A high-flavor soybean milk processing method based on soybean quality evaluation index optimization

Abstract

This invention relates to the field of food processing technology, specifically disclosing a high-flavor soy milk processing method based on optimized soybean quality evaluation indicators. The steps include: selecting soybeans, washing them, and soaking them in water, controlling the soaking temperature at 15-25℃, the time at 7-9 hours, and the pH value at 6-8; then grinding the soaked soybeans, controlling the grinding time at 280-320 seconds; finally, boiling the soy milk to inactivate enzymes, filtering, and cooling to sterilize to obtain the finished product. This invention, without adding any external substances, effectively reduces the formation of beany odor substances such as hexanal by optimizing the soaking and grinding process parameters, significantly improving the uniformity of soy milk particle size distribution and system stability, resulting in a rich, smooth, and balanced natural flavor. Furthermore, the process parameters are precisely quantified, making it easy to automate in industrial applications.

Description

Technical Field

[0001] This invention relates to the field of food processing technology, specifically to a high-flavor soybean milk processing method based on optimized soybean quality evaluation indicators. Background Technology

[0002] Soybeans, as an important source of plant protein and oil, are popular among consumers for their processed product, soy milk, which is rich in nutrients and easily digestible. However, a common misconception exists that higher protein content equates to better soy milk quality, leading to a market preference for high-protein soybean varieties. In reality, soybean quality indicators and soy milk sensory quality are not simply linearly positively correlated. Fat, as a crucial component, plays a decisive role in soy milk's flavor development, emulsification stability, and the richness and smoothness of its texture. Therefore, developing suitable processing techniques for soybean varieties with high fat and moderate / low protein content is key to improving soy milk quality.

[0003] However, existing soy milk processing technologies lack dynamic control mechanisms based on soybean quality, and mostly adopt traditional, one-size-fits-all processes. The soaking process is only considered a pretreatment step to soften the soybeans, and grinding also relies heavily on empirical parameters. This traditional process, lacking precise control, often leads to: 1) unreasonable soaking conditions, insufficient protein dissolution or excessive dissolution of endogenous substances, resulting in soy milk that is thin, has low viscosity, a strong fishy smell, or a bland flavor; 2) inaccurate grinding parameters, resulting in uneven particle size, leading to a coarse texture and easy separation and sedimentation.

[0004] To improve the flavor of soy milk and reduce its beany taste, various technologies have been explored, but most have not optimized the basic processing based on the inherent quality indicators of soybeans. For example: 1) Patent CN115226848A discloses a method for inactivating lipoxygenase by stepwise high-temperature heat shock (70℃-100℃). Although this method can reduce the beany taste, the high-temperature heat shock process has high equipment requirements and high energy consumption. Furthermore, excessive heat treatment may damage the natural properties of soybeans and cause the inactivation of some endogenous enzymes while inactivating LOX, resulting in the loss of flavor precursor substances.

[0005] 2) Patent CN116869126A discloses a method for chelating hexanal and inhibiting lipid oxidation by soaking in vinegar, fermenting with compound bacteria, and adding antioxidants such as blueberry juice and beetroot powder. Although this method effectively removes the fishy smell, it introduces additional chemical or biological additives, which not only makes the process complex and costly, but also masks or changes the original natural soy flavor of the soy milk.

[0006] In summary, for soybean varieties with high fat content and moderate or low protein content, how to overcome the problems of thin texture, insufficient consistency and poor flavor caused by traditional processes by optimizing basic processing parameters without introducing external substances or undergoing extreme heat treatment is a technical problem that urgently needs to be solved in the current soybean milk processing industry. Summary of the Invention

[0007] The purpose of this invention is to provide a high-flavor soybean milk processing method based on optimized soybean quality evaluation indicators, so as to solve the problems of thin texture, insufficient consistency and poor flavor caused by existing soybean milk processing technology.

