High-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed counter-current chromatography separation

Abstract

The application provides a ginsenoside high-purity extraction process based on supercritical fluid extraction and high-speed counter-current chromatography separation, under preset extraction conditions, supercritical fluid extraction device is used to extract ginseng powder to obtain ginsenoside crude extract, and the extraction conditions are determined according to the target ginsenoside component monomer type combination; the ginsenoside crude extract is dissolved in a solvent to obtain a liquid to be separated; under preset separation conditions, a high-speed counter-current chromatography device is used to separate the liquid to be separated to obtain a fraction of the target ginsenoside component monomer type, and the separation conditions are determined according to the target ginsenoside component monomer type. The application dynamically determines the conditions of supercritical fluid extraction, realizes targeted extraction according to the polarity distribution and solubility behavior of the target ginsenoside in the extraction stage, improves the separation efficiency, and further improves the purity of the target ginsenoside monomer.

Description

Technical Field

[0001] This invention relates to the field of supercritical fluid extraction technology, and in particular to a high-purity extraction process for ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography separation. Background Technology

[0002] Traditional ginsenoside extraction processes mainly employ organic solvent reflux extraction or water extraction, which suffer from problems such as solvent residue, low extraction efficiency, and saponin decomposition. Existing technologies have several issues that severely restrict the development of the ginsenoside extraction and separation industry. First, traditional extraction methods result in solvent residue and environmental pressure: While traditional alcohol or water extraction achieves high extraction rates, it leaves significant organic solvent residues, increasing the difficulty of post-processing and posing safety risks. Second, traditional extraction methods suffer from inefficient component separation and low purity: Traditional silica gel column chromatography suffers from sample adsorption and deactivation issues, high costs, and difficulty in rapidly separating high-purity pharmaceutical-grade ginsenoside monomers. Therefore, a new extraction process is urgently needed to improve the extraction efficiency and purity of ginsenoside monomers. Summary of the Invention

[0003] This invention provides a high-purity extraction process for ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography (HSCLC). It aims to address the problems of low extraction efficiency, insufficient monomer purity, and long separation time in existing technologies, thereby improving the extraction efficiency and purity of ginsenoside monomers. This invention dynamically determines the supercritical fluid extraction conditions based on the combination of target ginsenoside monomer types, avoiding a wide range of extraction parameters and improving the extraction efficiency of ginsenoside monomers. High-speed countercurrent chromatography accurately separates each target ginsenoside monomer, improving their purity. Since the extraction stage achieves targeted extraction based on the polarity distribution and solubility behavior of the target saponins, the crude extract has a higher relative content of target saponins and a simpler impurity profile, not only improving separation efficiency but also further enhancing the purity of the target ginsenoside monomers.

[0004] In a first aspect, embodiments of the present invention provide a high-purity extraction process for ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography, comprising: Step S1: Under preset extraction conditions, ginseng powder is extracted using a supercritical fluid extraction device to obtain crude ginsenoside extract. The extraction conditions are determined based on the combination of monomer types of the target ginsenoside components. Step S2: Dissolve the crude ginsenoside extract in a solvent to obtain the liquid to be separated. The solvent is determined based on the residual components in the crude ginsenoside extract. Step S3: Under the preset separation conditions, high-speed countercurrent chromatography is used to separate the liquid to be separated by high-speed countercurrent chromatography to obtain the fraction of the target ginsenoside monomer type. The separation conditions are determined according to the target ginsenoside monomer type, or the separation conditions are dynamically adjusted according to the stage parameters of the supercritical fluid extraction process.

[0005] In this embodiment of the invention, under preset extraction conditions, ginseng powder is extracted using a supercritical fluid extraction device to obtain a crude ginsenoside extract. The extraction conditions are determined based on the combination of target ginsenoside monomer types. The crude ginsenoside extract is dissolved in a solvent to obtain a liquid to be separated. Under preset separation conditions, a high-speed countercurrent chromatography (HSCLC) device is used to separate the liquid to obtain a fraction of the target ginsenoside monomer type. The separation conditions are determined based on the target ginsenoside monomer type. This invention dynamically determines the supercritical fluid extraction conditions based on the combination of target ginsenoside monomer types, avoiding a wide range of extraction parameters and improving the extraction efficiency of ginsenoside monomers. High-speed countercurrent chromatography can accurately separate each target ginsenoside monomer, improving the purity of the ginsenoside monomers. Since the extraction stage has already achieved targeted extraction based on the polarity distribution and solubility behavior of the target saponins, the crude extract has a higher relative content of target saponins and a simpler impurity profile, which not only improves the separation efficiency but also further improves the purity of the target ginsenoside monomers. Attached Figure Description

[0006] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a flowchart of a high-purity extraction process for ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography, provided by an embodiment of the present invention. Detailed Implementation

[0007] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0008] This invention provides a high-purity extraction process for ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography, comprising: Step S1: Under preset extraction conditions, ginseng powder is subjected to supercritical fluid extraction using a supercritical fluid extraction device to obtain crude ginsenoside extract. The extraction conditions are determined based on the combination of monomer types of the target ginsenoside components.

[0009] In this embodiment of the invention, ginseng roots are washed, sliced, dried, and then pulverized to a suitable particle size for extraction (usually 20-40 mesh) to remove metallic impurities, resulting in ginseng powder. The ginseng powder is then loaded into the extraction vessel of a supercritical fluid extraction device. The loading amount does not exceed 80% of the effective volume of the extraction vessel to ensure uniform fluid penetration.

[0010] Before formal extraction, the preset extraction conditions can be determined based on the pre-selected combination of target ginsenoside monomers. Specifically, the polar parameters (such as logP value, hydrophilic-lipophilic balance value) and solubility behavior in supercritical carbon dioxide of each target saponin monomer can be queried or calculated (obtained through empirical formulas, molecular simulations, or literature data). These parameters are then combined into an index representing the overall polarity of the target saponin combination (a comprehensive polarity index), for example, by weighting the polarity parameters of each monomer according to the target proportion. This comprehensive polarity index is then matched with a pre-stored extraction condition database. This database stores recommended operating ranges corresponding to different polarity index ranges, including the target density range of supercritical carbon dioxide (e.g., 0.6-0.9 g / mL). Using the equation of state between carbon dioxide density and temperature and pressure (such as the Span-Wagner equation or an empirically modified Peng-Robinson equation), the initial values ​​of extraction temperature and pressure required to reach this density range can be derived. Based on the solubility requirements of the strongest or weakest polar component in the target saponin composition, determine whether an organic auxiliary agent (such as ethanol, acetone, ethyl acetate, etc.) is needed, and set the volume ratio of the auxiliary agent to carbon dioxide. The type of auxiliary agent should be selected to increase the solubility of the target saponin by at least 50%, while avoiding irreversible effects on subsequent separation. If the target saponin composition contains multiple monomers with significant differences in polarity and nonpolarity (e.g., containing both Rb1 and Rg1), a segmented extraction strategy should be initiated: the extraction process is divided into multiple stages, each using different temperatures and pressures (and possibly different entrainer ratios), to sequentially enrich saponin components of different polarities.

[0011] After adding ginseng powder to the supercritical fluid extraction apparatus, the extraction vessel is closed, and the high-pressure carbon dioxide pump and entrainer pump are started to pressurize the system to the preset pressure and simultaneously heat it to the preset temperature to maintain the supercritical state. Carbon dioxide circulates, ensuring sufficient contact with the raw material. During the extraction process, samples of the extract can be collected periodically via an online sampling valve to detect the concentration of the target saponins. Extraction automatically stops when the concentration increase rate falls below the preset termination threshold.

[0012] After extraction, the extract is introduced into a separation vessel, where the carbon dioxide is vaporized and released by depressurization (below the critical pressure of carbon dioxide). Simultaneously, some of the entrainer evaporates, and the residue is collected as the crude ginsenoside extract. The gaseous carbon dioxide is recovered by condensation and recycled.

[0013] Step S2: Dissolve the crude ginsenoside extract in a solvent to obtain the liquid to be separated. The solvent is determined based on the residual components in the crude ginsenoside extract.

[0014] In this embodiment of the invention, the crude ginsenoside extract obtained in step S1 is taken and its quality, target ginsenoside content, type and content of residual organic adjuvants, and whether there is carbon dioxide residue are detected.

[0015] Depending on the properties of the crude extract, select one or more solvents (e.g., water, methanol, ethanol, n-butanol, or mixtures thereof) for dissolution. The polarity of the solvent should match the polarity of the main saponins in the crude extract, and it should also be miscible with any of the two-phase solvent systems in the subsequent high-speed countercurrent chromatography to avoid precipitation or phase disturbance during injection.

[0016] If step S1 uses fractional extraction, the crude extracts obtained in each stage are dissolved separately, and the most suitable solvent system is selected for each crude extract (for example, a higher proportion of organic solvent can be used for crude extracts in the low polarity stage, and a higher proportion of water can be used for crude extracts in the high polarity stage).

[0017] If residual carbon dioxide in the crude extract is detected to exceed acceptable levels (e.g., causing the solution to foam), degassing is performed before dissolution (e.g., ultrasonic degassing or vacuum degassing), and the polarity of the dissolving solvent is fine-tuned to compensate for the solubility changes caused by degassing.

[0018] The concentration of each target saponin in the solution is determined. If the concentration is too high (exceeding the HSCCC injection capacity), it is diluted with the same solvent. If the concentration is too low, it is appropriately concentrated by rotary evaporation to finally obtain a clear, homogeneous, particle-free liquid to be separated for later use.

[0019] Step S3: Under preset separation conditions, high-speed countercurrent chromatography is used to separate the liquid to be separated by high-speed countercurrent chromatography to obtain a fraction of the target ginsenoside monomer type. The separation conditions are determined according to the target ginsenoside monomer type or dynamically adjusted according to the stage parameters of the supercritical fluid extraction process.

[0020] In this embodiment of the invention, before formal separation, the partition coefficient K value of each target saponin monomer is determined according to the combination of monomer types of the target saponin component for each candidate two-phase solvent system (e.g., n-hexane-ethyl acetate-water, chloroform-methanol-water, ethyl acetate-n-butanol-water, etc.). The determination method can be the test tube partition method or the analytical HSCCC rapid extrapolation method.

[0021] The K-values ​​of each monomer are compared with a preset effective separation range. This effective separation range is defined as the range of K-values ​​that, under selected rotation speed and flow rate conditions, allow adjacent saponin chromatographic peaks to achieve baseline separation (resolution ≥ 1.5). At least one two-phase solvent system and its volume ratio are selected such that the K-values ​​of all target saponins fall within this range, and the ratio of K-values ​​of adjacent components is not less than a certain minimum threshold.

[0022] Calculate the degree of difference in K values ​​among the monomers in the target saponin combination (e.g., range, standard deviation, or minimum of adjacent ratios). If the degree of difference is small (i.e., all K values ​​are close to each other), an isocratic elution mode is selected; if the degree of difference is large (some adjacent K value ratios are less than an acceptable threshold), a gradient elution mode is selected.

[0023] If gradient elution is used, the K-value variation curves of each monomer under different mobile phase compositions can be further measured, and a K-value-mobile phase composition relationship diagram can be plotted. By overlaying the curves, the initial and final compositions that maximize the difference in K-values ​​between adjacent monomers can be identified. The elution process is divided into multiple stages based on the curve slope, and the rate of change in each stage is proportional to the slope of the monomer with the fastest K-value change in that stage. The rate of change is reduced in the curve intersection or convergence regions to prolong the separation time, while the rate of change is increased in the uniformly spaced regions to shorten the total time. Through iterative optimization, a function curve (gradient and rate of change) of mobile phase composition as a function of elution time is finally determined.

[0024] The selected two-phase solvent system is mixed in proportion, shaken thoroughly, and allowed to stand to separate the upper and lower phases. The lower phase (or upper phase, depending on the design) is used as the mobile phase, and the other phase is used as the stationary phase. The high-speed countercurrent chromatography apparatus is started, and the stationary phase is pumped into the spiral column at a certain flow rate. Then, the motor is started to make the column rotate and revolve to reach the preset speed. After the stationary phase has stabilized, the mobile phase is pumped in at the set flow rate, and the detector (such as an ultraviolet detector or an evaporative light scattering detector) is turned on to monitor the baseline.

