Method for deducing and analyzing baijiu odor compound and ratio limit thereof through ratio analysis and logic

Abstract

The invention provides a method for analyzing baijiu odor compounds and a ratio limit thereof through ratio analysis and logic derivation, which comprises the following steps: S1, calculating the ratio of the contents of the compounds in a baijiu sample, and performing Spearman correlation analysis on the compound contents, the compound content ratio and an odor score, the compound and the ratio are positively correlated with the peculiar smell; classifying the compounds through ratio logic derivation and flavor analysis, and screening potential odor compounds; s2, on the basis of the potential peculiar smell compounds, verifying and determining a final peculiar smell compound through an aroma reconstruction experiment and an aroma deficiency experiment; and S3, carrying out statistical analysis on the ratio of the final peculiar smell compound to ethyl hexanoate, and determining the limit ratio of the final peculiar smell compound. According to the method, a novel ratio analysis and logic derivation strategy is established, the odor compounds in the white spirit can be effectively identified and regulated, and a theoretical basis is provided for white spirit flavor control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of liquor identification, in particular to a method for analyzing off-odor compounds of liquor and their ratio limits through ratio analysis and logical deduction. BACKGROUND

[0002] English abbreviation explanation: COC: candidate off-odor compounds (potential off-odor compounds) CBC: candidate burnt-bitter-aroma compounds (potential burnt-bitter-aroma compounds) SC: synergistic compounds (synergistic compounds) MC: masking compounds (masking compounds) NEC: no effective compounds (no effective compounds) R: Ratio (ratio) UC: unsure compounds (unsure compounds) GC-MS: gas chromatography-mass spectrometry HPLC: high performance liquid chromatography BBA: burnt-bitter-aroma (burnt-bitter-aroma) OS: off-odor score (off-odor score) BBAS: burnt-bitter-aroma score (burnt-bitter-aroma score) ρ: spearman correlation coefficient (spearman correlation coefficient) OPLS-DA analysis (orthogonal partial least squares discriminant analysis) VIP (variable importance projection) Duo-Trio Test (two-three-point test method) SD (standard deviation) IQR (interquartile range) Chinese liquor is one of the six major distilled liquors in the world. Its brewing history is long, and twelve types of liquor have been formed, including Luzhou-flavor, Maotai-flavor, and Qing-flavor. The style characteristics of each type of liquor are mainly determined by the brewing process and regional environment. Among them, Luzhou-flavor liquor is famous for its outstanding cellar aroma, mellow and harmonious taste, and long aftertaste, and accounts for more than 50% of the domestic market share. However, due to its open solid-state fermentation process, it is easily affected by various controllable and uncontrollable factors during production, often producing off-flavors such as bitterness, sourness, astringency, bitter-burnt taste, muddy smell, and astringency, which not only reduces the quality of the liquor body, but also causes economic losses to the enterprise.

[0003] Currently, the regulation of off-flavor in Baijiu is mainly achieved by liquor body blending, process optimization, and raw material quality control. However, a complete understanding of the compounds responsible for off-flavor is required to effectively control it. In addition, there are complex flavor interactions between compounds, and the interaction between off-flavor-related compounds and other aroma components significantly affects sensory perception. In light-flavor Baijiu, acetic acid, butyric acid, hexanoic acid, lactic acid, and their corresponding ethyl esters together form the skeleton components that affect flavor, and their content plays a decisive role in the overall flavor quality (Lin Yuan, Yunjie Li, Liuyan Zheng, Yiyan Qin, Xin Zhang, Lijuan Ma, Huan Zhang, Liping Du. (2025). Impact of organic acids on aroma release in light-flavor Baijiu: A focus on key aroma-active compounds. Food Bioscience, 65. https: / / doi.org / 10.1016 / j.fbio.2025.106071). The formation of off-flavor may not only come from the flavor contribution of a single compound, but also from the imbalance of the relative proportion of multiple compounds. For example, when the ratio of n-propanol to ethyl hexanoate is 0.84:1, the fruitiness of Baijiu is about 8 points. However, as the content of n-propanol increases, the ratio increases to 3.36:1, and the fruitiness decreases to about 4 points, resulting in a significant decrease in flavor quality (Han Zhao, Lijuan Chai, Wei Zhang, Xiaojuan Zhan, Zhenming Lu, Songtao Wang, Caihong Shen, Jinsong Shi, Zhenghong Xu. (2025). Changes in flavor profile of sauce-flavor baijiu: Perceptual interactions between 1-propanol and aroma compounds. Food Chemistry: X, 25. https: / / doi.org / 10.1016 / j.fochx.2024.102153).

