A method for industrial hemp metabolite separation based on reversible boronate ester reaction

Abstract

The application belongs to the field of chemical separation and functional materials, and provides an industrial hemp metabolite extraction method based on dynamic borate reversible reaction. The application adopts a core-shell mesoporous functional adsorbent design. The adsorbent comprises a mesoporous silica core and an organic shell layer containing a pincer-type double-boron recognition cavity. The distance from boron to the boron center is 0.32-0.50 nm. The selective adsorption of ortho-diphenol type industrial hemp metabolites is realized under the condition that the volume fraction of ethanol is 85-96 vol%, and the temperature is 30-40 DEG C. The borate bond is reversibly broken by reducing the acidity and / or introducing a glycol competitive agent to complete mild desorption, realizing the performance of a selectivity coefficient of greater than or equal to 100, a recovery rate of greater than or equal to 95%, an activity retention rate of greater than or equal to 90% after 100-200 cycles, and a boron leaching amount of less than or equal to 1 ppm. The triple contradiction between the shell layer cross-linking and recognition cavity accessibility, high liquid hourly space velocity and ethanol-water microenvironment regulation, mild regeneration and site chemical integrity is solved, and the application has wide application value.

Description

Technical Field

[0001] This invention relates to the field of chemical separation and functional materials, specifically to a method for extracting industrial hemp metabolites based on a dynamic borate ester reversible reaction. Background Technology

[0002] Industrial hemp metabolites, especially ortho- and oxo-diphenol cannabidiols, have attracted much attention due to their unique physiological activities in the pharmaceutical, health product, and functional food fields. High-purity separation of these metabolites is a key prerequisite for their high-value applications. In crude extracts, target metabolites coexist with sugars, terpenes, flavonoids, and other phenols, exhibiting complex compositions and similar properties. Traditional solvent extraction, resin adsorption, or chromatographic separation methods often face challenges such as insufficient selectivity, high solvent consumption, high energy consumption, and difficulty in achieving pharmaceutical-grade purity. The ortho- and oxo-diphenol structures endow these metabolites with specific coordination chemistry properties, and designing identification and separation systems based on these structural characteristics has become an important approach to overcoming existing technological bottlenecks. Boric acid and its derivatives can form dynamically reversible borate ester bonds with ortho- and oxo-diphenols. These covalent bonds can be rapidly formed and broken under appropriate pH and solvent conditions, providing a molecular basis for constructing highly selective, recyclable, and mild separation systems. Immobilizing boric acid sites on a porous carrier and constructing a core-shell structure with microenvironment regulation through an organic shell can achieve stable recycling of the adsorbent while maintaining high selectivity. This is of great significance for promoting the large-scale preparation and industrial application of industrial hemp metabolites, and also provides a reference technical route for the separation of other natural products containing ortho-diol structures.

[0003] Although adsorption separation methods based on borate ester chemistry have been applied in the separation of sugars and catechol compounds, they still have significant limitations in the complex system of industrial hemp metabolites. For example, Chinese patent CN105312038B discloses a formylphenylboronic acid modified magnetic nanoparticle and its preparation and application (201410366030.4), but its single borate site has a limited binding constant for ortho-diols, and it is prone to non-specific adsorption in highly polar solvents, with a selectivity coefficient usually below 50, making it difficult to meet the separation requirements of complex extracts. Another example is Chinese patent CN104788602A, which discloses a phenylboronic acid modified covalent affinity ultra-high crosslinked resin and its preparation and application, but the regeneration process requires strongly alkaline conditions or high concentrations of organic acids, resulting in high energy consumption, large solvent consumption, and chemical degradation and leaching of borate sites, leading to insufficient cycle stability. Furthermore, existing technologies often employ randomly distributed single-boron structures for borate sites, lacking a design for synergistic recognition of ortho- and tho-dihydroxyl groups. They also fail to effectively regulate the dielectric constant and water activity within the adsorption microenvironment, limiting the geometric matching and thermodynamic driving force for borate ester bond formation. Simultaneously, under high solids content conditions, the pore structure and shell design of existing adsorbents fail to balance mass transfer flux and selectivity barriers, as well as the contradiction between crosslinking lock-in strength and recognition cavity accessibility. This results in fundamental limitations in achieving synergistic optimization of multiple objectives, including high selectivity, high flux, low leaching, and long cycle life. Summary of the Invention

[0004] The purpose of this invention is to provide a method for extracting industrial hemp metabolites based on the reversible reaction of dynamic borate esters, aiming to resolve the three coupling contradictions in the industrial hemp metabolite separation system based on clamp-type dual boron recognition: First, shell cross-linking and locking improve cycle stability and low leaching, but compressing the accessibility and diffusion flux of the recognition cavity makes it difficult to balance selectivity and flux; Second, enlarging the pore mass transfer and reducing the pressure drop are beneficial for high liquid hourly space velocity operation, but weaken the dielectric and water activity regulation required for borate ester formation in the ethanol-water microenvironment, reducing the coordination geometry matching for ortho-diol; Third, using CO2 and diol competing agents to achieve mild regeneration can reduce energy consumption and equipment load, but there is tension between the desorption driving force and the chemical integrity of the site, which easily leads to a conflict between high-efficiency regeneration and long-term low leaching and site density retention.

