Method for judging end point of column chromatography elution process of chrysanthemum morifolium and application thereof

Abstract

The present application belongs to the technical field of analytical chemistry and pharmaceutical analysis method, and relates to a method for judging the end point of column chromatography elution process of Hangzhou white chrysanthemum and application. The method comprises the following steps: column chromatography elution of Hangzhou white chrysanthemum extract, time sequence determination of the content of flavonoid components in the column chromatography elution process of Hangzhou white chrysanthemum extract, and collection of near-infrared spectrum of the column chromatography elution process; NIR spectrum pretreatment is performed on the near-infrared spectrum to obtain a pretreated spectrum; an MBSD model is established by using the pretreated spectrum to judge the end point of the elution process; and the end point of the elution process is verified according to the detection result of the content of flavonoid components. The method can quickly and accurately judge the end point of the elution process of Hangzhou white chrysanthemum, and improves the efficiency of judging the end point of the column chromatography elution process.

Description

Technical Field

[0001] This invention belongs to the technical field of analytical chemistry and pharmaceutical analysis methods, and relates to a method and its application for determining the endpoint of the column chromatography elution process of Chrysanthemum morifolium. Background Technology

[0002] The main active ingredients of Hangzhou white chrysanthemum (Chrysanthemum morifolium) are flavonoids (such as luteolin and apigenin), chlorogenic acid, volatile oils, polysaccharides, amino acids, and trace elements. Among these, flavonoids are considered the hallmark components and core medicinal substances of Hangzhou white chrysanthemum. Flavonoids (especially luteolin) have significant antioxidant, anti-inflammatory, and antiviral activities, and their ability to scavenge free radicals is more than five times that of vitamin C. This is the core material basis for the heat-clearing, detoxifying, vision-improving, and liver-protecting effects of Hangzhou white chrysanthemum. In addition, chlorogenic acid, as another key component, plays an important role in antibacterial, hypoglycemic, and immune-regulating effects, forming a synergistic effect with flavonoids. Together, they constitute the modern scientific basis for the "medicinal and edible homology" of Hangzhou white chrysanthemum.

[0003] In the extraction and separation process of Hangzhou white chrysanthemum, endpoint determination is crucial. Misjudgment not only wastes resources but may also affect downstream operations, ultimately impacting the quality of the final product. However, commonly used analytical techniques, such as traditional chromatography, are insufficient for effective monitoring of the pharmaceutical process due to their lengthy analysis cycles and significant manpower and material costs. Therefore, there is an urgent need to develop a rapid, non-destructive, and accurate endpoint determination method for the column chromatography elution process of Hangzhou white chrysanthemum.

[0004] Chinese patent CN113125375A discloses a method for detecting boiling time in an intelligent manufacturing extraction process. The detection technology used is near-infrared spectroscopy online detection technology. The specific steps are as follows: Step 1, install a near-infrared online detection system; Step 2, collect online near-infrared spectra during the extraction process and record the boiling time; Step 3, use the moving window standard deviation (MBSD) and principal component analysis-moving window standard deviation (PCA-MBSD) methods to establish a correspondence between the near-infrared spectral data obtained in Step 2 and the boiling time, and represent this correspondence using a mathematical model; Step 4, use the near-infrared online detection system with the mathematical model entered in Step 3 to detect the boiling time in the intelligent manufacturing extraction process. However, its application is not specifically for the process steps of determining the endpoint of Hangzhou white chrysanthemum and its column chromatography, so its modeling method and judgment criteria are not applicable to the process steps of determining the endpoint of Hangzhou white chrysanthemum and its column chromatography.

[0005] Wu Sijun, in his research on endpoint determination methods for traditional Chinese medicine manufacturing processes based on near-infrared spectroscopy, disclosed the use of NIR spectroscopy to monitor the endpoint of the mixing process of rhubarb soda tablets. First, by changing the content level of rhubarb extract and the ratio of four excipients in the prescription, 18 batches of samples were prepared. NIR spectral data including pressure variables were collected from 15 batches of samples after homogenization. Simultaneously, the content of active ingredients (emodin and emodin methyl ether) in the samples was determined using HPLC. Quantitative correction models and pressure-insensitive models for the content of each component were established using PLS regression algorithm, and the robustness differences between the two models were compared. Furthermore, principal component analysis combined with moving block standard deviation (PCA-MBSD) algorithm was used to qualitatively determine the mixing endpoint. However, the PCA-MBSD model used in this study exhibited significant errors when applied to the determination of the elution endpoint in column chromatography of Hangzhou white chrysanthemum.

