CN111157646B - High-throughput screening method and application of high-speed counter-current chromatography solvent system - Google Patents

Abstract

The invention belongs to the technical field of separation and purification, and relates to a high-throughput screening method and application of a high-speed counter-current chromatography solvent system. Different linear prediction models are respectively established by three different C18 chromatographic columns, two different solvent system selection strategies are provided aiming at the conditions of directional separation of specified compounds, multi-component separation of complex systems and the like, and the selection process of a solvent system is simplified into one-time high performance liquid chromatography analysis and one-time shake flask test. The invention also provides the application of the screening method.

Description

High-throughput screening method and application of high-speed counter-current chromatography solvent system

Technical Field

The invention belongs to the technical field of separation and purification, and relates to a high-throughput screening method and application of a high-speed counter-current chromatography solvent system.

Background

The high-speed counter-current chromatography is a novel separation and purification technology based on continuous liquid-liquid extraction, the stationary phase and the mobile phase of the high-speed counter-current chromatography are common laboratory reagents which can be recycled, the chromatographic conditions are mild, and the preparation amount is large. Compared with other chromatographic techniques, the method does not cause irreversible adsorption to the sample like other adsorption chromatographs, greatly protects the original chemical components in the sample in the separation process, and has obvious advantages. At present, high-speed countercurrent chromatography is widely applied to the separation and preparation of important active natural products such as flavone, alkaloid, phenolic acid, terpenes, lignans, saponin, protein, fructo-oligosaccharide and the like. In particular, the method has incomparable advantages in the aspects of enrichment and purification of active compounds, separation and analysis of enantiomers and the like. Unfortunately, screening for a suitable solvent system has long been a significant amount of time and effort by the experimenter, limiting its use.

At present, the selection method of the solvent system of the high-speed counter-current chromatography which is most widely applied is to perform a shake flask test by utilizing a known solvent system of a compound with a similar structure referred by a literature, measure the distribution coefficient (K) of a sample in an upper phase and a lower phase, and adjust the solvent system in the literature according to the K value until a satisfactory separation effect is obtained. But for compounds that are not The basis of investigation or unknown compounds, it can be determined by The solvent family of solvent systems, which includes: in 1991, oka et al used five solvents such as n-hexane, ethyl acetate, methanol, n-butanol and water as components, and arranged and combined into 16 solvent systems according to polarity order; abbott et al designed 13 solvent systems of different compositions and classified them according to polarity into three types, liposoluble, moderately polar, and polar. In 1992, ito established a screening method for a n-hexane-ethyl acetate-methanol/n-butanol-water solvent system, with the partition coefficient (K) of the target compound between 0.5 and 2 as an evaluation criterion, and introducing chloroform-methanol-water (2. Several important solvent system screening theories were reported in a book, centrifugal Partition Chromatography, published in 1995: foucault combines four solvents such as n-heptane, n-butanol, acetonitrile, water and the like to construct an HBAW solvent system. Margraff establishes ArIZNA solvent system selection theory with n-heptane, ethyl acetate, methanol and water. The above solvent system selection theories emphasize that a solvent system suitable for the polarity of the target compound is selected, and the process of selecting the solvent system needs to be adjusted by trial and error continuously until a suitable solvent system is selected, so that the process is complex, and the practical application is difficult.

In order to reduce the number of times of shake flask tests, simplify the selection process of a high-speed counter-current chromatography solvent system and reduce the screening difficulty, a plurality of innovative solvent system selection methods are continuously provided in recent years. For ARIZONA type solvent system with large polar span, wide application and strong regularity, J.Brent Friesen and Guido F.Pauli utilize 21 standard compounds to discuss that the methanol or n-hexane content in an n-hexane-ethyl acetate-methanol-water system and lgK of a sample to be separated have strong linearity, and the K value of the sample can meet the condition that the K is more than or equal to 0.4 and less than or equal to 2.5 by adjusting the solvent proportion, thus obtaining the proper solvent system. On the basis, a G.U.E.S.S. method for screening a solvent system by taking the specific displacement value of 21 standard substances as reference is established by adopting thin layer chromatography. The method greatly changes the traditional solvent system selection idea, improves the screening efficiency, but needs to purchase expensive standard products.

