Targeted extraction and activity stabilization method for dihydromyricetin in ampelopsis grossedentata
By employing nanoemulsion encapsulation and intelligent temperature-controlled fermentation technology, the problems of activity degradation, microbial contamination, and storage oxidation of dihydromyricetin in vine tea during enzymatic hydrolysis have been solved, achieving efficient extraction and stable preservation, improving extraction efficiency and activity stability, and reducing production costs.
Patent Information
- Application Number
- CN202511555014.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-29
- Publication Date
- 2026-01-23
AI Technical Summary
In the existing technology, the activity of vine tea dihydromyricetin is severely degraded during enzymatic hydrolysis, microbial contamination leads to enzyme inactivation, and oxidation and mold growth occur during storage, affecting its preservation and application.
By employing nanoemulsification encapsulation, low-temperature membrane sterilization, and intelligent temperature-controlled fermentation technologies, combined with supramolecular host-guest assembly, magnetic nano-separation, and biomimetic mineralization microcapsule encapsulation, a system for preserving the activity of dihydromyricetin and stabilizing tea beverages was constructed. The extraction and storage environment was controlled through multi-objective particle swarm optimization and deep learning.
It achieves efficient extraction and stable preservation of dihydromyricetin, increasing extraction yield by 40%, purity by 32%, reducing activity decay during storage, lowering the risk of oxidative degradation and microbial contamination, increasing production efficiency by 50%, and reducing costs by 42%.
Abstract
Description
Technical Field
[0001] This invention relates to the field of automated preparation technology, specifically to a method for targeted extraction and activity stabilization of dihydromyricetin from vine tea. Background Technology
[0002] Currently, the preservation of dihydromyricetin in vine tea and the preparation of enzyme tea face three major technical bottlenecks: Severe degradation of activity during enzymatic hydrolysis: When the temperature of traditional enzymatic hydrolysis exceeds 40℃, the loss rate of dihydromyricetin due to polyphenol oxidase catalysis reaches 35%. After a certain company adopted enzymatic hydrolysis at 45℃, the activity retention rate dropped from 32% of the raw material to 19%, which could not meet the high activity requirements.
[0003] Microbial contamination leads to enzyme inactivation: During the fermentation process of enzyme tea, the contamination rate of miscellaneous bacteria reaches 20%, resulting in a decrease of more than 50% in protease activity, a reduction in the conversion efficiency of dihydromyricetin, and a fluctuation of more than ±25% in the content of effective components in the finished enzyme tea.
[0004] Oxidation and mold growth during storage: Existing packaging technology cannot completely block oxygen and moisture. After 3 months of storage, the activity of dihydromyricetin in vine tea enzyme tea decreases to 15%, the mold rate reaches 12%, and a rancid odor appears. The short shelf life restricts industrial application. Summary of the Invention
[0005] This invention constructs a system for preserving the activity of dihydromyricetin and stabilizing the preparation of tea beverages through nano-emulsion encapsulation, low-temperature membrane sterilization, and intelligent temperature-controlled fermentation technology, solving the problems of high-temperature inactivation, severe precipitation, and microbial contamination in traditional processes.
[0006] This invention provides a method for targeted extraction and activity stabilization of dihydromyricetin from vine tea, comprising the following steps: Supramolecular host-guest assembly extraction: Preparation of sulfonated β-cyclodextrin-graphene quantum dot supramolecular assemblies (particle size 50-80 nm) based on a multi-objective particle swarm optimization (MOPSO) algorithm model. The extraction temperature was controlled at 25-30℃, the supramolecular assembly concentration at 0.8%, and the extraction time at 45 min. Y_1 represents the yield, Y_2 represents the activity retention rate, and T represents the extraction time. Fe3O4@SiO2 magnetic nanoparticle separation and purification: Amino-functionalized Fe3O4@SiO2 magnetic nanoparticles (saturation magnetization 65 emu / g) were used, based on the magnetic separation kinetics model. The magnetic field strength was controlled at 0.3T and the adsorption time at 15min. Biomimetic mineralization microcapsule encapsulation: Chitosan-sodium alginate-calcium phosphate biomimetic mineralization microcapsules were prepared by a layer-by-layer self-assembly method, and the adaptive neurofuzzy reasoning system (ANFIS) model was utilized. , control the microcapsule wall thickness 5-8 μm, embedding rate ≥92%.
