Method for strengthening characteristic flavor of liquor based on liquor distillation connection point selection

Abstract

The application discloses a method for strengthening characteristic flavor of liquor based on liquor distillation connection cut point selection, and belongs to the technical field of liquor distillation process.The method comprises the following steps: (1) obtaining the change rule of flavor substances in the distillation process through detection; (2) analyzing the correlation between the flavor substances and ethanol; obtaining target flavor substances according to the correlation between the flavor substances and ethanol and the change rule of the flavor substances in the distillation process; (3) performing characteristic dynamics modeling on the target flavor substances by using a first-order kinetics model; (4) selecting target flavor substances to construct a multi-objective optimization model based on the dynamics model, and obtaining corresponding Pareto front; and (5) selecting the connection cut point position with the highest recovery rate of key target flavor substances based on the Pareto front, that is, obtaining the accurate connection time.The method has the advantages of good accuracy, low trial and error cost, high stability and efficiency and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for strengthening the characteristic flavor of liquor based on the selection of the cut point of liquor distillation, belonging to the technical field of liquor distillation process. BACKGROUND

[0002] Liquor has the characteristics of mellow body, long-lasting aroma retention after emptying the cup, and long aftertaste. This unique flavor is closely related to its special and ancient brewing process. Among them, the liquor distillation process is the key step to convert the fermented grains into liquor. Through steam heating contact, ethanol and a large amount of flavoring substances in the fermented grains are extracted. In order to extract ethanol and flavoring substances from solid matrix, liquor uses a unique solid distillation device called a retort, which can complete the extraction of liquor in one step. During the solid distillation process of liquor, a high-concentration ethanol solution (70% vol) and more than 1000 flavoring compounds can be obtained. The liquor operation is based on the fact that the concentration of flavoring compounds changes constantly during the distillation process. By selecting two cut points before and after the distillation time, the intermediate fraction segment is obtained as the base liquor segment. By adjusting the cut point position, the proportion of flavor concentration in the base liquor segment (i.e., different recovery rates) is changed, and ultimately determines the aroma characteristics and quality of the liquor. The traditional cut point selection is completed by looking at the flower (i.e., by observing the shape of the alcohol flower to determine the alcohol content and flavor characteristics of the fraction segment). However, this method relies too much on the experience of production personnel and lacks scientific process analysis and optimization methods. With the current mechanization and intelligentization of the liquor brewing process, it is urgent to develop a digital standard and control method to predict and optimize the solid distillation liquor process.

[0003] Currently, there is little research on liquor solid distillation process, such as:

[0004] CN 106018651 B discloses a method for identifying the flavor differences of different fraction segments using gas chromatography and mathematical statistics analysis. Specifically, it identifies different fraction segments in the distillation process by flavoring substances. However, this identification method cannot directly provide a cut point scheme for liquor operation, and there are too many flavoring substances to consider, making it difficult to consider the flavor in all directions.

[0005] CN 113720797 A discloses a method for segmenting base liquor using near-infrared spectrometers and pattern recognition classification algorithms. Although this method can quickly identify the characteristics of different fraction segments in the distillation process, this identification method has high instability and cannot quantitatively describe the flavor changes in the distillation process. The characteristics summary is more macroscopic. Moreover, this method cannot directly guide the liquor distillation process.

[0006] The wine receiving operation is an important means of controlling the wine yield and flavor in the white spirit distillation process, and has an important influence on the recovery rate of white spirit flavor substances, for example: the regulation of pyrazine substances or the regulation of acid substances in Maotai-flavor liquor. The wine receiving operation can determine the quality and yield of the distilled base liquor to some extent. However, the selection of the cut point will also affect the recovery rate of other flavor substances, thereby changing the flavor concentration ratio in the whole base liquor section. Because there are many flavor substances and they are different, the strengthening of some flavor compounds will inevitably lead to the reduction of the recovery rate of some other flavor compounds; it is difficult to consider the regulation of multiple compounds at the same time through intuitive change expression or monitoring. Therefore, a scientific and rigorous method is needed to quantify the target flavor change in the wine receiving operation and make an optimal judgment on the wine receiving. SUMMARY

[0007] [Technical problem]

[0008] The conventional selection of the cut point for wine receiving is through wine receiving by looking at the flower, but this method is too dependent on the experience of the production personnel and lacks a scientific process analysis and optimization method.

