Hassan' has a high content of ethyl acetate, 1-propanol, pentanal, 2-pentanone, and 4-methyl-2-pentanone; 'Zuoshaner' has a high content of 1-penten-3-ol, 2,5-dimethylfuran, ethyl isobutyrate, and methyl acetate. The content of substances such as 1-butanol is higher in 'Zuoyouhong'; 1-pentanol, 1-hydroxy-2-propanone, 2-Methylpropanal, and isobutyl acetate are higher in Beta; acetone, 1-hexanol, 2-butanol, Ethyl hexanoate, hexyl acetate, Ethyl octanoate, isoamyl acetate, Ethyl butanoate, Ethyl propanoate, and isobutyl butyrate, in 'Zijingganlu'. Acetaldehyde is present in higher amounts; ethyl lactate and acetic acid are present in higher amounts in 'Syrah'.

Two-Dimensional Mapping of Wine Samples of Different Varieties
Significant differences were observed in the fingerprints of the nine wine samples ( Figure 5). The differences were mainly in the content, with the color representing the concentration of the substance, white indicating a lower concentration, red a higher concentration, and darker colors indicating a greater concentration. HS-GC-IMS well separated the volatile substances in the nine wine samples, and the differences can be visualized.
Significant differences were observed in the fingerprints of the nine wine ( Figure 5). The differences were mainly in the content, with the color represe concentration of the substance, white indicating a lower concentration, red a hi centration, and darker colors indicating a greater concentration. HS-GC-IMS w rated the volatile substances in the nine wine samples, and the differences can ized. Using the Hassan variety of wine as a reference, the rest of the spectrum tracted from the signal peaks in the Hassan to obtain a spectrum of the difference the two (Figure 6). The blue area indicates that the substance is lower in this sam in the Hassan wine, while the red area indicates that the substance is more prese sample than in the Hassan wine. Again, the darker the color, the more significan ference. The difference spectrum shows that ethyl acetate, 1-propanol, pentanal none, and 4-methyl-2-pentanone are present in higher amounts in 'Hassan' th other varieties of wine. The entire graph has a blue background, and the red vertical line at the horizontal coordinate 1.0 is the RIP peak (reactive ion peak, normalized). Each point on either side of the RIP peak represents a volatile organic compound.
Using the Hassan variety of wine as a reference, the rest of the spectrum was subtracted from the signal peaks in the Hassan to obtain a spectrum of the difference between the two ( Figure 6). The blue area indicates that the substance is lower in this sample than in the Hassan wine, while the red area indicates that the substance is more present in this sample than in the Hassan wine. Again, the darker the color, the more significant the difference. The difference spectrum shows that ethyl acetate, 1-propanol, pentanal, 2-pentanone, and 4-methyl-2-pentanone are present in higher amounts in 'Hassan' than in the other varieties of wine.

Analysis of Volatile Matter Components
Aroma is one of the essential sensory characteristics of wine. A total of 5 volatile compounds were detected by qualitative analysis of the volatile comp the wine samples using the NIST database built into HS-GC-IMS and the IMS ( Table 6). The retention index is the calculation using N-ketones and views qualit quantitative analytical spectra and data with VO-Cal. with the most significant n species detected being 20 esters, followed by 16 alcohols, 8 aldehydes, 4 ketones, 1 and 1 furan. The nine wine varieties had the same types of volatile aroma com

Analysis of Volatile Matter Components
Aroma is one of the essential sensory characteristics of wine. A total of 52 typical volatile compounds were detected by qualitative analysis of the volatile components in the wine samples using the NIST database built into HS-GC-IMS and the IMS database ( Table 6). The retention index is the calculation using N-ketones and views qualitative and quantitative analytical spectra and data with VO-Cal. with the most significant number of species detected being 20 esters, followed by 16 alcohols, 8 aldehydes, 4 ketones, 1 terpene, and 1 furan. The nine wine varieties had the same types of volatile aroma compounds detected, but the levels varied significantly. Of the nine wine varieties, 'Hassan' had the highest total volatile compound content of 58,160.24 µg/L, followed by 'Beta ' 53,287   Means with different letters in the same column express significant differences (Duncan's test p < 0.05). _M and _D, which are the Monomer and Dimer of the same substance. Esters are the most abundant compounds detected in each variety. Some important esters, such as ethyl butyrate, isoamyl acetate, and ethyl caproate, contribute to the desirable fruit organoleptic characteristics of the wine, including fruity aromas such as banana, strawberry, and green apple [65][66][67]. Among the esters detected were ethyl isovalerate, which contributes a fruity aroma with wine notes, isoamyl acetate with a banana odor, ethyl acetate with a sweet fruit flavor, and isobutyl propionate with a rum odor, and ethyl lactate with a pungent odor, etc. The aroma descriptions of the esters show that the esters are mainly fruit flavors, with ethyl acetate best reflecting the fruit aroma.