[0006] Therefore, there is an urgent need to develop a rapid, non-destructive, and accurate method for determining the endpoint of the column chromatography elution process of chrysanthemum. Summary of the Invention

[0007] This invention addresses the problems existing in the prior art by providing a method and application for determining the endpoint of the column chromatography elution process of Hangzhou white chrysanthemum. The method of this invention can quickly and accurately determine the endpoint of the elution process of Hangzhou white chrysanthemum, thereby improving the efficiency of endpoint determination in column chromatography elution.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] In a first aspect, the present invention provides a method for determining the endpoint of the column chromatography elution process of Hangzhou white chrysanthemum, comprising the following steps:

[0010] (1) The extract of Chrysanthemum morifolium was eluted by column chromatography. The content of flavonoids in the extract of Chrysanthemum morifolium during the column chromatography elution process was determined over time, and the near-infrared spectrum of the column chromatography elution process was collected.

[0011] (2) Perform NIR spectral preprocessing on the near-infrared spectrum obtained in step (1) to obtain the preprocessed spectrum;

[0012] (3) Use the pre-processed spectrum obtained in step (2) to establish an MBSD model and determine the endpoint of the elution process;

[0013] (4) Based on the flavonoid content detection results of step (1), the endpoint of the elution process in step (3) is verified.

[0014] Further, the column chromatography elution method in step (1) is elution with 60-70% ethanol by mass concentration.

[0015] Preferably, the column chromatography elution in step (1) is elution with 70% ethanol by mass.

[0016] Furthermore, the time-series determination is as follows: the content of flavonoid components in the extract of Hangzhou white chrysanthemum is collected once every 30 seconds during the column chromatography elution process.

[0017] Furthermore, the wavelength range of the near-infrared spectrum in step (1) is 900-1700 nm, and the near-infrared spectrum is collected once every 20-40 seconds.

[0018] Preferably, the near-infrared spectroscopy detection flow cell has an optical path of 0.8 mm, a resolution of 0.5 nm, and a scan count of 32.

[0019] Further, the preprocessing method in step (2) is at least one of Savitzky-Golay smoothing, standard canonical transformation, multivariate scattering correction, Savitzky-Golay with first derivative, and Savitzky-Golay with second derivative.

[0020] Preferably, the preprocessing method in step (2) is multivariate scattering correction.

[0021] Furthermore, the MBSD model described in step (3) is the H2O-MBSD model.

[0022] Furthermore, the method for establishing the H2O-MBSD model is as follows: extract water matrix characteristic bands related to water molecules as input data for the MBSD model, and establish the H2O-MBSD model with a window value of 15 as a parameter.

[0023] Furthermore, the characteristic wavelength bands of the water matrix are 1347 nm, 1364 nm, 1373 nm, 1387 nm, 1415 nm, 1427 nm, 1440 nm, 1453 nm, 1466 nm, 1479 nm, 1490 nm and 1514 nm.

[0024] Further, the endpoint of the elution process in step (3) is determined by the baseline change rate fluctuation being less than or equal to 0.5%.

[0025] Furthermore, the baseline change rate ( The calculation method for ) is as follows:

[0026]

[0027] t represents time, in minutes;

[0028] MBSD t Representative: Standard deviation at time t;

[0029] MBSD t-1Representative: Standard deviation at time t-1.

[0030] Preferably, the criterion for determining the end point of the elution process is: when m consecutive points (m≥3) All satisfy: This is the endpoint of elution.

[0031] Further, the verification in step (4) is carried out with the endpoint being that the content of flavonoids in the elution process of Hangzhou white chrysanthemum extract is lower than the maximum value of the elution process by 0.5-1%.

[0032] Secondly, the present invention provides the application of the method in pharmaceutical quality monitoring.

[0033] The technical effects achieved by this invention are:

[0034] This invention determines the endpoint of the elution process by analyzing the standard deviation (MBSD) of the near-infrared absorbance of the eluent during chromatography and the influence of water spectrometry combined with MBSD on the concentration of flavonoids during elution. It establishes a green manufacturing method for determining the endpoint of chromatographic elution processes, improving the efficiency and accuracy of endpoint determination and providing a research foundation for intelligent production of this process. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of a near-infrared online detection system for the column chromatography elution process.