[0007] Further, in the supermolecular host-guest assembly extraction, the mass ratio of sulfonated β-cyclodextrin to graphene quantum dots is 3:1, the supermolecular assembly forms a 1:1 host-guest complex with dihydromyricetin, and the yield of dihydromyricetin after extraction is ≥2.5g / 100g, and the activity retention rate is ≥90%.
[0008] Further, in the Fe3O4@SiO2 magnetic nanometer separation and purification, the particle size of the magnetic nanoparticles is 100-200 nm, the separation purity of dihydromyricetin is ≥95%, the recovery rate of magnetic material is ≥98%, and the separation period is ≤15 min.
[0009] Further, in the biomimetic mineralization microcapsule embedding, the mass ratio of chitosan-sodium alginate is 2:1, the thickness of the calcium phosphate mineralization layer is 1-2 μm, and the half-life of the microcapsule optimized by the ANFIS model under light conditions (600 lux) is ≥360 days.
[0010] Further, it further includes the steps of intelligent light-oxygen environment response regulation: constructing a deep learning-proportional integral derivative (DL-PID) composite controller, real-time monitoring and regulating the storage environment light intensity (300-600 lux), oxygen concentration (0.1-5%), control accuracy ±0.2%, and maintaining a low-oxygen-light-avoiding microenvironment around the microcapsule.
[0011] Further, it further includes the steps of full-process quality digital twinning: deploying a near-infrared spectrum online detection module, establishing a Transformer-ResNet fusion model to predict the activity of dihydromyricetin, with a prediction error of ≤0.8%, and combining digital twinning technology to replicate the entire production process.
[0012] Further, in the supermolecular host-guest assembly extraction, the ultrasonic assisted extraction power is 300W, the solvent is 0.02mol / L phosphate buffer solution pH5.5, and the solid-liquid ratio is 1:20g / mL. The oxidation rate of dihydromyricetin is ≤5%.
[0013] Further, the yield of dihydromyricetin in the full-process control is ≥2.5g / 100g, the purity is ≥95%, the activity attenuation is ≤8% after 12 months of storage, and no oxidative degradation products are generated.
[0014] Further, the system of the method comprises: The supermolecular extraction unit: integrates the MOPSO algorithm controller and the ultrasonic assisted device, the temperature control accuracy is ±0.2℃, and the concentration of the supermolecular assembly is controlled in the range of 0.1-2%; Magnetic separation and purification system: containing Fe3O4@SiO2 particle synthesis module and magnetic field control device, magnetic field strength 0.1-1T, separation efficiency monitoring accuracy ±1%; Bionic mineralization embedding workstation: Equipped with a layer-by-layer self-assembly reactor and ANFIS parameter optimization module, with microcapsule particle size control accuracy of ±2μm.
[0015] 10. The system according to claim 9 is characterized in that it further includes an intelligent environmental response storage chamber DL-PID controller and a photo-oxygen sensor array, with an oxygen concentration control range of 0-1% and a near-infrared spectrometer wavelength range of 900-1700nm for the digital twin quality platform, a data storage capacity of ≥5 million sets, and full-process parameters are tamper-proof and traceable through blockchain technology.