Aldehydes
The aldehyde content of each species ranged from 3.23% to 3.75%, with the highest content being 2645.91 µg/L for 'Beta'.

Principal Component Analysis (PCA) of Wine Samples
In order to better present and distinguish the differences between the different varieties of wine samples, the volatile compounds identified by HS-GC-IMS were analyzed by PCA. The nine samples were well differentiated according to their aroma characteristics and varietal. The unsupervised multidimensional statistical analysis method (PCA) was applied to the samples to discriminate the magnitude of variability between groups of samples, between subgroups, and between samples within groups for the different wines. The contribution of PC1 was 46.8%, and that of PC2 was 20.7%, and the nine groups of samples showed a clear trend of separation on the two-dimensional plot, with no outlier samples and good clustering of samples of the same wine type. The PCA results reflected a significant overall difference in aroma matter between the nine groups of samples and differentiated them from each other. As shown in Figure 7, the 'Zuoshaner', 'Beibinghong', 'Beta', 'Shuanghong', and 'Zijingganlu' samples were closer to each other, while the 'Hassan', 'Zuoyouhong', 'Cabernet Sauvignon', and 'Syrah' samples were farther apart, indicating a significant difference between the aromatic characteristics of the different samples.

Analysis of the Key Aroma Compounds OAV in Wine Samples of Different Varieties
It is generally accepted that components with an OAV greater than one may directly influence the overall flavor. Based on the qualitative and quantitative results of GC-IMS, the literature was used to find the threshold values of the corresponding aroma compounds in water and to calculate their OAV values [68][69][70][71][72]. It was calculated that a total of 14 aroma compounds with OAV greater than one were detected in nine wine samples (Table 7), and the study showed that the OAV values were proportional to the contribution of aroma. The highest number of aroma compounds with OAV >1 was found in the lipid group, with 10 species: Ethyl hexanoate-M, Isoamyl acetate-M, Ethyl 3-methylbutanoate-M, Ethyl 3-methylbutanoate-D, Isoamyl acetate-D, Ethyl butanoate-D, Ethyl isobutyrate, Ethyl Acetate, Ethyl hexanoate-D, and Ethyl octanoate; the three aldehydes are 2-Methylpropanal, Acetaldehyde, and 2-Methylbutanal; and one furan is 2,5-dimethylfuran. The key compounds' OAV values in the nine wine samples varied. However, in general, the esters had higher OAV values than the other compounds, with isoamyl acetate-M among the esters having the most significant OAV values of 55.07-75.62 and contributing more to the overall aroma. The predominance of fruit flavors in the esters likewise indicates that esters are one of the most crucial compound groups contributing to the overall aroma and that fruit flavors are an essential aromatic feature in wine aromas.