[0036] Figure 2 The images show online NIR spectra of the column chromatography elution process of Chrysanthemum morifolium. (A) is the original spectrum, (B) is the average spectrum of 4 batches, (C) is the spectrum after SNV pretreatment, (D) is the spectrum after MSC pretreatment, (E) is the spectrum after first derivative pretreatment, and (F) is the spectrum after second derivative pretreatment.

[0037] Figure 3 The following are the concentration curves of flavonoid components during the elution process for four batches: a-luteolin, b-luteolin-7-O-β-D-glucuronide, c-isochlorogenic acid B, d-isochlorogenic acid A, e-apigenin-7-O-glucosinolate, f-isochlorogenic acid C, g-apigenin 7-O-(6”-O-malonyl)-β-D-glucosinolate, h-scutellarin, (A) batch 1, (B) batch 2, (C) batch 3, and (D) batch 4.

[0038] Figure 4 The MBSD values ​​of the NIR spectra of four batches of Hangzhou white chrysanthemum under different pretreatment methods during column chromatography elution were modeled in the range of 830-2410 nm with a window value of 15. (A) Batch 1, (B) Batch 2, (C) Batch 3, (D) Batch 4.

[0039] Figure 5 The MBSD values ​​of the online NIR spectra of Hangbai chrysanthemum column chromatography elution process in different modeling bands were obtained. The spectral preprocessing was SG+SNV with a window value of 15. (A) Batch 1, (B) Batch 2, (C) Batch 3, (D) Batch 4.

[0040] Figure 6 The following are the score curves and loading curves for PC1 and PC2: (A) PC1 score curve during elution, (B) PC1 loading curve during elution, (C) PC2 score curve during elution, and (D) PC2 loading curve during elution.

[0041] Figure 7 For online NIR spectral MBSD models with different window values, (A) NIR-MBSD value of elution process with window value of 5, (B) NIR-MBSD value of elution process with window value of 10, (C) NIR-MBSD value of elution process with window value of 15, and (D) NIR-PCA-MBSD value of elution process with principal component 2 (window value of 15).

[0042] Figure 8 The NIR-MBSD values ​​are based on the elution process using 12 water matrix bands. Detailed Implementation

[0043] The following non-limiting embodiments are intended to enable those skilled in the art to gain a more comprehensive understanding of the present invention, but do not limit the invention in any way. The following description is merely an exemplary illustration of the scope of protection of the present invention, and those skilled in the art can make various changes and modifications to the invention based on the disclosed content, which should also fall within the scope of protection of the present invention.

[0044] The present invention will be further described below by way of specific embodiments. Unless otherwise specified, all chemical reagents used in the embodiments of the present invention were obtained through conventional commercial means. Unless otherwise specified, all contents mentioned below are mass contents. Unless otherwise specified, it is understood that the process was carried out at room temperature.

[0045] Comparative Example 1: A method for determining the endpoint of the column chromatography elution process of Hangzhou white chrysanthemum

[0046] (1) Sample preparation: The preparation method of Hangzhou white chrysanthemum column chromatography sample includes key process flow such as extraction and concentration, macroporous resin activation and sample loading.

[0047] Extraction and concentration of Hangzhou white chrysanthemum: First, accurately weigh 500 g of Hangzhou white chrysanthemum raw material, add 7500 ml of 70% ethanol solution at a material-to-liquid ratio of 1:15, heat with an electric heating mantle, reflux for 2 hours, cool to room temperature, and pour out the extract; then perform a second extraction, following the same steps; next, mix the two Hangzhou white chrysanthemum extracts and concentrate by rotary evaporation at a temperature of 55℃ and a rotation speed of 20 r / min; finally, adjust the concentrated Hangzhou white chrysanthemum extract to a 20% ethanol concentration for loading.

[0048] Macroporous resin activation: First, the macroporous resin was soaked in 95% ethanol for 24 hours and then wet-packed. It was then washed with deionized water until no alcohol odor remained. Second, the column was passed through 2 column volumes (BV) of 5% NaOH solution, allowed to stand for 30 min, and then washed with water until neutral. Next, the column was passed through 2 BV of 10% glacial acetic acid solution, allowed to stand for 30 min, and then washed with water until neutral. Finally, the column was passed through 2 BV of 20% ethanol solution before sample loading. Sample loading process: The concentrated extract of Hangzhou white chrysanthemum was pumped using a flow pump at a flow rate of 30 mL / min for approximately 3 hours. After sample loading, the column outlet valve was closed and the column was allowed to stand for 30 min to elute. A total of four batches were prepared according to the above sample preparation procedure. Due to differences in the content of the Hangzhou white chrysanthemum raw material, there were some differences in the elution samples among the four batches.