[0025] After the baseline stabilizes, the liquid to be separated prepared in step S2 is injected into the system through the injection valve. The sample enters the spiral column with the mobile phase and repeatedly undergoes liquid-liquid partitioning with the stationary phase under the action of centrifugal force. Saponin monomers with different K values ​​migrate at different speeds and elute sequentially from the tail of the column. The detector records the signal in real time to form a chromatogram.

[0026] Based on the detector signal and preset collection rules, the fraction collection valve is automatically or manually switched to collect the elution peak corresponding to the target saponin monomer. The collected fraction is the solution of the target ginsenoside monomer type. If gradient elution is used, the fraction collection window needs to be dynamically adjusted according to the real-time peak start and end points. The collected fraction can be subjected to purity testing, purification, or drying according to preset post-processing methods.

[0027] In one possible embodiment, the high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography can be implemented as follows: 1. Raw material pretreatment: Ginseng root is crushed to 20-40 mesh to remove metal impurities and dried until the moisture content is ≤10%.

[0028] 2. Supercritical fluid extraction (SFE): Extraction conditions: Carbon dioxide (CO2) is used as the extraction fluid, supplemented with an appropriate amount of organic additives (such as ethanol).

[0029] Temperature and pressure: Within the temperature range of 40-80℃, the optimal extraction temperature is 65℃; within the pressure range of 150-300 bar, the optimal pressure is 250 bar. Under these conditions, carbon dioxide has the highest density and the strongest solubility.

[0030] The ratio of auxiliary agents is 10:1 in volume between CO2 and ethanol, and the ethanol content is controlled at 5%-10% (v / v). This can improve the solubility of polar saponins without affecting the subsequent liquid-liquid separation.

[0031] Number of cycles: After 30 minutes of preheating, perform 6-8 cycles of extraction to ensure that the saponins in the raw materials are fully dissolved.

[0032] Operating procedure: The pretreated raw material is loaded into the extraction vessel, and supercritical state control is performed to maintain a stable flow rate for extraction. After extraction, the extract is separated from CO2 by vacuum separation, and the recovered CO2 is recycled.

[0033] 3. Crude separation and desolventization: The extract is subjected to rotary evaporation or freeze drying to remove CO2 and organic additives, yielding crude ginsenoside extract.

[0034] 4. High-speed countercurrent chromatography (HSCCC) separation: Solvent system selection: Select a suitable two-phase system (such as n-hexane / ethyl acetate / water system) based on the polarity of the target saponin.

[0035] Two-phase system: n-hexane / ethyl acetate / water (1:1:1, v / v / v). This system has a stable partition coefficient (K value) at pH 5.5 and remains relatively static in countercurrent mode.

[0036] Rotation speed and flow rate: The rotor speed was controlled at 800 rpm to maintain a high centrifugal force; the flow rate was controlled at 1.0 mL / min to ensure a narrow and clear peak shape and to collect fractions with different retention times.

[0037] Predict the partition coefficient K value using thin-layer chromatography (TLC) or high-performance liquid chromatography (HPLC), ensuring that the K value is between 3 and 10 to achieve efficient separation.

[0038] Separation Procedure: A gradient elution mode is employed to address the distribution characteristics of ginsenosides. Elution Mode: Gradient elution is used, gradually transitioning from an initial ratio of 70% lower phase / 30% upper phase to 50% lower phase / 50% upper phase to differentiate between lipophilic and hydrophilic saponins.

[0039] 5. Refining and Purification: For target saponins (such as Rg1, Rb1, Re), preparative UPLC can be used for further purification to remove trace impurities.

[0040] For isomers that are difficult to separate, high performance liquid chromatography (HPLC) can be used for supplementary separation.

[0041] 6. Quality Control (QC): UPLC-MS / MS detection: UPLC-MS / MS technology was used for rapid identification and quantification of the isolated and purified monomers. A multiple reaction monitoring (MRM) mode was established, with specific precursor and daughter ion pairs set to achieve highly sensitive detection.

[0042] Method validation: The linear range, limit of detection (LOD), limit of quantitation (LOQ), recovery, and precision (RSD) were validated. The linearity of each ginsenoside was ensured to be good within the concentration range of 0.02–50 µg / mL, and the correlation coefficient (R) was [value missing]. 2 All are greater than 0.99.

[0043] Chromatographic conditions: An Acquity BEH C18 column (2.1×100 mm, 1.7 µm) was used. The mobile phase was A: 0.1% formic acid in water; B: 0.1% formic acid in acetonitrile. Gradient elution: 0-1 min 5% B; 1-4 min 5%→95% B; 4-5 min 95% B.

[0044] Mass spectrometry parameters: Multiple reaction monitoring (MRM) mode in positive ion mode, with the following settings for the parent ion-daughter ion pairs: Rg1 (m / z 819.5 → 207.1), Rb1 (m / z 961.5 → 205.1), Re (m / z 945.5 → 207.1).

[0045] In this embodiment of the invention, under preset extraction conditions, ginseng powder is extracted using a supercritical fluid extraction device to obtain a crude ginsenoside extract. The extraction conditions are determined based on the combination of target ginsenoside monomer types. The crude ginsenoside extract is dissolved in a solvent to obtain a liquid to be separated. Under preset separation conditions, a high-speed countercurrent chromatography (HSCLC) device is used to separate the liquid to obtain a fraction of the target ginsenoside monomer type. The separation conditions are determined based on the target ginsenoside monomer type. This invention dynamically determines the supercritical fluid extraction conditions based on the combination of target ginsenoside monomer types, avoiding a wide range of extraction parameters and improving the extraction efficiency of ginsenoside monomers. High-speed countercurrent chromatography can accurately separate each target ginsenoside monomer, improving the purity of the ginsenoside monomers. Since the extraction stage has already achieved targeted extraction based on the polarity distribution and solubility behavior of the target saponins, the crude extract has a higher relative content of target saponins and a simpler impurity profile, which not only improves the separation efficiency but also further improves the purity of the target ginsenoside monomers.

[0046] The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography provided by this invention can significantly improve the extraction purity and purification efficiency of ginsenosides. Supercritical fluid extraction achieves green extraction with no solvent residue, and high-purity pharmaceutical-grade monomers are separated by high-speed countercurrent chromatography, providing high-quality raw materials for the research and development of traditional Chinese medicine preparations and functional foods.

[0047] Optionally, the extraction conditions are determined based on the combination of monomer types of the target ginsenoside components, specifically including: determining the polarity parameters and solubility behavior of each target ginsenoside monomer in supercritical carbon dioxide; calculating the comprehensive polarity index of the target ginsenoside combination based on the polarity parameters and solubility behavior; determining the target density range of supercritical carbon dioxide by matching a preset extraction condition database based on the comprehensive polarity index; back-calculating and setting the initial values ​​of extraction temperature and extraction pressure based on the equation of state relationship between carbon dioxide density and temperature and pressure; determining the type of organic auxiliaries and their volume ratio with carbon dioxide based on the solubility requirements of the strongest or weakest polar component in the target ginsenoside combination; if the target ginsenoside combination contains multiple monomers with large differences in polarity, a segmented extraction strategy is adopted, adjusting the temperature and pressure in stages during the extraction process to sequentially enrich ginsenoside components of different polarities.

[0048] In this embodiment of the invention, before starting the extraction, the polarity parameters and solubility behavior of each monomer in the target saponin combination are first obtained to obtain a vector describing the polarity-solubility characteristics of each monomer in the target saponin combination.

[0049] Obtain the logP values ​​of each monomer by consulting literature databases (such as PubChem and Reaxys) or using computational chemistry software (such as ChemDraw and Molinspiration). If accurate values ​​cannot be obtained, reversed-phase HPLC can be used: employ a C18 column, elute with an acetonitrile-water gradient, and record the retention time of each saponin; a longer retention time indicates lower polarity. Compare the retention times with known standards and normalize to obtain the relative polarity index.

[0050] If the database already contains solubility data for the saponin under different conditions (temperature, pressure, type and proportion of additives), it can be retrieved directly. If no data is available, a dynamic method is used for preliminary experiments: the pure saponin is mixed with an inert support (such as glass beads) and packed into a column. Supercritical carbon dioxide is then passed through the column at a specific temperature and pressure, and the curve of the outlet concentration changing with the conditions is measured. The key outputs of the solubility behavior are: the trend of solubility under pure CO2 with temperature and pressure, and the factor by which solubility increases after adding different proportions of additives.

[0051] Based on polarity parameters and solubility behavior, the comprehensive polarity index of the target saponin combination is calculated. The comprehensive polarity index is obtained by weighting the polarity parameters (e.g., logP values) of each monomer according to their expected content proportion in the target combination. .

[0052] Examples (for illustration only, not specific values): If the target combination contains three saponins A, B, and C, the target mass ratio is: The polarity parameters are respectively ,but: ; If the solubility behavior of a monomer in the composition shows that it is highly dependent on the auxiliaries (i.e., almost insoluble in pure CO2), then the polarity parameter of that monomer is given a higher weight in the weighting process to reflect the actual extraction requirements.

[0053] In another embodiment, response surface methodology can be used. By designing several virtual combinations of different polarity indices for extraction experiments, the optimal comprehensive polarity index can be fitted with the total extraction rate as the response.

[0054] Based on the comprehensive polarity index and matched with the preset extraction condition database, the target density range of supercritical carbon dioxide is determined. For example, if logP=1.0 falls within the range of 0.8-1.2, then a carbon dioxide density of 0.7-0.8 g / mL is recommended, corresponding to a temperature of approximately 55-65℃ and a pressure of approximately 200-250 bar (specific values ​​need to be calculated using the equation of state).

[0055] The calculated comprehensive polarity index Used as the query key, it retrieves data from a database based on preset extraction criteria. A typical database structure is as follows: ; The density ranges in the database were determined based on extensive preliminary experiments, ensuring that the extraction rate of the target saponin combination corresponding to the polarity index range is the highest and the impurities are the fewest at that density.

[0056] During matching, if If it falls exactly on the boundary of a certain interval, the previous or next interval can be selected according to actual needs, or an interpolation method (such as linear interpolation) can be used to determine the intermediate density value.

[0057] Target density range for supercritical carbon dioxide output .

[0058] Based on the equation of state relationship between carbon dioxide density and temperature and pressure, the initial values ​​of extraction temperature and extraction pressure are deduced and set.

[0059] Solving for a given target density range using state equations (such as the Peng-Robinson equation) The possible combinations of temperature and pressure are considered. Due to the compressibility of supercritical carbon dioxide, the same density may typically correspond to multiple (T,P) points. The selection principle is to prioritize moderate temperatures (e.g., between 40-70°C) to avoid saponin degradation caused by high temperatures; to prioritize lower pressures (e.g., between 150-250 bar) to reduce equipment investment and energy consumption; if additives are added, the equation of state of the mixture must be used, considering the influence of the additives on the critical parameters.

[0060] The calculation process can utilize specialized thermodynamic software (such as Aspen Plus, NIST REFPROP) or call open-source libraries. The final output is a set of initial extraction temperatures. and extraction pressure This initial value serves as the setpoint for the supercritical fluid extraction device. This initial value can be fine-tuned during actual operation based on online monitoring results.

[0061] Based on the solubility requirements of the most or least polar component in the target saponin combination, determine the type of organic auxiliaries and their volume ratio to carbon dioxide.

[0062] If the solubility of all monomers in the target saponin composition in pure supercritical carbon dioxide is sufficiently high (e.g., the minimum solubility at the target density is higher than the threshold required by the process), then no additives are needed. Otherwise, additives are required.

[0063] For moderately polar saponins (such as Rb1 and Rg1), ethanol can be selected as an adjuvant; for strongly polar saponins (such as Re and Rf), a mixture of ethanol and a small amount of water (e.g., 95% ethanol + 5% water) can be selected to further increase the polarity; for weakly polar saponins (such as aglycones), ethyl acetate or acetone may be more effective.

[0064] The selection criterion can be the additive that increases solubility by the greatest factor in the preliminary experiment.

[0065] First, determine the target solubility value of the most polar component after adding the additive (usually a process design value, such as the minimum solubility required for extraction >90%). Using the "additive ratio-solubility" curve established in preliminary experiments, find the minimum additive ratio required to achieve this target. Simultaneously, check the increase in solubility of the weakest polar component at this additive ratio to ensure that non-target impurities are not over-extracted. Finally, select an equilibrium point, typically with a volume ratio between 1:20 and 1:5 (additive:CO2). The type of additive and its volume ratio to carbon dioxide are also important considerations. (For example, 1:10).