[0005] (2) Technical solution To achieve the above objectives, the present invention provides the following technical solution: A method for separating industrial hemp metabolites based on a reversible borate ester reaction, comprising the following steps: S1. Provides a core-shell mesoporous functional adsorbent, comprising a mesoporous silica core and an organic shell, wherein the organic shell contains a clamp-type dual-boron recognition cavity, and the distance between the boron centers is 0.32-0.50 nm; S2. The feed solution containing industrial hemp metabolites is contacted with the adsorbent in an alcohol-water mixed solvent with an ethanol volume fraction of 85-96 vol% at a temperature of 30-40℃, and selective adsorption is carried out under the condition of reversible formation of borate ester bonds; S3. Reduce system acidity and / or introduce diol competitors to break borate ester bonds and desorb metabolites; S4. Collect the desorption liquid and rinse the adsorbent with alcohol and water and dry it to achieve circulation; During fixed-bed operation, the liquid hourly space velocity is 0.5-2.0 h⁻¹. -1 Alternatively, the batch contact time is 30-60 minutes.

[0006] Furthermore, the feed solution is a solution of industrial hemp extract in ethanol and water, and the industrial hemp metabolites are selected from cannabidiols and their analogues; and the operating pH of the adsorption section is 7.0-7.6.

[0007] Furthermore, the desorption fractionation includes a first stage of lowering the pH of the system to 7.0-7.2 with carbon dioxide, and a second stage of further lowering the pH of the system to 6.0-6.5 or adding 1-3 wt% of a diol competitor to the solution, wherein the diol competitor is 1,2-propanediol or glycerol carbonate.

[0008] Furthermore, the method performance meets the following requirements: selectivity coefficient ≥100, recovery rate ≥95%, activity retention rate ≥90% after 100-200 cycles, and boron leaching amount ≤1ppm after every 100 cycles.

[0009] Furthermore, based on elemental boron, the total boron site density is 1.2-2.4 mmol·g. -1 The proportion of 20-60% of the sites are boron sites; and the relative site density retention rate is 90-100% after cross-linking and locking treatment.

[0010] Furthermore, the mesoporous silica core is SBA-15 or KIT-6 ordered mesoporous silica with an average pore size of 1.5-3.0 nm; and the organic shell thickness is 3-8 nm or characterized by a total organic layer mass increment of 2-8 wt%.

[0011] Furthermore, the clamp-type dual-boron recognition cavity is formed by grafting a dual-boron module with end-group functionalization. The dual-boron module is derived from 1,4-phenyldiboronic acid or 1,8-naphthalenediboronic acid and is fixed to the alkyne-containing functional segment of the organic shell through end-group coupling; the organic shell contains a polyethylene glycol methacrylate brush.

[0012] As a concept of this invention, the design of a core-shell mesoporous functional adsorbent is primarily used to enhance the selective recognition and efficient separation performance of industrial hemp metabolites. The mesoporous silica core is selected from ordered mesoporous materials such as SBA-15 or KIT-6, whose regular mesoporous channels with an average pore size of 1.5-3.0 nm provide a low-resistance mass transfer path, ensuring a liquid hourly space velocity of 0.5-2.0 h⁻¹. -1 Rapid adsorption kinetics are achieved through fixed-bed operation or batch contact within 30-60 min. The organic shell, constructed by free radical polymerization grafting of an ethanolophilic polymer and an alkyne-containing functional polymer, creates a microenvironment with high affinity for ethanol-water mixed solvents. Under conditions of 85-96 vol% ethanol, the solvation of the shell regulates the local dielectric constant and water activity, maintaining suitable coordination activity of the boric acid sites and promoting the formation of borate ester bonds. The clamp-type dual-boron recognition cavity originates from a dual-boron module functionalized with end groups of 1,4-phenylenediboric acid or 1,8-naphthalenediboric acid. It is precisely anchored to the alkyne-containing functional segment via azide-alkyne click chemistry. The boron-to-boron center distance of 0.32-0.50 nm closely matches the spatial spacing of the dihydroxyl groups in cannabidiol molecules, achieving synergistic recognition at two sites and significantly improving the binding constant and selectivity coefficient. The shell cross-linking and locking maintain a total boron site density of 1.2-2.4 mmol·g⁻¹. -1 Simultaneously, the flexible segments of the polyethylene glycol methacrylate brush maintain the dynamic accessibility of the recognition cavity, avoiding mass transfer hindrance caused by excessive cross-linking. CO2 is bubbled in to adjust the pH to 7.0-7.2 or further to 6.0-6.5, and 1-3 wt% of diol competitors 1,2-propanediol or glycerol carbonate are introduced. Under mild conditions, reversible cleavage of boronic acid ester bonds and efficient desorption of metabolites are achieved, with a recovery rate ≥95%. At the same time, chemical degradation of boronic acid sites under strong alkaline or acidic conditions is avoided, ensuring an activity retention rate ≥90% and boron leaching ≤1 ppm after 100-200 cycles. This achieves synergistic optimization of high selectivity, high throughput, low leaching, and long cycle life.