[0004] Despite the above understanding, there is still a relative lack of systematic research on the relationship between the content ratio of each compound and the formation of off-flavor in Baijiu and other types of liquor. SUMMARY

[0005] Therefore, the application provides a method for analyzing liquor odor compounds and their limit ratios by ratio analysis and logical deduction, so as to quickly and accurately determine the odor compounds and their content ratios in liquor.

[0006] To achieve the above object, the application adopts the following technical scheme: A method for analyzing liquor odor compounds and their limit ratios by ratio analysis and logical deduction, comprising: S1: calculating the content ratios of compounds in liquor samples, and performing Spearman correlation analysis on the compound content, the compound content ratio and the odor score (OS) to obtain the compounds and ratios positively correlated with odor; classifying the compounds by ratio logical deduction and flavor analysis, and screening potential odor compounds; S2: verifying and determining the final odor compounds based on the potential odor compounds through aroma reconstruction experiments and aroma loss experiments; S3: determining the limit ratio of the final odor compounds by statistical analysis of the ratio of the final odor compounds to ethyl hexanoate.

[0007] Further, the classification of the compounds by ratio logical deduction and flavor analysis in S1 to screen potential odor compounds specifically comprises: S1.1: first, for the compounds positively correlated with the odor score, according to the flavor characteristics of the compounds, the compounds with unpleasant flavor are classified as the first part of potential odor compounds (COC-1), and the compounds with aroma or no odor are classified as the first part of potential synergistic compounds (SC-1); S1.2: second, for the ratios positively correlated with the odor score, the ratio R1 list containing the CBC-1 compound in the numerator or denominator is screened out from the ratios, and the compounds corresponding to the numerator and the denominator of each ratio in the ratio R1 list are classified as COC (potential odor compounds), SC (synergistic compounds), MC (masking compounds) or NEC (ineffective compounds) according to the five types of combinations of COC / COC, COC / SC, COC / MC, COC / NEC and SC / COC; S1.3: screening out the ratio R2 list containing the SC-1 compound in the numerator or denominator, and classifying the compounds corresponding to the numerator and the denominator of each ratio in the ratio R2 list as COC, SC, MC or NEC according to the five types of combinations of COC / SC, SC / COC, SC / SC, SC / MC and SC / NEC; S1.4: further divide the potential off-flavor compounds obtained in steps S1.2 and S1.3 according to their flavor characteristics, specifically, identify the compounds presenting unpleasant flavor as second part of potential off-flavor compounds (COC-2), and classify the compounds with aroma or no odor as "uncertain compounds (UC)" category; S1.5: according to the compounds in COC-2 in S1.4, perform steps S1.2 and S1.4, thereby obtaining CBC-3; S1.6: in the same way, according to the newly identified COC-n, repeat the steps S1.2 and S1.4 until all ratios are classified, wherein n = 1, 2, 3,....

[0008] Further, the step S1.6 further comprises, if no new COC is found, the classification cycle is terminated, at this time, it is judged whether the flavor properties of the compounds in the remaining ratios have unpleasant flavor, if not, the process is completed; otherwise, the previous operation is checked back to check whether it is correct.

[0009] Further, the step S1 further comprises: S1.7: combine the screened COC-1, COC-2,..., and COC-n to construct a COC total list; for the compounds with multiple labels and cannot be clearly classified, they are uniformly divided into "uncertain compounds", finally, three compound category lists are obtained: COC, SC and UC.

[0010] Further, the potential off-flavor compounds include butyric acid, heptanoic acid, hexanoic acid, acetic acid, isovaleric acid, valeric acid, furfuryl alcohol, n-octanol, isoamyl alcohol, isobutyl alcohol, n-butyl alcohol, n-hexyl alcohol, n-pentyl alcohol, sec-butyl alcohol, isoamyl aldehyde, benzaldehyde, nonanal diethyl acetal, p-cresol, 2,4-di-tert-butyl phenol and ethyl furfuryl ether.

[0011] Further, the step S2 specifically comprises: S2.1: aroma reconstruction experiment: add the potential off-flavor compounds screened in step S1 to the reconstruction matrix to prepare an off-flavor reconstruction model, and score the intensity of the flower aroma, fruit aroma, sweet aroma, sour aroma, cellar aroma, alcohol aroma and bitter taste, and determine whether there is an off-flavor compound in the potential off-flavor compound according to its flavor profile; S2.2: aroma loss experiment: classify the potential off-flavor compounds into acid, alcohol, aldehyde, acetal, phenol and ether, each time systematically remove one category or a single compound from the reconstruction model, and identify whether the loss has a significant impact on the flavor profile of the reconstruction model by two-three point test method, thereby determining seven final off-flavor compounds: n-butyl alcohol, n-hexyl alcohol, sec-butyl alcohol, furfuryl alcohol, 2,4-di-tert-butyl phenol, p-cresol and ethyl furfuryl ether.