[0016] Beneficial effects Supramolecular targeted extraction mechanism: Sulfonated β-cyclodextrin forms a 1:1 host-guest complex with dihydromyricetin, which, combined with the fluorescence recognition function of graphene quantum dots, increases the extraction yield by 40% and avoids organic solvent residue. Magnetic nano-high-efficiency separation technology: The specific interaction between amino groups and supramolecular groups on the surface of Fe3O4@SiO2 particles shortens the purification cycle to 15 minutes, which is 8 times more efficient than the traditional resin method; Biomimetic mineralization dual protection: The calcium phosphate mineralization layer and the biopolymer membrane work together to form a "physical barrier-chemical buffer" dual protection structure, which improves light stability by 3 times; DL-PID intelligent control system: Deep learning predicts environmental change trends, PID corrects control quantities in real time, oxygen concentration fluctuation in storage chamber is ≤0.1%, and response speed is improved by 50%; Digital twin end-to-end optimization: The digital twin model, which integrates near-infrared spectroscopy and process parameters, can provide early warning of activity decay risk up to 24 hours in advance, with an accuracy rate of 99.2%. Detailed Implementation
[0017] Example 1 I. Raw material pretreatment Two-year-old Ampelopsis lanceolata tea leaves from Tongren, Guizhou (initial dihydromyricetin content 31.2g / 100g) were selected, and after removing impurities, they were crushed and passed through a 60-mesh sieve. They were then frozen and stored at -20℃ for later use. The moisture content of the raw materials after pretreatment was controlled below 7.5% (Kal Fischer method detection).
[0018] II. Core Process Steps and Parameter Control Supramolecular host-guest assembly extraction Preparation of supramolecular assembly: sulfonated β-cyclodextrin (degree of substitution 1.2) and graphene quantum dots (particle size 5 nm) were dissolved in 0.02 mol / L phosphate buffer (pH 5.5) at a mass ratio of 3:1, and magnetic stirring was performed at 600 rpm for 30 min to form an assembly with a particle size of 72 nm (dynamic light scattering instrument detection).
[0019] MOPSO algorithm optimization: taking yield, activity retention rate, and time as the objective function, the optimal parameters were determined after 50 generations of particle optimization: liquid ratio 1:20 (g / mL), ultrasonic power 300 W, temperature 28℃, and extraction time 45 min.
[0020] Extraction effect: the yield of dihydromyricetin was 2.61 g / 100 g, the activity retention rate was 90.3% (DPPH method for determining antioxidant activity), and HPLC detection showed that the peak area of oxidation products accounted for ≤4.8%.
[0021] Fe3O4@SiO2 magnetic nanoseparation and purification Preparation of magnetic nanoparticles: Fe3O4 core (particle size 80 nm) was synthesized by co-precipitation, SiO2 layer (thickness 20 nm) was coated by sol-gel method, and after functionalization with aminopropyl triethoxysilane, the saturation magnetization was 64.8 emu / g (detected by vibrating sample magnetometer).
[0022] Separation process: 0.5% (w / v) magnetic nanoparticles were added to the extract, and adsorption was performed by oscillation at 25℃ for 15 min (rotation speed 200 rpm), followed by separation under a magnetic field of 0.3 T for 5 min. After discarding the supernatant, 80% ethanol was used for desorption 3 times.
[0023] Purification effect: the purity of dihydromyricetin was increased from 35.2% in the crude extract to 95.7%, the recovery rate of magnetic particles was 98.2%, and the single separation cycle was 15 min (reduced by 87.5% compared with the resin method).
[0024] Biomimetic mineralization microencapsulation Preparation of microcapsules: the purified dihydromyricetin solution (10 mg / mL) was mixed with 2% chitosan solution at a volume ratio of 1:3, and then dropped into 1% sodium alginate solution (containing 5 mmol / L CaCl2) to form primary capsules. After immersion in 0.1 mol / L disodium hydrogen phosphate solution for 2 h, a calcium phosphate outer layer was formed.
[0025] ANFIS model optimization: input sodium alginate concentration, mineralization time, and stirring rate, output embedding rate and wall thickness, after 100 groups of sample training, the optimized parameters are sodium alginate 1.2%, mineralization time 2 h, and rotation speed 300 rpm, the final embedding rate is 92.5%, and the wall thickness is 6.8 μm (observed by scanning electron microscope).
[0026] Stability test: under the condition of 30℃ and 600lux light, the activity of non-embedded sample decreased by 42% in 30 days, while the activity of microcapsule-embedded sample decreased by only 5.3%.
[0027] Intelligent light-oxygen environment response storage DL-PID controller settings: light threshold 400lux, oxygen concentration threshold 1%, when the sensor detects that the light intensity rises to 450lux, the light shielding layer is automatically started (response time ≤2s); when the oxygen concentration is more than 1.2%, the oxygen adsorbent (activated carbon-sodium sulfite composite) is released.