[0013] Furthermore, the preparation method of the adsorbent includes the following steps: A1. Detemplating and drying of mesoporous silica carriers with an average pore size of 1.5-3.0 nm; A2. An amino site is introduced by silanization with 3-aminopropyltriethoxysilane, and the amino site is converted into a free radical polymerization initiation site by α-bromoisobutyryl bromide; A3. Grafting alcohol-loving polymers with alkyne-containing functional segments via free radical polymerization; A4. To enable end-group coupling between the clamp-type double boron module with azide end groups and the alkyne-containing functional segment; A5. After single-boron site replacement and construction with polyethylene glycol methacrylate brushes, crosslinking and locking were implemented. The final total boron site density was 1.2–2.4 mmol·g. -1 The proportion of double boron sites is 20-60%.

[0014] Furthermore, the silanization temperature is 20-80℃ and the time is 0.5-4h; the grafting temperature of the alkyne-containing functional segment is 20-60℃ and the time is 0.5-2h; the end-group coupling temperature is 20-35℃ and the time is 4-8h; and the alcohol-loving segment polymer is selected from polyethylene glycol methacrylate with an average molecular weight of 200-600, while the alkyne-containing functional segment polymer is formed from propynyl methacrylate monomer.

[0015] As another aspect of this invention, the application of the aforementioned core-shell mesoporous functional adsorbent in the selective adsorption and desorption of industrial hemp metabolites primarily enhances the operational stability and economy of the separation process. Under preferred conditions of 85-96 vol% ethanol, pH 7.0-7.6, and temperature 30-40°C, the organic shell of the adsorbent exhibits good compatibility with the ethanol-water mixed solvent, with a moderate degree of shell solvation. This ensures both the exposure and accessibility of the clamp-type double-boron recognition cavity and maintains the steric shielding of the shell against non-target components, resulting in a selectivity coefficient ≥100. During fixed-bed operation, the liquid hourly space velocity is 0.5-2.0 h⁻¹. -1 The design balances mass transfer efficiency with equipment processing capacity, achieving high-throughput continuous operation while ensuring sufficient adsorption kinetics. The batch contact time of 30-60 minutes is suitable for small-batch or intermittent production scenarios, offering high flexibility. The desorption process employs a first stage where carbon dioxide is bubbled in to adjust the pH to 7.0-7.2, weakening the stability of borate ester bonds and initially desorbing some weakly bound metabolites. The second stage further lowers the pH to 6.0-6.5 or adds 1-3 wt% of the diol competitor 1,2-propanediol or glycerol carbonate, completely breaking the borate ester bonds through competitive coordination or further pH reduction, achieving a recovery rate ≥95%. This staged desorption strategy is completed under mild conditions, avoiding the chemical erosion of borate sites caused by traditional strong alkali or strong acid regeneration, ensuring boron leaching ≤1 ppm, maintaining a site density retention rate of 90-100%, and ensuring an activity retention rate ≥90% after 100-200 cycles. A regeneration process involving ethanol-water rinsing and 2-hour vacuum drying at 60°C is performed every 50 cycles to further remove accumulated non-specific adsorbates and maintain the long-term stability of the adsorbent. This application is not only suitable for the large-scale separation of industrial hemp metabolites, but can also be extended to the purification of other natural products containing ortho-diphenol structures, such as catechins and proanthocyanidins, showing broad industrial application prospects.

[0016] Furthermore, in the fixed-bed mode, the adsorption section of the method is cycled 100-200 times, and a regeneration procedure of ethanol-water washing and vacuum drying at 60°C for 2 hours is performed every 50 cycles.

[0017] Furthermore, the thickness of the organic shell layer of the adsorbent is maintained in the range of 3-8 nm during the cycling process, and the total organic layer mass increment is maintained in the range of 2-8 wt%.

[0018] Furthermore, the ethanol volume fraction in the feed solution is 85-96 vol%, the temperature is 30-40℃, and the batch contact time is 30-60 min.

[0019] Furthermore, the proportion of diboron sites in the total sites of the adsorbent is 20-60%, and the density of total boron sites in the adsorbent is 1.2-2.4 mmol·g. -1 .

[0020] Furthermore, desorption is performed by bubbling carbon dioxide to adjust the pH. When the pH drops to 7.0-7.2, the bubbling is stopped, and the mixture is allowed to stand for 10-30 minutes before proceeding to the second stage.

[0021] Furthermore, the diol competitor is 1,2-propanediol or glycerol carbonate, added in an amount of 1-3 wt%, based on the solution mass.

[0022] Furthermore, the mesoporous silica core is selected from SBA-15 or KIT-6, with an average pore size of 1.5-3.0 nm.

[0023] Furthermore, the end-group coupling is carried out by azide-alkyne click reaction in the presence of a copper (I) catalyst. The copper (I) catalyst is selected from at least one of the copper sulfate / sodium ascorbate reduction system, cuprous iodide, and cuprous bromide. The amount of catalyst used is 1-10 mol% of the molar amount of the azide group, the temperature is 20-35℃, and the time is 4-8 h.