[0009] The method for segmenting the distilled liquor is difficult to consider all the flavor component contents in all directions.

[0010] [Technical solution]

[0011] In order to solve the above at least one problem, the present application firstly detects the change rule of flavor substances in the solid state distillation process of white liquor, and then selects representative compounds as target flavor substances, realizes the purpose of strengthening the target flavor of white liquor by wine receiving selection through the change characteristic expression of flavor substances and multi-objective optimization selection of wine receiving. The method of the present application has the advantages of good accuracy, low trial and error cost, high stability and efficiency, etc.

[0012] The first object of the present application is to provide a method for strengthening the characteristic flavor of white liquor based on the cut point selection of white liquor distillation wine receiving, comprising the following steps:

[0013] (1) Instantaneous change detection of flavor substances:

[0014] The concentration of flavor substances in the instantaneous sample in the solid state distillation process of white liquor is detected to obtain the change rule of flavor substances in the distillation process;

[0015] (2) Target flavor component selection:

[0016] The correlation between flavor substances and ethanol is analyzed to obtain the correlation between flavor substances and ethanol; according to the correlation between flavor substances and ethanol and the change rule of flavor substances in the distillation process, the target flavor substances are obtained;

[0017] (3) Kinetic model construction:

[0018] A first-order kinetic model is used to model the characteristic kinetics of the target flavor substances, and a kinetic model is obtained;

[0019] (4) Multi-objective optimization model construction:

[0020] Based on the kinetic model, a multi-objective optimization model (MOO) is constructed by selecting a combination of target flavor substances, and a corresponding Pareto frontier (PF) is obtained;

[0021] (5) Selection of wine receiving cut point:

[0022] Based on the Pareto frontier, the position of the wine receiving cut point with the highest recovery rate of key target flavor substances is selected, i.e. the accurate time of wine receiving is obtained.

[0023] In an embodiment of the present application, the liquor used in step (1) is one of Jiangxiang liquor, Qingxiang liquor, Nongxiang liquor, Fengxiang liquor, Jianxiang liquor, Dongxiang liquor, rice-flavor liquor, sesame-flavor liquor, Shixiang liquor, Laobaigan liquor, Fuxi liquor, and Texiang liquor.

[0024] In an embodiment of the present application, the pressure for liquor distillation in step (1) is 1.2-1.5 MPa.

[0025] In an embodiment of the present application, in step (1), the instantaneous sample is sampled once every 3-6 minutes during the distillation process as an instantaneous sample until the end of the distillation.

[0026] In an embodiment of the present application, in step (1), the flavor substance instantaneous change detection is detected by using the separation method of GC-FID combined with HS-SPME-GC-MS.

[0027] In an embodiment of the present application, in step (2), the method used for correlation analysis is Pearson correlation analysis.

[0028] In an embodiment of the present application, in step (2), the target flavor substance is one or more of 2,3,5,6-tetramethylpyrazine, 3-methylbutanol, acetal, ethyl acetate, dimethyl trisulfide, propionic acid, phenethyl alcohol, benzaldehyde, 3-phenylpropyl acetate, 3-hydroxy-2-butanone, and phenol.

[0029] In an embodiment of the present application, in step (3), the kinetic model is:

[0030] C t =C e (1-e -Kt );

[0031] Wherein, Ct is the accumulation amount (μg / L) of flavor substances; Ce is the accumulation amount (μg / L) of aroma compounds in the equilibrium state; K is the distillation constant (1 / min); t is the distillation time.

[0032] In an embodiment of the present application, the algorithm in the multi-objective optimization model in step (4) adopts the NSGA-II multi-objective genetic algorithm; the NSGA-II multi-objective genetic algorithm is constructed by using Python language.