[0066] If the target saponin combination contains multiple monomers with significant polarity differences, a fractional extraction strategy is employed. The difference in polarity parameter (ΔP) between the most polar and least polar monomers in the target saponin combination is calculated. If ΔP exceeds a preset threshold (provided by a database, e.g., ΔP>0.3), it is determined that the polarity differences are significant, necessitating the use of a fractional extraction strategy.

[0067] In a segmented extraction strategy, the process involves dividing the mixture into stages based on the polarity of each monomer, resulting in 2-3 subgroups. For example: Stage 1: Subgroups with the lowest polarity (e.g., saponins or low-polarity saponins); Stage 2: Subgroups with moderate polarity; Stage 3: Subgroups with the highest polarity (e.g., high-polarity saponins).

[0068] The conditions for each stage are determined by calculating the comprehensive polarity index, target density, initial temperature and pressure values, and types and proportions of additives for each subgroup. For example, in the first stage, a lower proportion of additives (or pure CO2) is used, and the temperature and pressure settings are biased towards low polarity conditions; in the second stage, the proportion of additives is appropriately increased, and the temperature and pressure are adjusted; in the third stage, the highest proportion of additives is used (water is added if necessary), and the temperature and pressure may be further adjusted.

[0069] During the extraction process, the saponin composition in the extract is monitored online. When the rate of increase in the concentration of the target saponin in a certain subgroup decreases significantly, the system automatically switches to the next stage. During the switch, the temperature and pressure can be changed stepwise (direct jump) or gradually (slowly increase / decrease temperature / decrease pressure). The proportion of the entrainer is also adjusted accordingly.

[0070] The extracts obtained from each stage can be collected separately (to obtain crude extracts of different polarities) or combined for the next step. Separate collection is more beneficial for subsequent separation because the impurity profile of each crude extract is simpler. If the polarity difference among the target saponin combinations is not significant (ΔP is below the threshold), fractional extraction is not used, and a single condition is used directly.

[0071] Optionally, the separation conditions are determined based on the monomer type of the target ginsenoside component, specifically including: determining the partition coefficient K value of the target ginsenoside monomer in the candidate two-phase solvent system; selecting at least one two-phase solvent system and its volume ratio based on whether the partition coefficient K value falls within a preset effective separation range; determining whether a gradient elution mode needs to be adopted based on the difference in K values ​​of each monomer in the target saponin combination; if gradient elution is adopted, setting the gradient and rate of change of the composition of the mobile phase during the elution process based on the distribution behavior change curve of each monomer in the stationary phase and the mobile phase.

[0072] In this embodiment of the invention, 2-5 candidate two-phase solvent systems are selected from known literature or experience based on the polarity range of the target saponin combination. For example, for moderately polar ginsenosides, the n-hexane-ethyl acetate-water system is preferred; if this system is not suitable, the chloroform-methanol-water or ethyl acetate-n-butanol-water system can be tried.

[0073] Mix the three solvents according to a preset volume ratio (e.g., 1:1:1), shake thoroughly, and allow to stand to separate the upper and lower phases. Record the volumes of the upper and lower phases separately, and determine the densities of the two phases if necessary.

[0074] For each candidate system, the partition coefficient K value of each monomer in the target saponin combination was determined. The determination method used was the test tube partition method. Take a certain volume of the upper and lower phases of the system (e.g., 5 mL each) and place them in a stoppered test tube; add a known amount of saponin monomer (approximately 1 mg), seal tightly, and shake in a constant temperature water bath (usually 25 ± 1 °C) for at least 30 minutes to allow the distribution to reach equilibrium; after standing and separating the layers, separately pipette the upper and lower phase solutions and determine the saponin concentration using high performance liquid chromatography or ultraviolet spectrophotometry; if the upper phase is the stationary phase and the lower phase is the mobile phase, then If the lower phase is the stationary phase and the upper phase is the mobile phase, then For each candidate system, a list of K values ​​for all target saponins is recorded. For example, in a hexane / ethyl acetate / water (1:1:1) system, the K value of ginsenoside Rg1 was approximately 0.8 (with the upper phase as the stationary phase and the lower phase as the mobile phase), the K value of Rb1 was approximately 1.5, and the K value of Re was approximately 0.6, as determined by the tube partition method.

[0075] If the K values ​​of all target saponins in a candidate system are between 0.2 and 5.0 (approximate range), and the ratio of K values ​​of adjacent saponins is greater than 1.2, then the system is retained for the next screening step; otherwise, it is excluded.

[0076] Before formal screening, the effective separation range of the selected HSCCC device at typical rotational speeds (e.g., 800 rpm) and flow rates (e.g., 1.0 mL / min) is established experimentally. Specifically, a set of standards with known K values ​​(e.g., dyes or known saponins) are selected, eluted under isocratic conditions, and their resolution is measured; the resolution is plotted. Contour plot showing the variation of K value and the ratio of adjacent K values; take The range of K values ​​and adjacent ratios corresponding to the baseline separation is considered the effective separation interval. This interval is device-dependent, but once established, it can be reused and therefore stored locally. For example, under conditions of 800 rpm rotation speed and 1.0 mL / min flow rate, experiments have shown that when the K value is between 0.5 and 2.0 and the ratio of adjacent K values ​​is ≥1.5, the resolution can reach above 1.5.

[0077] For each candidate two-phase solvent system retained, check whether the K value of each target saponin monomer falls within the effective separation range (e.g., 0.5-2.0); sort the K values ​​of each monomer from smallest to largest, and calculate the K value ratio of adjacent saponins in turn. Check all Are both of these conditions greater than or equal to the minimum ratio threshold required for an effective separation interval (e.g., 1.5)? If both conditions are met, the system and its current volume ratio are deemed valid, and the corresponding list of K values ​​and ratios are recorded.

[0078] If multiple effective systems exist, select the system that maximizes the minimum ratio of all adjacent K values ​​(i.e., maximizes the separation margin); if the ratios are similar, select the system with higher stationary phase retention or simpler operation. The final output should include at least one two-phase solvent system and its volume ratio (e.g., n-hexane / ethyl acetate / water = 1:1:1, with the upper phase as the stationary phase and the lower phase as the mobile phase).

[0079] If, at the initial volume ratio, the K value is close to but not entirely within the effective range, the K value can be re-measured by fine-tuning the solvent ratio (e.g., changing the ratio of n-hexane:ethyl acetate:water from 1:1:1 to 1:2:1 or 2:1:1) until the conditions are met. This fine-tuning process can be performed iteratively.

[0080] Using the list of K values ​​for each target saponin in the final two-phase solvent system selected from the screening, the ratio of all adjacent K values ​​was calculated. minimum value ; or the difference between the maximum and minimum K values, ΔK; or the coefficient of variation (CV) of the K values.

[0081] A pre-defined acceptable threshold for isocratic elution is defined as the level at which isocratic elution, under selected rotational speed and flow rate conditions, enables all adjacent target saponin peaks to achieve the target separation (e.g., ...). The minimum allowed ratio of adjacent K values ​​(denoted as ) The acceptable threshold for isocratic elution can be determined experimentally. Under isocratic conditions, a mixture of standard substances with varying K-value ratios is used to measure the change in resolution with respect to the ratio, and the critical ratio at which the resolution equals 1.5 is found.

[0082] like If all adjacent saponins can be baseline separated under isocratic conditions, then isocratic elution mode is adopted. If at least one pair of adjacent saponins cannot achieve baseline separation under isocratic conditions, a gradient elution mode can be used to improve separation. For example, if the target saponins are Rg1 (K=0.8) and Rb1 (K=1.5), and the ratio is 1.875 > 1.5, isocratic elution can be used; if the target saponins are Rg1 (K=0.8), Rc (K=1.0), and Rb1 (K=1.5), and the adjacent ratios are 1.25 and 1.5 respectively, and the minimum ratio is 1.25 < 1.5, then gradient elution is required.

[0083] In one possible embodiment, an absolute K value range can also be considered. If the maximum K value is greater than a certain upper limit (e.g., 5), gradient elution is recommended to shorten the elution time, even if the ratio meets the requirements.

[0084] The output decision result is either isocratic elution or gradient elution.

[0085] If gradient elution is used, the gradient and rate of change of the composition of the mobile phase during the elution process are set according to the distribution behavior curves of each monomer in the stationary and mobile phases.

[0086] For decisions using gradient elution, the K-value of each target saponin monomer is determined as a function of the mobile phase composition for the selected two-phase solvent system. Specifically, the overall solvent composition of the two-phase solvent system is fixed, but the volume fraction of the strong eluting solvent in the mobile phase is varied. For example, if the system is hexane / ethyl acetate / water, and the lower phase is the mobile phase, the ratio of ethyl acetate to water in the lower phase can be changed (or the mixing ratio of the upper and lower phases can be changed). Five to ten different mobile phase composition points (from the initial composition to the final composition) are selected, and the K-value of each saponin is determined using the tube partitioning method at each composition. A curve is plotted for each saponin with the mobile phase composition as the x-axis (e.g., the volume fraction of the strong eluting solvent) and the K-value as the y-axis. These curves are collectively referred to as partition behavior change curves.

[0087] Determine the start and end points of the gradient elution. Start point: Select a mobile phase composition that ensures the K value of the monomer with the smallest K value among all saponins is not lower than a certain lower limit (e.g., 0.2). Typically, the start point ensures that the most easily eluted saponins are still partially retained, avoiding overlap with the solvent front. End point: Select a mobile phase composition that ensures the K value of the monomer with the largest K value among all saponins is not higher than a certain upper limit (e.g., 2.0). Typically, the end point ensures that the most difficult-to-elute saponins are eluted within a reasonable time.

[0088] In one possible embodiment, the behavior curves of each individual unit can be superimposed to observe the spacing between adjacent curves. The starting point should be selected in the component region that maximizes the sum of the spacing between all adjacent curves; the ending point should be selected in the component region where the spacing between all adjacent curves no longer increases significantly.

[0089] The entire elution process is divided into several stages based on the slope of the behavior curves (i.e., dK / d(mobile phase composition)). The slope of each curve at each mobile phase composition is calculated. For any mobile phase composition interval, the maximum absolute value of the slopes of all curves is taken as the characteristic slope of that interval. Consecutive intervals with similar characteristic slopes are merged into one stage. Within each stage, the rate of change of the mobile phase composition is proportional to the characteristic slope of that stage: the larger the slope (the more sensitive K is to compositional changes), the lower the rate of change should be to avoid loss of separation; the smaller the slope, the higher the rate of change can be to shorten the elution time.

[0090] The proportionality coefficient is determined through prior experiments, and the rate of change is increased as much as possible while ensuring that the worst separation is not lower than the target value (e.g., 1.5).

[0091] In regions where the K-value curves of any two adjacent monomers intersect (i.e., the two lines intersect) or approach each other (with a spacing less than a preset spacing threshold), the risk of separation is highest. In such cases, the rate of change of the mobile phase composition should be significantly reduced, or even the change should be paused (maintained at an isocratic level for a period of time) to maintain sufficient separation.

[0092] Set a separation window near the intersection point, set the rate of change within the window to 10%~30% of the normal rate, and keep the window width above the safe value until the distance between the two curves is restored.

[0093] In regions where there are no intersections and the spacing between all adjacent curves is greater than the safety threshold, the rate of change of the mobile phase composition can be appropriately increased to accelerate elution. The increase should be limited to prevent the spacing between the closest adjacent curves from falling below the safety threshold.

[0094] The initial gradient and rate of change are input into the HSCCC system for simulation or actual pre-separation. Feedback adjustments are made based on the obtained chromatogram (peak retention time, resolution, peak width). If the resolution of a pair of adjacent saponins is insufficient, the rate of change in that region is reduced or the gradient time is extended. If the total elution time is too long, the rate of change is increased in the region with sufficient resolution.

[0095] After several iterations, the final gradient function is obtained. and rate of change curve This curve represents the gradient and rate of change of the mobile phase composition during the elution process.

[0096] The output gradient elution program includes the initial mobile phase composition, the final mobile phase composition, the gradient shape (linear, step, convex / concave curve), the duration and rate of change of each stage. For example, for the above combination of Rg1, Rc, and Rb1, the distribution behavior curves show that as the upper phase volume fraction in the mobile phase increases from 30% to 50%, the K value of Rg1 decreases from 1.2 to 0.5, Rc from 2.0 to 0.9, and Rb1 from 3.5 to 1.6. A steeper slope is set in the 30%-40% range, resulting in a lower rate of change (e.g., 0.5% / min); a gentler slope is set in the 40%-50% range, resulting in a higher rate of change (e.g., 2% / min). The final gradient program is obtained after one iteration of optimization.