In addition, for n-hexane/EtOAc/MeOH/H ₂ The O system has strong regularity, scholars at home and abroad establish various mathematical prediction models aiming at the directional separation of target compounds, elisabeth Hopmann et al compares the experimental data and the prediction data of five model solute systems by combining a quantum chemistry method of statistical thermodynamics, and proves the potential of using COSMO-RS as a screening tool of a screening solvent system. The model prediction methods are only used for determining necessary parameters through a plurality of shake flask tests aiming at a specified compound so as to establish a K value prediction model, thereby realizing accurate optimization of a specific compound solvent system. Liangjunling doctor according to n-hexane/EtOAc/MeOH/H ₂ The linear relationship between the methanol or n-hexane content in the O system and the lgK of the sample to be separated was designed to cover a more comprehensive range of n-hexane/EtOAc/MeOH/H ₂ 9X 9 diagram of O System, a mathematical model was developed to quickly model the target compound by 2 shake flask experiments to determine the optimal n-hexane/EtOAc/MeOH/H for a given compound ₂ O system. The method does not need complex knowledge such as phase diagram and the like, is simple to operate, has accurate results, and is the most convenient and effective solvent system selection method so far. However, in the method, the mathematical model is only suitable for one compound, and a plurality of compounds can be applied only by establishing a plurality of mathematical models, so that the workload of solvent system selection can be increased rapidly along with the increase of the number of separated components, and meanwhile, the accuracy of the result can be influenced due to the problem of inaccurate K value measurement caused by the limitation of liquid chromatography separation degree, thereby influencing the prediction of an optimal solvent system and increasing the selection difficulty coefficient.

Disclosure of Invention

Aiming at the problems existing in the selection of the high-speed counter-current chromatography solvent system at the present stage, the invention provides a high-throughput screening method of the high-speed counter-current chromatography solvent system, wherein different linear prediction models are respectively established by three different C18 chromatographic columns, two different solvent system selection strategies are provided aiming at the conditions of directional separation of specified compounds, multi-component separation of complex systems and the like, and the selection process of the solvent system is simplified into one-time high-performance liquid chromatography analysis and one-time shake flask test. The invention also provides the application of the screening method.

A high-throughput screening method for a high-speed counter-current chromatography solvent system comprises the following steps:

(1) High performance liquid chromatography

The mobile phase is water (A) -methanol (B), the gradient elution is 10% by weight A-100% by weight B,45min;100% by weight, B,10min. Flow rate: 1mL/min; column temperature: 30 ℃; the sample injection amount is 10 mu L;

(2) Determination of the proportion B% of methanol in the Mobile phase

Gradient delay time of 0.8min, gradient slope of 2%/min, initial gradient of 10% B, methanol proportion at the time of peak B% = (t _R -0.8)% 2% +10%, where t _R Is the compound peak time; t is the gradient delay time; bo% is the initial methanol ratio; r is gradient slope;

(3) Establishment of mathematical model

HEMW solvent System n-hexane/EtOAc/MeOH/H ₂ In the step (10-X), the size of the P 'value represents the polarity of a solvent system and the composition of solvent components, the polarity of a target sample is represented by the proportion of methanol in a mobile phase, and a mathematical model is established with the average polarity parameter P' of the solvent system;

after HPLC analysis, the B% and P' of the standard compound show a complete linear relationship when the peak appears, then the proportion X of methanol or normal hexane in the HEMW system and the B% show a complete linear relationship, namely the linear relationship between X and B% can be directly analyzed to construct a solvent system prediction model; and obtaining a prediction model formula.

(4) Determination of solvent System

And calculating the proportion of X, namely methanol or n-hexane according to a solvent system prediction model formula, and verifying the K value by shaking the bottle.