[0033] In an embodiment of the present application, the multi-objective optimization model in step (4) is as follows:

[0034] Min F(x i )=[f1(x i ),f2(x i ),...,f n (x i )]

[0035]

[0036] g j (x i )≥0,j=1,2,...,J

[0037] h k (x i )=0,k=1,2,...,k

[0038]

[0039] Wherein, x represents a decision vector in the D-dimensional decision space X, that is, a solution of the optimization problem;

[0040] F(x i ) is the nth optimization objective (function) in the multi-objective optimization problem;

[0041] is a target function vector in the N-dimensional target space Y;

[0042] y is a mapping function from the D-dimensional decision space X to the N-dimensional target space;

[0043] N is the total number of target functions to be optimized (the number of optimization objectives);

[0044] g i (x i )≥0 is the ith inequality constraint function;

[0045] h k (x i ) is the jth equality constraint function.

[0046] In one embodiment of the present application, the key target flavoring substance of step (5) is selected according to the type of baijiu, such as: the key target flavoring substance of Jiangxiang baijiu is 2,3,5,6-tetramethylpyrazine.

[0047] In one embodiment of the present application, the cut point time obtained in step (5) can be converted into alcohol concentration as needed.

[0048] The second object of the present application is the application of the method of the present application in the field of baijiu processing.

[0049] [beneficial effects]

[0050] (1) The present application uses a kinetic model to solve the problem that the change rule of flavoring substances in the distillation process is complex and cannot be comprehensively and accurately summarized.

[0051] (2) The present application uses correlation analysis to select representative flavoring substances, simplifying the target flavoring substance content that needs to be considered when selecting the cut point.

[0052] (3) The present application uses a multi-objective optimization model to screen and optimize the cut point selection in the distillation process of flavoring substances, and constructs a set of methods suitable for the solid-state distillation process of baijiu to strengthen the characteristic flavor of baijiu by using the cut point selection of receiving wine. Once the model is constructed, it can be called at any time when selecting the cut point of receiving wine, without the need to repeat the construction, and has the advantages of simple operation, high accuracy of discrimination, etc. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 Cut points of baijiu obtained in Examples 1-10. DETAILED DESCRIPTION

[0054] The preferred embodiments of the present application are described below, and it should be understood that the embodiments are used to better explain the present application and are not used to limit the present application.

[0055] Test method:

[0056] The calculation formula of the recovery rate of flavoring substances is as follows:

[0057]

[0058] Wherein, r is the recovery rate, C t1 and C t2 are the cumulative amounts of flavoring substances obtained according to the kinetic model at t1 and t2 (where t2>t1), C tmax is the total cumulative amount of flavoring substances at the end of distillation.

[0059] Raw materials used in the examples:

[0060] Distillation is carried out in a distillery.

[0061] The method used in the examples is derived from:

[0062] Multi-objective optimization model: the algorithm uses NSGA-II multi-objective genetic algorithm; the NSGA-II multi-objective genetic algorithm is constructed by using Python language;

[0063] Specifically as follows:

[0064] Min F(x i )=[f1(x i ),f2(x i ),...,f n (x i )]

[0065]

[0066] g j (x i )≥0,j=1,2,...,J

[0067] h k (x i )=0,k=1,2,...,k

[0068]

[0069] Wherein, x represents a decision vector in the D-dimensional decision space X, that is, a solution of the optimization problem;

[0070] F(x i ) is the nth optimization objective (function) in the multi-objective optimization problem;

[0071] is a target function vector in the N-dimensional target space Y;

[0072] y is a mapping function from the D-dimensional decision space X to the N-dimensional target space;

[0073] N is the total number of optimization objectives (the number of optimization objectives);

[0074] g j (x i )≥0 is the ith inequality constraint function;

[0075] h k (x i ) is the jth equality constraint function;

[0076] Reference is made to the literature (Bradford, E.; Schweidtmann, A. M.; Lapkin, A. Efficient multiobjective optimization employing Gaussian processes, spectral sampling and a genetic algorithm. Journal of Global Optimization 2018, 71, 407-438, doi:10.1007 / s10898-018-0609-2.).

[0077] Example 1

[0078] A method for strengthening the characteristic flavor of liquor based on the cut point selection of liquor distillation, comprising the following steps:

[0079] (1) Instantaneous change detection of flavor substances:

[0080] Take the solid fermented grains of Maotai-flavor liquor for distillation, with a total filling weight of 700 kg of fermented grains of mixed grains, and the inlet air pressure is kept at 1.5 MPa during the distillation process. The instantaneous distillate is collected every 6 minutes using a 250 mL glass bottle for sampling until the end of the distillation process. Each batch of distillation takes 54 minutes, resulting in 10 distillation samples. Six batches (upper, middle, and lower layers of fermented grains, each with 2 batches) of Maotai-flavor liquor solid distillation process are collected, totaling 60 samples.