[0024] Furthermore, after cross-linking and locking, the site density retention rate is 90-100%, which is calibrated by dual methods of B element XPS atomic fraction and titration.

[0025] As a senior expert in materials chemistry, the mechanism and synergistic effect of the mesoporous silica core and organic shell in core-shell mesoporous functional adsorbents of this invention are reflected in the following aspects. The mesoporous silica core, as a rigid framework, focuses on providing well-ordered mesoporous channels and high specific surface area. The two-dimensional hexagonal or three-dimensional bicontinuous mesoporous structure of SBA-15 or KIT-6 enables the pore network with an average pore size of 1.5-3.0 nm to have low tortuosity and high connectivity, ensuring that industrial hemp metabolites and coexisting components in the feed solution can rapidly diffuse to the adsorption sites, reducing mass transfer resistance, thereby achieving a liquid hourly space velocity (LHSV) of 0.5-2.0 h⁻¹ in the fixed bed. -1High-throughput adsorption is achieved under batch contact conditions of 30-60 min, supporting high liquid hourly space velocity (LISH) operation and rapid kinetic performance. The main function of the organic shell is to construct a selective recognition microenvironment and an anchored clamp-type dual-boron recognition cavity. The shell is formed by a composite of ethanolophilic polymers such as polyethylene glycol methacrylate and alkyne-containing functional polymers. The thin-layer design, with a thickness of 3-8 nm or a mass increment of 2-8 wt%, maintains unobstructed mesoporous channels while regulating the local distribution of the ethanol-water mixed solvent through the solvation of polymer segments. This maintains a suitable dielectric constant and water activity within the shell, promotes the coordination activity of borate sites and the formation of borate ester bonds, thereby enhancing the selective recognition ability of ortho-diphenol metabolites and achieving a selectivity coefficient ≥100. The synergistic effect of the mesoporous core and shell is reflected in the core providing a rapid mass transfer channel and mechanical strength, while the shell provides a selective barrier and recognition sites. The combination of the two achieves synergistic optimization of mass transfer flux and selectivity. The clamp-type dual-boron recognition cavity is derived from 1,4-phenylenediboric acid or 1,8-naphthalenediboric acid. The boron-to-boron center distance of 0.32-0.50 nm precisely matches the spacing of the ortho-dihydroxyl groups. The dual-boron sites synergistically coordinate with the ortho-dihydroxyl groups to form a stable bidentate chelate structure, with a binding constant much higher than that of single-boron sites, significantly improving the selectivity coefficient. The shell cross-linking and locking mechanism uses chemical cross-linking bonds to lock the polymer chain segments, preventing swelling and peeling of the shell and leaching of boron sites during solvent washing, ensuring boron leaching is ≤1 ppm. Simultaneously, the flexible chain segments of the polyethylene glycol methacrylate brush maintain a certain degree of freedom of chain movement within the cross-linking network, maintaining the dynamic accessibility of the recognition cavity. This ensures an activity retention rate ≥90% and a site density retention rate of 90-100% after 100-200 cycles, achieving a synergy between stability and accessibility, overcoming the contradiction between cross-linking and locking and the accessibility of the recognition cavity.

[0026] (3) Beneficial technical effects 1. Achieving High Selectivity in Recognition and Separation: Through the precise design of the clamp-type dual-boron recognition cavity, the boron-to-boron center distance of 0.32-0.50 nm is highly matched with the spacing between the two hydroxyl groups of ortho-dihydroxyl-type industrial hemp metabolites. The dual-boron sites coordinate to form stable bidentate chelate borate bonds, with a binding constant much higher than that of single-boron sites, resulting in a selectivity coefficient ≥100. This is significantly better than the selectivity coefficient of single-boronic acid site adsorbents in existing technologies, which is usually <50. This effectively distinguishes target metabolites from coexisting components such as sugars, terpenes, and flavonoids, achieving efficient enrichment of target compounds in complex extracts.

[0027] 2. Achieving high throughput and rapid kinetics: The ordered mesoporous structure of the SBA-15 or KIT-6 mesoporous silica core provides low-resistance mass transfer channels with an average pore size of 1.5-3.0 nm. Combined with a thin-layer design of 3-8 nm organic shell, the space velocity in the fixed-bed liquid is 0.5-2.0 h⁻¹. -1Rapid adsorption equilibrium can be achieved under batch contact conditions of 30-60 minutes, with mass transfer flux increased by 2-3 times compared to existing polystyrene-based boric acid affinity resins, meeting the needs of high-throughput continuous operation in industrial applications and reducing equipment investment and operating costs.

[0028] 3. Achieving mild regeneration and long-term cycle stability: A staged desorption strategy using CO2 bubbling for acid adjustment and diol as a competing agent is adopted. The borate ester bonds are gently broken within the pH range of 7.0-6.0, avoiding the chemical degradation of borate sites caused by traditional strong alkali or strong acid regeneration. The recovery rate is ≥95%, the boron leaching amount is ≤1ppm, and the activity retention rate is ≥90% and the site density retention rate is 90-100% after 100-200 cycles. This is far superior to the problem of significant activity decline after 30-50 cycles in existing technologies, significantly reducing the frequency of adsorbent replacement and operating costs.