[0012] Further, the upper limit of the ratio of the seven final off-flavor compounds to ethyl hexanoate is 0, 0.000084, 0.0018, 0.2862, 0.0943, 0.1559, and 0.0443, respectively.

[0013] Compared with the prior art, the application has the following beneficial effects: 1. The application establishes a new proportion analysis strategy, which can effectively identify and regulate off-flavor compounds in liquor, and provides a theoretical basis for flavor control of liquor.

[0014] 2. Although the proposed proportion limit is based on a limited sample set, future research can use this proportion logic derivation in a larger population to improve the blending standard through meta-analysis. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the application, and therefore should not be considered as a limitation to the scope. For those skilled in the art, other related drawings can also be obtained without creative labor.

[0016] Figure 1 Figure 1 is a flowchart of a method for analyzing and logically deriving off-bitter compounds and their ratio limits in liquor by ratio analysis.

[0017] Figure 2 Figure 2 (A) is a network diagram of the Spearman correlation coefficient between off-bitter compounds, off-bitter compound ratios, and off-bitter scores, Figure 2 Figure 2 (B) is a flowchart for classifying compounds in liquor. Figure 2 Figure 2 (C) is 20 potential off-bitter compounds screened out.

[0018] Figure 3 Figure 3 (A) is a flavor profile diagram of the reconstructed matrix and the reconstructed model, Figure 3 Figure 3 (B) is a flavor profile diagram of the reconstructed model and the off-bitter liquor sample.

[0019] Figure 4 Figure 4 is a ratio analysis diagram of off-bitter compounds and ethyl hexanoate. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the embodiments of the application more clear, the following will combine the drawings in the embodiments of the application to clearly and completely describe the technical solutions in the embodiments of the application. Obviously, the described embodiments are some of the embodiments of the application, not all the embodiments.

[0021] In the specific embodiment, the off-flavor is a bitter taste, and in other embodiments, the off-flavor can also be other unpleasant odors, which are not limited in the present application.

[0022] The present application determines 89 compounds in liquor samples by GC-MS (gas chromatography-mass spectrometry) and HPLC (high performance liquid chromatography). Most of the compounds are present in higher concentrations in normal liquor samples. Notably, 2,4-di-tert-butyl phenol is unique to off-flavor liquor. Through OPLS-DA analysis (orthogonal partial least squares discriminant analysis), 13 compounds with a VIP (variable importance projection) value greater than 1 are screened out, including ethyl hexanoate, ethyl acetate, ethyl lactate, ethyl valerate, ethyl butyrate, hexyl acetate, lactic acid, hexanoic acid, 1,1-diethoxy-3-methylbutane, n-butanol, furfuryl alcohol, n-hexanol, and 3-methyl-1-butanol. Except for 3-methyl-1-butanol and lactic acid, the remaining 11 compounds in normal liquor reach 1.12-3.23 times the content of bitter liquor. Notably, although n-butanol, furfuryl alcohol, n-hexanol, and other compounds are associated with off-flavor, their ratios to the four main esters are low in normal liquor, while their ratios to the four main acids show the opposite trend. This indicates that there are limitations in determining off-flavor components based solely on absolute content, and the formation of off-flavor may involve complex flavor interactions, which requires comprehensive consideration of the ratios between compounds.

[0023] Therefore, the present application innovatively introduces a ratio-based analysis method to identify off-flavor compounds. In the specific embodiment, 1830 ratios between 89 compounds are calculated using Python software, and then the compound content, compound ratio, and off-flavor score are subjected to Spearman correlation analysis, and 9 compounds and 454 ratios are positively correlated with the off-flavor score. Through logical deduction based on flavor contribution categories (such as direct contribution, synergistic effect, masking effect, and no effect) and flavor characteristic analysis, 20 potential off-flavor compounds are finally screened out. Through aroma reconstruction loss experiments, it is determined that furfuryl alcohol, n-butanol, sec-butanol, n-hexanol, 2,4-di-tert-butyl phenol, p-cresol, and ethyl furfuryl ether are the contributors to the bitter taste (off-flavor) in strong-flavor liquor. Notably, ethyl furfuryl ether is mentioned for the first time in the formation of bitter taste (off-flavor). To control the bitter taste (off-flavor), it is recommended that the upper limits of the ratios of 2,4-di-tert-butyl phenol, p-cresol, ethyl furfuryl ether, furfuryl alcohol, n-butanol, sec-butanol, and n-hexanol to ethyl hexanoate be controlled below 0, 0.000084, 0.0018, 0.2862, 0.1559, 0.0443, and 0.0943, respectively.