[0008] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: A method for processing high-flavor soybean milk based on optimized soybean quality evaluation indicators is provided, including the following steps: S1. Select soybeans, wash them, and soak them in water. Control the soaking temperature to be 15-25℃, the soaking time to be 7-9 hours, and the soaking pH value to be 6-8. S2. Mix the soaked soybeans with grinding water and grind them together, controlling the grinding time to 280-320 seconds, to obtain raw soybean milk; S3. Heat the raw soy milk to 95-100℃ to boil and inactivate the enzymes, filter, and sterilize the filtrate after cooling to obtain high-flavor soy milk.

[0009] Preferably, in step S1, the soybeans are high-fat soybean varieties with a fat content ≥23g / 100g and a protein content ≤38g / 100g, and moderate or low protein content.

[0010] Preferably, in step S1, the weight ratio of soybeans to soaking water is 1:2 to 4. Preferably, in step S2, the weight ratio of soaked soybeans to grinding water is 1:7 to 9.

[0011] Preferably, the parameter combination for steps S1 and S2 is as follows: controlling the soaking temperature to 20°C, the soaking time to 8 hours, the soaking pH value to 7, and controlling the grinding time to 300 seconds.

[0012] Preferably, in step S3, the temperature of the boiled slurry is raised to 95-100°C and maintained for 8-12 minutes; the filtration is performed using a 180-220 mesh sieve.

[0013] The beneficial effects of this invention are as follows: 1. Significantly Reduces Soybean Odor While Preserving Natural Properties: This invention, without adding any external substances or undergoing extreme heat treatment, effectively inhibits lipid oxidation through the synergistic regulation of room temperature (20℃) soaking combined with precise time-limited grinding. Experimental data shows that this method reduces the content of hexanal, a hallmark of soybean odor, in soy milk by 56% compared to the comparative process, significantly reducing off-odors; at the same time, it retains and enhances key substances with sweet and fruity aromas, such as 2,3-pentanedione and 2-heptenal, resulting in a more balanced, richer, and more umami-rich flavor in the soy milk.

[0014] 2. Significantly Improved Soy Milk Texture and System Stability: Optimized soaking conditions (20℃, 8h, pH 7) and grinding time (300s) effectively improved the microstructure of the soy milk system. Experimental data showed that the resulting soy milk had a more concentrated particle size distribution (D50 of 0.52μm) and a distribution width (SPAN value) of only 26.88, significantly better than the comparative process. This indicates that the soy milk particle system has excellent uniformity, effectively solving the problems of thin, rough texture and easy stratification caused by traditional processes, resulting in a richer and smoother soy milk texture.

[0015] 3. Precisely quantified process parameters, easy for industrial control: This invention establishes a complete set of quantitative process parameters from raw material selection and soaking to grinding for soybean varieties with high fat and moderate / low protein content. This method eliminates the need for complex chemical additives or special step-by-step heat shock equipment. The process is simple, the parameters are precise, and the energy consumption is low. It has the advantages of being environmentally friendly and cost-effective, greatly improving production operability and making it suitable for industrial automation control. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the process flow for the high-flavor soy milk processing method of the present invention; Figure 2 The graph shows the effects of different single-factor conditions (soaking temperature, soaking time, soaking pH, and grinding time) on the sensory scores of soy milk. Figure 3 Radar images of the electronic tongue taste characteristics of soy milk prepared in Examples 1 and 2; Figure 4 The particle size distribution frequency curves of the soy milk prepared in the examples and comparative examples 1-2 are shown. Figure 5 Cluster analysis heatmaps of the volatile flavor compound content in the soy milk prepared in Examples 1 and 2 and Comparative Examples 1-2; Figure 6 The graph shows the total content of flavor substances and the percentage of each flavor substance in the soy milk prepared in the examples and comparative examples 1-2. Detailed Implementation

