Comparison of Potent Odorants in Traditional and Modern Types of Chinese Xiaoqu Liquor (Baijiu) Based on Odor Activity Values and Multivariate Analyses

Predominant odorants in modern and traditional types of Chinese xiaoqu liquor (Baijiu) were identified and compared by the combined use of gas chromatography−olfactometry, odor activity values (OAVs), and multivariate analyses. A total of 79 aroma compounds were identified in a typical modern type xiaoqu Baijiu (M) and a typical traditional type xiaoqu Baijiu (T), 42 of them had OAV > 1 in both M and T samples. The main differences between the two samples were obtained for the concentration of 23 aroma-active compounds. A total of 22 samples made by different brewing processes were analyzed to confirm the differences. Partial least squares discriminant analysis confirmed that 20 compounds could be used as potential markers for discrimination between modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu. Their difference in content is between 1.5 and 17.9 times for modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu. The results showed the aroma characteristics of modern and traditional type xiaoqu Baijiu clearly and comprehensively, which will provide guidance for modern Baijiu quality control and evaluation.


Introduction
Traditional fermented foods and beverages (TFFB) have been an important part of the human diet since the beginning of civilization [1]. TFFB remain widely favored by people because of their unique flavor and high nutritional value [2]. TFFB are primarily produced through largely uncontrolled spontaneous inoculation methods [3]. However, many drawbacks in traditional fermentation are directly or indirectly caused by a lack of control. These include low production efficiency, lack of consistency in product quality, and lengthy fermentation times. Thus, modern industrial development is essential in TFFB. Traditional fermentation process evolution and modernization have succeeded in many TFFB, such as soy sauce in the East [4] and cheese in the West [5].
Baijiu, Chinese liquor, is one of the most popular TFFB in China [6], which has thousands of years of history [7]. The annual production of Baijiu reached 7.41 billion liters and the sales revenue of Baijiu reached about 90 billion dollars in 2020, which played an important role in China's beverage industry [8]. The traditional fermentation and manufacturing methods of Baijiu have existed for centuries and strongly rely on individual operation skills and experience [6]. Traditional type Baijiu is produced using spontaneous fermentation processes in an open environment, and the fermentation temperature varies with the season. With the development of numerous technological innovations, the modern brewing technological process has been applied to Baijiu production, which not only improved the quality of Baijiu but also reduced energy consumption [9,10]. Mechanized operations are used in all stages of the brewing process and information technology is used to control the production environment in a relatively closed and stable environment, which maximally reduces the impact of the environment and operators during the brewing process.
There are very few publications comparing modern type and traditional type Baijiu. Wang et al. [11] separately studied the microbial succession and metabolism changes during Baijiu fermentation of traditional and modern workshops. The study showed that the microbial abundance of the traditional workshop is higher and the environmental microbiota have an important influence on the flavor of Baijiu. Gong et al. [12] explored the differences of volatile and nonvolatile compounds between traditional and mechanical raw Baijiu of sesame-like aroma-type Baijiu. The result showed that except for acids, phenols, and lactones, the total concentrations of the other compounds in traditional raw Baijiu were higher than those in mechanically produced raw Baijiu. Sun et al. [13] characterized 33 and 32 odor-active compounds in traditional and modern type xiaoqu Baijiu, respectively, by aroma extract dilution analysis and odor activity values (OAVs). However, the study was focused on only two samples and did not reveal the difference between the aroma compounds of the two types of xiaoqu Baijiu. The aroma is the most important characteristic of alcoholic beverages, and the production process is an important factor affecting the aroma profile [14]. So, the characterization of the aroma difference between traditional and modern-type Baijiu is of interest for improving modern brewing technology.
Xiaoqu Baijiu, one of the oldest and most popular Baijiu in China, is the first and most successful one to which the fully mechanized modern brewing process has been applied. Therefore, it is necessary to understand fully the similarities and differences of aromatic compounds in the Baijiu manufactured by different production processes. The primary objectives of this study were: (1) to identify the major aroma-active compounds in traditional and modern type xiaoqu Baijiu using gas chromatography-olfactometry (GC-O) and gas chromatography-mass spectrometry (GC-MS), (2) to confirm the contribution of those aroma compounds by quantitative analysis and their OAVs, and (3) to verify the aroma difference between the two types of xiaoqu Baijiu using multivariate statistical techniques.