The shaking bottle result of the step (4) needs to meet 0.25< -K < -2.5.

In the step (4), if K is less than 0.25, the predicted initial solvent system HEMW = X (10-X): X (10-X) is obtained by subtracting 1 from X, namely, the distribution coefficient of target compounds in the solvent system is satisfied by 0.25-K-P2.5;

if K >2.5, the predicted initial solvent system HEMW = X (10-X): X (10-X) requires that X be increased by 1 to meet the target compound partition coefficient in the solvent system of 0.25-K-Ap-2.5.

Taking three chromatographic columns such as Agilent Eclipse XDB-C18 (5 μm, 4.6X 250 mm), agilent extended-C18 (5 μm, 4.6X 250 mm) and Diamonsil C18 (5 μm, 4.6X 250 mm) as examples, samples are analyzed under specified chromatographic conditions, the methanol content B% in the mobile phase when the sample peaks is calculated, and three different linear prediction models are constructed by analyzing the linear relation between X and B%. These models are in substantial agreement with theoretical expectations. Although none of the three mathematical models can predict the solvent system of the sample with K =1 very accurately, the K value is not required to be controlled accurately in the high-speed countercurrent chromatography separation process, and only 0.25-K-Ap-2.5 is required to be satisfied to achieve good separation effect.

And for the directional separation of the specified compound, calculating the content B% of methanol in the mobile phase when the sample is subjected to peak separation by a chromatographic column, and predicting to obtain an initial solvent system by substituting different chromatographic columns into corresponding linear prediction models. Shake flask experiments were conducted to determine whether the partition coefficients of target compounds in solvent systems met 0.25< -K < -2.5. Approximately 70% of the compounds can be predicted directly to give a suitable solvent system, but approximately 30% of the compounds need to be corrected for the predicted outcome:

if K <0.25, indicating that the target compound is more polar than the average polarity of the solvent system, then the predicted initial solvent system HEMW = X (10-X): X (10-X) requires that X be subtracted from 1, i.e. such that the partition coefficient of the target compound in the solvent system satisfies 0.25-K- < -2.5;

if K >2.5, indicating that the target compound is less polar than the average polarity of the solvent system, the predicted initial solvent system HEMW = X (10-X): X (10-X) would require an addition of 1 to X, i.e. such that the partition coefficient of the target compound in the solvent system would satisfy 0.25-K-straw 2.5.

In the application of multi-component separation of complex systems, each separation method has certain application range and limitation, and similarly, in high-speed counter-current chromatography, only one solvent system cannot enable all compounds with different structural types to have distribution coefficients in a proper range. In order to achieve a better separation effect and reduce the separation difficulty, an average solvent system of a plurality of components can be selected for separation. And calculating the content B% of methanol in the mobile phase when the sample is subjected to peak separation by the chromatographic column, calculating the average value of the content B%, and substituting the average value into a corresponding linear prediction model according to different chromatographic columns to directly obtain an average solvent system for predicting various components.

Drawings

FIG. 1 is a P' value for the HEMW solvent system in the 9X 9 chart;

FIG. 2 is a graph showing the relationship between the ratio of methanol or n-hexane in the HEMW system and P';

FIG. 3 is a model of the linear relationship of X and B% in an Agilent Eclipse XDB-C18 chromatographic column;

FIG. 4 is a model of the linear relationship between X and B% in an Agilent extended-C18 column;

FIG. 5 is a model of the linear relationship between X and B% in a Diamonsil C18 column;

FIG. 6 is an HPLC analysis chromatogram of an Asclepiadaceae sample;

FIG. 7 is a chromatogram of imperatorin on Diamonsil C18;

FIG. 8 is a chromatogram of kuhseng neol F on Diamonsil C18;

FIG. 9 is a chromatogram of imperatorin on Agilent extended-C18;

FIG. 10 is a chromatogram of Sophora flavescens neoalcohol F on Agilent extended-C18.