[0097] Optionally, the step of selecting at least one set of two-phase solvent systems and their volume ratios based on whether the partition coefficient K value falls within a preset effective separation range specifically includes: establishing a mathematical model between resolution and partition coefficient K value based on the baseline separation requirements of the target ginsenoside monomer in high-speed countercurrent chromatography; determining the partition coefficient K value of the target ginsenoside monomer in the candidate two-phase solvent system and calculating the K value ratio between adjacent ginsenoside components; determining that the partition coefficient K value falls within the effective separation range when the K value ratio of adjacent components is greater than or equal to the minimum ratio threshold required to meet baseline separation; determining the lower limit of the K value ratio corresponding to the resolution of adjacent chromatographic peaks reaching a preset value under selected rotation speed and flow rate conditions; if the target ginsenoside combination contains three or more monomers, then taking the minimum value of the K value ratio of all adjacent components, and only when the minimum value is not lower than the threshold value, the corresponding solvent system is selected as an effective system.

[0098] In this embodiment of the invention, a preset resolution value is determined for this separation task. Typically, the preparation of pharmaceutical-grade ginsenoside monomers requires that adjacent impurity peaks achieve baseline separation from the target peak, i.e., the resolution... Users can also set a higher separation degree (e.g., 2.0) or a more relaxed degree (e.g., 1.2) according to specific purity requirements. This preset value is denoted as... .

[0099] Before establishing a mathematical model, the rotational speed and flow rate of the high-speed countercurrent chromatography apparatus need to be pre-selected. These conditions can be determined based on a combination of the apparatus's mechanical properties, stationary phase retention rate, and separation efficiency. For example, preliminary experiments can be conducted to select the highest rotational speed that ensures a stationary phase retention rate greater than 50%, and the maximum flow rate at which the column pressure does not exceed the allowable value. Once selected, these rotational speed and flow rate will serve as fixed conditions for all subsequent experiments and calculations.

[0100] One possible implementation, using an experimental empirical model, is as follows: A set of standard substances with known K values ​​(e.g., known saponins) were selected and prepared into mixture pairs with different K value ratios. Isocratic HSCCC separation was performed under selected rotational speed and flow rate conditions, and the resolution of each pair was measured. .by As the independent variable, Using the dependent variable, perform curve fitting (e.g., power function, logarithmic function, or quadratic polynomial). This yields an empirical formula. This formula is the required mathematical model.

[0101] In one possible implementation, a theoretical semi-empirical model is used as follows: According to the retention time formula of HSCCC Combined with theoretical plate number Export the separation expression: ; in This is the volume ratio of the stationary phase to the mobile phase (which can be determined experimentally). Under the selected rotational speed and flow rate conditions... and Substitute in, and you will get and and The relationship. For the target saponin combination, It is usually between 0.5 and 2.0.

[0102] The model's predictive accuracy was verified using at least two sets of standard substances not involved in the fitting. If the prediction error exceeded an acceptable range (e.g., relative error > 10%), additional experimental points were added and the model was refitted.

[0103] For each candidate two-phase solvent system (different types or different volume ratios), the partition coefficient K value of each monomer in the target saponin combination is determined using the tube partition method or the analytical HSCCC method. During the determination, it is necessary to determine which phase is used as the stationary phase and which is used as the mobile phase. Record the K values ​​of all monomers.

[0104] Arrange the K values ​​of each monomer measured in the same system in ascending order to obtain an ordered sequence. ,in The number of target saponin monomers.

[0105] For each adjacent pair, calculate the ratio: ; Record all data in this solvent system. value When the ratio of the K values ​​of adjacent components is greater than or equal to the minimum ratio threshold required to meet baseline separation, the allocation coefficient K value is determined to fall into the effective separation range. The preset separation value corresponding to the baseline separation requirement. Substitute into the model and solve for the required solution. Minimum value. That is, solve the equation: ; get .Should This refers to the minimum K-value ratio threshold required to achieve baseline separation under selected rotational speed and flow rate conditions. For the current candidate solvent system, if... If the K values ​​of the two components in the adjacent pair fall within the effective separation range, then the two components can achieve baseline separation in the current system. If the adjacent pair does not meet the baseline separation requirement, the corresponding K value distribution does not fall within the effective separation interval.

[0106] When all adjacent pairs satisfy When the K-value distribution of the entire target saponin combination falls within the effective separation range, the solvent system is marked as a valid candidate. If any one of these K-values ​​exists... If the solvent system is invalid, it needs to be adjusted (by changing the type or volume ratio of the solvent) or replaced with another candidate system.

[0107] Minimum ratio threshold Under selected rotational speed and flow rate conditions, the critical ratio corresponding to achieving a preset resolution (e.g., 1.5) is determined. At the selected rotational speed and flow rate, a series of standard substance pairs with K-value ratios gradually approaching 1 are prepared, and their resolutions are measured. The ratio at which the resolution exactly equals the preset value is the lower limit of the ratio.

[0108] It should be noted that if the rotation speed or flow rate is changed, the lower limit of the ratio needs to be redefined, because these two parameters affect the theoretical plate number and the retention rate of the stationary phase, thereby changing the separation capacity.

[0109] If the target saponin combination contains three or more monomers, the minimum value of the ratio of K values ​​of all adjacent components is taken. Only when the minimum value is not lower than the threshold value is the corresponding solvent system screened as an effective system. Calculating the minimum ratio involves three or more monomers, after calculating all... Then, find the minimum value among them: ; Will and The minimum ratio threshold required for baseline separation is compared, if... If the solvent system is deemed effective, all adjacent pairs meet or exceed baseline separation requirements, and therefore the entire system can be separated. If the solvent system is invalid, it is determined that at least one pair of adjacent components (i.e., the pair corresponding to the minimum ratio) cannot achieve baseline separation.

[0110] For all candidate solvent systems, only those that satisfy the following conditions are retained. The system. If multiple effective systems exist, further adjustments can be made based on... Choose the optimal system based on the size of the solvent (the larger the size, the greater the separation margin) or the retention rate of the stationary phase. Output at least one set of two-phase solvent systems that meet the conditions and their volume ratios.

[0111] In one possible embodiment, if no candidate system meets the conditions, it is necessary to expand the range of candidate systems (e.g., try different solvents or adjust the volume ratio), or adjust the rotation speed and flow rate conditions (e.g., increase the rotation speed to increase the theoretical plate number, thereby reducing the impact on the theoretical plate number). (According to the requirements), and then repeat the screening process until an effective system is selected.

[0112] By screening out the two-phase solvent system and its volume ratio, the accuracy of high-speed countercurrent chromatography separation can be improved, and ginsenoside monomers with higher purity can be obtained.

[0113] Optionally, determining whether a gradient elution mode needs to be used based on the difference in K values ​​among the monomers in the target saponin combination specifically includes: calculating the partition coefficient K values ​​of all monomers in the target saponin combination in a selected two-phase solvent system; arranging the K values ​​of each monomer in ascending order, and sequentially calculating the K value ratio or K value difference between adjacent monomers; comparing each calculated adjacent ratio or difference with a preset acceptable threshold for isocratic elution; if all adjacent ratios or differences are not lower than the acceptable threshold for isocratic elution, then isocratic elution mode is determined to be used; if any set of adjacent ratios or differences is lower than the acceptable threshold for isocratic elution, then gradient elution mode needs to be used; the acceptable threshold for isocratic elution is the minimum K value ratio or minimum K value difference allowed to achieve the target resolution by experimental determination under set rotation speed and flow rate conditions.

[0114] In this embodiment of the invention, one or more two-phase solvent systems and their volume ratios that can cause the target saponin K value to fall within the effective separation range are determined. If multiple effective systems exist, the one with the largest separation margin or the simplest operation can be selected as the selected system for this step.

[0115] For the selected system, the partition coefficient K value of each monomer in the target saponin combination is determined using the tube partition method or the analytical HSCCC method. During the determination, the stationary phase and mobile phase must be clearly defined and maintained in the same phase partitioning as in subsequent separation operations. Record the K value corresponding to each saponin monomer and indicate the determination conditions (temperature, solvent batch, etc.).

[0116] If the K value of the same saponin monomer varies between batches (e.g., within ±10%) under the same system, the average value can be taken. If the fluctuation is too large, the accuracy of the solvent preparation or the determination method should be checked, and the determination should be repeated if necessary.

[0117] Arrange the K values ​​of all the target saponin monomers in ascending order to obtain an ordered sequence: ; in The number of target saponin monomers ( If there are cases where K values ​​are equal (different saponins have the same partition coefficient), then the order of adjacent components during sorting can be arbitrary.

[0118] Depending on the order of magnitude and distribution range of the K values, choose either the K-value ratio or the K-value difference as the measure of dissociation. If all K values ​​are between 0.2 and 5.0, the ratio is preferred because it has a more direct relationship with separation. If the K value range is wide (e.g., minimum 0.1, maximum 10), or if there are extremely large (>5) or extremely small (<0.2) K values, the ratio may be distorted; in this case, the difference can be used. Alternatively, both the ratio and the difference can be calculated simultaneously and compared with their corresponding thresholds, as long as one of the indicators meets the isocratic elution requirement (but the ratio is usually the primary criterion).

[0119] Calculate the adjacent ratios. If a ratio is selected, for... ,calculate: ; Calculate the difference between adjacent values. If the difference is selected, for... ,calculate: ; Record all or The value of .

[0120] The acceptable threshold for isocratic elution can be determined experimentally beforehand. Under selected rotational speed and flow rate conditions (same as the subsequent actual separation operation), prepare a series of standard substance pairs with known K-value ratios (or substance pairs with known K-value differences). Perform isocratic HSCCC separation separately and determine the resolution of each pair. Find the resolution that exactly reaches the target resolution (e.g., ...). The ratio of the minimum K values ​​when () is denoted as ; or the difference of the minimum K value, denoted as If the rotational speed or flow rate changes, the threshold needs to be re-measured.

[0121] If the ratio is used for determination: compare each... and The size of the value. If the difference is used for determination: compare each value sequentially. and The size. Record the comparison results: each or Whether it is valid or not.

[0122] If all adjacent ratios or differences are not lower than the acceptable threshold for isocratic elution, then isocratic elution mode is used; if any set of adjacent ratios or differences is lower than the acceptable threshold for isocratic elution, then gradient elution mode is required. If all (or All are full (or If the separation is 0, it means that under isocratic elution conditions, each pair of adjacent saponins can achieve or exceed the target separation. In this case, isocratic elution mode is selected. If at least one... Make (or If the result is zero, it indicates that at least one pair of adjacent saponins cannot achieve the target separation degree under isocratic conditions. Isocratic elution will lead to overlap or insufficient separation of this pair of saponins. In this case, it is determined that a gradient elution mode should be used.

[0123] If all All greater than However, if the maximum K value exceeds a certain upper limit (e.g., >5), the total time for isocratic elution may be too long. In this case, even if the ratio meets the requirements, gradient elution can be selected based on the actual time cost.

[0124] If there are multiple candidate solvent systems, the decision results for each system can be calculated separately. The system that can be eluted with isocratic elution should be selected first, because isocratic operation is simpler.

[0125] Output isocratic elution or gradient elution.

[0126] In one possible embodiment, a set of chemically stable standard substances (e.g., acetophenone, phenylacetone, naphthalene, etc., or pure ginsenosides with known K values) are selected. Standard substance pairs with different K value ratios are prepared, ranging from close to 1 (e.g., 1.05) to larger ratios (e.g., 2.5), with at least 5-6 different ratios prepared.

[0127] Under the selected rotational speed and flow rate conditions, isocratic HSCCC separation was performed on each standard reference pair, and the retention time and peak width of the two peaks were recorded to calculate the resolution. .

[0128] by or As the x-axis, with Plot a scatter plot with the ordinate as the vertical axis and fit a curve (e.g., linear, logarithmic, or power function). Find the target separation (e.g., ...) on the fitted curve. The x-coordinate value corresponding to ) is For example, if the fitted equation is ,but Similarly, for the difference threshold, ... Experiments and fitting were performed on the x-axis to obtain... .