[0028] Storage effect: under the condition of 40℃ / 75%RH for 12 months, the activity of dihydromyricetin is retained by 89.7%, the microcapsule morphology is complete (no rupture is observed by transmission electron microscopy), and no rancid odor is detected.
[0029] Full-process digital twin monitoring Near-infrared detection: collect 500-2500nm spectral data, the correlation coefficient between the predicted activity value by Transformer-ResNet model and the actual measured value by HPLC is 0.992, and the prediction error is 0.78%.
[0030] Digital twin platform: real-time display of 200+ process parameters such as extraction temperature, magnetic separation time, and embedding rate, production status at any time node can be traced back, and the accuracy rate of process abnormality warning is 99.2%.
[0031] Supramolecular host-guest assembly extraction: prepare sulfonated β-cyclodextrin-graphene quantum dot supramolecular assembly (particle size 50-80nm), based on multi-objective particle swarm optimization (MOPSO) algorithm model , control the extraction temperature at 25-30℃, the supramolecular assembly concentration at 0.8%, and the extraction time at 45min, where Y_1 is the yield, Y_2 is the activity retention rate, and T is the extraction time; Fe3O4@SiO2 magnetic nanometer separation and purification: use amino-functionalized Fe3O4@SiO2 magnetic nanoparticles (saturation magnetization 65emu / g), according to the magnetic separation kinetics model , control the magnetic field strength at 0.3T and the adsorption time at 15min; Biomimetic mineralization microcapsule embedding: prepare chitosan-sodium alginate-calcium phosphate biomimetic mineralization microcapsule by layer-by-layer self-assembly method, use adaptive neuro-fuzzy inference system (ANFIS) model , control the microcapsule wall thickness at 5-8μm and the embedding rate at ≥92%.
[0032] III. Product indicators and process scalability Product core indicators Active ingredients: dihydromyricetin content ≥95%, specific optical rotation [α]20 D-82.3° (consistent with standard), infrared spectrum at 3350 cm -1 -1 (phenolic hydroxyl), 1610 cm
[0033] Safety: Heavy metal content (Pb ≤ 0.15 mg / kg, Cd ≤ 0.05 mg / kg) meets GB1886.308-2022 standard, solvent residue ≤ 0.005% (headspace gas chromatography detection).
[0034] Economy: The production cost per ton is reduced by 42% compared with traditional process, among which the solvent cost is reduced by 60% and the labor cost is reduced by 55%.
[0035] Process scalability verification In Zhangjiajie rattan tea base (variety "Xiangteng 2") in Hunan, the concentration of supramolecular assembly was adjusted to 0.9%, and the yield of dihydromyricetin was 2.58 g / 100 g with a purity of 95.1%, proving that the process is suitable for different rattan tea producing areas. Currently, it has been applied in 3 production lines, with a daily output of 200 kg per line, and the product has been used as a functional food additive (addition amount 0.1-0.5%), with an antioxidant performance improved by 30% compared with similar products.
[0036] Beneficial effects Extraction efficiency: The yield of dihydromyricetin reaches 2.6 g / 100 g with a purity of 95.3%, which is increased by 44% and 32% respectively compared with traditional process; Active stability: After biomimetic mineralization embedding, the activity retention rate is 89.7% after 12 months of storage, and the half-life is extended to 360 days; Production efficiency: The extraction-purification cycle is shortened from 12 hours to 1.5 hours, the energy consumption is reduced by 45%, and the solvent consumption is reduced by 60%.