[0029] 4. Achieving synergistic optimization of selective barrier and recognition cavity accessibility: The organic shell, through the composite design of ethanol-loving polymer segments and alkyne-containing functional polymer segments, constructs a microenvironment highly compatible with ethanol-water mixed solvents under conditions of 85-96 vol% ethanol volume fraction. This regulates the local dielectric constant and water activity to promote the formation of borate ester bonds. At the same time, the flexible segments of the polyethylene glycol methacrylate brush after shell crosslinking and locking maintain the dynamic accessibility of the recognition cavity. This overcomes the dilemma in the existing technology where excessive crosslinking leads to recognition cavity blockage or insufficient crosslinking leads to site leaching, achieving dual optimization of stability and accessibility.

[0030] 5. Wide range of applications: The method of this invention is not only applicable to the separation of industrial hemp metabolites, but can also be extended to the purification of other natural products containing ortho-diphenol structures such as catechins, proanthocyanidins, gallic acid derivatives, etc., as well as the separation of ortho-diphenol-type drug intermediates in the biopharmaceutical field. It has important application value in many fields such as natural product extraction, fine chemicals, and drug synthesis, and the technology is highly versatile. Attached Figure Description

[0031] Figure 1 The effect of total boron site density on selectivity coefficient and adsorption capacity.

[0032] Figure 2 The effect of the proportion of 2-boron sites on selectivity coefficient and adsorption capacity.

[0033] Figure 3 The effect of organic shell thickness on selectivity coefficient and adsorption capacity.

[0034] Figure 4 The effect of ethanol volume fraction on selectivity coefficient and adsorption capacity.

[0035] Figure 5 The effect of pH on selectivity and adsorption capacity in the adsorption section.

[0036] Figure 6 This is a high-resolution XPS B1s fitting image of Example 1 of the present invention (comparison between fresh sample and sample 150 cycles).

[0037] Figure 7 This is a multi-axis plot of the cyclic stability of Example 1 of the present invention, showing the adsorption capacity and boron leaching level for cycles 1–150. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0039] Example 1 I. Adsorbent Preparation A1. Pretreatment of mesoporous carriers Weigh 10.0g of SBA-15 ordered mesoporous silica (average pore size 2.2nm, specific surface area 650m²). 2 ·g -1 Commercially available), placed in a tube furnace, and heated at 5°C / min in an air atmosphere. -1 The temperature was rapidly increased to 550℃ and held at that temperature for 6 hours to remove the template agent. After cooling to room temperature, the sample was transferred to a vacuum drying oven and dried under vacuum at 120℃ for 4 hours to obtain an activated mesoporous silica support.

[0040] A2. Silanization and Initiation Site Construction The activated support was dispersed in 150 mL of anhydrous toluene (analytical grade, ≥99.5%, commercially available), and 6.5 mL of 3-aminopropyltriethoxysilane (analytical grade, ≥98%, commercially available) was added. The mixture was refluxed at 60 °C for 2.5 h under nitrogen protection. After cooling to room temperature, the mixture was filtered and washed three times each with toluene and ethanol, and then dried under vacuum to obtain the aminated support. This support was redispersed in 100 mL of anhydrous dichloromethane (analytical grade, ≥99.8%, commercially available), and a mixed solution of 4.2 mL of α-bromoisobutyryl bromide (chemically pure, ≥98%, commercially available) and 8.5 mL of triethylamine (analytical grade, ≥99%, commercially available) was slowly added dropwise over an ice bath for 30 min. After the addition was complete, the mixture was heated to room temperature and reacted for 3 h. The mixture was then filtered, washed, and dried under vacuum to obtain a support with free radical polymerization initiation sites on its surface.

[0041] A3. Polymer grafting Weigh 5.0 g of the above-mentioned initiator-modified carrier and disperse it in a mixed solvent of 80 mL anhydrous methanol (analytical grade, ≥99.9%, commercially available) and 20 mL deionized water. Add 12.0 g of polyethylene glycol methacrylate (average molecular weight 360, chemically pure, commercially available), 2.8 g of propyne methacrylate (chemically pure, ≥97%, commercially available), 120 mg of 2,2'-bipyridine (analytical grade, ≥99%, commercially available), and 60 mg of cuprous bromide (I) (chemically pure, ≥98%, commercially available). After bubbling the system with nitrogen to remove oxygen for 15 min, react it magnetically at 40 °C (300 rpm) for 1.2 h. After the reaction is complete, wash with methanol until the filtrate is colorless, and dry under vacuum to obtain the intermediate with the grafted organic layer. At this point, the total organic layer mass increase is approximately 5.2 wt%.