Analysis of the Key Aroma Compounds OAV in Wine Samples of Different Varieties
It is generally accepted that components with an OAV greater than one may directly influence the overall flavor. Based on the qualitative and quantitative results of GC-IMS, the literature was used to find the threshold values of the corresponding aroma compounds in water and to calculate their OAV values [68][69][70][71][72]. It was calculated that a total of 14 aroma compounds with OAV greater than one were detected in nine wine samples (Table 7), and the study showed that the OAV values were proportional to the contribution of aroma. The highest number of aroma compounds with OAV > 1 was found in the lipid group, with 10 species: Ethyl hexanoate-M, Isoamyl acetate-M, Ethyl 3-methylbutanoate-M, Ethyl 3-methylbutanoate-D, Isoamyl acetate-D, Ethyl butanoate-D, Ethyl isobutyrate, Ethyl Acetate, Ethyl hexanoate-D, and Ethyl octanoate; the three aldehydes are 2-Methylpropanal, Acetaldehyde, and 2-Methylbutanal; and one furan is 2,5-dimethylfuran. The key compounds' OAV values in the nine wine samples varied. However, in general, the esters had higher OAV values than the other compounds, with isoamyl acetate-M among the esters having the most significant OAV values of 55.07-75.62 and contributing more to the overall aroma. The predominance of fruit flavors in the esters likewise indicates that esters are one of the most crucial compound groups contributing to the overall aroma and that fruit flavors are an essential aromatic feature in wine aromas.
Heat Map Analysis, PCA Analysis, and Correlation Analysis in Aromatic Compounds with OAV > 1 in Nine Grapefruits of Compounds with OAV > 1 in Fruits of Different Varieties.
Hierarchical analysis was used to cluster the concentrations of volatile aroma compounds with OAV values greater than one in the nine sample wines, as seen from the heat map analysis of each fruit sample (Figure 8), with red indicating high expression of that aroma compound component in the sample and blue indicating low expression of that aroma compound in the sample. The concentrations of volatile aroma compounds with OAV values greater than one varied considerably among the various samples.
Principal component analysis (PCA) is a multivariate statistical analysis technique. By identifying several principal component factors to represent the many complex and hard-to-find variables in the original sample, regularities and differences between samples are then assessed based on the contribution of the principal component factors in different samples [73]. The PCA results clearly show (Figure 9) that, in a relatively independent space, PCA analysis of the concentrations of volatile aroma compounds with OAV values greater than one in nine wine samples extracted a total of two principal components, with cumulative contributions of up to 68.6% for PC1 and PC2. Among the different varietal wines, 'Zijingganlu' was in the first quadrant of the score, with positive values on both PC1 and PC2. The 'Shuanghong', 'Zuoyouhong', and 'Zuoshaner' samples were located in the second quadrant, with positive scores on PC1 and negative scores on PC2. 'Beibinghong' is located at the junction of quadrants one and two and is positive on PC1 and negative on PC2. 'Beta' is located at the junction of quadrants two and three and is negative on PC2. 'Hassan' is in the third quadrant and is negative on PC1 and PC2. The 'Cabernet Sauvignon' and 'Syrah' samples were in quadrant four, with negative numbers on PC1 and positive numbers on PC2. This indicates that compounds with OAV values greater than one for volatile aromatic content vary considerably between wine samples of different varieties. Heat Map Analysis, PCA Analysis, and Correlation Analysis in Aromatic Compounds with OAV > 1 in Nine Grapefruits of Compounds with OAV > 1 in Fruits of Different Varieties Hierarchical analysis was used to cluster the concentrations of volatile aroma compounds with OAV values greater than one in the nine sample wines, as seen from the heat map analysis of each fruit sample (Figure 8  Principal component analysis (PCA) is a multivariate statistical analysis technique. By identifying several principal component factors to represent the many complex and hard-to-find variables in the original sample, regularities and differences between samples are then assessed based on the contribution of the principal component factors in different samples [73]. The PCA results clearly show (Figure 9) that, in a relatively independent space, PCA analysis of the concentrations of volatile aroma compounds with OAV values greater than one in nine wine samples extracted a total of two principal components, with cumulative contributions of up to 68.6% for PC1 and PC2. Among the different varietal wines, 'Zijingganlu' was in the first quadrant of the score, with positive In the correlation analysis of Figure 10, the red box's Pearson correlation coefficient was significantly correlated. 2-Methylpropanal was strongly correlated with Isoamyl acetate-M, Ethyl 3-methylbutanoate-M, Ethyl 3-methylbutanoate-D, Ethyl butanoate-D, Ethyl isobutyrate, and Ethyl octanoate; Acetaldehyde is strongly correlated with Ethyl Acetate and Isoamyl acetate-D; 2-Methylbutanal is strongly correlated with Ethyl butanoate-D; Ethyl hexanoate-M is strongly correlated with 2,5 Dimethylfuran, Ethyl hexanoate-D, and Isoamyl acetate-D; Isoamyl acetate-M is strongly correlated with Ethyl isobutyrate, Ethyl butanoate-D, Ethyl 3-methylbutanoate-M and Ethyl 3-methylbutanoate-D; Ethyl 3-methylbutanoate-M is strongly correlated with Ethyl isobutyrate, Ethyl butanoate-D, and Ethyl 3-methylbutanoate-D; Ethyl 3-methylbutanoate-D is strongly correlated with Ethyl isobutyrate; Ethyl butanoate-D is strongly correlated with Ethyl octanoate and Ethyl isobutyrate; Ethyl hexanoate-D is strongly correlated with D-2,5 Dimethylfuran.
values on both PC1 and PC2. The 'Shuanghong', 'Zuoyouhong', and 'Zuoshaner' samples were located in the second quadrant, with positive scores on PC1 and negative scores on PC2. 'Beibinghong' is located at the junction of quadrants one and two and is positive on PC1 and negative on PC2. 'Beta' is located at the junction of quadrants two and three and is negative on PC2. 'Hassan' is in the third quadrant and is negative on PC1 and PC2. The 'Cabernet Sauvignon' and 'Syrah' samples were in quadrant four, with negative numbers on PC1 and positive numbers on PC2. This indicates that compounds with OAV values greater than one for volatile aromatic content vary considerably between wine samples of different varieties. In the correlation analysis of Figure 10, the red box's Pearson correlation coefficient was significantly correlated. 2-Methylpropanal was strongly correlated with Isoamyl acetate-M, Ethyl 3-methylbutanoate-M, Ethyl 3-methylbutanoate-D, Ethyl butanoate-D, Ethyl isobutyrate, and Ethyl octanoate; Acetaldehyde is strongly correlated with Ethyl Acetate and Isoamyl acetate-D; 2-Methylbutanal is strongly correlated with Ethyl butanoate-D; Ethyl hexanoate-M is strongly correlated with 2,5 Dimethylfuran, Ethyl hexanoate-D, and Isoamyl acetate-D; Isoamyl acetate-M is strongly correlated with Ethyl isobutyrate, Ethyl butanoate-D, Ethyl 3-methylbutanoate-M and Ethyl 3-methylbutanoate-D; Ethyl 3-methylbutanoate-M is strongly correlated with Ethyl isobutyrate, Ethyl butanoate-D, and Ethyl 3-methylbutanoate-D; Ethyl 3-methylbutanoate-D is strongly correlated with Ethyl isobutyrate; Ethyl butanoate-D is strongly correlated with Ethyl octanoate and Ethyl isobutyrate; Ethyl hexanoate-D is strongly correlated with D-2,5 Dimethylfuran.
The Pearson correlation coefficients for the green boxes are significantly negatively correlated. 2-methylpropanal was significantly negatively correlated with Ethyl Acetate, Ethyl hexanoate-m was significantly negatively correlated with Acetaldehyde and Isoamyl acetate M, and 2-methylbutanal was significantly negatively correlated with Isoamyl acetate D. Isoamyl acetate-M was significantly negatively correlated with Ethyl acetate, Ethyl 3-methylbutanoate-M was significantly negatively correlated with Ethyl acetate.  Ethyl 3-methylbutanoate-D was significantly negatively correlated with Ethyl Acetate, and Ethyl isobutyrate was significantly negatively correlated with Ethyl Acetate. Figure 10. Analyses of the correlation between aroma compounds and OAV greater than 1 in nine grape wine varieties.