[0024] As shown in Figure 1 The present application provides a method for analyzing off-flavor compounds in liquor and their ratio limits through ratio analysis and logical deduction, comprising: S1: Calculate the ratio of compounds in the liquor sample, and perform Spearman correlation analysis on the compound content, compound ratio and off-flavor score (as shown in FIG. Figure 2 A), obtain the compounds and ratios that are positively correlated with the off-flavor score; classify the compounds through ratio logical deduction and flavor analysis, and screen potential off-flavor compounds; Specifically, the ratio of the content of all compounds in the liquor sample is calculated using Python software, Spearman correlation analysis is performed on the compound content, ratio and off-flavor score, and the compounds and ratios that are positively correlated with the off-flavor score are obtained. The compound that is positively correlated with the off-flavor score refers to a compound whose Spearman correlation coefficient between the compound content and the bitter taste score (BBAS) is greater than 0.5. The ratio that is positively correlated with the bitter taste score refers to a ratio whose Spearman correlation coefficient between the compound content ratio and the bitter taste score is greater than 0.5. For the compounds that are positively correlated with the off-flavor score, classification is performed through flavor characteristic analysis, thereby screening potential off-flavor compounds. For the ratios that are positively correlated with the off-flavor score, logical deduction and flavor characteristic analysis are performed on the compounds corresponding to the numerator and denominator in the ratio, thereby screening potential off-flavor compounds. S2, based on the potential off-flavor compounds, verify and determine the final off-flavor compounds through aroma reconstruction experiments and aroma loss experiments; S3: Determine the limit ratio of the final off-flavor compounds by statistical analysis of the ratio of the final off-flavor compounds to ethyl hexanoate.

[0025] The S1 specifically includes (as shown in FIG. Figure 2 B): S1.1: First, for the compounds that are positively correlated with the off-flavor score, according to the flavor characteristics of the compounds, the compounds with unpleasant flavor are classified as the first part of the potential bitter taste compounds (CBC-1), and the compounds with aroma or no odor are classified as the first part of the synergistic compounds (SC-1); S1.2: Second, for the ratios that are positively correlated with the bitter taste score, the ratios R1 list containing CBC-1 compounds are screened from the ratios, and the compounds corresponding to the numerator and denominator of each ratio in the R1 list are classified as CBC, SC, MC or NEC according to the five types of combinations of CBC / CBC, CBC / SC, CBC / MC, CBC / NEC and SC / CBC. S1.3: Screening out the ratio R2 list of the molecule or denominator containing SC-1 compounds, and classifying the compounds corresponding to the molecule and denominator of each ratio in the R2 list into CBC, SC, MC or NEC according to the five types of combinations of CBC / SC, SC / CBC, SC / SC, SC / MC and SC / NEC; S1.4: Further dividing the potential bitter taste compounds (CBC) obtained in steps S1.2 and S1.3 according to their flavor characteristics, and the specific division method is: identifying the compounds with unpleasant flavor as the second part of the potential bitter taste compounds (CBC-2), and classifying the compounds with aroma or no odor as the "uncertain compound (UC)" category; S1.5: According to the compounds in CBC-2 in S1.4, performing steps S1.2 and S1.4 to obtain CBC-3; S1.6: Similarly, according to the newly identified CBC-n, repeat steps S1.2 and S1.4 until all ratios are classified, and n = 1, 2, 3,....

[0026] Further, in step S1.6, if no new CBC is found, the classification cycle is terminated, and at this time, it is judged whether the flavor properties of the compounds corresponding to the remaining ratios have unpleasant flavor or not, and if not, the process is completed; otherwise, the previous operation is checked back.

[0027] Further, the step S1 further comprises: S1.7: Combining the screened CBC-1, CBC-2, …, and CBC-n to construct a CBC total list; for the compounds with multiple labels and cannot be clearly classified, they are uniformly divided into "uncertain compounds (UC)", finally, three compound category lists are obtained: CBC, SC and UC.

[0028] Further, the potential off-taste compounds include butyric acid, heptanoic acid, hexanoic acid, acetic acid, isovaleric acid, valeric acid, furfuryl alcohol, n-octanol, isoamyl alcohol, isobutyl alcohol, n-butyl alcohol, n-hexyl alcohol, n-pentyl alcohol, sec-butyl alcohol, isoamyl aldehyde, benzaldehyde, nonanal diethyl acetal, p-cresol, 2,4-di-tert-butyl phenol and ethyl furfuryl ether.