[0042] A4. End-base coupling fixed dual-boron module Preparation of azido-functionalized bisboron module solution: Dissolve 1.5 g of 1,4-phenyldiboronic acid (pinacol ester) (chemically pure, ≥95%, commercially available) in 50 mL of anhydrous dimethylformamide (analytical grade, ≥99.8%, commercially available), add 0.8 g of 3-azidopropylamine (chemically pure, ≥98%, commercially available) and a condensing agent, and react to obtain bisboron modules with azido end groups. Add 3.0 g of the above grafting intermediate to this solution, and then add catalysts: 150 mg of copper sulfate (analytical grade, ≥99%, commercially available) and 300 mg of sodium ascorbate (analytical grade, ≥99%, commercially available). The amount of catalyst is 5 mol% of the molar amount of the azido group. React at 25 °C with magnetic stirring (200 rpm) for 6 h, and add sodium ascorbate every 2 h to maintain the Cu(I) concentration. After the reaction, wash with dimethylformamide and methanol successively, and dry under vacuum to obtain the bisboron functionalized product.

[0043] A5. Single Boron Complementation and Crosslinking Lock-in 2.5 g of the above product was redispersed in 60 mL of methanol, and 1.2 g of phenylboronic acid (pinacol ester) (chemically pure, ≥98%, commercially available) was added to supplement single boron sites. The mixture was stirred at room temperature (150 rpm) for 4 h. Subsequently, 0.5 g of polyethylene glycol methacrylate (average molecular weight 300, chemically pure, commercially available) and 50 mg of potassium persulfate (analytical grade, ≥99%, commercially available) were added as initiators, and the mixture was reacted at 50 °C for 2 h to construct a polyethylene glycol methacrylate brush. Finally, 0.3 g of N,N'-methylenebisacrylamide (chemically pure, ≥99%, commercially available) was added as a crosslinking agent, and crosslinking was performed at 65 °C for 1.5 h to achieve immobilization. After washing and drying, a core-shell mesoporous functional adsorbent was obtained, with an organic shell thickness of approximately 5.5 nm, a total organic layer mass increment of 6.0 wt%, and a total boron site density of 1.8 mmol·g. -1(Dual calibration by titration and XPS atomic fraction), 40% of the sites are boron sites, the boron-to-boron center distance is 0.40 nm, and the site density retention rate is 95%.

[0044] II. Separation Process S1. Feed solution preparation and adsorbent loading Industrial hemp extract (containing o-cannabidiol type, commercially available) was dissolved in a 90 vol% alcohol-water mixture of ethanol (food grade, ≥95 vol%) and deionized water to prepare a feed solution. 30 g of the prepared core-shell mesoporous functional adsorbent was packed into a glass fixed bed column with an inner diameter of 2.5 cm and a height of 30 cm, resulting in a bed height of approximately 25 cm.

[0045] S2. Selective adsorption The feed solution is fed through a peristaltic pump at a liquid hourly space velocity (LISH) of 1.2 h⁻¹. -1 The adsorbent is fed through a fixed bed at a temperature controlled at 35°C (using a jacketed water bath for constant temperature circulation). The pH of the adsorption section is adjusted to 7.3 using a trace amount of sodium bicarbonate buffer. Under the alcohol-water mixed solvent and pH conditions of this embodiment, the clamp-type double boron recognition cavity on the adsorbent surface forms a coordination complex with the ortho-dihydroxy structure of ortho-dihydroxyl-type industrial hemp metabolites through reversible borate ester bonds, achieving selective adsorption. Feeding continues until the target component is detected to have penetrated through the column at the outlet (UV detection wavelength 280 nm).

[0046] S3. Desorption and staged recovery After stopping the feed, the bed is first flushed for 30 minutes with an alcohol-water mixed solvent of the same composition as the adsorption section to remove unadsorbed components. Desorption is carried out in two stages: First stage: Introduce carbon dioxide (industrial purity, ≥99.5%, commercially available) at a flow rate of 80 mL / min. -1 Bubbling is introduced into the bottom of the column to slowly lower the pH of the system to 7.1. After stopping the bubbling, the system is allowed to stand for 20 minutes, and the first part of the desorption solution is collected.

[0047] The second stage: An alcohol-water solution containing 2.0 wt% (by solution mass) of 1,2-propanediol (analytical grade, ≥99.5%, commercially available) was added to the system. 1,2-propanediol acts as a diol competitor, competing with the borate ester bond to promote the complete desorption of the target metabolite. The desorption was carried out at a liquid hourly space velocity (LISH) of 0.8 h⁻¹. -1 The second portion of the desorbed liquid is collected through the bed.

[0048] S4. Recycling After desorption, the bed was flushed sequentially with 95 vol% ethanol-water solution and 85 vol% ethanol-water solution, each flushing for one bed volume. Then, it was purged with nitrogen (industrial purity, ≥99.9%, commercially available) for 30 minutes. The bed was then transferred to a vacuum drying oven at 60°C and dried for 2 hours before the next cycle. In this embodiment, the adsorbent maintained an activity retention rate of ≥90% after 150 cycles, and the boron leaching amount was ≤1 ppm after every 100 cycles.