[0129] Based on the difference in K values ​​among the monomers in the target saponin combination, it can be determined whether a gradient elution mode is needed, which can improve the separation accuracy of different target saponin combinations.

[0130] Optionally, the step of setting the gradient and rate of change of mobile phase composition during elution based on the distribution behavior curves of each monomer in the stationary and mobile phases specifically includes: measuring the distribution coefficient K value of each monomer in the target saponin combination under different mobile phase composition ratios, and plotting the behavior curve of K value changing with mobile phase composition; superimposing the behavior curves of each monomer to identify the mobile phase composition interval that maximizes the difference in K value between adjacent monomers, which serves as the starting and ending points of gradient elution; dividing the entire elution process into several stages based on the slope of the behavior curves, with the rate of change of mobile phase composition in each stage being proportional to the slope of the monomer with the fastest K value change in that stage; reducing the rate of change of mobile phase composition and extending the elution time in regions where the K value curves of any two adjacent monomers intersect or approach each other to maintain sufficient separation; increasing the rate of change of mobile phase composition in regions where there is no intersection and the K value spacing of each monomer is uniform to shorten the total elution time; and determining the final gradient and rate of change curves through iterative optimization, with the goal of achieving the shortest elution time for all target monomers while meeting the preset separation requirements.

[0131] In this embodiment of the invention, the range of variation of the mobile phase composition is determined based on the selected two-phase solvent system. For example, if the system is n-hexane / ethyl acetate / water, and the lower phase is the mobile phase, then the mobile phase composition can be determined by the volume fraction of ethyl acetate in the lower phase (…). The range should be expressed as follows: (This should be used to indicate that all target saponins are strongly retained). Starting with a larger composition, until all saponins are rapidly eluted ( The composition ends at the smaller (smaller) fraction. Typically, the starting composition is chosen to be the one from which the saponins are most difficult to elute under isocratic conditions. The composition with a value approximately equal to 2-3; the termination composition is selected from the most easily eluted saponins. The composition when the value is approximately 0.2-0.5.

[0132] Within the range of variation, select 5-10 representative component points, whose distribution should cover the entire range, with increased density in areas where the expected curve slope changes drastically. For example, select... (Volume fraction).

[0133] For each selected mobile phase composition, prepare a corresponding two-phase solvent system (Note: Changing the mobile phase composition will also change the stationary phase composition; the proportions of the entire system should be adjusted accordingly). Determine the partition coefficient of each target saponin monomer under that composition using the tube partition method or the analytical HSCCC method. Record the composition of each saponin at each composition point. value.

[0134] Composed of mobile phase As the x-axis, with or Plot a smooth curve (obtainable through interpolation or fitting) for each saponin, with the ordinate as the ordinate. Preferably, use... Drawing, because and The relationship is often exponential, but the linearity is better after taking the logarithm.

[0135] The distribution behavior curves of all obtained saponins were plotted on the same coordinate system to obtain an overlay graph.

[0136] For each pair of adjacent saponins (in the order of elution, i.e.) (Values ​​in ascending order) on the x-axis Calculate at each point Or calculate directly The greater the difference, the easier it is to separate the two components under that composition.

[0137] For each pair of adjacent saponins, find the one that makes (or (reaching the maximum value) The interval. This interval is located where the separation between the two curves is widest.

[0138] Starting point Choose the option that results in the latest peak ( The largest) of saponins The value is still greater than a certain lower limit (e.g.) This refers to the components where the differences between all adjacent curves have not yet begun to narrow sharply. Or, more simply, select all... The components with relatively large values ​​(strong retention) and reasonable differences were used as the starting point.

[0139] End point : Select the option that produces the earliest peak ( The smallest) saponin, The value is less than a certain upper limit (e.g.) ), while the differences between all adjacent curves have not yet become too small.

[0140] In one possible implementation, the objective is to maximize the minimum difference between all adjacent curves. Select the start and end intervals on the axis. Specifically, calculate the minimum difference between all adjacent pairs as... The changing curve is selected from the range above a certain threshold, and its left and right boundaries are the starting point and the ending point, respectively.

[0141] The initial mobile phase composition of the output gradient elution Composed of the terminated flow phase .

[0142] exist Within the interval, calculate the value of each saponin. right derivative This derivative reflects the sensitivity of saponins to changes in the composition of the mobile phase.

[0143] In each Take all saponins from the source. The maximum value is taken as the characteristic slope of that point. The steeper the characteristic slope, the more likely there is at least one saponin in the vicinity of that composition. The value changes very drastically, and the rate of change needs to be carefully controlled.

[0144] Based on the characteristic slope Follow The changes divide the entire interval into several stages. Regions with gradual changes and similar values ​​are merged into one stage, and the dividing point can be chosen at... The location of the mutation.

[0145] Within each stage, the rate of change of the mobile phase composition is set. (Volume fraction / minute). Rate of change versus average characteristic slope of this phase. Inversely proportional (because the steeper the slope, the slower the change needs to be), or, in other words, the rate of change is negatively correlated with the characteristic slope. Specific proportionality coefficients... Preliminary experiments determined that, under the condition of a feature slope of 1 (unit), a rate of change that enables the separation to reach the target value should be selected. For other stages, (After normalization).

[0146] In one possible embodiment, the feature slope can be directly set to a larger slope, resulting in a lower rate of change, and a smaller slope, resulting in a higher rate of change. For example, the feature slope can be divided into three levels: high, medium, and low, corresponding to three different rates of change.

[0147] In one possible implementation, instead of dividing the process into stages, the rate of change can be allowed to continuously follow the... change: ,in This is a normalization constant. It can be generated through numerical integration. function.

[0148] In the superimposed behavioral curves, examine the curves of each pair of adjacent saponins to identify the presence of overlapping regions. Make Near this point, the two curves intersect. Although they do not intersect in the approaching region, The area, in which For a preset security threshold (e.g.) , corresponding to (Ratio < 1.22). Merging all intersections and proximity regions yields a set of high-risk intervals. .

[0149] Within the high-risk range, the rate of change is reduced to 10%–30% of the normal rate (or to a preset low rate, such as 0.1% / min). Simultaneously, the elution time in this range can be extended (i.e., maintaining a low-rate change for a longer period), allowing sufficient time for the two peaks to separate. The specific extension time can be determined through simulation or preliminary experiments.

[0150] If the intersection point happens to be near the start or end point, adjustments can be considered. or To avoid intersections.

[0151] By using the methods described above to reduce the rate of change, the negative impact of crossover can be minimized.

[0152] exist After excluding the identified high-risk areas, the remaining region is considered a safe zone. Within the safe zone, the distance between all adjacent curves is greater than the safety threshold, and there is no intersection.

[0153] Within the safe region, the rate of change can be further increased based on the established rate of change, as long as the worst-case separation remains above the target value. Experiments with different rates of change can be conducted beforehand within a representative component of the safe region to find the critical rate at which the separation begins to decrease. The critical rate is then multiplied by a safety factor (e.g., 0.8) to determine the maximum permissible rate for that region. If the safe region is wide, it can be further subdivided, using higher rates in larger, more spaced subregions.

[0154] The main purpose of increasing the rate of change is to shorten the total elution time, but it is essential to ensure that the low-rate time in the high-risk region is not excessively compressed. The overall optimization objective is to achieve the following while ensuring that the separation degree of all adjacent saponins is ≥ the target value: Minimize, that is, minimize the total time.

[0155] Construct a gradient program, including: initial components Termination of composition The rate of change (or continuous change function) at each stage (This includes) special handling for high-risk intervals. The gradient procedure is then converted into a time function. .

[0156] Perform an actual high-speed countercurrent chromatography separation using this gradient program (or simulate it using chromatographic simulation software), and record the retention time, resolution, and peak width for each target saponin. Check whether the resolution of all adjacent saponins is ≥ the preset target value (e.g., ...). ).

[0157] If the separation of a pair of adjacent saponins is insufficient, locate the region corresponding to that pair of saponins on the behavior curve, further reduce the rate of change in that region, or expand the low-rate range of that region, and then re-verify.

[0158] If all separations meet the requirements, but the total elution time is significantly longer than expected (e.g., more than 50% longer than the theoretical minimum time), try to appropriately increase the rate of change within the safe range, or reduce the width of the low-speed range during the merging phase in the region with sufficient separation.

[0159] If the separation meets the requirements and the total time is acceptable, proceed to the next step.

[0160] The separation degree of all adjacent saponins is within 1.0 to 1.2 times the target value (neither lower than the target value nor excessively higher, thus avoiding wasted time); the total elution time change between two consecutive iterations is less than 5%; the gradient curve is smooth without abrupt changes (or abrupt changes can be performed by the device).

[0161] Output a function showing the final determined composition of the mobile phase as a function of time. (The changing gradient) and its derivative (Rate change curve). This gradient program can be directly used for separation operations in preparative high-speed countercurrent chromatography.

[0162] The final gradient program is associated with and stored in conjunction with the target saponin combination, the selected two-phase solvent system, rotation speed, flow rate, and other conditions to form a standard operating parameter library, which can be directly called upon for subsequent batch production of the same target combination.

[0163] Optionally, the separation conditions are dynamically adjusted based on the stage parameters of the supercritical fluid extraction process, specifically including: predicting the polarity distribution range and impurity spectrum characteristics of the target saponin combination in the crude extract based on the stage parameters of the supercritical fluid extraction process; using the polarity distribution range and impurity spectrum characteristics as input constraints for screening the two-phase solvent system, and quantitatively adjusting the components of the two-phase solvent system separated by high-speed countercurrent chromatography through the input constraints; and dynamically adjusting the gradient of gradient elution based on the different polarity of organic auxiliaries or the temperature and pressure combinations at different stages during the supercritical fluid extraction process.

[0164] In this embodiment of the invention, the stage parameters of the extraction are read from the control system or operation log of the supercritical fluid extraction device. If staged extraction is used, the temperature, pressure, auxiliary agent ratio, duration, and crude extract yield of each stage are recorded. If staged extraction is not used, the parameters of a single stage are recorded.

[0165] Based on the temperature, pressure, and additive ratio during the extraction stage, the polarity range of the saponins mainly extracted in that stage can be inferred using solubility models or empirical rules. For example, pure CO2 at lower pressures mainly extracts non-polar components, adding 5% ethanol can extract moderately polar saponins, and adding 10% ethanol and a small amount of water can extract highly polar saponins.

[0166] If the known polarity parameters (logP values) of the monomers in the target saponin combination differ significantly, and a segmented extraction strategy is adopted, the crude extracts collected in each stage will have different polarity distributions: the crude extracts in the early stage are mainly low-polarity saponins, while those in the later stage are mainly high-polarity saponins.

[0167] Output the expected polarity range (e.g., minimum logP to maximum logP) of the target saponin combination in the crude extract.

[0168] The known polarity range of other saponins (such as non-target saponins) and fat-soluble components (such as fatty acids and sterols) in ginseng raw materials, excluding the target saponin; Under given extraction conditions, these impurities tend to co-dissolve with the target saponin (e.g., low-polarity impurities may appear together with low-polarity target saponins in the early stages). If historical data exists, a correlation model between extraction conditions and impurity profiles can be established. By inputting the parameters for this stage, the expected types and relative contents of impurities can be output.

[0169] Output a list of impurity spectral characteristics, including the possible types of impurities, their polarity range, and a predicted degree of similarity to the target saponin K value.

[0170] Based on the polarity range of the target saponin combination, select 1-3 initial candidate systems from a known library of two-phase solvent systems. For example, for moderately polar saponins, choose n-hexane / ethyl acetate / water (1:1:1); for highly polar saponins, choose ethyl acetate / n-butanol / water (or chloroform / methanol / water).

[0171] The obtained polarity distribution range and impurity spectrum characteristics are added as additional constraints to the solvent system screening process. Specific constraints include: Polarity coverage constraint: The selected system should ensure that the K values ​​of the target saponins all fall within the effective separation range (0.5-2.0), while the K values ​​of impurities at both ends of the polarity distribution range (the most polar and the least polar) are as far away from this range as possible. For example, the K values ​​of low-polarity impurities should be less than 0.2, and the K values ​​of high-polarity impurities should be greater than 5, so as to separate them from the target saponins in the early or late stages of elution.

[0172] Impurity exclusion constraints: For impurities that are predicted to have a K value close to that of the target saponin, the ratio of their K value to the K value of the nearest target saponin in the selected system must be less than 0.8 or greater than 1.25 (i.e., far from the K value of the target saponin).