Claims
1. A method for targeted extraction and activity stabilization of ampelopsis meadownii dihydromyricetin, characterized in that, The method comprises the following steps: Supermolecular host-guest assembly extraction: preparation of sulfonated β-cyclodextrin-graphene quantum dot supramolecular assembly (particle size 50-80 nm), based on multi-objective particle swarm optimization (MOPSO) algorithm model , control the extraction temperature 25-30℃, the supramolecular assembly concentration 0.8%, the extraction time 45min, wherein Y_1 is the yield, Y_2 is the activity retention rate, T is the extraction time; Fe3O4@SiO2 magnetic nanoseparation and purification: using amino functionalized Fe3O4@SiO2 magnetic nanoparticles (saturation magnetization 65 emu / g), according to the magnetic separation kinetics model , control the magnetic field strength 0.3 T, adsorption time 15 min; Biomimetic mineralization microcapsule embedding: chitosan-sodium alginate-calcium phosphate biomimetic mineralization microcapsules were prepared by layer-by-layer self-assembly method, and adaptive neuro-fuzzy inference system (ANFIS) model was used to control the microcapsule wall thickness of 5-8 μm and the embedding rate of ≥92%.
2. The method of claim 1, wherein, In the supermolecular host-guest assembly extraction, the mass ratio of sulfonated β-cyclodextrin to graphene quantum dots is 3:1, the supermolecular assembly and dihydromyricetin form a 1:1 host-guest complex, the yield of dihydromyricetin after extraction is ≥2.5 g / 100 g, and the activity retention rate is ≥90%.
3. The method of claim 1, wherein, In the Fe3O4@SiO2 magnetic nanoseparation and purification, the particle size of the magnetic nanoparticles is 100-200 nm, the separation purity of dihydromyricetin is ≥95%, the recovery rate of the magnetic material is ≥98%, and the separation period is ≤15 min.
4. The method of claim 1, wherein, In the biomimetic mineralization microcapsule embedding, the mass ratio of chitosan-sodium alginate is 2:1, the thickness of the calcium phosphate mineralization layer is 1-2 μm, and the half-life of the microcapsule optimized by the ANFIS model is ≥360 days under light conditions (600 lux).
5. The method of claim 1, wherein, The method further comprises an intelligent light-oxygen environment response regulation step: a deep learning-proportional integral derivative (DL-PID) composite controller is constructed to monitor and regulate the light intensity (300-600 lux) and oxygen concentration (0.1-5%) of the storage environment in real time, the control accuracy is ±0.2%, and a low-oxygen-light-avoiding microenvironment around the microcapsule is maintained.
6. The method of claim 1, wherein, Also include the whole process quality digital twin steps: deploy near infrared spectrum online detection module, establish Transformer-ResNet fusion model Predict the activity of dihydromyricetin, the prediction error is ≤0.8%, and the whole process is reproduced by combining digital twin technology.
7. The method of claim 1, wherein, In the supermolecular host-guest assembly extraction, the ultrasonic auxiliary extraction power is 300 W, the solvent is 0.02 mol / L phosphate buffer solution pH 5.5, the solid-liquid ratio is 1:20 g / mL, and the oxidation rate of dihydromyricetin is ≤5%.
8. The method according to any one of claims 1 to 7, characterized in that, The dihydromyricetin extraction yield is ≥2.5 g / 100 g, the purity is ≥95%, the activity attenuation is ≤8% after 12 months of storage, and no oxidative degradation product is generated.
9. A system for implementing the method of any one of claims 1 to 8, characterized in that, The method comprises the following steps: The supermolecular extraction unit: an MOPSO algorithm controller and an ultrasonic auxiliary device are integrated, the temperature control accuracy is ±0.2℃, and the concentration of the supermolecular assembly is controlled in the range of 0.1-2%; The magnetic separation and purification system: the Fe3O4@SiO2 particle synthesis module and the magnetic field regulation device with a magnetic field strength of 0.1-1 T are included, and the separation efficiency monitoring accuracy is ±1%; The biomimetic mineralization embedding workstation: a layer-by-layer self-assembly reaction kettle and an ANFIS parameter optimization module are provided, and the microcapsule particle size control accuracy is ±2 μm.
10. The system of claim 9, wherein, The method further comprises a DL-PID controller of an intelligent environment response storage cabin and an array of light-oxygen sensors, an oxygen concentration control range of 0-1%, and a digital twin quality platform near-infrared spectrometer wavelength range of 900-1700 nm, a data storage capacity of ≥5 million groups, and the realization of the whole-process parameter non-tamperable traceability through the blockchain technology.