[0081] The samples are detected using the separation method of GC-FID combined with HS-SPME-GC-MS;

[0082] The specific operation of GC-FID is as follows:

[0083] The sample liquor is adjusted to 50% vol using anhydrous ethanol, and the dilution ratio is recorded. Take 1 mL of the adjusted liquor sample to a 2 mL chromatographic sample bottle and add 50 μL of a mixed internal standard containing 102.57 mg / L of 2-ethylbutyric acid, 125.50 mg / L of n-pentyl acetate, and 111.04 mg / L of tert-pentanol. Mix well and take 1 μL for injection;

[0084] The temperature program is set as follows: initial temperature 35℃, stable operation for 5 min, temperature rise to 100℃ at 4℃ / min, maintain operation for 2 min, then temperature rise to 150℃ at 8℃ / min, finally temperature rise to 200℃ at 15℃ / min, maintain operation for 25 min;

[0085] Split ratio was set to 20:1 for split injection, injection port and detector temperature was 250 °C, helium (>99.999%) was used as carrier gas at a flow rate of 1 mL / min; the mixed standard solution was diluted into a series of 50% vol ethanol aqueous solutions with a concentration of 2 times dilution method to establish a standard correction curve; each sample was repeated three times;

[0086] The specific operation of HS-SPME-GC-MS was as follows:

[0087] Each sample was adjusted to a volume of 5 mL and an ethanol content of 10% vol with purified water;

[0088] 5 mL of the sample saturated with 1.5 g of NaCl was placed in a 20 mL sampling bottle, 40 μL of an internal standard mixture (1.197 mg / L of 2,2-dimethylpropionic acid, 0.294 mg / L of 2-octanol, 0.146 mg / L of 2-phenylethyl acetate-D3, 0.322 mg / L of n-hexyl-D13 alcohol, and 0.307 mg / L of 2-methoxyphenol-D3) was added; then the bottle was sealed with a polytetrafluoroethylene / silicone septum and a screw cap; all analyses were performed in triplicate;

[0089] SPME automatic headspace sampling system (CTC Analytics AG, Switzerland) and 120 μm divinylbenzene / carbon wide range / polydimethylsiloxane (DVB / CAR WR / PDMS) fiber (CTC Analytics AG, Switzerland) were used for extraction and injection; the sample was equilibrated at 45 °C for 5 minutes, then extracted at the same temperature with stirring at a speed of 250 rpm for 45 minutes. After extraction, the fiber was automatically inserted into the gas chromatograph injection port (250 °C) for 5 minutes to allow the analytes to be desorbed;

[0090] The sample was separated on a DB-FFAP chromatographic column (60 m x 0.25 mm i.d.; 0.25 μm film thickness, Agilent Technologies); the oven temperature was kept at 50 °C for 2 minutes, then increased to 230 °C at a rate of 6 °C / min and kept at 230 °C for 15 minutes;

[0091] The mass spectrometer was operated in 70 eV electron ionization mode with an ion source temperature of 230 °C; in full scan mode, the mass range was set to 35 to 350 amu;

[0092] The aroma peaks were determined by comparison with the mass spectral database of the National Institute of Standards and Technology (NIST) and the retention index (RI, determined by n-alkanes C7-C40);

[0093] The SPME fiber head after extraction was inserted into the sample port without shunting to desorb at 250℃ for 300s, and analyzed by GC-MS; the chromatographic column was composed of a DB-FFAP capillary column; the carrier gas was helium with a flow rate of 2mL / min; the oven temperature was initially maintained at 45℃ for 2min, then increased to 230℃ at a rate of 4℃ / min, and finally maintained at 230℃ for 15min. In the electron ionization mode, the ionization energy of the mass spectrometer was 70eV; the temperature of the ion source was 230℃, and the recorded m / z range was 30-350amu;