The invention has the advantages of

The research develops a simple high-throughput screening method for a high-speed counter-current chromatography solvent system, namely different linear prediction models are respectively established by three different C18 chromatographic columns, two different solvent system selection strategies are provided aiming at the conditions of directional separation of specified compounds, multi-component separation of complex systems and the like, and the selection process of the solvent system is simplified into one-time high-performance liquid chromatography analysis and one-time shake flask test. Compared with the g.u.e.s.s. Method established by similar j.breesen and Guido f.pauli, the purchase of valuable standard compounds is avoided and the sensitivity and accuracy of the selection method is improved.

Compared with other solvent system mathematical prediction models, the method has the advantages that although the prediction precision is reduced to some extent, the model does not need to be reconstructed aiming at the compound to be separated, the selection process is simple, and the method is more suitable for high-throughput screening of a multi-component solvent system in a complex system. Although the 9X 9 table covers a more comprehensive HEMW system than the two-point method established by wainscot, the diagonally-located solvent system [ HEMW = X (10-X): X (10-X) ] in this study is still the most commonly used part of the two-point method, considering the solubility to the sample and the span of polarity of the solvent system. In addition, in practical application, at least two shaking tests are needed to establish a specific mathematical model aiming at a certain target compound, and for multi-component separation in a traditional Chinese medicine complex system, a plurality of mathematical models need to be established.

DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION

EXAMPLE 1 establishment of mathematical model

1. Method for measuring K value

And adding the reagents in the high-speed counter-current solvent system into a separating funnel according to a ratio, fully and uniformly mixing, and standing and balancing. After full balance, precisely transferring 2mL of the upper phase and the lower phase into 10mL sample bottles by using a pipette, adding 3mg of samples, fully shaking the sample bottles to dissolve, standing, after full balance, precisely transferring 1.2mL of the upper phase and the lower phase into the 10mL sample bottles respectively by using a pipette gun, and drying by air. Precisely transferring 2mL of methanol dissolved sample after the reagent is completely volatilized, filtering into a liquid phase small bottle by adopting an organic filter head with the diameter of 0.45 mu m, and adopting the chromatography under the item of 3.2Conditions were analyzed by HPLC. The peak areas of the upper phase sample and the lower phase sample are respectively marked as A ₁ And A ₂ Then the partition coefficient K = a ₁ /A ₂ 。

2. Method for determining solvent system of standard compound

The flask test was first carried out with a solvent system HEMW = 1. If K =1, the optimal solvent system for the sample is HEMW = 1. If K ≠ 1, then the solvent system for the next experiment is selected based on the size of the K value. If K >1, selecting a low polarity HEMW = 7; if K <1, then a polar HEMW = 3. And (3) solving a linear equation of the proportion (X) of lg K and methanol or normal hexane by a two-point method according to the K value calculated by the two-time shake flask test, so as to calculate the proportion of the methanol or normal hexane in the optimal solvent system of the sample when K =1, and obtaining the optimal solvent system of the sample. Similarly, the proportion of methanol or n-hexane in the solvent system when the sample distribution coefficient K =0.25 or K =2.5 can be calculated according to a linear equation.

3. Model building

3.1 solution systems prediction

According to the determination method of the K value in the step (1) and the determination method of the solvent system of the standard compounds in the step (2), the solvent systems of 14 standard compounds such as benzoylpaeoniflorin, forsythin, arctigenin, apigenin, diosmetin, scutellarin, wogonin, curcumin, atractylenolide III, isoimperatorin, magnolol, morin, adantin aglycone, resveratrol and the like are determined, and the prediction results of the solvent systems at K =1 are verified, and the results are shown in Table 1. The predicted K values for the solvent systems are all around 1.