[0017] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0018] Example 1. Raw material selection Extensive literature review and previous quality evaluation studies by our research group revealed that the relationship between soybean quality indicators and the sensory quality of soy milk is not a simple linear positive correlation (Zhu GY, Huang X, Peng XY, et al. A deep learning-based quantitative prediction model for the processing potentials of soybeans as soymilk raw materials[J / OL]. Food Chemistry, 2024, 453: 139671. DOI:10.1016 / j.foodchem.2024.139671.). High protein content does not necessarily lead to a high sensory score, while fat plays a decisive role in the flavor, emulsification stability, and richness of soy milk. As shown in Table 1 below, Heihe 43 (fat 23.89g / 100g, protein 34.21%) had the highest overall sensory score (77.6 points). Therefore, this superior variety was selected as the benchmark raw material for subsequent process optimization in this invention.

[0019] Table 1 Sensory scores of soy milk made from different soybean varieties 2. Specific methods and steps This embodiment provides a high-flavor soybean milk processing method based on the above-mentioned preferred raw materials, serving as a benchmark process for subsequent optimization experiments. The method specifically includes the following steps: (1) Selection of materials and soaking Selected soybeans for soy milk are used. After washing, the soybeans are soaked in purified water at a constant temperature and pH of 20℃ for 8 hours with a water-to-solid weight ratio of 1:3. This step aims to fully soften the soybeans by absorbing water and improve protein dissolution efficiency by inhibiting the formation of undesirable flavor compounds in the early stages.

[0020] (2) Grinding Soaked soybeans are mixed with grinding water at a weight ratio of 1:8, and the grinding time is controlled at 300 seconds. This step aims to ensure that the soybeans are broken down to the optimal fineness by precisely controlling the grinding time, thus ensuring that the solids are fully dissolved.

[0021] (3) Finished product Raw soy milk and soy pulp are boiled under low pressure and rapidly heated to 95-100℃, maintained for 10 minutes to inactivate LOX and other anti-nutritional factors. The mixture is then filtered through a 200-mesh sieve, rapidly cooled with ice water, and bottled for sterilization, resulting in soy milk with a low beany taste and a rich, smooth texture.

[0022] 3. Single-factor experiment To clarify the independent effects of each process parameter on the quality of soy milk and to determine the factor level range for response surface methodology optimization experiments, single-factor experiments were conducted based on the above processing methods. These experiments involved varying the soaking temperature (4℃, 25℃, 35℃, 50℃), soaking time (6h, 8h, 10h, 12h), soaking pH (5, 6, 7, 8), or grinding time (100s, 200s, 300s, 400s). Sensory evaluations were also performed to screen for factors that significantly affect the flavor of soy milk.

[0023] Depend on Figure 2 It can be seen that the soaking and grinding processes of soybeans have a significant impact on the sensory score of soy milk (p < 0.05). Analysis shows that, when soaking time is considered as a single factor, the sensory score initially increases and then decreases with increasing soaking time, reaching its highest point at 8 h (31.8 points), followed by 10 h. When soaking temperature is considered as a single factor, samples soaked at 25 ℃ have the highest sensory score (34.6 points). When soaking pH is considered as a single factor, samples soaked at pH 7 have the highest sensory score (32.73 points). When grinding time is considered as a single factor, there are significant differences among each group of variables, with the sensory score reaching its highest value of 34 points at 300 s grinding.

[0024] In summary, the key factors that significantly affect the sensory quality of soy milk and their optimal levels were identified through single-factor experiments: soaking time of 8 h, soaking temperature of 25 ℃, soaking pH of 7, and grinding time of 300 s.

[0025] 4. Response Surface Optimization Experiment Based on the results of single-factor experiments, to examine the interactions among factors and seek the optimal combination of process parameters, factors with significant impact on sensory scores were selected, with the optimized sensory score of soy milk used as the response value. Each factor was set with two levels: low (-1) and high (+1). The factors included: A (soaking temperature), B (soaking time), C (soaking pH), and D (grinding time). Using Design-Expert 13 software, a four-factor, three-level Box-Behnken design was employed to optimize the sensory evaluation of the soy milk, as shown in Table 2 below.