Baijiu Samples
A total of 24 xiaoqu Baijiu raw liquors were analyzed in this study, including 12 modern type of xiaoqu Baijiu and 12 traditional type xiaoqu Baijiu. All samples were manufactured by Jing Brand Co., Ltd. (Daye, China) in 2019. The representative modern and traditional type Baijiu samples were chosen by the sensory panel composed of five national Chinese Baijiu tasting judges in Jing Brand Co., Ltd. Among the obtained xiaoqu Baijiu samples, one modern type xiaoqu Baijiu sample (M) and one traditional type xiaoqu Baijiu sample (T) were selected for liquid-liquid extraction (LLE) and GC-O analysis. All the samples were stored at 4 • C until analysis. The brand name is mentioned here for clear labeling, not for advertising purposes.

Extraction of Volatile Compounds
The extraction method was modified from a previous study [15]. The M and T samples (100 mL each) were diluted to 10 vol% with boiled ultrapure water, saturated with NaCl, and then extracted three times using freshly distilled CH 2 Cl 2 (100 mL each time) in a separatory funnel. The combined extracts were further separated into acidic/water-soluble and neutral/basic fractions with the following method.
The combined extract (about 300 mL) was washed three times with Na 2 CO 3 (50 mL each time, 0.2 mol/L, pH 10.0) and then washed with 30 mL of saturated NaCl solution. The obtained organic phase, containing the neutral/basic aroma compounds, was named the NBF fraction. The combined aqueous phase was acidified to pH 2.0 with HCl (4.0 mol/L) and extracted three times with freshly distilled CH 2 Cl 2 (50 mL each time). The combined extract (about 150 mL), containing the remaining acidic aroma compounds, was named the AF fraction. Afterward, each fraction was dried with anhydrous Na 2 SO 4 overnight and concentrated to a final volume of 0.5 mL under a gentle stream of nitrogen. These concentrated NBF and AF were stored at −20 • C before the GC-O analysis.

Identification of Aroma Compounds Using GC-MS and GC-O
GC-MS and GC-O analyses were performed on an Agilent 6890N GC equipped with an Agilent 5975 mass selective detector and a sniffing port (ODP 2, Gerstel, Württemberg, Germany). The sample was analyzed on both a DB-FFAP column (60 m × 0.25 mm i.d.; 0.25 µm film thickness, Agilent Technologies, Inc., Santa Clara, CA,) and a DB-5 column (30 m × 0.25 mm i.d.; 0.25 µm film thickness, Agilent Technologies, Inc.). High-purity helium (>99.999%) was used as the carrier gas at a constant flow rate of 2 mL/min. The concentrated extract (1 µL) was injected into the GC injector in splitless mode. The GC injector temperature was 250 • C. The oven temperature was held at 50 • C for 2 min, then raised to 230 • C at a rate of 6 • C/min and held for 15 min. The mass spectrometer (MS) was operated in electron ionization mode at 70 eV. The time of solvent delay was 5 min, and the temperature of the ion source was 230 • C. The mass range was set from 35 to 350 amu in full-scan mode.
Four well-trained panelists (two females and two males, 22 years old on average) from the Laboratory of Brewing Microbiology and Applied Enzymology at Jiangnan University were selected for the GC-O study. During a GC run described above, a panelist placed his/her nose close to the sniffing port, responded to the aroma intensity of the stimulus, and recorded the retention time, descriptors, and intensities of the odor peak for each compound. The Osme values reflecting the aromatic intensity of the stimulus based on a six-point scale ranging from 0 to 5 was used for intensity judgment: 0 = none, 1 = very weak, 2 = weak, 3 = moderate, 4 = strong, and 5 = very strong. Every sample was sniffed twice by each panelist, if the aroma was detected six or more times on each column, an odor location was identified. The Osme value for aroma intensity was an average result of the four panelists.
The aroma compounds of the odor location were identified by comparing the aroma, mass spectrum (MS), and RIs to the pure standards (S). RIs were calculated based on the linear retention times of the n-alkanes (C 6 -C 30 ) in both the DB-FFAP and DB-5 columns under the same chromatographic conditions.