[0173] Additive compatibility constraints: If the crude extract contains certain organic additives (such as ethanol), solvent components that are immiscible with the additive or that may cause emulsification of the two-phase system should be avoided.

[0174] For each initial candidate system, quantitative adjustments are made based on input constraints, rather than testing only a few fixed ratios. The adjustable parameters are determined, typically the volume ratios of the solvents in the two-phase system. For example, for a hexane / ethyl acetate / water system, the adjustable parameters are the ratio of hexane to ethyl acetate (maintaining overall volume balance between water and organic phases) or the ratio of ethyl acetate to water. An adjustment step size is set, for example, adjusting the ratio by 0.05 (volume fraction) each time. For each adjusted ratio, the K values ​​for the target saponin and key impurities are determined or estimated (either via rapid tube partitioning or using a previously established K-value prediction model).

[0175] Determine if all constraints are met. Iterate and adjust until at least one solvent system and its volume ratio that satisfy all constraints are found. If the constraints cannot be met within a certain range (e.g., n-hexane:ethyl acetate:water = 0.5-2:0.5-2:1), then replace the initial candidate system.

[0176] Output the composition of the two-phase solvent system after quantitative adjustment (i.e., the specific volume ratio of each solvent), as well as the corresponding stationary phase and mobile phase specifications.

[0177] After determining the solvent system, further examine whether the types of organic auxiliaries and the combination of temperature and pressure used in the supercritical fluid extraction process differ significantly from historical standard conditions. If the extraction conditions are consistent with the standard crude extract conditions used when establishing the gradient elution program, the existing gradient program can be used directly; if there are differences (e.g., different auxiliaries were used or different fractionation combinations were adopted), the gradient needs to be dynamically adjusted.

[0178] Residual organic additives in the crude extract can alter the actual partitioning behavior of the HSCCC two-phase system. For example, if ethanol was used during extraction, residual ethanol in the crude extract (even after rotary evaporation, trace amounts may remain) increases the polarity of the aqueous phase, resulting in a slightly lower K value for the target saponin compared to that measured in the pure solvent system. To compensate for this effect, the initial mobile phase composition of the gradient elution can be shifted towards a stronger retention capacity (i.e., reducing the elution intensity of the mobile phase), for example, by reducing the volume fraction of the strong eluting solvent in the initial composition by a predetermined offset (this offset is related to the concentration of residual additives and can be calibrated through pre-experiments). Alternatively, while keeping the total gradient program time constant, the entire gradient curve can be shifted backward (i.e., extending the initial isocratic phase).

[0179] The adjustment amount can be determined by adding the same concentration of residual adjuvant as the crude extract to a blank solvent system, re-measuring the change in the K value of the target saponin, and then calculating the correction value of the gradient initial composition.

[0180] If supercritical fluid extraction employs fractional extraction, the composition ratio of target saponins in the crude extracts from different stages (if collected and injected separately) or in the combined crude extracts will differ from that in the crude extracts under single extraction conditions. For example, low-polarity saponins may be relatively enriched in the first-stage crude extract, while high-polarity saponins may be relatively enriched in the second-stage crude extract.

[0181] For separate injections, a gradient program can be set for each stage of the crude extract. For samples enriched with low polarity, the gradient starting composition can be more non-polar (i.e., lower elution intensity), and the total gradient time can be shortened; for samples enriched with high polarity, the starting composition should be more polar (higher elution intensity).

[0182] For cases involving combined injections, the overall polarity distribution of the target saponin combination can be recalculated based on the mass ratio of the crude extracts at each stage. If the overall polarity distribution is more biased towards lower polarity than the standard case, the gradient starting composition should have its elution intensity appropriately reduced, while the time in the low rate of change region should be extended to ensure the separation between low-polarity components; conversely, the opposite should be done.

[0183] Dynamic adjustment can be performed automatically before each separation. The obtained stage parameters are input into a pre-trained adjustment model (e.g., based on response surface methodology or machine learning), and the model outputs the corrected gradient program parameters (initial composition, termination composition, and rate of change at each stage). Alternatively, a rule base approach can be used, where several gradient program templates corresponding to typical extraction conditions are pre-defined. The closest template is selected based on the degree of matching of the current conditions, and linear interpolation is performed for adjustment.

[0184] When performing the first separation according to the dynamically adjusted gradient procedure, it is recommended to perform a rapid analytical HSCCC pre-separation (small injection volume) to check whether the retention time and resolution of the target saponin are consistent with expectations. If deviations occur, fine-tune the parameters (e.g., shift the overall retention time or adjust the rate of change locally) and save the fine-tuned parameters to the database for further optimization of the model.

[0185] The output is the final gradient elution program determined after dynamic adjustment, including the function of the change of mobile phase composition over time (the changing gradient) and the rate of change at each stage.

[0186] By quantitatively adjusting the composition of the two-phase solvent system separated by high-speed countercurrent chromatography and dynamically adjusting the gradient of gradient elution, the separation process can correspond to the extraction process, thereby improving the accuracy of separation in a targeted manner.

[0187] Optionally, the high-speed countercurrent chromatography separation using a high-speed countercurrent chromatography device to separate the liquid to be separated includes: during the separation process, real-time acquisition of the post-column effluent and determination of the actual partition coefficient K value of the target saponin monomer in the current two-phase solvent system; comparison of the actual measured K value with a preset effective separation range; if the actual K value deviates from the effective separation range, automatic adjustment of the volume ratio of each solvent in the two-phase solvent system according to a preset step size; rebalancing the chromatographic system after adjustment and measuring the K value again until the K value falls into the effective separation range; if the K value still cannot fall into the effective separation range after multiple adjustments, the current solvent system is determined to be unsuitable, and the system is automatically switched to an alternative two-phase solvent system.

[0188] In this embodiment of the invention, a rapid analytical pre-separation (small injection volume) is performed before the formal separation begins to assess the suitability of the current solvent system.

[0189] During the formal separation process, standards are injected periodically (e.g., after each batch is separated or after each fixed run time) or the K value is determined using the target saponin peak that has been separated.

[0190] Measurements are triggered when the detector signal becomes abnormal (such as peak distortion or retention time drift).

[0191] First, the dead time was measured. Inject a substance that is not retained at all in the selected two-phase system (such as glucose or tartrazine, depending on the solvent system) and record its elution time.

[0192] Then inject the standard of the target saponin (or use the known peak of the target saponin in the crude extract) and record its retention time. .

[0193] Determining the volume of the stationary phase and mobile phase volume The concentration can be determined by stopping the pump at equilibrium, emptying the liquid from the column, and weighing the liquid, or by calculating the flow difference. For simplicity, a pre-calibrated stationary phase retention rate can also be used. (The proportion of the stationary phase volume to the total column volume), then .

[0194] Calculate the actual K value: ; The formula is accurate in isocratic elution mode; in gradient elution mode, the K value under the equivalent composition at the midpoint of the gradient can be approximated, or a more complex integration method can be used. For ease of real-time adjustment, it is generally recommended to perform the measurement under isocratic conditions (the gradient can be paused and switched to isocratic mode).

[0195] The measured actual K value is stored together with information such as the current solvent system's identifier, timestamp, and temperature for subsequent analysis and model optimization.

[0196] Retrieve the pre-stored effective separation interval from the database under the current target saponin combination, current rotation speed, and flow rate conditions. This interval includes: Allowable range of a single K value For example, 0.5 to 2.0); if multiple target saponins are involved, the minimum allowable value of the ratio of adjacent K values ​​also needs to be stored. .

[0197] If the measured K value is in If the K value falls within the effective separation range, and (for multiple target saponins) the adjacent ratio condition is also met, then the K value is considered to fall within the effective separation range, and no adjustment is needed. If the K value is less than... If the K value is too low, it is determined that the value is too weak (too weak to retain). If the K value is greater than... If so, the K value is determined to be too high (too strong to retain).

[0198] If multiple target saponins exist, it is also necessary to check whether their K values ​​are ordered in the same order and whether the adjacent ratio is less than 1. If the ratio is too small, even if a single K value is within the range, it will be considered a deviation.

[0199] If the K value is too low (retention is too weak), it indicates that the target saponin is distributed excessively in the mobile phase, requiring an increase in the retention capacity of the stationary phase, i.e., a decrease in the elution strength of the mobile phase. This can be achieved by reducing the volume fraction of strong eluting solvents in the mobile phase or increasing the volume fraction of weak eluting solvents. For example, in a hexane / ethyl acetate / water system (with the lower phase as the mobile phase), a low K value can be addressed by reducing the proportion of ethyl acetate in the lower phase (i.e., increasing the proportion of water).

[0200] If the K value is too high (retention is too strong), then make the opposite adjustment: increase the volume fraction of the strong elution solvent in the mobile phase.

[0201] If the adjacent ratio is too small (i.e. the K values ​​of the two saponins are too close), it is necessary to change the selectivity of the solvent system. Usually, the type or ratio of medium polar solvent is adjusted, rather than simply the strength of the strong elution solvent.

[0202] For example, in a hexane / ethyl acetate / water system, adjusting the ratio of hexane to ethyl acetate can alter the selectivity for saponins.

[0203] The step size can be a fixed value (e.g., adjusting the volume ratio by 0.05 each time) or it can be adaptively adjusted according to the degree of deviation; the greater the deviation, the larger the step size. The step size should be pre-calibrated experimentally: under typical conditions, the change in K value caused by a unit step size (sensitivity). To avoid over-adjustment, the step size is usually set so that the expected change in K value is 30% to 50% of the current deviation.

[0204] The control system sends commands to the solvent delivery pump, changing the flow rate ratio of each solvent in preset steps. For example, if the original ratio is hexane:ethyl acetate:water = 1:1:1 (volume ratio), and it is necessary to reduce the elution intensity of the lower phase, the ratio can be adjusted to 1:0.9:1.1 (maintaining overall volume balance between the organic and aqueous phases, or directly changing the composition of the mobile phase). After adjustment, the new mixed solvent immediately enters the column system.

[0205] After adjustment, maintain the rotational speed and continue flushing the column system with the newly composed mobile phase until the detector baseline stabilizes. The equilibration time is typically 3-5 times the mobile phase transit time of the column volume. During this period, changes in stationary phase retention can be monitored (by observing column pressure or effluent volume).

[0206] After equilibration, repeat the K-value determination method (either by re-injecting the standard or using an existing peak) to obtain a new actual K-value. Compare the newly measured K-value with the effective separation range. If it falls within the effective range, stop adjustment, record the current solvent system composition, and continue or begin formal separation. If it still deviates, adjust again according to the new direction of deviation (either by maintaining the same step size or adjusting the step size), then reequilibrate and measure. Set a maximum number of adjustments (e.g., 5 times) to prevent infinite looping.

[0207] The current solvent system is deemed unsuitable when any of the following conditions are met: the number of adjustments reaches the preset maximum number (e.g., 5 times), and the last measured K value still deviates from the effective range; after two consecutive adjustments, the direction of change of the K value is opposite to the expected direction (indicating abnormal system response); or the stationary phase retention rate drops to an unacceptable level (e.g., below 20%) during the adjustment process.

[0208] Select the next system from a pre-stored list of alternative solvent systems. Alternative systems should have significantly different polarity or selectivity from the current system. The selection order can be: prioritizing systems with the second-best separation performance in previous screenings based on pre-assessed separation potential; or recommending alternative systems with corresponding polarity based on the characteristics of the current deviation (e.g., consistently low K-values ​​indicating a need for a weaker polarity system).

[0209] Stop the current separation operation and drain the existing solvent from the column. Flush the column system with the two-phase solvent of the new alternative system to re-establish the stationary phase. Re-determine the dead time and K value of the target saponin to confirm that it falls within the effective range. Record the switchover log and automatically update the process parameters.

[0210] If all alternative systems have been tried and still fail to meet the requirements, the system will issue an alarm, prompting the operator to check the raw materials, solvents, or equipment status, and will suspend automatic operation.

[0211] Optionally, the dissolving solvent is determined based on the residual components in the crude ginsenoside extract, including: detecting the types and contents of residual organic auxiliaries in the crude ginsenoside extract; selecting a dissolving solvent based on the types and contents of residual organic auxiliaries; if a segmented extraction strategy is adopted, dissolving the crude ginsenoside extracts collected at each stage separately, and selecting different dissolving solvent systems for the crude extracts at different stages; determining the total amount and polarity distribution of each target saponin monomer in the crude ginsenoside extract, and determining the volume and composition ratio of the dissolving solvent; if the residual carbon dioxide in the crude extract exceeds a preset threshold, degassing is performed before dissolution, and the polarity of the dissolving solvent is adjusted.