[0026] Table 2 Response Surface Design Table The experimental design was carried out using Design-Expert 13 statistical analysis software. Sensory scores were used as the response value, and soaking temperature, soaking time, soaking pH and grinding time were used as independent variables. A four-factor, three-level experimental design for sensory scoring of soy milk was established, as shown in Table 3 below.

[0027] Table 3 Experimental Design and Results A variance statistical analysis was performed on the mathematical model for predicting the sensory scores of soy milk. The F-value is an important indicator for evaluating the influence of each variable on the response value. The larger the F-value, the higher the contribution of the relevant model component to the response. Among the single factors, soaking temperature has the greatest influence on the sensory scores of soy milk. When the significance test probability P < 0.05, the variable has a significant effect on the response value; when the significance test probability P < 0.01, the variable has a highly significant effect on the response value, which is statistically significant. As shown in Table 4, A, B, C, BC, A2, B2, C2, and D2 have significant or highly significant effects on the sensory scores of soy milk. The model P < 0.0001 indicates that the model is statistically significant, and the correction coefficient R is [value missing]. 2 The value is 0.9913, indicating a high model fit and a high R-value. 2 The adj=0.9826 value explains 98.26% of the experimental response value variation and has a correlation coefficient R with the prediction. 2 The similarity between pred and pred indicates that the model has high prediction accuracy.

[0028] Table 4. Analysis results of sensory scores for soy milk (Note: * indicates statistically significant difference, P<0.05; ** indicates highly significant difference, P<0.01) To further accurately determine the optimal solution of the process, with the maximum sensory score of the soy milk as the optimization objective, based on the results of the Design-Expert 13 software operation, the optimal sensory score under the combined influence of A soaking temperature (°C), B soaking time (h), C soaking pH, and D grinding time (s) is as follows: soaking temperature is 20.13°C, soaking time is 8.23 ​​h, soaking pH is 7.04, grinding time is 297.745 s, and the sensory score predicted by the model under these conditions is 43.

[0029] Based on the feasibility of actual production, the theoretically optimal conditions were modified as follows: soaking temperature at room temperature (20℃), soaking time of 8 hours, soaking pH of 7, and grinding time of 300 seconds. This combination of parameters not only meets the optimal range predicted by the model but also facilitates industrial production control.

[0030] 4. Verification and comparison experiment To verify the superiority of the above-mentioned optimal process parameters, two comparative examples were set up for comparative analysis.

[0031] Example: Soaking temperature was 20℃ (room temperature), soaking time was 8h, soaking pH was 7, and grinding time was 300s.

[0032] Comparative Example 1: Soaking temperature was 4℃, soaking time was 6 h, soaking pH was 7, and grinding time was 400 s (This group aimed to investigate the effects of low-temperature short-time soaking and extended grinding time on the flavor and texture of soy milk, simulating the effects of non-optimal parameter combinations). Comparative Example 2: Soaking temperature was room temperature, soaking time was 8 h, soaking pH was 8, and grinding time was 300 s (this group aimed to investigate the effect of deviation from the optimal pH value on the stability and flavor of the system). Table 5. Cumulative particle size distribution characteristics of soybean milk from three examples. (1) Taste and flavor analysis Flavor encompasses not only aroma but also taste; therefore, an electronic tongue sensor was used to characterize the taste fingerprint of soy milk. For example... Figure 3 As shown, AHS is a sensor sensitive to sourness, CTS is a sensor sensitive to saltiness, and NMS is a sensor sensitive to umami. It can be seen that the umami flavor of the example is the highest, the sourness flavor of Comparative Example 1 is the highest, and the saltiness flavor of Comparative Example 2 is the highest. Umami flavor is the most important indicator in the taste of soy milk, therefore the example has the best taste.