Table 1 two-point method for predicting solvent systems for standard compounds (K =0.25, K =1, K = 2.5)

3.2 determination of B% when sample peaks in HPLC

The chromatographic conditions are as follows: the mobile phase is water (A) -methanol (B), gradient elution 10% by weight B-100% by weight B,45min;100% by weight, B,10min. Flow rate: 1mL/min; column temperature: 30 ℃; the sample injection amount is 10 mu L;

detection wavelength: wogonin, forsythiaside, arctigenin, curcumin, magnolol, diosmetin and apigenin are all 210nm; isoimperatorin, atractylenolide III and adaptogenin are all 222nm; the barbaloin and benzoylpaeoniflorin are both 232nm; the content of morin is 270nm; resveratrol was 312nm.

The B% of the 14 standard compounds after HPLC analysis was determined by HPLC analysis using three different types of columns of Agilent Eclipse XDB-C18 (5 μm, 4.6X 250 mm), agilent extended-C18 (5 μm, 4.6X 250 mm), and Diamonsil C18 (5 μm, 4.6X 250 mm), respectively, using 3.3 HPLC peak B% determination methods. The results are shown in Table 2.

TABLE 2 HPLC analysis of the standard compounds gives the B% peak

3.3 creation of mathematical model

Liangjunling for n-hexane/EtOAc/MeOH/H ₂ O (HEMW) was plotted in a 9X 9 chart, and the average polarity parameter P 'was calculated for each solvent system in the chart, and the results are shown in FIG. 1, where it can be seen that the magnitude of the P' value can be used to characterize a ARIZNA type solvent system [ HEMW = X (10-X): X (10-X)]The polarity of the medium solvent system and the solvent components; also as shown in FIG. 2, the ratio of P' to methanol or n-hexane in the solvent system HEMW is completely linear (r ² ＝1)。

Therefore, if the standard compound is analyzed by HPLC to show a completely linear relationship between B% and P', the ratio X of methanol or n-hexane to B% in the HEMW system will necessarily be completely linear. Namely, the linear relation between X and B% can be directly analyzed to complete the construction of the solvent system prediction model. The samples were analyzed using 3 different types of C18 columns from different manufacturers to obtain three different linear equations, as shown in FIGS. 3-5.

Samples were analyzed on an Agilent Eclipse XDB-C18 (5 μm, 4.6X 250 mm) column, and a linear relationship model of X and B% was constructed with X =0.0865B% -1.4209 ² ＝0.6457；

Samples were analyzed on an Agilent extended-C18 (5 μm, 4.6X 250 mm) column, and a model of the linear relationship between X and B% was constructed as X =0.0821B% -0.8779,r ² ＝0.657；

Samples were analyzed on a Diamonsil C18 (5 μm, 4.6X 250 mm) column, and a linear model of X vs. B% was constructed with X =0.0944B% -2.1254,r ² ＝0.6438，r ² ＝0.649。

The above 3 linear models indicate that there is a strong positive correlation between X and B% (r > 0.8).

Example 2

Preparation of samples

Collecting 23.8kg of fresh Japanese metaplexis herb, cutting the Japanese metaplexis herb into 2-5 cm small sections by using scissors, immediately adding 90L of 80% ethanol, extracting for 5 times at room temperature in a leakage barrel, each time for 24 hours, combining filter liquor, filtering, recovering the solvent under reduced pressure until no alcohol smell exists, sequentially extracting by using reagents such as n-hexane, dichloromethane, ethyl acetate and the like with different polarities (1, 5 times), filtering out 87.8g of chlorophyll precipitate floating in extract liquor, recovering an organic reagent to obtain an n-hexane part (207.3 g), a dichloromethane part (28.38 g), an ethyl acetate part (22.8 g) and a water part (1850.6 g). After sampling the dichloromethane fraction 28.24g through 80-100 mesh silica gel 1, a primary separation is performed by silica gel column chromatography (silica gel: 600g, 80-100 mesh; Φ 8cm × H35 cm; column volume: 1800 mL) with a n-hexane-ethyl acetate system (100. After TLC detection of the eluent, 14 fractions were combined: LM-CH2Cl2-1 to 14. And (3) cutting a LM-CH2Cl2-11 sample into segments by adopting HSCCC, wherein a solvent system is HEMW =4.