[0029] In order to identify the potential bitter taste compounds, the application not only bases on the absolute content, but also combines the ratio relationship between the compounds, and calculates the ratio between 89 kinds of compounds by using Python software, and a total of 1830 effective ratios are screened out. For example Figure 2(A) showed that 454 ratios were positively correlated with scorched-bitter taste scores, 764 and 612 ratios were negatively correlated and not significantly correlated, respectively. In addition, 65 and 15 of the 89 compounds were negatively correlated and not significantly correlated, respectively, while 9 compounds were positively correlated (p > 0.5), including isoamyl acetate, ethyl linoleate, ethyl palmitate, ethyl decanoate, ethyl nonanoate, isoamyl alcohol, lactic acid, isoamyl aldehyde, and 2,4-di-tert-butyl phenol. Among them, isoamyl alcohol, isoamyl aldehyde, and 2,4-di-tert-butyl phenol have unpleasant odors, and 2,4-di-tert-butyl phenol has been confirmed to be related to the scorched-bitter taste flavor of strong-flavor liquor. These three compounds are strongly suggested as key contributors to scorched-bitter taste and are classified as the first part of the potential off-flavor compounds (CBC-1, as shown in Figure 2 (B)). The remaining six positively correlated compounds are lactic acid and five esters, which do not present unpleasant odors themselves but can act as synergistic enhancers of off-flavors, and are therefore classified as synergistic compounds (SC-1).

[0030] The 454 ratios positively correlated with scorched-bitter taste scores were analyzed, and the specific process is shown in Figure 2 (C). Higher positive correlation coefficients indicate that the imbalance in the ratio of the compounds corresponding to the numerator and denominator in these ratios can lead to the generation of off-flavors. The flavor contribution of a single compound to off-flavors can be divided into four categories: scorched-bitter taste direct flavor contribution (CBC), synergistic compounds (SC), masking compounds (MC), and no-effect compounds (NEC). When the ratios of compounds involving these four types of effects are positively correlated with scorched-bitter taste scores, the possible combinations of compounds (16 types) in the numerator and denominator are shown in Table S1. Among them, eight types of ratio combinations can be positively correlated with scorched-bitter taste scores, including CBC / CBC, CBC / SC, CBC / MC, CBC / NEC, SC / CBC, SC / SC, SC / MC, and SC / NEC. For example, when the content of the compound causing scorched-bitter taste (CBC) is high, and the content of the compound masking the odor (MC) is low, the scorched-bitter taste intensity of the liquor sample is often high.

[0031] Therefore, referring to Table 1, based on the CBC-1 and SC screened above, a new list of potential bitter compounds (CBC) can be established by logical deduction and flavor profile analysis. For example, if the denominator of a certain ratio belongs to CBC-1, the numerator compound of the ratio may belong to one of the following categories: CBC or SC; if the numerator of a certain ratio belongs to CBC-1, the denominator compound of the ratio may belong to one of the following categories: CBC, MC, SC or NEC. If the compound presents an unpleasant flavor, it is classified as CBC-2; otherwise, it is classified as an uncertain compound category (UC). Similarly, taking CBC-2 as a new hypothesis basis, iterative logical deduction and flavor profile analysis are performed until no new CBC is identified, thereby completing the classification. Finally, 89 compounds are classified into three categories: 20 potential bitter compounds (C) in the middle (C), 6 synergistic compounds, and 63 compounds with unclear flavor contributions (B). Figure 2 Figure 2

[0032] Table 1 Ratio likelihoods positively correlated with bitter taste

[0033] Further, the step S2 specifically comprises: S2.1: aroma reconstruction experiment: adding the potential bitter compounds screened in step S2 to a reconstruction matrix to prepare a bitter taste reconstruction model, and scoring the intensity of the floral, fruity, sweet, sour, cellar, alcoholic and bitter tastes of the model to determine whether there is a bitter compound among the potential bitter compounds according to the flavor profile; S2.2: aroma deletion experiment: classifying the potential bitter compounds into acids, alcohols, aldehydes, acetals, phenols and ethers, systematically removing one category or a single compound from the reconstruction model each time, and identifying whether the deletion has a significant impact on the flavor profile of the reconstruction model by a two- or three-point test method, thereby determining the seven final bitter compounds: n-butanol, n-hexanol, sec-butanol, furfuryl alcohol, 2,4-di-tert-butyl phenol, p-cresol and ethyl furfuryl ether.

[0034] Specifically, the aroma reconstruction experiment in step S2 specifically comprises: As described above, based on the quantitative results of different compounds in bitter liquor and normal liquor and ratio analysis, the present application screens and classifies 20 potential bitter compounds, including 2,4-di-tert-butyl phenol, isoamyl alcohol, isoamyl aldehyde, benzaldehyde, butyric acid, heptanoic acid, hexanoic acid, furfuryl alcohol, n-octanol, acetic acid, iso-butyl alcohol, iso-valeric acid, n-butanol, n-hexanol, n-pentanol, 1,1-diethoxynonane, p-cresol, valeric acid, ethyl furfuryl ether and sec-butanol. According to the actual concentration of these compounds in bitter liquor, they are added to the reconstruction matrix for verification. ​​