[0094] The change rule of flavor substances in the distillation process was obtained by detecting the concentration of flavor substances in the instantaneous distillate in the distillation process of Baijiu;

[0095] (2) Target flavor component selection:

[0096] Pearson correlation analysis was used to analyze the correlation between flavor substances and ethanol, and the correlation between flavor substances and ethanol was obtained;

[0097] According to the correlation between flavor substances and ethanol and the change rule of flavor substances in the distillation process, the target flavor substance was obtained;

[0098] Specifically as shown in Table 1:

[0099] Table 1 Correlation between target flavor substance and ethanol

[0100]

[0101] (3) Kinetic model construction:

[0102] A first-order kinetic model was used to model the characteristics of the target flavor substance, and a kinetic model was obtained;

[0103] Specifically:

[0104] C t =C e (1-e -kt )

[0105] Wherein, C t is the accumulation amount of flavor substances (μg / L); C e is the accumulation amount of flavor substances in the equilibrium state (μg / L); K is the distillation constant (1 / min); t is the distillation time;

[0106] Table 2 Kinetic parameters of flavor substances in the distillation process

[0107]

[0108] (4) Multi-objective optimization model construction:

[0109] Based on the kinetic model, 2, 3, 5, 6-tetramethylpyrazine and acetal were selected to construct a multi-objective optimization model (MOO), and the corresponding Pareto frontier (PF) was obtained;

[0110] (5) Selection of wine cutting point:

[0111] Based on the Pareto frontier, the highest recovery rate (14.53%) of the key target flavor substance (2, 3, 5, 6-tetramethylpyrazine) was selected as the wine cutting point position, i.e. the accurate cutting point of the wine (3.9 minutes and 19.8 minutes) was obtained.

[0112] The content of other target flavor substances in the obtained cutting point section is: the recovery rates of 3-methylbutanol, acetal, ethyl acetate and dimethyltrisulfide are 45.19%, 44.54%, 49.58% and 46.80%, respectively; the recovery rates of propionic acid, phenethyl alcohol, benzaldehyde, 3-phenylpropyl acetate, 3-hydroxybutanone and phenol are 16.58%, 16.51%, 16.63%, 20.51%, 17.21% and 7.65%, respectively.

[0113] Example 2

[0114] In step (5) of Example 1, the construction substance of the multi-objective optimization model was adjusted to 2, 3, 5, 6-tetramethylpyrazine and 3-methylbutanol, and the others were consistent with Example 1.

[0115] Example 3

[0116] In step (5) of Example 1, the construction substance of the multi-objective optimization model was adjusted to 2, 3, 5, 6-tetramethylpyrazine and ethyl acetate, and the others were consistent with Example 1.

[0117] Example 4

[0118] In step (5) of Example 1, the construction substance of the multi-objective optimization model was adjusted to 2, 3, 5, 6-tetramethylpyrazine and dimethyltrisulfide, and the others were consistent with Example 1.

[0119] Example 5

[0120] In step (5) of Example 1, the construction substance of the multi-objective optimization model was adjusted to 2, 3, 5, 6-tetramethylpyrazine and propionic acid, and the others were consistent with Example 1.

[0121] Example 6

[0122] In step (5) of Example 1, the construction substance of the multi-objective optimization model was adjusted to 2, 3, 5, 6-tetramethylpyrazine and phenethyl alcohol, and the others were consistent with Example 1.

[0123] Example 7

[0124] The construction material of the multi-objective optimization model in step (5) of Example 1 was adjusted to 2,3,5,6-tetramethylpyrazine, benzaldehyde, and the other conditions were the same as Example 1.

[0125] Example 8

[0126] The construction material of the multi-objective optimization model in step (5) of Example 1 was adjusted to 2,3,5,6-tetramethylpyrazine, 3-phenylpropyl acetate, and the other conditions were the same as Example 1.

[0127] Example 9

[0128] The construction material of the multi-objective optimization model in step (5) of Example 1 was adjusted to 2,3,5,6-tetramethylpyrazine, 3-hydroxy-2-butanone, and the other conditions were the same as Example 1.

[0129] Example 10

[0130] The construction material of the multi-objective optimization model in step (5) of Example 1 was adjusted to 2,3,5,6-tetramethylpyrazine, phenol, and the other conditions were the same as Example 1.