Quantitation of Aroma-Active Compounds by HS-SPME-Arrow Combined with GC-MS
The aroma compounds in all xiaoqu Baijiu samples were quantitated using headspace solid-phase microextraction arrow (HS-SPME-Arrow) combined with GC-MS, as described previously [16]. An SPME-Arrow automatic headspace sampling system (CTC Analytics AG, Switzerland) with a 120 µm divinylbenzene/carbon wide range/polydimethylsiloxane (DVB/CAR WR/PDMS) fiber (CTC Analytics AG, Basel, Switzerland) was used for extraction and injection. A total of 5 mL of the diluted xiaoqu Baijiu sample (10 vol%) was transferred into a 20 mL headspace glass vial, saturated with 1.5 g NaCl, and spiked with 40 µL mixed ISs (final concentration: IS1, 146 µg/L; IS2, 323 µg/L; IS3, 1 197 µg/L; IS4, 294 µg/L; IS5, 307 µg/L). Tightly capped with a Teflon-faced silicone septum, the sample was equilibrated at 45 • C for 5 min and extracted for 45 min at the same temperature under stirring of 250 rpm. After extraction, the fiber was immediately inserted into the injection port of the GC for desorption at 250 • C for 5 min. The samples were separated on a DB-FFAP column with splitless mode. The oven temperature was held at 40 • C for 2 min, increased at 4 • C/min to 150 • C, then held for 2 min, raised at 6 • C/min to 230 • C and held for 5 min.
Every standard stock solution was prepared by dissolving a standard compound in the model synthetic solution (10% alcohol by volume at pH 4.0) and then diluting to 10 different concentrations. The same amount of ISs as in the xiaoqu Baijiu samples was added to the standard solutions, and then the working standard solutions were extracted using HS-SPME-Arrow and detected using GC-MS with the same conditions as for sample analysis. The standard curves were created by plotting the ratio of the peak area of the reference compounds to the corresponding IS against their concentration ratio. The analytical limit of detection (LOD) of aroma compounds was obtained from the lowest concentrations of the analyte standard solutions based on a signal-to-noise ratio of 3. The limits of quantitation (LOQ) was defined as the lowest concentration of the calibration curve based on a signal-to-noise ratio of 10. The accuracy is reported as the percentage recovery of a known amount of target analyte added to the xiaoqu Baijiu sample. All analyses were performed in triplicate.

Statistical Analysis
Principal component analysis (PCA) was performed to explore the clustering of the xiaoqu Baijiu samples in terms of their aroma compounds. To maximize the separation among samples and to identify the aroma compounds responsible for the separation, partial least squares discriminant analysis (PLS-DA) was used. PCA and PLS-DA were conducted using SIMCA version 14 software (Umetrics, Umearing, Sweden). The heatmap and hierarchical clustering analysis (HCA) were performed using HemI 1.0 (The CUCKOO Workgroup, http://hemi.biocuckoo.org, accessed on 4 September 2021).

Identification of Aroma Compounds in Xiaoqu Baijiu Using GC-O
By extraction with CH 2 Cl 2 , an organic phase was obtained that exactly showed the typical xiaoqu Baijiu aroma when sniffed on a filter paper by the panelists. Thus, it was appropriate and reliable to be used to analyze the profile of aroma compounds in xiaoqu Baijiu. To facilitate GC-O analysis and compound identification, the extract obtained was further separated into AF and NBF fractions to reduce complexity as previously described [15]. To identify the constituents responsible for the odors, aroma areas were first located using GC-O. The RI of each odorant regions on the DB-FFAP and DB-5 columns was calculated. Then, the aroma quality and odor descriptor of each odorant were determined by comparison to a Chinese Baijiu flavor database (an in-house database consisting of more than 900 odor-active standard compounds). The identities of the tentatively assigned aroma compounds were finally confirmed by comparing their MS data (EI mode) with those of the standard reference compounds.
As shown in Table 1, a total of 79 odorants were located in the M and T samples in this study. The results showed that both liquors had a similar aroma profile, but the Osme values varied ( Figure 1). Most of these components are well-known odorants in Baijiu [7,17,18], but 36 of them have not been reported in a previous study of xiaoqu Baijiu [13]. Among the 79 aroma compounds, 26 esters, 15 alcohols, 12 aldehydes and ketones, nine acids, six phenols, five terpenoids, four furans, and two sulfur compounds were detected in the M and T samples.    Table 2, the abscissa represents the retention index of the compounds on the DB-FFAP chromatographic column, and the ordinate represents the aroma intensity of the compounds. The second dominant group of aroma compounds identified were alcohols. Among the alcoholic compounds identified in the M sample, 3-methylbutanol, 1-octen-3-ol, and β-phenethyl alcohol showed the highest Osme values of 3.65, contributing whiskey, mushroom, and honey odors, respectively. However, 2-methylpropanol with the Osme value of 3.52 was the most important aroma contributor in the T sample.
Based on GC-O and GC-MS, eight aroma aldehydes were detected. Aromatic aldehydes seemed to be important in xiaoqu Baijiu, where 2-methylpropanal (Osme values =  Table 2, the abscissa represents the retention index of the compounds on the DB-FFAP chromatographic column, and the ordinate represents the aroma intensity of the compounds. The second dominant group of aroma compounds identified were alcohols. Among the alcoholic compounds identified in the M sample, 3-methylbutanol, 1-octen-3-ol, and β-phenethyl alcohol showed the highest Osme values of 3.65, contributing whiskey, mushroom, and honey odors, respectively. However, 2-methylpropanol with the Osme value of 3.52 was the most important aroma contributor in the T sample.