[0049] (2) Analysis of HPLC quantitative detection results of flavonoid components

[0050] Chromatographic conditions: A Waters CORTECS T3 C18 column (4.6 mm × 150 mm, 2.7 μm) was used for analysis of the eluent samples; column temperature was 25℃; detection wavelength was 327 nm; flow rate was 1.0 mL; mobile phase A was 0.1% phosphoric acid aqueous solution, and mobile phase B was acetonitrile; gradient elution: 0 min, 10% B; 0–15 min, 10%–20% B; 15–25 min, 20–22% B; 25–35 min, 22%–35% B; 35–35.1 min, 35%–95% B; flow rate was set to 1 mL / min. Injection volume was 10 μL.

[0051] The concentration curves of flavonoid components during the elution process of 4 batches are as follows: Figure 3 As shown, substances with higher concentrations are mainly a, d, e, and f; while the remaining four substances, b, c, g, and h, have relatively lower concentrations. Due to the differences in flavonoid content between different batches, the elution time endpoint shifts. Therefore, there is an urgent need to develop a real-time, non-destructive, rapid detection method to help laboratory operators accurately determine the endpoint of the elution process.

[0052] (3) Near-infrared spectral acquisition during column chromatography elution process

[0053] A schematic diagram of the near-infrared online detection system for the column chromatography elution process is shown below. Figure 1 As shown: The sample, contained in a beaker, was loaded into the chromatography column using a peristaltic pump and collected in another beaker via a flow cell. The online NIR fiber optic probe of the near-infrared spectrometer was connected to the flow cell at the eluent outlet of the Hangzhou white chrysanthemum chromatography process, and the near-infrared spectrometer was connected to a computer to output the spectra. Elution was performed with 3 BV of 70% ethanol. Online near-infrared spectra of the eluent were acquired every 30 seconds. 201 near-infrared spectra were obtained per batch, for a total of 804.

[0054] (4) Online NIR spectral characteristics analysis of the column chromatography elution process of Hangzhou white chrysanthemum

[0055] Figure 2 (A) is the online NIR raw spectrum of the column chromatography elution process of Hangzhou white chrysanthemum. The characteristic peaks of the online NIR raw spectra of the 804 samples are basically consistent, with 4 obvious characteristic intervals.

[0056] Figure 2 (B) shows the average spectral curves of the four batches. It can be seen that the four characteristic ranges are 1300-1600nm, 1692-1790nm, 1800-2000nm and 2200-2400nm, respectively.

[0057] The 1300-1600 nm range is the characteristic range of water molecule matrix groups. The strong absorption at 1400-1450 nm is a combination of the fundamental and overtone absorption of the symmetric and antisymmetric stretching vibrations of water molecules. As the elution process continues, the absorption peak intensity of the eluent in this range decreases linearly because the water molecule content in the solution continuously decreases from 80% to 30%.

[0058] The strong and broad absorption in the 1850-2000 nm range is a combination of antisymmetric stretching and bending vibrations of water molecules. In the early stages of elution, the characteristic peaks in this range exhibit overload, possibly due to the high concentration of water molecules. In the 1600-1800 nm and 2200-2400 nm ranges, the intensity of the characteristic peaks increases linearly with elution, possibly because the flavonoids adsorbed on the macroporous resin are continuously eluted and their concentration increases. Furthermore, the NIR spectral distribution is relatively dense in the early (0-25 min) and late (55-100 min) stages of elution; in the middle (25-55 min), the NIR spectral distribution is more uniform, indicating the feasibility of online NIR spectral characterization of the state properties of the column chromatography elution process. NIR spectra are pretreated using SNV (Surface Noise Reduction) during the elution process. Figure 2 (C); MSC pretreatment, Figure 2(D) First derivative preprocessing, Figure 2 (E), and second derivative preprocessing, Figure 2 The (F) values ​​all indicate that there is a lot of noise in the online NIR spectra from 1850 to 2000 nm.