High performance liquid chromatography

Chromatographic conditions are as follows: and (3) chromatographic column: agilent Eclipse XDB-C18 (5 μm, 4.6X 250 mm), mobile phase water (A) -methanol (B), gradient elution 10% A-100% B,45min;100% by weight, B,10min. Flow rate: 1mL/min; column temperature: 30 ℃; the sample injection amount is 10 mu L; detection wavelength: 313nm. As shown in FIG. 6, the high performance liquid chromatogram obtained was mainly focused on the peaks at 31.389min, 36.649min and 38.371min in this experiment.

The methanol ratio B% when the peak appears after HPLC column separation

The proportion B% of methanol when the compound is separated by an HPLC chromatographic column to generate a peak is greatly related to factors such as the type of the chromatographic column, the column temperature, the flow velocity of a mobile phase, the gradient slope, the initial concentration of a gradient, the delay volume of the gradient and the like, and fixed chromatographic conditions are adopted for eliminating the factors to interfere the result. According to the formula B% = (t) _R -0.8) × 2% +10%, where t is _R Compound peak time.

To obtain: methanol ratio at peak time of B% = (t) _R -0.8) × 2% +10%, the B% of the chromatographic peak at 31.389min is 71.178. Substituting B% into the linear relationship model for X =0.0874B% -1.5126, X of the chromatographic peak at 31.389min is 4.7, i.e. the solvent system is HEMW = 4.7.

The B% of the chromatographic peak at 36.649min was 81.692. Substituting B% into the linear relationship model for X =0.0874B% -1.5126, X of the chromatographic peak at 36.649min is 5.6, i.e. the solvent system is HEMW = 5.6.

The B% of the chromatographic peak at 38.371min was 85.142. Substituting B% into the linear relationship model for X =0.0874B% -1.5126, X of the chromatographic peak at 38.371min is 5.9, i.e. the solvent system is HEMW = 5.9.

Verification of predicted results

Using the method for determination of K value under 3.4, the partition coefficient of the chromatographic peak at 31.389min in solvent system HEMW =4.7 is 1.4, the partition coefficient of the chromatographic peak at 36.649min in solvent system HEMW =5.6 is 2 in 4.4.

Example 3

Sample information

Imperatorin, purchased from vkkci biotechnology limited, sikawa, lot No.: wkq16121908.

Kuhseng neo-alcohol F, purchased from vkkci biotechnology limited, sichuan, under batch No.: wkq16102703.

High performance liquid chromatography

Chromatographic conditions are as follows: a chromatographic column: agilent extended-C18 (5 μm, 4.6X 250 mm), diamonsil C18 (5 μm, 4.6X 250 mm), mobile phase water (A) -methanol (B), gradient elution 10% A-100% B,45min;100% by weight, B,10min. Flow rate: 1mL/min; column temperature: 30 ℃; the sample injection amount is 10 mu L; detection wavelength of imperatorin: 212nm, detection wavelength of kuh-seng neo-alcohol F: 300nm. The obtained high performance liquid chromatogram is shown in FIG. 7-FIG. 10. On a Diamonsil C18 chromatographic column, the peak emergence time of imperatorin is 35.667min, and the peak emergence time of kushenol F is 39.115min; on Agilent extended-C18 chromatographic column, the peak time of imperatorin is 33.805min, and the peak time of kuhseng neol F is 37.324min; the methanol ratio B% when the peak appears after HPLC column separation

In an Agilent extended-C18 (5 μm, 4.6X 250 mm) column:

according to the formula B% = (t) _R -0.8) × 2% +10%, where t is _R Compound peak time.

To obtain: methanol ratio at peak time of B% = (t) _R -0.8) × 2% +10%, the peak time of imperatorin is 33.805min, then B% is 76.01. Substituting B% into the linear relationship model for X =0.0821B% -0.8779, X for imperatorin is 5.4, i.e. the solvent system is HEMW = 5.4.