Quantitation and OAVs of Aroma-Active Compounds in Xiaoqu Baijiu
To gain a deeper insight into the aroma-active compounds of xiaoqu Baijiu, the 79 odorants detected using GC-O were quantitated by constructing standard curves for the M and T samples. The quantitative ion, IS, calibration curve, recovery, and LOQ of each aroma compound are shown in Table 2 and Figure 2. Quantitative results showed that 1,1dimethoxyethane, ethyl acetate, 1-propanol, 2-butanol, 2-methylpropanol, 3-methylbutanol, and acetic acid showed relatively high concentrations (above 100 mg/L) in both samples.  Figure 2, A is the concentration difference between ester compounds, and B is the concentration difference between others.

Confirmation of the Key Compounds Related to the Aroma Profile Differences between Modern and Traditional Type Xiaoqu Baijiu
As concluded above, the odorant compositions of M and T infusions were similar, but the Osme value and OAV of each substance were different. However, these were just the results based on two samples, not enough to represent all the aroma characteristics of modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu. To determine if the differences between M and T samples can also be found in an additional 22 different types of xiaoqu Baijiu samples, which were randomly selected from the modern and traditional workshops, the 79 aroma compounds were quantitated. Then, PCA and PLS-DA analyses It is known that the contribution to Baijiu depends not only on the concentration of an odorant but also its odor threshold. So, for further confirmation of the contribution of aroma-active compounds, their OAVs were calculated. The OAV was obtained from a compound concentration divided by its odor threshold [22]. Herein, the major-ity of the threshold values were taken from the literature [23][24][25]. As shown in Table 3 . These results revealed the important aroma contribution of esters, especially ethyl esters, for xiaoqu Baijiu liquors. Esters, which contribute a strong fruity odor, were also one of the most important aroma groups for other alcoholic beverages, such as wine [26], rice wine [27], and other types of Baijiu [7,15,25,28]. Besides, 3-methylbutanal and acetal ranked at the second and fifth places, respectively, according to their OAV. It is worth noting that the content of β-damascenone and dimethyl trisulfide was not high but the OAVs ranked at second and third places among all aroma compounds in xiaoqu Baijiu. This is because their thresholds are present at trace amounts in Baijiu [23,25]. β-Damascenone giving a honey note was previously reported as a key odorant with high OAV in the light aroma type of Baijiu [25,29]. Dimethyl trisulfide, exhibiting cabbage notes, was one of the key aroma compounds in Zhima aroma-type Baijiu and Moutai aroma-type Baijiu [24,30]. The other important odoractive volatiles with OAVs greater than or equal to 10