[0212] In this embodiment of the invention, a representative sample (e.g., approximately 0.1-1 gram, depending on the total yield) is taken from the collected crude ginsenoside extract. If a fractional extraction strategy is employed, samples are taken from the crude extract at each stage.

[0213] Qualitative and quantitative analysis of the samples was performed using gas chromatography (GC) or gas chromatography-mass spectrometry (GC-MS).

[0214] The crude extract sample is dissolved in a volatile organic solvent (such as acetone or diethyl ether), and insoluble matter is removed by filtration. The sample is then injected into a gas chromatograph using a suitable column (such as a polyethylene glycol column or a nonpolar column) and detector (flame ionization detector FID or mass spectrometer). The types of organic auxiliaries are identified by comparing their retention times with those of standards and their mass spectra. The content of each auxiliary is calculated (usually expressed as a mass fraction) using the external standard method or the internal standard method.

[0215] Record the test results in the batch log, including the name of the additive, its concentration, and the testing conditions. This information will be used for subsequent selection of the dissolving solvent.

[0216] Different residual auxiliaries have different effects on the solubility of saponins in crude extracts: If the residual auxiliaries are ethanol or methanol (highly polar), the crude extract may already have some hydrophilicity, and water or a water-alcohol mixture should be preferred for dissolution. If the residual auxiliaries are ethyl acetate or acetone (moderately polar), the crude extract may be more soluble in moderately polar organic solvents (such as n-butanol or ethyl acetate), and direct dissolution with water may cause saponins to precipitate. If the residual auxiliaries are present in high concentrations (e.g., >2%), solvents miscible with the residual auxiliaries should be used as much as possible to avoid layering or precipitation.

[0217] Based on the type and content of residual adjuvants, and considering the polarity of the target saponin, solvents that are the same as or similar to the residual adjuvants should be preferentially selected as the main component of the dissolving solvent to enhance compatibility. For example, if the residual adjuvant is ethanol, an ethanol-water mixture can be used as the dissolving solvent. If the residual adjuvant is immiscible with one phase of the subsequent HSCCC two-phase solvent system, another solvent that can simultaneously dissolve the crude extract and is compatible with the HSCCC system should be selected. If the content of the residual adjuvant is extremely low (below 0.1%), its effect on dissolution behavior is negligible. In this case, a general-purpose dissolving solvent (such as methanol-water or ethanol-water) should be selected based on the polarity of the target saponin.

[0218] Output the selected solvent type (e.g., "95% ethanol - 5% water" or "methanol:water = 1:1").

[0219] Check the records to see if a fractional extraction strategy was enabled. If multiple stages exist, independently test the crude extract collected from each stage for residual auxiliaries and select a dissolving solvent, then perform dissolution operations separately. Considering that the crude extract from the early stages (low polarity enrichment) may contain more non-polar auxiliaries (such as ethyl acetate), a lower polarity dissolving solvent should be selected; the crude extract from the later stages (high polarity enrichment) may contain more ethanol or water, so a higher polarity dissolving solvent should be selected.

[0220] After dissolving the crude extract at each stage, separate liquids are obtained for separation. These liquids are labeled and then separately introduced into subsequent HSCCC separation processes (sequential separation or parallel separation using a multi-channel system). Separate separation significantly improves separation efficiency and purity because the impurity profile at each stage is simpler.

[0221] The dissolution conditions (solvent type, volume, ratio) at each stage are associated with the corresponding crude extract and stored for subsequent traceability and optimization.

[0222] Take a small amount of crude extract (e.g., 10-50 mg), dissolve it completely in a suitable solvent (e.g., methanol), and determine the concentration of each target saponin monomer by UPLC-MS / MS or HPLC-UV. Then, calculate the mass fraction of each saponin in the crude extract based on the dilution factor. Finally, calculate the absolute mass (mg or g) of each target saponin based on the total mass of the crude extract.

[0223] The weighted average logP and the minimum and maximum logP can be calculated using the known logP values ​​of each target saponin (obtainable from literature or calculation) and their relative contents in the crude extract. Alternatively, reversed-phase HPLC analysis (C18 column, acetonitrile-water gradient) can be performed on the crude extract, and the relative polarity of each saponin can be determined based on its retention time (the shorter the retention time, the stronger the polarity). The distribution of retention times is the polarity distribution.

[0224] The volume of the solvent should be such that the total concentration of the target saponins in the final liquid to be separated is within the optimal injection concentration range for preparative separation using HSCCC. This range depends on the capacity of the HSCCC column, the injection volume, and the separation scale. For analytical or small preparative HSCCCs (column volume 10-50 mL), the injection concentration is typically 1-20 mg / mL (total of all saponins). For preparative HSCCCs (column volume 100-1000 mL), the injection concentration can be increased to 20-100 mg / mL.

[0225] Based on the total mass of each target saponin in the crude extract and the desired injection concentration, the required solvent volume is calculated using the following formula: ; Meanwhile, considering the contributions of other impurities and non-target saponins in the crude extract, the volume can be adjusted appropriately to avoid sample overload.

[0226] Based on the polarity distribution of the crude extract and the selected solvent, determine the volume ratio of each component in the solvent: If the target saponin is predominantly nonpolar (weighted average logP > 1.5), the solvent should contain a higher proportion of organic solvent (such as ethanol or n-butanol), for example, ethanol:water = 80:20. If the target saponin is predominantly polar (weighted average logP < 0.5), the proportion of the aqueous phase in the solvent should be increased, for example, ethanol:water = 30:70. If the polarity distribution is broad, a gradient dissolution with mixed solvents can be used (first add a low-polarity solvent to dissolve the low-polarity portion, then add a high-polarity solvent to dissolve the high-polarity portion), but usually a compromise composition is chosen to ensure that all target saponins are completely dissolved.

[0227] This can be determined through small-scale testing. Take a small amount of crude extract, add different proportions of solvent, observe the dissolution (whether it becomes clear or precipitates), and select the composition that can completely dissolve and produce a homogeneous solution.

[0228] Output the volume (or volume range) of the solvent and the specific composition ratio (e.g., "ethanol:water = 70:30, total volume 50 mL").

[0229] Residual carbon dioxide can be detected by taking a certain amount of crude extract, heating it to 60-80℃ in a sealed container, collecting the released gas, measuring it with a gas volume meter, and converting it into carbon dioxide mass. Alternatively, the crude extract can be placed in a headspace vial, heated to equilibrium, and the headspace gas can be injected into a gas chromatograph, and the carbon dioxide concentration can be determined using a thermal conductivity detector (TCD). Another method is to drop the crude extract sample into acidic water and observe the amount of bubbles produced (semi-quantitative).

[0230] Compare the measured value with a preset threshold (e.g., 0.5% by mass). If it is below the threshold, skip the degassing step; if it is equal to or exceeds the threshold, place the crude extract in a suitable container for degassing. The container can be placed in an ultrasonic bath and sonicated at room temperature or slightly above room temperature (e.g., 40°C) for 10-30 minutes, while applying a slight negative pressure (e.g., using a vacuum pump). Alternatively, connect the container to a vacuum system and evacuate to a pressure below 100 Pa at room temperature, maintaining this for 5-15 minutes with appropriate stirring to promote gas escape. Alternatively, nitrogen or argon gas can be introduced into the crude extract at a moderate flow rate for 5-10 minutes with stirring. After degassing, residual carbon dioxide can be tested again to confirm it has dropped below the threshold.

[0231] Degassing may alter the polar environment of the crude extract because carbon dioxide has a certain degree of polarity (especially in the dissolved state). After degassing, the effective polarity of the crude extract may decrease slightly. To maintain the solubility of the target saponin, the composition of the solvent needs to be fine-tuned. If precipitation or turbidity occurs in the crude extract after degassing, it indicates a decrease in solubility, and the proportion of polar components in the solvent should be appropriately increased (e.g., increasing the volume fraction of water by 5%-10%). If the solution remains clear after degassing, no adjustment is needed or only minor adjustments are required (e.g., increasing the polar component by 1%-2%).

[0232] You can conduct a small-scale test again by adding different proportions of solvent to the degassed crude extract, observing the dissolution process, and selecting the optimal composition. Record the adjusted solvent composition for use in the final dissolution test.

[0233] Add the final determined dissolving solvent (type, composition ratio, volume) to the crude extract, stir or use ultrasound to aid dissolution, until a clear, homogeneous, particle-free liquid is formed to be separated.

[0234] Perform the following checks on the prepared liquid to be separated: the solution should be clear and transparent, without suspended matter or layering. Dilute a small amount of the solution and determine the concentration of the target saponin by UPLC-MS / MS to confirm that the deviation from the expected value is within an acceptable range (e.g., ±10%). If it fails to meet the requirements, adjust the solvent volume or composition according to the deviation and redissolve the solution.

[0235] Optionally, after obtaining the fraction of the target ginsenoside monomer type, the method further includes: collecting the fraction of the target ginsenoside monomer type obtained by high-speed countercurrent chromatography separation, and performing online or offline purity detection on the fraction of the target ginsenoside monomer type; comparing the detected purity with a preset release purity threshold; if the purity is not lower than the release purity threshold, drying the fraction of the target ginsenoside monomer type to obtain the finished product of the target ginsenoside monomer type; if the purity is lower than the release purity threshold, when the purity deviation is less than a preset refining deviation threshold, sending the fraction of the target ginsenoside monomer type to preparative liquid chromatography for secondary purification, setting a peak cutting window according to the impurity spectrum characteristics in the fraction of the target ginsenoside monomer type during the purification process, collecting the high-purity central portion and then drying it; and when the purity deviation reaches or exceeds the refining deviation threshold, marking the fraction of the target ginsenoside monomer type as a non-conforming product and triggering a re-separation or discard instruction.

[0236] In the embodiments of the present invention, according to the detector signal, within the retention time window corresponding to the target saponin, the effluent is collected by an automatic fraction collector or manually. The collection container should be clean and free of contamination, and a suitable material (such as glass or polypropylene) should be selected according to subsequent detection requirements. If there are multiple target saponins in the same batch of separation, they should be collected separately.

[0237] The system is equipped with an online purity detection function (such as peak purity inspection by a diode array detector or multiple reaction monitoring ratio of mass spectrometry), which can evaluate the purity of the fraction in real time while collecting. For UV detection, by comparing the spectral similarity of different points on the peak, if the similarity is higher than a threshold value (such as 0.99), it is determined that the fraction is spectroscopically pure. For mass spectrometry detection, by monitoring the ratio of the characteristic ions of the target saponin to the total ion current, if the ratio is higher than a preset value, it is determined that the purity is high. The online detection results can be used as a basis for rapid screening, but usually still need to be confirmed offline.

[0238] For offline purity detection, a representative sample (such as 1 - 5 mL) is taken from the collected fraction, and quantitative analysis is carried out using UPLC-MS / MS or HPLC-UV. The sample is appropriately diluted and injected, and the peak area of the target saponin and the total area of all impurity peaks are recorded.

[0239] Calculate the purity: ; For mass spectrometry detection, the external standard method can also be used to calculate the content and then compare it with the total solids.

[0240] Record the purity value and detection conditions.

[0241] Associate and store the detection results (purity, impurity profile, detection method) with the fraction number and label them on the collection container.

[0242] Read the preset release purity threshold for the current product (type and grade of target saponin) from the process parameter database. This threshold is determined by the product quality standard. For example, the pharmaceutical grade threshold is usually 98% or 99%.

[0243] Compare the measured actual purity with the release purity threshold. If the actual purity ≥ threshold, it is determined to be qualified. If the actual purity < threshold, it is determined to be unqualified, and calculate the deviation degree = threshold - actual purity.

[0244] For the fractions determined to be qualified, directly enter the drying step. If multiple fractions come from the same target saponin (such as multiple collections), they can be combined and then dried.

[0245] The appropriate drying method should be selected based on the thermal stability and solvent properties of the target saponin. If the saponin is heat-sensitive (e.g., some protopanaxadiol-type saponins may degrade at high temperatures), freeze drying is preferred. If the solvent is a low-boiling-point organic solvent (such as ethanol or methanol) and the saponin has good thermal stability, rotary evaporation can be selected. For large-scale production, spray drying can be used.