[0059] (5) Comparison of MBSD modeling based on different NIR spectral preprocessing, band ranges and window values

[0060] Online NIR spectral MBSD models with different preprocessing methods, such as Figure 4 As shown, the modeling band is 833–2409 nm, and the window value is 15.

[0061] The results showed that the first and second derivative MBSD values ​​fluctuated significantly during the chromatographic elution process, indicating that the reliability of the online NIR first and second derivative MBSD models was low. The MBSD values ​​of the preprocessed SNV and MSC spectra differed in intensity from the original spectra at 30-40 min, but the peak shapes were consistent; around 50-60 min at the end of the elution phase, the MBSD values ​​tended to be consistent.

[0062] Furthermore, the MBSD values ​​of MSC and SNV spectral pretreatments were compared. The MBSD value of the MSC-pretreated spectrum showed a lower trough during the elution stage. Therefore, MSC was chosen as the pretreatment method to establish an online NIR spectral MBSD model. In summary, based on the MBSD model results of four batches with different pretreatments, the MSC-MBSD model exhibits stronger stability and is more suitable for establishing an MBSD endpoint determination model for the elution process of Hangzhou white chrysanthemum column chromatography.

[0063] Online NIR spectral MBSD models based on different modeling intervals, such as Figure 5 As shown, the spectral preprocessing is MSC with a window value of 15.

[0064] The results showed that the MBSD values ​​of different batches using the 1350-2409 nm band fluctuated significantly, while the MBSD values ​​of the 1350-1701 nm, 1705-2146 nm, and 2153-2409 nm bands fluctuated less. Since the MBSD values ​​of 1350-1701 nm and 2153-2409 nm were more stable and gradual, these bands were selected as the modeling bands to establish an online NIR spectrum MBSD model.

[0065] Since the threshold of the MBSD model is also affected by the window value, online NIR spectral MBSD models with different window values ​​are as follows: Figure 7 As shown, the spectral preprocessing was performed using MSC, with modeling bands of 1350-1701 nm and 2153-2409 nm. The results show that when the window value is 5, as...Figure 7 As shown in (A), the MBSD values ​​of the four batches fluctuated significantly, making them unsuitable for determining the endpoint of the elution process. When the window value is 10, as... Figure 7 As shown in (B), the results indicate that a window value of 10 is superior to the MBSD model with a window value of 5, but this model exhibits greater fluctuations during peak elution periods. Furthermore, when the window value is 15, as... Figure 7 As shown in (C), the overall fluctuations of the four batches were relatively small. Batch 1 and 4 completed elution in approximately 40 minutes, while batches 2 and 3 completed elution in approximately 35 minutes. In summary, using MSC as the spectral preprocessing method, with modeling bands of 1350-1701 nm and 2153-2409 nm and a window value of 15, the online NIR spectral MBSD model can accurately characterize the endpoint of the column chromatography elution process. This enables quality control of the process for determining the endpoint of the Hangzhou white chrysanthemum column chromatography elution process, reduces energy consumption, and ultimately achieves energy-saving and emission-reducing green manufacturing of traditional Chinese medicine.

[0066] (6) Methods for determining the elution endpoint of column chromatography based on PCA analysis and PCA-MBSD method:

[0067] First, principal component analysis is used to perform eigenvalue decomposition on the NIR spectrum within the window. The specific method is as follows:

[0068] The moving block standard deviation (MBSD) and PCA-MBSD methods were used. When both MBSD and PCA-MBSD values ​​fell below a set threshold, it indicated that the chromatographic elution process of Hangzhou white chrysanthemum had reached its endpoint. MBSD was determined by comparison... n The endpoint of the process is determined by the standard deviation of the continuously recorded spectra. When the standard deviation tends to be consistent or falls below a specified value, the process is considered to have reached its endpoint. The formula for the standard deviation of the moving block is shown below:

[0069] (1)

[0070] (2)

[0071] In equation (1), wavelength i The first point j The absorbance values ​​of the spectrum, wavelength i Continuous n The square of the average absorbance of the spectrum. In equation (2), m This is the sum of the selected wavenumbers. The oldest spectrum in the selected spectra is removed sequentially, and a new spectrum is added in each order for recalculation. and .when When the process gradually stabilizes, it can be considered to have reached its end.

[0072] PCA-MBSD: Matrix X n*m After standardizing the mean, principal component analysis (PCA) is performed, and a PCA analysis model is established.