[0131] The cut points of the white liquor obtained in Examples 1-10 are shown in Table 2 and Table 3. Figure 1

[0132] Table 32,3,5,6-tetramethylpyrazine and other aromatic compounds in the cut point of the distillation formula

[0133]

[0134] Note: Other flavor substances are 3-methylbutanol, acetal, ethyl acetate, dimethyl trisulfide, propionic acid, phenethyl alcohol, benzaldehyde, 3-phenylpropyl acetate, 3-hydroxy-2-butanone, phenol.

[0135] From Figure 1 and Table 3, it can be seen that the cut point of Example 1 obtained the highest recovery rate of 2,3,5,6-tetramethylpyrazine in white liquor, and could maximize the retention of other flavor substances; while other combinations, it is difficult to achieve the balance of key target components and other components.

[0136] Comparative Example 1

[0137] Five experienced workers in the factory used their experience to obtain the cut point by looking at the liquor, and the white liquor was tested.

[0138] ​Results show that: 5 workers get the performance of liquor is not exactly the same, it is difficult to ensure the content of flavoring substances at the same time; visible, through the artificial method relies on strong subjectivity, leading to the quality of liquor is not stable.

[0139] The method of the embodiment realizes the key target flavor recovery rate is higher while ensuring the other flavor loss rate is reduced, basically maintains the flavor balance, and the quality of liquor calculated by the model each time is stable, and does not need to rely on mature workers.

[0140] Although the present application has been disclosed with the preferred embodiments as above, it is not intended to limit the present application, and any person skilled in the art can make various modifications and modifications without departing from the spirit and scope of the present application, therefore the protection scope of the present application should be defined by the claims.

Claims

1. A method for enhancing the characteristic flavor of liquor based on the selection of liquor distillation cut point, characterized in that, Comprising the following steps: (1) Instantaneous change detection of flavor substances: Detecting the concentration of flavor substances in the instantaneous sample in the solid distillation process of Baijiu, and obtaining the change rule of flavor substances in the distillation process; Wherein, the Baijiu is Jiangxiang Baijiu; (2) Target flavor component selection: Analyzing the correlation between flavor substances and ethanol to obtain the correlation between flavor substances and ethanol; according to the correlation between flavor substances and ethanol and the change rule of flavor substances in the distillation process, obtaining the target flavor substance; Wherein, the method for correlation analysis is Pearson correlation analysis; the target flavor substance is one or more of 2,3,5,6-tetramethylpyrazine, 3-methylbutanol, acetal, ethyl acetate, dimethyltrisulfide, propionic acid, phenethyl alcohol, benzaldehyde, 3-phenylpropyl acetate, 3-hydroxy-2-butanone and phenol; (3) Kinetic model construction: Using a first-order kinetic model to model the characteristics of the target flavor substance, and obtaining the kinetic model; Wherein, the kinetic model is: wherein Ct is the accumulation of the flavor substance, pg / L; Ce is the accumulation of the flavor substance in the equilibrium state, pg / L; K is the distillation constant, 1 / min; t is the distillation time; (4) Multi-objective optimization model construction: Based on the kinetic model, a multi-objective optimization model MOO is constructed by selecting a combination of target flavor substances, and a corresponding Pareto frontier is obtained; Wherein, the algorithm in the multi-objective optimization model is NSGA-II multi-objective genetic algorithm; the NSGA-II multi-objective genetic algorithm is constructed by using Python language; (5) Selection of the cut-off point for receiving wine: Based on the Pareto frontier, the cut-off point position with the highest recovery rate of key target flavor substances is selected, i.e. the accurate time for receiving wine is obtained.

2. The method of claim 1, wherein, The pressure of Baijiu distillation in step (1) is 1.2-1.5 MPa.

3. The method of claim 1, wherein, The instantaneous change detection of flavor substances in step (1) is detected by using the separation method of GC-FID combined with HS-SPME-GC-MS.

4. The method of claim 1, wherein, The key target flavor substance in step (5) is selected according to the type of Baijiu.

5. The method of any one of claims 1-4 for use in the field of Baijiu processing.

Images