[0246] The choice of drying method can be determined by the operator according to preset rules, or automatically recommended by the system based on the type of solvent.

[0247] The dried ginsenoside monomer powder is packed into a sealed container, labeled with information such as name, batch number, purity, quality, and production date, and then stored for sale.

[0248] Deviation level = Release purity threshold - Actual purity (absolute value). For example, if the threshold is 98% and the actual purity is 96%, the deviation level is 2%.

[0249] Retrieves a preset refining deviation threshold (e.g., 5%) from the database. This threshold can be preset based on the recovery rate and cost of the refining process.

[0250] If the deviation is less than the refinable deviation threshold, the product is determined to be refinable. If the deviation is greater than or equal to the refinable deviation threshold, the product is transferred to non-conforming product processing.

[0251] Secondary purification injection preparation: Concentrate the substandard fraction to a concentration suitable for preparative liquid chromatography injection (typically 10-50 mg / mL). Chromatographic condition selection: Select a suitable column (e.g., C18 reversed-phase column) and mobile phase (e.g., acetonitrile-water or methanol-water) based on the properties of the target saponin and the characteristics of the impurity spectrum. Conditions can be determined based on the separation behavior known in step S3 or through rapid screening. Peak cutoff setting: Monitor the peak shape of the target saponin using an online detector (e.g., UV detector) during the preparative liquid chromatography run.

[0252] The automatic cutting window is set to collect only the central part of the main peak, for example: starting collection at 50% of the peak height and ending at 90% of the peak height (symmetric peak); or avoiding the leading edge and tail part according to the asymmetry of the peak shape.

[0253] The width and position of the cutting window can be dynamically adjusted based on the separation degree between impurities and the main peak: if the impurities are close to the main peak, the window is narrowed; if the impurities are far away, the window can be widened appropriately to improve the recovery rate.

[0254] Collection and post-processing: Collect the fraction within the cutting window, repeat the above steps, and retest the purity. If it meets the standard, dry it; if it still does not meet the standard, re-evaluate. Usually, one purification process is sufficient to achieve the required purity.

[0255] Record purification conditions (column, mobile phase, cut window) and results (purity after purification, recovery rate) to optimize subsequent purification parameters.

[0256] In the batch record, the fraction is marked as non-compliant, and the purity, degree of deviation, and testing method are noted. The processing method is automatically selected according to preset rules. If the absolute content of the target saponin in the fraction is still high (e.g., greater than 50% of total solids), and the impurities are mainly saponins with properties similar to the target saponin (which can be separated again via HSCCC), then re-separation is selected. If the content of the target saponin in the fraction is very low (e.g., less than 20%), or the impurities are difficult-to-remove pigments or degradation products, then disposal is selected.

[0257] An economic model can also be used to estimate the costs and expected benefits of re-separation; if the costs are too high, the separation should be abandoned.

[0258] If re-separation is selected, the system will collect the non-conforming sample, which can then be concentrated and reinjected into the high-speed countercurrent chromatography unit as the liquid to be separated, using the same or finely adjusted separation conditions for further separation. Alternatively, the non-conforming sample can be combined with the next batch of crude extract for processing.

[0259] If the system chooses to discard the defective product, it will discharge the defective product into a waste liquid collection container or automatically open the waste liquid valve to discharge the liquid into the waste liquid treatment system. Disposal operations must comply with safety and environmental protection regulations. Simultaneously, the system records the reason for and quantity of the disposal. If multiple batches of defective products occur consecutively, the system will issue an alarm, prompting operators to check upstream processes (such as extraction and separation conditions) for any abnormalities.

[0260] Secondly, embodiments of the present invention provide a ginsenoside formulation, wherein the formulation is prepared by mixing ginsenoside monomers obtained by extraction, separation and purification through the high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in the embodiments of the present invention with excipients in a certain proportion, and is suitable for preparing health food or traditional Chinese medicine preparations.

[0261] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A high-purity extraction process for ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography, characterized in that, Includes the following steps: Step S1: Under preset extraction conditions, ginseng powder is extracted using a supercritical fluid extraction device to obtain crude ginsenoside extract. The extraction conditions are determined based on the combination of monomer types of the target ginsenoside components. Step S2: Dissolve the crude ginsenoside extract in a solvent to obtain the liquid to be separated. The solvent is determined based on the residual components in the crude ginsenoside extract. Step S3: Under the preset separation conditions, high-speed countercurrent chromatography is used to separate the liquid to be separated by high-speed countercurrent chromatography to obtain the fraction of the target ginsenoside monomer type. The separation conditions are determined according to the target ginsenoside monomer type, or the separation conditions are dynamically adjusted according to the stage parameters of the supercritical fluid extraction process.

2. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in claim 1, characterized in that, The extraction conditions are determined based on the combination of monomer types of the target ginsenoside components, specifically including: Determine the polarity parameters and solubility behavior of each target ginsenoside monomer in supercritical carbon dioxide; Based on the polarity parameters and solubility behavior, the comprehensive polarity index of the target saponin combination is calculated. Based on the comprehensive polarity index, a preset extraction condition database is matched to determine the target density range of supercritical carbon dioxide; Based on the equation of state relationship between carbon dioxide density and temperature and pressure, the initial values ​​of extraction temperature and extraction pressure are deduced and set. Based on the solubility requirements of the most or least polar component in the target saponin combination, determine the type of organic auxiliaries and their volume ratio to carbon dioxide. If the target saponin combination contains multiple monomers with significant differences in polarity, a segmented extraction strategy is adopted, in which temperature and pressure are adjusted in stages during the extraction process to sequentially enrich saponin components with different polarities.

3. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in claim 1, characterized in that, The separation conditions are determined based on the monomer type of the target ginsenoside component, and specifically include: The partition coefficient K of the target ginsenoside monomer in the candidate two-phase solvent system was determined; Based on whether the partition coefficient K value falls within the preset effective separation range, at least one set of two-phase solvent systems and their volume ratios are selected. Based on the difference in K values ​​among the monomers in the target saponin combination, determine whether a gradient elution mode is needed; If gradient elution is used, the gradient and rate of change of the composition of the mobile phase during the elution process are set according to the distribution behavior curves of each monomer in the stationary and mobile phases.

4. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in claim 1, characterized in that, The step of selecting at least one set of two-phase solvent systems and their volume ratios based on whether the partition coefficient K value falls within a preset effective separation range specifically includes: Based on the baseline separation requirements of the target ginsenoside monomers in high-speed countercurrent chromatography, a mathematical model is established between the resolution and the partition coefficient K. The partition coefficient K of the target saponin monomer in the candidate two-phase solvent system was determined, and the ratio of K values ​​between adjacent saponin components was calculated. When the ratio of the K values ​​of adjacent components is greater than or equal to the minimum ratio threshold required to meet baseline separation, the allocation coefficient K value is determined to fall into the effective separation range. Under selected rotation speed and flow rate conditions, the lower limit of the ratio of K values ​​corresponding to the preset value when the resolution of adjacent chromatographic peaks reaches the preset value; If the target saponin combination contains three or more monomers, the minimum value of the ratio of K values ​​of all adjacent components is taken. Only when the minimum value is not lower than the threshold value is the corresponding solvent system screened as an effective system.

5. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in claim 3, characterized in that, The determination of whether a gradient elution mode is needed based on the difference in K values ​​among the monomers in the target saponin combination specifically includes: Calculate the partition coefficient K values ​​of all monomers in the target saponin combination in a selected two-phase solvent system; Arrange the K values ​​of each monomer in ascending order, and calculate the ratio or difference of K values ​​between adjacent monomers in turn. Each of the calculated adjacent ratios or differences is compared with the preset acceptable threshold for isocratic elution. If all adjacent ratios or differences are not lower than the acceptable threshold for isocratic elution, then isocratic elution mode is adopted. If any set of adjacent ratios or differences is lower than the acceptable threshold for isocratic elution, then it is determined that gradient elution mode needs to be used. The acceptable threshold for isocratic elution is the minimum K-value ratio or minimum K-value difference that allows adjacent chromatographic peaks to achieve the target resolution, as determined experimentally under set rotation speed and flow rate conditions.

6. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in claim 3, characterized in that, The step of setting the gradient and rate of change of the mobile phase composition during elution based on the distribution behavior curves of each monomer in the stationary and mobile phases specifically includes: The partition coefficient K of each monomer in the target saponin combination was determined under different mobile phase composition ratios, and the behavior curve of K value as a function of mobile phase composition was plotted. By superimposing the behavior curves of each monomer, the mobile phase composition interval that maximizes the difference in K value between adjacent monomers is identified and used as the starting and ending points of gradient elution. Based on the slope of the behavior curve, the entire elution process is divided into several stages. The rate of change of the mobile phase composition in each stage is proportional to the slope of the monomer with the fastest change in K value in that stage. In the region where the K-value curves of any two adjacent monomers intersect or approach each other, reduce the rate of change of the mobile phase composition and extend the elution time to maintain sufficient separation. In regions where there is no overlap and the K values ​​of each monomer are evenly spaced, the rate of change of the mobile phase composition is increased to shorten the total elution time. Through iterative optimization, with the goal of achieving the shortest elution time for all target monomers while meeting the preset separation requirements, the final gradient and rate of change curve are determined.

7. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in claim 3, characterized in that, The separation conditions are dynamically adjusted based on the stage parameters of the supercritical fluid extraction process, specifically including: Based on the stage parameters of the supercritical fluid extraction process, the polarity distribution range and impurity spectrum characteristics of the target saponin combination in the crude extract are predicted. The polarity distribution range and impurity spectrum characteristics are used as input constraints for screening two-phase solvent systems. The composition of the two-phase solvent system separated by high-speed countercurrent chromatography is quantitatively adjusted by using the input constraints. The gradient of gradient elution is dynamically adjusted based on the combination of organic auxiliaries of different polarities or different stages of temperature and pressure during supercritical fluid extraction.

8. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in claim 3, characterized in that, The high-speed countercurrent chromatography separation using a high-speed countercurrent chromatography apparatus, for separating the liquids to be separated, includes: During the separation process, the post-column effluent is collected in real time to determine the actual partition coefficient K value of the target saponin monomer in the current two-phase solvent system. The actual measured K value is compared with the preset effective separation interval; If the actual K value deviates from the effective separation range, the volume ratio of each solvent in the two-phase solvent system will be automatically adjusted according to the preset step size. After adjustment, rebalance the chromatographic system and measure the K value again until the K value falls within the effective separation range. If the K value still cannot fall into the effective separation range after multiple adjustments, the current solvent system is determined to be unsuitable, and the system is automatically switched to the alternative two-phase solvent system.

9. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in any one of claims 1 to 8, characterized in that, The dissolving solvent is determined based on the residual components in the crude ginsenoside extract, including: The types and contents of residual organic adjuvants in crude ginsenoside extract were detected. Select the solvent based on the type and content of the residual organic additives; If a segmented extraction strategy is adopted, the crude ginsenoside extracts collected at each stage are dissolved separately, and different dissolution solvent systems are selected for the crude extracts at different stages. The total amount and polarity distribution of each target saponin monomer in the crude ginsenoside extract were determined to determine the volume and composition ratio of the dissolving solvent. If the residual carbon dioxide in the crude extract exceeds a preset threshold, degassing is performed before dissolution, and the polarity of the dissolving solvent is adjusted.

10. The high-purity extraction process of ginsenosides based on supercritical fluid extraction and high-speed countercurrent chromatography as described in any one of claims 1 to 8, characterized in that, After obtaining the fraction of the target ginsenoside monomer type, the process also includes: Collect the monomeric fractions of the target ginsenoside components obtained by high-speed countercurrent chromatography, and perform online or offline purity detection on the monomeric fractions of the target ginsenoside components. The detected purity is compared with the preset release purity threshold; If the purity is not lower than the release purity threshold, the fraction of the target ginsenoside monomer type is dried to obtain the finished product of the target ginsenoside monomer type. If the purity is lower than the release purity threshold, when the purity deviation is less than the preset refining deviation threshold, the target ginsenoside monomer type will be sent to a preparative liquid chromatography system for secondary purification. During the purification process, a peak cutting window will be set according to the impurity spectrum characteristics in the fraction of the target ginsenoside monomer type, and the high-purity central fraction will be collected and then dried. When the purity deviation reaches or exceeds the refining deviation threshold, the fraction of the target ginsenoside monomer type will be marked as unqualified, and a re-separation or disposal instruction will be triggered.

Images