[0033] (2) Particle size and stability analysis Soy milk is a multiphase complex dispersion system, exhibiting characteristics of suspensions, colloids, and oil-in-water emulsions. The system contains particles of varying sizes, and their size distribution directly affects the physical stability and mouthfeel of the soy milk. For example... Figure 4As shown in the diagram, the horizontal axis of the frequency distribution graph represents particle size, and the vertical axis represents the proportion of particles within each size range. A sharp curve indicates concentrated (uniform) particle size distribution, while a flat curve indicates dispersion (large differences). Therefore, a higher peak in the particle size distribution graph means a larger proportion of particles near that size, a more concentrated particle size distribution, better uniformity, and a smoother texture in soy milk. Figure 4 It can be seen that the peak of the three soy milk samples is highest at 0.486 μm, with the highest peak in the example sample, followed by Comparative Example 2, and then Comparative Example 1. Therefore, the soy milk sample from the example sample is the most stable and uniform. D10 is the particle size (fine particle end boundary) when the cumulative distribution reaches 10%. D50 is the median particle size, i.e., the particle size when the cumulative distribution reaches 50%. D90 is the particle size (coarse particle end boundary) when the cumulative distribution reaches 90%. SPAN is a dimensionless index characterizing the width and uniformity of the particle size distribution. The smaller the SPAN, the more concentrated the particle size and the more uniform the system; the larger the SPAN value, the greater the difference in particle size, the wider the distribution, and the stronger the polydispersity. Therefore, as shown in Table 5, the stability and uniformity of the example sample are better than the other two comparative examples, confirming the positive control effect of the process parameters of the present invention on the microstructure.

[0034] (3) Analysis of volatile flavor compounds The volatile flavor compounds in soy milk were detected using headspace solid-phase microextraction-gas chromatography-mass spectrometry (HS-SPME-GC-MS). A total of 37 flavor compounds were detected, including 7 alcohols, 16 aldehydes, 7 ketones, 1 ester, 2 acids, and 4 furans.

[0035] To visually demonstrate the impact of different processes on flavor compounds, cluster analysis was performed on the detected volatile components. The results are as follows: Figure 5 As shown in the image. Cluster heatmaps can be used to cluster the three examples of soy milk; gray indicates a high content of the compound in the soy milk, while purple indicates a low content of the compound. Figure 5Soy milk can be divided into two categories: Comparative Example 1 and Examples / Comparative Example 2, indicating that Examples and Comparative Example 2 have similar flavors. Comparative Example 1 has a large area of ​​gray, Comparative Example 2 has a large area of ​​purple, and Examples have a richer color palette. Comparative Example 1 shows the highest overall abundance of volatile flavor compounds and a high content of aroma compounds, but it is also accompanied by high levels of beany and other off-flavor components. This group has the highest degree of lipid hydrolysis and oxidation, with a large accumulation of rancid and beany off-flavor compounds such as valeric acid, hexanoic acid, and short-chain saturated aldehydes, resulting in poor flavor harmony. Comparative Example 2 has the lowest overall content of various volatile compounds, but it lacks aroma compounds and has a bland flavor. It inhibits both lipid oxidation and Maillard reactions. In Examples, the content of characteristic off-flavor short-chain fatty acids (valeric acid and hexanoic acid) is significantly reduced (the color blocks are pink with low abundance), while Maillard sweet aromas such as 2,3-pentanedione and 2-heptenal are prominent, with fewer off-flavors and a more balanced, harmonious, and fragrant overall flavor. Furthermore, the total content of flavor compounds and the proportion of each flavor compound are also analyzed. Figure 6 It can be seen that the total flavor compounds are the least in ratio 2, and the flavor is the most bland. Alcohol compounds usually have pleasant aromas such as light fragrance, floral and fruity aromas, and have the highest proportion in the examples. Furans and ketones are also important flavor compounds, belonging to aroma compounds and exhibiting flavors such as fruity aroma, floral aroma, oily aroma and soy milk aroma. These also have the highest proportion in the examples.