Analysis of Volatile Wine Compounds 0PLS-DA
OPLS-DA is a supervised statistical method for discriminant analysis that enables the identification of sample differences and the acquisition of characteristic markers of sample differences [74,75]. The contribution of each variable to wine flavor was further quantified based on the variable importance for the projection (VIP) in the OPLS-DA model, and volatile flavor compounds with VIP > 1 were screened as potential characteristic volatile markers [76]. In general, variables with VIP values > 1 are considered metabolites that cause differences between groups. Most studies use the content of volatile compounds as an evaluation indicator for OPLS-DA analysis, but some volatile compounds, although high in content, also have high threshold values that are not easily perceived by the human sense of smell, so in this experiment, the OAV values of compounds with OAV values Figure 10. Analyses of the correlation between aroma compounds and OAV greater than 1 in nine grape wine varieties.
The Pearson correlation coefficients for the green boxes are significantly negatively correlated. 2-methylpropanal was significantly negatively correlated with Ethyl Acetate, Ethyl hexanoate-m was significantly negatively correlated with Acetaldehyde and Isoamyl acetate M, and 2-methylbutanal was significantly negatively correlated with Isoamyl acetate D. Isoamyl acetate-M was significantly negatively correlated with Ethyl acetate, Ethyl 3-methylbutanoate-M was significantly negatively correlated with Ethyl acetate. Ethyl 3-methylbutanoate-D was significantly negatively correlated with Ethyl Acetate, and Ethyl isobutyrate was significantly negatively correlated with Ethyl Acetate.