Confirmation of the Key Compounds Related to the Aroma Profile Differences between Modern and Traditional Type Xiaoqu Baijiu
As concluded above, the odorant compositions of M and T infusions were similar, but the Osme value and OAV of each substance were different. However, these were just the results based on two samples, not enough to represent all the aroma characteristics of modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu. To determine if the differences between M and T samples can also be found in an additional 22 different types of xiaoqu Baijiu samples, which were randomly selected from the modern and traditional workshops, the 79 aroma compounds were quantitated. Then, PCA and PLS-DA analyses were conducted based on the concentration of the aroma compounds in the 22 xiaoqu Baijiu samples.
PCA was performed to show a trend of intergroup separation on the scores plot of data obtained in both positive and negative modes. The PCA results showed that five principal components (PCs) with eigenvalues higher than one were obtained from the data sets, whose cumulative variance proportion was 74.5%. PC1 accounted for 46.7% of the total variance while PC2 explained 7.02%, which explained 55.1% of the total variance in the data set. As shown in Figure 3A, it was found that 22 xiaoqu Baijiu samples were well separated. PC1 clearly distinguished modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu, traditional type xiaoqu Baijiu was on the positive side of PC1, while modern type xiaoqu Baijiu appeared on the negative side, indicating that there were obvious differences in aroma compounds' features.   [22,23]. c Odor thresholds taken from reference [24]. d Odor thresholds taken from reference [25]. e Odor thresholds taken from reference [31]. f Odor thresholds taken from reference [13]. g The concentration ratio of M to T. h The concentration ratio of T to M.Identified odorants with high OAVs can also be used as indicators to assess aroma differences objectively in the two samples. Among the compounds with OAVs ≥ 1, we found that the concentrations of ethyl octanoate, β-damascenone, 3-methylbutanal, ethyl 3-methylbutanoate, benzeneacetaldehyde, 1-octen-3-ol, guaiacol, 1,1-dimethoxyethane, 3-methylbutanol, ethyl dodecanoate, γ-nonanolactone, and β-phenethyl alcohol, were 1.50-17.9 times higher in the M than in the T sample. In contrast, the concentrations of ethyl butanoate, ethyl pentanoate, ethyl hexanoate, 1-butanol, 2-butanol, butanoic acid, 1,1,3-triethoxypropane, 4-ethylphenol, 4-methylguaiacol, ethyl 3-phenylpropanoate, and 4-ethylguaiacol were 1.52-24.77 times higher in the T sample than in the M sample. The 23 aroma compounds might be potential compounds responsible for the differences in the aroma profile between M and T samples. Figure 2 shows the concentration of the 12 aroma-active compounds (concentration ratio ≥ 2) in the M and T samples. the randomly permuted models ( Figure 3C). Variables with absolute variable importance for projection (VIP) values greater than 1 were selected as potential markers explaining the differences. As shown in Figure 3D, a total of 35 compounds had a VIP value greater than 1, and 20 out of the 35 compounds were important contributors to the aroma profile of xiaoqu Baijiu, which indicated that they were highly relevant to the difference between modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu samples. These 20 substances cover most of the different substances obtained by comparing the concentrations of the M and T samples. This showed that the multivariate statistical analyses of 22 samples verified the results of Section 3.2. To compare further the content of each substance in the two types of xiaoqu Baijiu, the concentrations of 20 potential aroma-active compounds in the samples from both the modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu were used to prepare HCA analysis, and the color (from light to dark) indicated the concentration change from low to high. As shown in Figure 4, the HCA clearly classified the aroma markers into two clusters. Cluster A consists of the aroma-active compounds with significantly higher concentrations in traditional type xiaoqu Baijiu than in modern type xiaoqu Baijiu, which included five ethyl esters (ethyl 3-phenylpropanoate, ethyl propanoate, ethyl butanoate, To maximize separation and identification of the aroma compound markers between the two types of xiaoqu Baijiu, a PLS-DA model was created for the 22 xiaoqu Baijiu samples based on the concentration data set. This type of supervised pattern recognition method provides a deeper analysis of the main matrix characteristics of samples and differs from PCA by the addition of grouping variables that indicate the category of samples. As shown in Figure 3B, the two types of xiaoqu Baijiu samples were well distinguished by a dependable PLS-DA model. The explained variation (R 2 Y) and the predictive capability (Q 2 Y) for the model were 0.993 and 0.978, respectively, which indicated that the model was excellent because Q 2 Y was close to 1. In addition, permutation tests (n = 200) were performed to evaluate whether the discriminant models were overfitting the data [32]. The permutation tests randomly rearranged the experiments by changing the sort order of the classification variables (Y) and randomly assigned Q 2 Y up to 200 times. The low intercepts (R 2 = 0.142, Q 2 = −0.331) were an indication of the validity of the original model. The R 2 and Q 2 values of the permutation test revealed that the initial model outperformed the randomly permuted models ( Figure 3C). Variables with absolute variable importance for projection (VIP) values greater than 1 were selected as potential markers explaining the differences. As shown in Figure 3D, a total of 35 compounds had a VIP value greater than 1, and 20 out of the 35 compounds were important contributors to the aroma profile of xiaoqu Baijiu, which indicated that they were highly relevant to the difference between modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu samples. These 20 substances cover most of the different substances obtained by comparing the concentrations of the M and T samples. This showed that the multivariate statistical analyses of 22 samples verified the results of Section 3.2.