[0036] OAV (Organic Acid Value) is used to assess the contribution of specific compounds to the overall aroma of soy milk samples. It is a ratio of the content of a volatile compound to its aroma threshold in water. Compounds with OAV ≥ 1 are considered to contribute significantly to the aroma of soy milk samples and are key flavor components in soy milk. Through OAV analysis, a total of 19 volatile components contributing to the overall aroma of soy milk were identified in three examples, as shown in Table 6. Hexanal, exhibiting grassy and fatty aromas, was the most prominent off-flavor component with a beany taste. In Comparative Example 1, hexanal was 504.86 μg·kg⁻¹. -1 In the example, hexanal was 220.06 μg·kg. -1 In Comparative Example 2, hexanal was 273.79 μg·kg⁻¹. -1 The hexanal content in this example was reduced by 56% compared to Comparative Example 1 and by 20% compared to Comparative Example 2. Excessive hexanal content would severely mask other aroma components; therefore, this example exhibited the lowest beany odor. 2-Hepanotone, as an aroma component, exhibited a fruity aroma, with a content of 13.54 μg·kg⁻¹ in Comparative Example 1. -1 In the example, it was 33.11 μg·kg. -1 In Comparative Example 2, the value was 25.14 μg·kg⁻¹. -1 The 2-heptanone content in the examples was increased by 144% compared to Comparative Example 1 and by 32% compared to Comparative Example 2.

[0037] Table 6 Key volatile flavor compounds and their thresholds, flavor characteristics descriptions Comparative analysis revealed that: Comparative Example 1 had the strongest flavor compounds, but also the most off-flavor components such as beany smell, which masked the aroma; Comparative Example 2 had the weakest flavor, with low levels of both off-flavor and aroma compounds; while the example sample had low levels of off-flavor compounds and high levels of aroma compounds, resulting in the most balanced and rich flavor.

[0038] In summary, the embodiments outperformed the comparative examples in terms of sensory evaluation, particle size distribution uniformity, and flavor compound coordination, demonstrating that the combination of soaking and grinding parameters provided by the present invention has a significant synergistic effect, effectively solving the problems of thin texture, insufficient consistency, and poor flavor caused by traditional soy milk processing, while improving product uniformity and stability.

[0039] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

Claims

1. A method for processing high-flavor soybean milk based on optimized soybean quality evaluation indicators, characterized in that, Includes the following steps: S1. Select soybeans, wash them, and soak them in water. Control the soaking temperature to be 15-25℃, the soaking time to be 7-9 hours, and the soaking pH value to be 6-8. S2. Mix the soaked soybeans with grinding water and grind them together, controlling the grinding time to 280-320 seconds, to obtain raw soybean milk; S3. Heat the raw soy milk to 95-100℃ to boil and inactivate the enzymes, filter, and sterilize the filtrate after cooling to obtain high-flavor soy milk.

2. The processing method according to claim 1, characterized in that, In step S1, the soybeans are high-fat soybean varieties with a fat content ≥23g / 100g and a protein content ≤38g / 100g, and moderate or low protein content.

3. The processing method according to claim 1, characterized in that, In step S1, the weight ratio of soybeans to soaking water is 1:2 to 4.

4. The processing method according to claim 1, characterized in that, In step S2, the weight ratio of soaked soybeans to grinding water is 1:7 to 9.

5. The processing method according to any one of claims 1 to 4, characterized in that, The parameter combination for steps S1 and S2 is as follows: the soaking temperature is controlled at 20℃, the soaking time is 8h, the soaking pH value is 7, and the grinding time is controlled at 300s.

6. The processing method according to claim 1, characterized in that, In step S3, the boiling pulp is heated to 95-100°C and maintained for 8-12 minutes; the filtration is performed using a 180-220 mesh sieve.

7. A high-flavor soy milk prepared by the processing method according to any one of claims 1-6.

Images