Analysis of Volatile Wine Compounds 0PLS-DA
OPLS-DA is a supervised statistical method for discriminant analysis that enables the identification of sample differences and the acquisition of characteristic markers of sample differences [74,75]. The contribution of each variable to wine flavor was further quantified based on the variable importance for the projection (VIP) in the OPLS-DA model, and volatile flavor compounds with VIP > 1 were screened as potential characteristic volatile markers [76]. In general, variables with VIP values > 1 are considered metabolites that cause differences between groups. Most studies use the content of volatile compounds as an evaluation indicator for OPLS-DA analysis, but some volatile compounds, although high in content, also have high threshold values that are not easily perceived by the human sense of smell, so in this experiment, the OAV values of compounds with OAV values > 1 in the composition of wine samples from different varieties were used as Y variables for OPLS-DA analysis (Figure 11), which can more accurately screen out differential volatile compounds. Wine aroma quality depends on the combined effect of several aroma compounds, and compounds with VIP values > 1 were screened. The results revealed that the compounds influencing aromatic differences might be related to ethyl isobutyrate, ethyl hexanoate-D, 2-methylpropanal, ethyl octanoate, ethyl butanoate-D, and Isoamyl acetate-D (Table 8).

Sensory Evaluation Characteristics of the Original Wine
The sensory evaluation of the quality of the nine wine samples was carried out using six descriptors: 'Floral', 'Fruity', 'Plant and Herb', 'Fermented', 'Tarry', and 'Candy'. Statistical analysis showed that the samples differed in each descriptor ( Figure 12). These significant differences suggest that the flavor intensity of each sample was significantly different. Although panelists were trained before sensory assessment, they may have sig-

Sensory Evaluation Characteristics of the Original Wine
The sensory evaluation of the quality of the nine wine samples was carried out using six descriptors: 'Floral', 'Fruity', 'Plant and Herb', 'Fermented', 'Tarry', and 'Candy'.
Statistical analysis showed that the samples differed in each descriptor ( Figure 12). These significant differences suggest that the flavor intensity of each sample was significantly different. Although panelists were trained before sensory assessment, they may have significantly influenced the descriptor rating results. This phenomenon is not uncommon in characterization analyses. It suggests that panelists applied different levels of qualitative scoring due to physiological differences in perceived intensity or personal preferences (e.g., central or extreme raters). However, no significant interactions between panelists and replication were found in the study, suggesting that all panelists were duplicated in triplicate for all descriptors scored. Similarly, there were no significant interactions between sample and replication and panel members, suggesting that the sensory data were valid and credible. It is also clear from the GC-IMS results that, apart from alcohols, esters make up the most significant proportion of the total volatile compounds detected. It is common sense that esters contribute to the desirable fruit organoleptic characteristics of the wine, with the 'fruity' and 'floral' descriptors being the primary and most fundamental parts of the global flavor of the wine. These two descriptors are therefore important indicators of the quality of a wine's aromas. Compared with other samples, 'Beta' showed a higher 'floral' and 'fruity' aroma, which was the same as the GC-IMS results. Among the nine wine varieties, the content of ester compounds in 'Beta' accounted for the highest proportion of the total content of aroma compounds. 'Shuanghong' shows high levels of vegetal and herbal aromas, with hexanal, valeraldehyde, (E)-2-heptanal, and (E)-2-hexenal probably closely related to the vegetal and herbal descriptors, i.e., aldehydes and alcohols are often associated with 'green', 'fresh grass', and 'green plants. A total of eight aldehydes were detected in the GC-IMS results, with acetaldehyde being the main aldehyde, 'Shuanghong' having a high acetaldehyde content of 1941.76 µg-L-1 among the nine varietal wines. The fermentation aromas were mainly produced by the fermentation and aging stages, with all varieties scoring low in tar aromas. The tar aromas were mainly produced by changes in tannins during fermentation, 'Zijingganlu' showing a higher tar and fermentation aroma. The sensory evaluation results also showed significant differences in the fruit aromas of the nine varieties, indicating that the results of the sensory evaluation were similar to the results of the data analysis. Combined with the aroma characteristics of several aroma compounds with VIP values greater than one calculated from the OPLS-DA analysis, it was found that the fruit aromas were the main compounds in the wine aroma substances, also indicating that the six compounds with VIP values greater than one calculated based on the OAV values were compounds that could influence the differences between groups, similar to the results of the sensory evaluation. vegetal and herbal aromas, with hexanal, valeraldehyde, (E)-2-heptanal, and (E)-2-hexenal probably closely related to the vegetal and herbal descriptors, i.e., aldehydes and alcohols are often associated with 'green', 'fresh grass', and 'green plants. A total of eight aldehydes were detected in the GC-IMS results, with acetaldehyde being the main aldehyde, 'Shuanghong' having a high acetaldehyde content of 1941.76 μg-L-1 among the nine varietal wines. The fermentation aromas were mainly produced by the fermentation and aging stages, with all varieties scoring low in tar aromas. The tar aromas were mainly produced by changes in tannins during fermentation, 'Zijingganlu' showing a higher tar and fermentation aroma. The sensory evaluation results also showed significant differences in the fruit aromas of the nine varieties, indicating that the results of the sensory evaluation were similar to the results of the data analysis. Combined with the aroma characteristics of several aroma compounds with VIP values greater than one calculated from the OPLS-DA analysis, it was found that the fruit aromas were the main compounds in the wine aroma substances, also indicating that the six compounds with VIP values greater than one calculated based on the OAV values were compounds that could influence the differences between groups, similar to the results of the sensory evaluation.