[0035] The reconstructed model is prepared by adding potential bitter compounds to a reconstructed matrix according to the content in the real off-flavor liquor sample. The reconstructed matrix is a base liquor for preparing the reconstructed model, which is usually prepared by diluting a normal liquor sample several times and then adding compounds in the bitter liquor sample. In the specific embodiment, the reconstructed matrix is prepared by diluting a normal liquor sample 4 times, and then backfilling the compounds with a content greater than 100 mg / L in the bitter liquor sample into the diluted 4 times normal liquor sample according to the concentration in the bitter liquor sample. The overall aroma characteristics of the reconstructed model are different from those of the reconstructed matrix (A), but similar to those of the original bitter liquor (B). The bitter and sour intensity of the reconstructed model is significantly higher than that of the reconstructed matrix (p<0.05), and the other four aroma intensities are basically the same or slightly lower. The reconstructed model and the original bitter liquor are close in terms of floral, fruity, wine and cellar aroma intensity. However, the bitter (3.35 / 4.1) and sweet (2.6 / 2.9) of the reconstructed model cannot reach the level of the original off-flavor liquor, and the sour (2.35 / 1.15) is significantly higher. The results show that the reconstructed model can successfully simulate the aroma profile of the original bitter liquor, but the sour after reconstruction is more prominent, indicating that among the 20 compounds, not only the key off-flavor components are included, but also some irrelevant compounds are mixed in, which still needs to be further verified by deletion experiments. Figure 3 Figure 3 Further, the aroma deletion experiment in step S2 specifically includes: The 20 potential bitter compounds are divided into acid, alcohol, aldehyde, acetal, phenol and ether, and one category or single compound is systematically removed from the reconstructed model each time, and whether the deletion has a significant effect on the flavor profile of the reconstructed model is identified by the two-three point test method, and the off-flavor compounds are determined accordingly.

[0036] The significance is determined by the two-three point test method (Duo-Trio Test) (the number of participants N=36). In the two-three point test method (Duo-Trio Test), the evaluator needs to identify one sample that is the same as the control sample from two samples. If there is no difference between the samples, the probability of correct selection is 0.5. According to the binomial distribution principle, when the number of participants is 36, the critical value of the number of correct answers is used to determine the significance:

[0037] <24: cannot reject the null hypothesis, there is no significant difference between the samples (p>0.05). Duo-Trio Test 24 ≤ Correct answer number ≤ 26: significant at the level of α=0.05, indicating that there is a difference between the samples (p≤0.05).

[0038]

[0039] 24 ≤ Correct answer number ≤ 26: significant at the level of α=0.05, indicating that there is a difference between the samples (p≤0.05).

[0040] ​​Correct number of answers > 26: significant at the level of a = 0.01, indicating that there is a very significant difference between samples (p < 0.01). N < 24 (—, not significant), 24 < N < 26 (*, a = 0.05), N > 26 (**, a = 0.01)

[0041] To verify the contribution of 20 compounds to the overall aroma, they were divided into acids, alcohols, aldehydes, acetals, phenols, and ethers. Each time, one category or single compound was systematically removed from the recombined model, and the two- to three-point test method was used to identify whether the absence had a significant impact on the flavor profile of the recombined model.

[0042] When the acid, aldehyde, and acetal compounds were removed from the recombined model, only 18 or fewer of the 36 panelists could perceive the flavor difference, indicating that there was no significant difference from the complete recombined model, which indicated that these three types of compounds had no significant impact on the bitter taste. However, the addition of acid compounds enhanced the sour taste of the recombined model, resulting in a significant difference in sour taste intensity between the recombined model and the bitter-tasting baijiu ( Figure 3 (B)). When the alcohol, phenol, and ether compounds were removed, there was a significant difference between the missing sample and the complete recombined model, indicating that alcohols, phenols, and ethers had an important impact on the bitter taste.

[0043] Not all alcohol and phenol compounds had an impact on the bitter taste. When furfuryl alcohol, n-butanol, sec-butanol, n-hexanol, 2,4-di-tert-butyl phenol, and p-cresol were removed from the recombined model, significant differences were perceived by the panelists; however, no significant differences were observed when n-octanol, n-pentanol, isoamyl alcohol, and isobutyl alcohol were removed.

[0044] Therefore, seven bitter-tasting compounds (furfuryl alcohol, n-butanol, sec-butanol, n-hexanol, 2,4-di-tert-butyl phenol, p-cresol, and ethyl furfuryl ether) were ultimately determined to be the compounds that caused the bitter taste in the strong-flavor baijiu. These substances exhibited caramel, bitter, cocoa, phenolic, leather, baked, and gasoline-like off-flavor characteristics, respectively. By controlling the concentration ratio of these compounds and the skeleton compounds, the bitter taste can be effectively reduced or eliminated. The results of the deletion experiments for the bitter-tasting recombined model are shown in Tables 2 and 3.