To compare further the content of each substance in the two types of xiaoqu Baijiu, the concentrations of 20 potential aroma-active compounds in the samples from both the modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu were used to prepare HCA analysis, and the color (from light to dark) indicated the concentration change from low to high. As shown in Figure 4, the HCA clearly classified the aroma markers into two clusters. Cluster A consists of the aroma-active compounds with significantly higher concentrations in traditional type xiaoqu Baijiu than in modern type xiaoqu Baijiu, which included five ethyl esters (ethyl 3-phenylpropanoate, ethyl propanoate, ethyl butanoate, ethyl pentanoate, and ethyl phenylacetate), three phenols (4-ethylphenol, 4-ethylguaiacol, and 4-methylguaiacol), pentanoic acid, β-ionone, and 2-butanol.  The ethyl esters in Baijiu were mainly produced in the fermentation or aging stages through the esterification of acids and alcohols [7,33]. This study found that the contents of propanoic acid, butanoic acid, and pentanoic acid in traditional type xiaoqu Baijiu were higher than those in modern type xiaoqu Baijiu, so the organic acids had a significant influence on the yield of aroma esters. Phenols, as an important trace component in Baijiu, play an important role in the aroma, taste, and stability of Baijiu [34]. It was reported that The ethyl esters in Baijiu were mainly produced in the fermentation or aging stages through the esterification of acids and alcohols [7,33]. This study found that the contents of propanoic acid, butanoic acid, and pentanoic acid in traditional type xiaoqu Baijiu were higher than those in modern type xiaoqu Baijiu, so the organic acids had a significant influence on the yield of aroma esters. Phenols, as an important trace component in Baijiu, play an important role in the aroma, taste, and stability of Baijiu [34]. It was reported that phenols come from the thermal degradation of lignin in brewing raw materials or are produced by amino acids and ferulic acid under the action of microorganisms [35,36]. The brewing materials of modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu are the same, therefore, it may be that the different types and abundance of microorganisms caused the higher content of phenols in traditional type xiaoqu Baijiu. β-Ionone with a floral aroma is also a very important aroma compound and bioactive substance in Baijiu [37]. Besides, 2-butanol is one of the higher alcohols in Baijiu, it is mostly synthesized by microorganisms (mainly Saccharomyces cerevisiae and bacteria) during fermentation via catabolic and anabolic pathways [38,39]. The fermentation temperature of traditional type xiaoqu Baijiu is highly susceptible to climatic conditions, but the fermentation temperature of modern type xiaoqu Baijiu is controlled at a suitable temperature, which could effectively inhibit the production of higher alcohols [13]. Cluster B consists of the aroma-active compounds with significantly higher concentrations in modern type xiaoqu Baijiu than in traditional type xiaoqu Baijiu, which included three esters (ethyl 3-methylbutanoate, ethyl octanoate, and ethyl decanoate), guaiacol, γ-nonanolactone, β-phenethyl alcohol, βdamascenone, benzeneacetaldehyde, and 1-octen-3-ol. Not surprisingly, the acids content corresponding to these three esters in modern type xiaoqu Baijiu was also higher than in traditional type xiaoqu Baijiu. β-Phenethyl alcohol with a rose-like aroma is a very important aroma compound in Baijiu. The processes by which the key odorant is formed have been well-established and found to be influenced by various yeast species present during the fermentation of liquor from cereal and legume materials [40]. β-Damascenone, presenting honey and floral aromas, was previously reported as an important odorant in whiskey, rum, brandy, and Baijiu [25,28,[41][42][43]. Luigi and Peter [43] assumed that cereals could contain the precursor (3-hydroxy-7,8-dihydro-β-ionol) leading to the generation of β-damascenone during fermentation and/or distillation.

Conclusions
In summary, a total of 79 aroma compounds were identified using GC-O and GC-MS in both M and T samples. Among them, 42 aroma compounds were further recognized as the important aroma-active compounds in both M and T owing to their relatively high OAVs. Twenty-three aroma-active compounds might be the potential compounds responsible for the differences in the aroma profile between M and T samples because of a concentration ratio higher than 1.5. Moreover, to determine if the 23 aroma-active compounds can represent the difference in aroma compounds' profile between traditional type xiaoqu Baijiu and modern type xiaoqu Baijiu, multivariate analyses of the 22 xiaoqu Baijiu samples made by different brewing processes were applied, including PCA, PLS-DA, and HCA. Twenty odor-active compounds were confirmed as potential flavor markers for the differentiation of modern type xiaoqu Baijiu and traditional type xiaoqu Baijiu. The results in this study will help improve the flavor quality and processing technology of xiaoqu Baijiu.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest:
The authors declare no conflict of interest.