[0073] (3)

[0074] In equation (3), The score matrix is ​​established by the first A principal components; This is the transpose of the load matrix; This is the residual matrix. When the number of principal components is 1, the first principal component loading is selected. Calculate the first principal component score vector for each spectrum. ,by Calculate the spectral MBSD value within the window; when the number of principal components is 2, calculate the score vector of the first principal component respectively. Second component score vector The MBSD value of the spectrum within the window is calculated by summing the two score vectors; similarly, other principal components are calculated in the same way.

[0075] Considering the influence of noise and dark current during near-infrared spectrum acquisition, this patent employs Savitzky-Golay smoothing (SG), standard normal variate (SNV), multiple scattering correction (MSC), Savitzky-Golay + first derivative (SG+1st), and Savitzky-Golay + second derivative (SG+2nd) to preprocess the online NIR raw spectrum.

[0076] To monitor water characteristic peaks associated with the active ingredient during column chromatography elution, PCA was applied to the average spectrum of the first batch in the 1300-1600 nm region after Savitzky-Golay smoothing. Figure 6 The loadings and fractions of the first two PCs are shown. The results indicate that the first PC explains more than 99.5% of the spectral variance, suggesting that most of the spectral information is contained in the first PC (PC1). The remaining PCs fluctuate around zero, which can be considered as variations caused by noise.

[0077] For PC1, the score continued to decrease, eventually entering a plateau phase, characterized by a broad peak centered at 1456 nm. With 99.5% variance explanation, PC1 can explain almost all significant spectral variations. PC2 accounts for 0.5% of the total variance, with a significant positive peak at 1409 nm, shifted from the central broad peak at 1456. It exhibits a negative peak at 1674 nm, characterized by fewer weakly polar hydrogen-bonded OH groups and more strongly polar hydrogen-bonded OH groups. Based on the changes in fraction and loading, combined with the material distribution, PC1 primarily reflects the transformation of the water structure. According to the changes in the PC1 score plot... Figure 6 The entire elution process can be divided into three stages: stage 1 (0-25 min), stage 2 (25-50 min), and stage 3 (50-100 min). The negative peak is related to the formation of the hydrogen bond network in the water molecule structure, while the positive peak is the opposite. The entire elution process can be viewed as a process of hydrogen bond network disruption, formation, and eventual dynamic stabilization. This indicates that the active components of chrysanthemum flavonoids have both structure-creating and structure-destroying effects on water, and these two functions are continuous. As the chromatography time increases, the concentration of active components in the eluent increases, continuously disrupting the water molecule network structure. The phenomenon of the water structure becoming stable in the final stage indicates that the chromatography process is nearing its end.

[0078] Since the PCA analysis results PC1 can represent 99.5% of the information in the spectrum, this study does not consider the influence of different principal component numbers on the online NIR spectral PCA-MBSD model. The PCA-MBSD model based on PC1 score uses MSC for spectral preprocessing, with modeling bands of 1350-1701 nm and 2153-2409 nm and a window value of 15.

[0079] In summary, MSC spectral preprocessing was used, with modeling bands of 1350-1701 nm and 2153-2409 nm, a window value of 15, a principal component count of 2, and a PCA-MBSD threshold of -1. The PCA-MBSD model exhibits a slight tendency to predict elution time too early overall.

[0080] Example 1: A method for determining the endpoint of column chromatography elution based on 12 water matrix bands combined with MBSD method

[0081] In the PCA analysis results of step (6) in Comparative Example 1, we found that the precipitation of active substances in the chromatography process is closely related to the structure of water molecules. Therefore, based on the comparative example, 12 strongly correlated characteristic bands of water molecules in the near-infrared band were selected for further verification in this example.

[0082] Steps (1)-(4) of Example 1 are the same as steps (1)-(4) of Comparative Example 1.

[0083] (5) The study of water spectromics takes water as the object and explores the interaction between water and light in various systems at near-infrared frequencies. Since the OH- bonds of water are easily changed by other molecules, the study can be carried out by detecting the solute changes in the water molecule system itself. Based on this, we screened four batches of NIR spectral information using the absorption bands of water molecules in the near-infrared region at 1347, 1364, 1373, 1387, 1415, 1427, 1440, 1453, 1466, 1479, 1490 and 1514 nm. Then, based on steps (5)-(6) of Comparative Example 1, the dimensionality reduction variable of PCA analysis, namely principal component 1 (PC1), was converted into 12 water matrix bands as feature variables and MBSD model analysis was performed to establish the H2O-MBSD model. This model can monitor the elution process of Hangzhou white chrysanthemum column chromatography in real time and judge that the elution process endpoint has been reached when the predicted baseline change rate fluctuation is less than or equal to 0.5. The results are as follows. Figure 8 As shown.