[0049] III. Product Characteristics The collected eluent was concentrated to obtain enriched industrial hemp metabolites. The main components of the product were o-dihydroxy cannabidiols and their analogues. The product was a pale yellow to yellow solid or semi-solid, and exhibited good solubility in an 85-96 vol% ethanol-water system.

[0050] IV. Features and Applicable Scenarios of the Implementation Examples Features: This embodiment uses a medium-pore size SBA-15 support and typical process parameters, with a moderate proportion of dual boron sites (40%), ensuring both selectivity and adsorption capacity. The fixed-bed operation mode offers good stability and is suitable for continuous production.

[0051] Applicable scenarios: Suitable for medium-scale continuous separation of industrial hemp metabolites, with good tolerance to fluctuations in feed composition and excellent cycle stability, recommended as a standard process parameter.

[0052] Example 2 I. Adsorbent Preparation A1. Pretreatment of mesoporous carriers Weigh 10.0g of KIT-6 ordered mesoporous silica (average pore size 3.0nm, specific surface area 780m²). 2 ·g -1 Commercially available), placed in a tube furnace, and heated at 5°C / min in an air atmosphere. -1 The temperature was rapidly increased to 550℃, and the template agent was removed by holding the temperature for 6 hours. After cooling to room temperature, the sample was transferred to a vacuum drying oven and dried under vacuum at 120℃ for 4 hours to obtain an activated mesoporous silica support.

[0053] A2. Silanization and Initiation Site Construction The activated support was dispersed in 150 mL of anhydrous toluene (analytical grade, ≥99.5%, commercially available), and 7.0 mL of 3-aminopropyltriethoxysilane (analytical grade, ≥98%, commercially available) was added. The mixture was refluxed at 80 °C for 4 h under nitrogen protection (high temperature and long duration treatment to fully cover the surface of the large-pore support). After cooling to room temperature, the mixture was filtered and washed three times each with toluene and ethanol, and then dried under vacuum to obtain the aminated support. The support was redispersed in 100 mL of anhydrous dichloromethane (analytical grade, ≥99.8%, commercially available), and a mixed solution of 4.5 mL of α-bromoisobutyryl bromide (chemically pure, ≥98%, commercially available) and 9.0 mL of triethylamine (analytical grade, ≥99%, commercially available) was slowly added dropwise over an ice bath for 30 min. After the addition was complete, the mixture was heated to room temperature and reacted for 3 h. The mixture was then filtered, washed, and dried under vacuum to obtain a support with free radical polymerization initiation sites on its surface.

[0054] A3. Polymer grafting Weigh 5.0 g of the above-mentioned initiator-modified support and disperse it in a mixed solvent of 80 mL anhydrous methanol (analytical grade, ≥99.9%, commercially available) and 20 mL deionized water. Add 15.0 g of polyethylene glycol methacrylate (average molecular weight 600, chemically pure, commercially available), 3.5 g of propyne methacrylate (chemically pure, ≥97%, commercially available), 150 mg of 2,2'-bipyridine (analytical grade, ≥99%, commercially available), and 75 mg of cuprous bromide (I) (chemically pure, ≥98%, commercially available). After bubbling the system with nitrogen to remove oxygen for 15 min, react it magnetically at 60 °C (300 rpm) for 2 h (higher temperature and longer time to construct a thicker shell). After the reaction is complete, wash with methanol until the filtrate is colorless, and vacuum dry to obtain the intermediate with the grafted organic layer. At this point, the total organic layer mass increase is approximately 7.5 wt%.

[0055] A4. End-base coupling fixed dual-boron module Preparation of azido-functionalized bis-boron module solution: Dissolve 2.0 g of 1,8-naphthalenediboric acid (pinacol ester) (chemically pure, ≥95%, commercially available) in 60 mL of anhydrous dimethylformamide (analytical grade, ≥99.8%, commercially available), add 1.0 g of 3-azidopropylamine (chemically pure, ≥98%, commercially available) and a condensing agent, and react to obtain bis-boron modules with azido end groups (the naphthalene ring structure makes the distance between boron to the boron center reach 0.50 nm). Add the above 3.0 g of grafting intermediate to this solution, and then add the catalyst: 200 mg of copper sulfate (analytical grade, ≥99%, commercially available) and 400 mg of sodium ascorbate (analytical grade, ≥99%, commercially available). The amount of catalyst is 6 mol% of the molar amount of the azido group. React at 30 °C with magnetic stirring (200 rpm) for 8 h (the longer coupling time ensures a high proportion of bis-boron sites), and add sodium ascorbate every 2 h to maintain the Cu(I) concentration. After the reaction, the product was washed successively with dimethylformamide and methanol, and then dried under vacuum to obtain the diboron-functionalized product.