[0045] Table 2 Deletion experiment results of the bitter-tasting recombined model (I)

[0046] Table 3 Deletion experiment results of the bitter-tasting recombined model (II)

[0047] S3: By statistically analyzing the ratio of off-flavor compounds to ethyl hexanoate, the limit ratio of the off-flavor compounds to ethyl hexanoate is determined.

[0048] Ethyl hexanoate always has the highest odor activity value in Luzhou-flavor liquor and is one of the compounds that constitute its flavor profile. Therefore, studying the ratio limit between the seven bitter and astringent compounds (BBAC) and ethyl hexanoate will help control the off-flavor in liquor.

[0049] Compounds with low olfactory thresholds, such as 2,4-di-tert-butyl phenol, p-cresol (0.167 mg / L), and ethyl furfuryl ether (0.46 mg / L), can bring some unpleasant flavors to liquor even in trace amounts. According to the actual data, the average concentrations of 2,4-di-tert-butyl phenol, p-cresol, and ethyl furfuryl ether in off-flavor liquor were 0.004 mg / L, 0.16 mg / L, and 7.18 mg / L, respectively, and the ratios of the three to ethyl hexanoate were all less than 0.02 ( Figure 4 In addition, the ratios of the three compounds to ethyl hexanoate in the reconstructed model were 0.0000011, 0.00011, and 0.0045, respectively, while the corresponding ratios in the reconstructed matrix were 0, 0.0000847, and 0.0018, respectively (Table 4). The off-flavor sensory intensity of bitter and astringent liquor, the reconstructed model, and the reconstructed matrix was 4.08, 3.33, and 0 ( Figure 4 ), respectively. In summary, it is recommended that the ratios of 2,4-di-tert-butyl phenol, p-cresol, and ethyl furfuryl ether to ethyl hexanoate be controlled below 0, 0.000084, and 0.0018, respectively.

[0050] Table 4 Ratios of three compounds to ethyl hexanoate in the reconstructed matrix and the reconstructed model

[0051] It was found that the ratios of furfuryl alcohol, n-hexanol, n-butanol, and sec-butanol to ethyl hexanoate in normal liquor were lower than those in off-flavor liquor. In addition, Figure 4 The box plot showed that the distribution of these ratios in normal liquor was significantly more concentrated, with significantly smaller standard deviation (SD), interquartile range (IQR), and range than in off-flavor liquor. For example, the standard deviation (0.02), interquartile range (0.054), and range (0.054) of the ratio of furfuryl alcohol in normal liquor were much smaller than those in bitter and astringent liquor (standard deviation = 0.51, interquartile range = 0.65, and range = 1.85). This finding provides a theoretical basis for the industry consensus that the compound distribution of high-quality liquor has similar rules. When the content of off-flavor compounds is controlled within a reasonable ratio range, they may not have a negative impact on the flavor of the liquor body.

[0052] Specifically, as shown in Table 5, the ratio of furfuryl alcohol / ethyl hexanoate in normal liquor is 0.2554-0.3093, and in off-flavor liquor is 0.2862-2.1376; the ratio of n-hexanol / ethyl hexanoate in normal liquor is 0.0833-0.0943, and in off-flavor liquor is 0-0.4373; the ratio of n-butanol / ethyl hexanoate in normal liquor is 0.1330-0.1559, and in off-flavor liquor is 0.1568-0.7939; the ratio of sec-butanol / ethyl hexanoate in normal liquor is 0.0359-0.0443, and in off-flavor liquor is 0.0475-0.2174. It is suggested that the ratio should be controlled below the minimum value observed in the liquor with bitter and harsh taste or the highest value found in normal liquor. Therefore, the ratio of furfuryl alcohol, n-hexanol, n-butanol and sec-butanol to ethyl hexanoate should be controlled below 0.2862, 0.0943, 0.1559 and 0.0443, respectively.

[0053] Table 5 Range of ratios of seven bitter and harsh taste compounds to ethyl hexanoate

[0054] The present application establishes a new ratio analysis strategy, which can effectively identify and regulate off-flavor compounds, and provides a theoretical basis for liquor flavor control. Although the proposed ratio limit is based on a limited sample set, future research can use this ratio logic derivation in a larger population to improve the standard of blending through meta-analysis.

[0055] The above is merely specific implementation of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for resolving the off-flavor compounds of baijiu and their ratio limits by ratio analysis and logical deduction, characterized in that, Comprise: S1: calculate the ratio of the content of the compounds in the liquor sample, and perform Spearman correlation analysis on the compound content, the ratio of the compound content and the off-flavor score to obtain the compounds and the ratio that are positively correlated with off-flavor; classify the compounds by ratio logical deduction and flavor analysis, and screen potential off-flavor compounds; S2, based on the potential off-flavor compounds, verify and determine the final off-flavor compounds through aroma reconstruction experiment and aroma loss experiment; S3: Through statistical analysis of the ratio of the final off-flavor compounds to ethyl hexanoate, the limit ratio of the final off-flavor compounds is determined.