The peak time of kurarinol F is 37.324min, and B% is 83.048. Substituting B% into the linear relationship model for X =0.0821B% -0.8779, X of kuhseng neo-alcohol F is 5.9, i.e. the solvent system is HEMW = 5.9.

In a Diamonsil C18 (5 μm, 4.6X 250 mm) column:

according to the formula B% = (t) _R -0.8) × 2% +10%, where t is _R Compound peak time.

To obtain: methanol ratio at peak time of B% = (t) _R -0.8) × 2% +10%, the peak time of imperatorin is 35.667min, then B% is 79.734. Substituting B% into the linear relationship model for X =0.0944B% -2.1254, imperatorin has X of 5.4, i.e. the solvent system is HEMW = 5.4.

The peak time of kurarinol F is 39.115min, and B% is 86.63. Substituting B% into the linear relationship model for X =0.0944B% -2.1254, X of kuhseng neo-alcohol F is 6.1, i.e. the solvent system is HEMW = 6.1.

Verification of prediction results

Using the method for determining K value under 3.4, imperatorin has a K value of 2.6 in HEMW = 5.4. The K value of kuhseng neoalcohol F is 0.33 in HEMW = 5.9.

Claims (2)

1. A high throughput screening method for a high-speed counter-current chromatography solvent system is characterized by comprising the following steps:

(1) High performance liquid chromatography analysis: using a C18 chromatographic column, wherein the mobile phase A is water, and the mobile phase B is methanol; gradient elution, performing high performance liquid chromatography analysis on the standard compound;

(2) Determination of the proportion B% of methanol at the time of appearance of the Standard Compound

Methanol ratio at peak emergence of standard compound B% = (C =: (C))t _R -t）*R+B _o % of, whereint _R Is the compound peak time;tis the gradient delay time; b is _o % is the initial methanol ratio; r is gradient slope;

(3) Establishment of mathematical model

The composition of the n-hexane-ethyl acetate-methanol-water system is H EMW = X (10-X) and X (10-X); wherein the ratio X of methanol or n-hexane to the polarity parameterP'The value is completely linear; b% when the standard compound peaks after HPLC analysis andP'if the relation is completely linear, the proportion X of methanol or normal hexane in a normal hexane-ethyl acetate-methanol-water system is completely linear with B percent, namely a solvent system prediction model can be constructed by directly analyzing the linear relation of X and B percent; determining the proportion X of methanol or n-hexane when the distribution coefficient K =0.25, 1 and 2.5 of a standard compound in an n-hexane-ethyl acetate-methanol-water system by a two-point method, and establishing a linear relation between X and B% based on X and B% data of a plurality of standard compounds to obtain a solvent system prediction model formula;

(4) Determination of solvent System

Carrying out high performance liquid chromatography analysis on a compound to be detected according to the step (1); determining the ratio B% of methanol in the mobile phase when the peak of the compound to be detected appears according to the step (2); substituting B% into the prediction model formula of the solvent system to calculate X, namely the proportion of methanol or n-hexane in the n-hexane-ethyl acetate-methanol-water system, and shaking the bottle to verify the K value;

the shaking bottle result of the step (4) needs to meet 0.25-K-plus 2.5;

in the step (4), if K is less than 0.25, the predicted initial solvent system HEMW = X (10-X), wherein X is reduced by 1 in the step (10-X), namely the distribution coefficient of the target compound in the solvent system is satisfied and is in a range of 0.25-K-Ap 2.5;

if K >2.5, the predicted initial solvent system HEMW = X (10-X): X (10-X) requires that X be increased by 1 to meet the target compound partition coefficient in the solvent system of 0.25-K-Ap-2.5.

2. The screening method according to claim 1, wherein the chromatographic conditions of step (1): column temperature: 30. DEG C; the amount of sample was 10. Mu.L.

Images