[0084] How to calculate the baseline rate of change ( ):

[0085]

[0086] t represents time, in minutes.

[0087] MBSD t Representative: Standard deviation at time t;

[0088] MBSD t-1 Representative: Standard deviation at time t-1.

[0089] Endpoint determination condition: When m consecutive points (m≥3) are reached... All satisfy:

[0090] %

[0091] This is determined as the elution endpoint.

[0092] Using the final eluted substance, e, apigenin-7-O-glucoside, as the standard, the elution endpoints for Example 1, Comparative Example 1, and component content detection are shown in Table 1.

[0093] Table 1

[0094]

[0095] The elution endpoint of Comparative Example 1 was much earlier than that of component content detection, which would lead to premature elution and ineffective elution of some products, resulting in waste.

[0096] When the mass concentration of flavonoids changes, the absorbance changes, which disturbs the covalent and hydrogen bonds in the elution, i.e., the structure of water, thus altering the water's spectrum. The H2O-MBSD model predicted the endpoint times for batches 1-4 to be approximately 54 min, 55 min, 50 min, and 44 min, respectively, in four batches. Near the end of the elution process, the prediction trends of the H2O-MBSD and PCA-MBSD models were largely consistent, but the overall predicted endpoint time was later, more closely resembling the results of flavonoid detection under HPLC, thus demonstrating greater accuracy and resolving the issue of the PCA-MBSD model predicting the elution endpoint too early.

[0097] The results show that water molecule spectroscopy can accurately characterize the changes in flavonoid components in the eluent. Therefore, while ensuring the stability of the model results, the H2O-MBSD model has better predictive ability than the PCA-MBSD model, saving model processing time and computational costs, and providing strong support for the online automated determination of the endpoint of the column chromatography elution process of industrial-grade Hangzhou white chrysanthemum.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for determining the endpoint of the column chromatography elution process of Hangzhou white chrysanthemum, characterized in that, Includes the following steps: (1) The extract of Chrysanthemum morifolium was eluted by column chromatography. The content of flavonoids in the extract of Chrysanthemum morifolium during the column chromatography elution process was determined over time, and the near-infrared spectrum of the column chromatography elution process was collected. (2) Perform NIR spectral preprocessing on the near-infrared spectrum obtained in step (1) to obtain the preprocessed spectrum; (3) Use the pre-processed spectrum obtained in step (2) to establish an MBSD model and determine the endpoint of the elution process; (4) Based on the detection results of the flavonoid content in step (1), the endpoint of the elution process in step (3) is verified; The preprocessing method described in step (2) is multivariate scattering correction; The MBSD model mentioned in step (3) is the H2O-MBSD model; The method for establishing the H2O-MBSD model is as follows: extract the water matrix characteristic bands related to water molecules as input data for the MBSD model, and establish the H2O-MBSD model with a window value of 15 as a parameter; the water matrix characteristic bands are 1347 nm, 1364 nm, 1373 nm, 1387 nm, 1415 nm, 1427 nm, 1440 nm, 1453 nm, 1466 nm, 1479 nm, 1490 nm and 1514 nm.

2. The method according to claim 1, characterized in that, The column chromatography elution method described in step (1) is: elution with 60-70% ethanol by mass concentration.

3. The method according to claim 1, characterized in that, The wavelength range of the near-infrared spectrum in step (1) is 900-1700nm, and the near-infrared spectrum is collected every 20-40s.

4. The method according to claim 1, characterized in that, The endpoint of the elution process in step (3) is determined by the baseline change rate fluctuation being less than or equal to 0.5%.

5. The method according to claim 1, characterized in that, The verification in step (4) is as follows: the verification is carried out with the content of flavonoids in the elution process of Hangzhou white chrysanthemum extract being 0.5-1% lower than the maximum value of the elution process as the endpoint.

6. The application of the method according to any one of claims 1-5 in the elution monitoring of Hangzhou white chrysanthemum column chromatography.

Images