Conclusions
In this study, nine grape samples from the National fruit tree germplasm Vitis amurensis nursery in Zuojia town were collected and used to produce the original grape wine. The wines' basic physical and chemical properties, color, and aromatic composition

Conclusions
In this study, nine grape samples from the National fruit tree germplasm Vitis amurensis nursery in Zuojia town were collected and used to produce the original grape wine. The wines' basic physical and chemical properties, color, and aromatic composition were examined, and the sensory evaluation of the nine varieties was carried out. The test results showed that 'Hassan' had the highest solids content, 'Zuoshaner' had the highest total acid, residual sugar, total anthocyanin, and total phenol content, and 'Shuanghong' had the highest tannin content. 'Cabernet Sauvignon' had the brightest color of the nine varieties, and 'Zuoshaner' had the most pronounced red and yellow hues, with the most excellent saturation. The HS-GC-IMS technique was used to analyze the variation in volatile flavor compounds of raw wine samples of different varieties. A total of 52 volatile flavor substances were identified, including 20 esters, 16 alcohols, 8 aldehydes, 4 ketones, and 1 terpene and furan each, with significant differences in the content of volatile flavor substances in different varieties of wine. The specific wine aroma characteristics were characterized based on the volatile compounds and by quantitative descriptive analysis of the data through multivariate statistical analysis. In contrast, the key volatile compounds affecting the wine aroma were screened by combining principal component analysis, OAV value analysis, and VIP value analysis. Fourteen volatile aroma compounds with OAV values greater than one were screened. In general, the OAV values of esters were higher than those of other compounds and contributed more to the overall aroma. The OAV values of compounds with OAV values greater than one in the composition of wine samples from different varieties were used as Y variables for OPLS-DA analysis to obtain characteristic markers of sample variation. The results revealed that compounds influencing aroma variation might be related to ethyl isobutyrate, ethyl hexanoate-D, 2-methylpropanal, ethyl octanoate, ethyl butanoate-D, and Isoamyl acetate-D. Through QDA analysis of the aroma components of the nine varieties, the sensory evaluation results were consistent with the analysis results. The main factors affecting the differences in wine flavor are variety, cultivation techniques, environment, cultivation techniques, environment, fermentation process, aging time and storage conditions, etc. This study reflects the differences between different varieties of wines to a certain extent through micro-wine making and provides a reference for wine development and promotion. In the future, the flavor of wine can be enhanced from different angles to develop wines with Chinese characteristics and increase the share of Chinese wines on the international market.  Data Availability Statement: All related data and methods are presented in this paper. Additional inquiries should be addressed to the corresponding author.