2. The method for resolving the off-flavor compounds of baijiu and their ratio limits by ratio analysis and logical deduction according to claim 1, characterized in that, The step S1 classifies the compounds by ratio logical deduction and flavor analysis, and screens potential off-flavor compounds, which specifically comprises: S1.1: First, for the compounds positively correlated with off-flavor score, according to the flavor characteristics of the compounds, the compounds with unpleasant flavor are classified as the first part of the potential off-flavor compounds (COC-1), and the compounds with aroma or no odor are classified as the first part of the synergistic compounds (SC-1); S1.2: Secondly, for the ratio positively correlated with off-flavor score, the ratio R1 list containing COC-1 compounds in the numerator or denominator is screened out from the ratio, and the compounds corresponding to the numerator and the denominator of each ratio in the ratio R1 list are classified as potential off-flavor compounds (COC), synergistic compounds (SC), masking compounds (MC) or ineffective compounds (NEC) according to the five types of combinations of COC / CBC, COC / SC, COC / MC, COC / NEC and SC / COC; S1.3: Screen out the ratio R2 list containing SC-1 compounds in the numerator or denominator, and classify the compounds corresponding to the numerator and the denominator of each ratio in the ratio R2 list as COC, SC, MC or NEC according to the five types of combinations of COC / SC, SC / COC, SC / SC, SC / MC and SC / NEC; S1.4: Further divide the potential off-flavor compounds obtained in steps S1.2 and S1.3 according to their flavor characteristics, and the specific division method is: the compounds that present unpleasant flavor are identified as the second part of the potential off-flavor compounds (COC-2), and the compounds with aroma or no odor are classified as "uncertain compounds (UC) category; S1.5: According to the compounds in COC-2 in S1.4, steps S1.2 and S1.4 are performed to obtain CBC-3; S1.6, in this way, according to the newly identified COC-n, steps S1.2 and S1.4 are repeatedly executed until all ratios are classified, and n=1, 2, 3, ….

3. The method for resolving the odor compounds and their ratio limits in Baijiu by ratio analysis and logical deduction according to claim 2, characterized in that, The step S1.6 further comprises: if no new COC is found, the classification cycle is terminated, at this time, it is judged whether the flavor properties of the compounds in the remaining ratios have unpleasant flavor or not, if not, the process is completed; otherwise, the previous operation is checked back.

4. The method for resolving the odor compounds and their ratio limits in Baijiu by ratio analysis and logical deduction according to claim 3, characterized in that, The step S1 further comprises: S1.7: The screened COC-1, COC-2, …, and COC-n are combined to construct a COC total list; for compounds with multiple labels and cannot be classified clearly, they are uniformly divided into "uncertain compounds", finally, three compound category lists are obtained: COC, SC and UC.

5. The method for resolving the off-flavor compounds and their ratio limits in Baijiu by ratio analysis and logical deduction according to claim 4, characterized in that, The potential off-flavor compounds include butyric acid, heptanoic acid, hexanoic acid, acetic acid, isovaleric acid, valeric acid, furfuryl alcohol, n-octanol, isoamyl alcohol, isobutyl alcohol, n-butyl alcohol, n-hexyl alcohol, n-pentyl alcohol, sec-butyl alcohol, isoamyl aldehyde, benzaldehyde, nonanal diethyl acetal, p-cresol, 2,4-di-tert-butyl phenol, and ethyl furfuryl ether.

6. The method for analyzing and deducing the off-flavor compounds and their ratio limits in Baijiu by ratio analysis and logical deduction according to any one of claims 1-5, characterized in that, The step S2 specifically comprises: S2.1: aroma reconstruction experiment: the potential off-flavor compounds screened in step S1 are added to the reconstruction matrix to prepare an off-flavor reconstruction model, and the flavor attributes and off-flavors thereof are scored, and whether there is an off-flavor compound in the potential off-flavor compounds is determined according to the flavor profile; S2.2: aroma loss experiment: the potential off-flavor compounds are classified, one category or single compound is systematically removed from the reconstruction model each time, and whether the loss has a significant effect on the flavor profile of the reconstruction model is identified by two-three point test method, thereby determining seven final off-flavor compounds: n-butyl alcohol, n-hexyl alcohol, sec-butyl alcohol, furfuryl alcohol, 2,4-di-tert-butyl phenol, p-cresol and ethyl furfuryl ether.

7. The method for resolving the odor compounds and their ratio limits in Baijiu by ratio analysis and logical deduction according to claim 6, characterized in that, The upper limit of the ratio of the seven final off-flavor compounds to ethyl hexanoate is 0, 0.000084, 0.0018, 0.2862, 0.0943, 0.1559 and 0.0443, respectively.

Images