[0056] A5. Single Boron Complementation and Crosslinking Lock-in 2.5 g of the above product was redispersed in 60 mL of methanol, and 1.5 g of phenylboronic acid (pinacol ester) (chemically pure, ≥98%, commercially available) was added to supplement single boron sites. The mixture was stirred at room temperature (150 rpm) for 4 h. Subsequently, 0.6 g of polyethylene glycol methacrylate (average molecular weight 300, chemically pure, commercially available) and 60 mg of potassium persulfate (analytical grade, ≥99%, commercially available) were added as initiator, and the mixture was reacted at 50 °C for 2 h to construct a polyethylene glycol methacrylate brush. Finally, 0.35 g of N,N'-methylenebisacrylamide (chemically pure, ≥99%, commercially available) was added as a crosslinking agent, and crosslinking was performed at 65 °C for 1.5 h to achieve immobilization. After washing and drying, a core-shell mesoporous functional adsorbent was obtained, with an organic shell thickness of approximately 8.0 nm, a total organic layer mass increment of 8.0 wt%, and a total boron site density of 2.4 mmol·g. -1 (Dual calibration by titration and XPS atomic fraction), 60% of the sites are boron sites, the boron-to-boron center distance is 0.50 nm, and the site density retention rate is 98%.

[0057] II. Separation Process S1. Feed solution preparation and adsorbent loading Industrial hemp extract (containing o-diphenol cannabidiols, commercially available) was dissolved in a 96 vol% alcohol-water mixture of ethanol (food grade, ≥95 vol%) and deionized water (high alcohol content improves the solubility of the target compound) to prepare a feed solution. 10 g of the prepared core-shell mesoporous functional adsorbent was mixed with 20 g of inert silica sand (analytical grade, commercially available) and packed into a glass column with an inner diameter of 1.5 cm and a height of 15 cm, resulting in a bed height of approximately 12 cm. Batch contact mode was used for the operation.

[0058] S2. Selective adsorption 100 mL of feed solution was added to an Erlenmeyer flask containing the adsorbent, and the flask was placed in a constant-temperature shaker at 40 °C with magnetic stirring (250 rpm) for batch contact. The pH of the adsorption section was adjusted to 7.6 with a small amount of sodium bicarbonate buffer (higher pH promotes the formation of borate ester bonds). The batch contact time was 60 min. Under the conditions of high alcohol content, high temperature, and high pH in this embodiment, the clamp-type double boron recognition cavity on the adsorbent surface forms a stable coordination complex with the ortho-dihydroxy structure of ortho-dihydroxyl industrial hemp metabolites through reversible borate ester bonds, achieving highly efficient and selective adsorption.

[0059] S3. Desorption and staged recovery After adsorption is complete, the adsorbent is separated by filtration and washed three times with an alcohol-water mixed solvent of the same composition as the adsorption section to remove unadsorbed components. Desorption is carried out in two stages: First stage: The adsorbent was redispersed in 50 mL of alcohol-water solution, and carbon dioxide (industrial grade, ≥99.5%, commercially available) was bubbled through at a flow rate of 100 mL / min. -1 Blow gas into the system until the pH slowly drops to 7.2. After stopping the blowing, let it stand for 30 minutes and then filter to collect the first part of the desorbed liquid.

[0060] Second stage: The adsorbent was redispersed in 50 mL of an alcohol-water solution containing 3.0 wt% (by solution mass) glycerol carbonate (chemically pure, ≥99%, commercially available) (glycerol carbonate has both diol and carbonate competing functions). The mixture was magnetically stirred (200 rpm) at 35 °C for 45 min. The second part of the desorption solution was collected by filtration and the two parts of the desorption solution were combined.

[0061] S4. Recycling After desorption, the adsorbent was washed three times each with 95 vol% ethanol-water solution and 85 vol% ethanol-water solution, then purged with nitrogen (industrial grade, ≥99.9%, commercially available) for 30 min, and then dried in a vacuum drying oven at 60°C for 2 h before the next cycle. In this embodiment, the adsorbent retained ≥90% of its activity after 200 cycles, and the boron leaching amount was ≤1 ppm after every 100 cycles.

[0062] III. Product Characteristics The collected eluent was concentrated to obtain highly purified and enriched industrial hemp metabolites. The main components of the product were o-dihydroxy cannabidiols and their analogues. The product was a yellow solid with good solubility in an 85-96 vol% ethanol-water system.

[0063] IV. Features and Applicable Scenarios of the Implementation Examples Features: This embodiment utilizes a large-pore KIT-6 carrier, the thickest organic shell (8 nm), and the highest proportion of double boron sites (60%), with the boron-to-boron center distance reaching the upper limit of 0.50 nm. Batch operation was conducted under conditions of high alcohol content (96 vol%), high temperature (40 °C), high pH (7.6), and longest contact time (60 min) to fully verify the feasibility of the limiting process parameters.

[0064] Applicable scenarios: Suitable for the fine separation of high-purity, small-batch industrial hemp metabolites, occasions requiring extremely high selectivity of target substances, and research on the performance limits of adsorbents.

[0065] Example 3 I. Adsorbent Preparation A1. Pretreatment of mesoporous carriers Weigh 10.0g of SBA-15 ordered mesoporous silica (average pore size 1.5nm, specific surface area 550m²). 2 ·g -1 Commercially available), placed in a tube furnace, and heated at 5°C / min in an air atmosphere. -1 The temperature was rapidly increased to 550℃, and the template agent was removed by holding the temperature for 6 hours. After cooling to room temperature, the sample was transferred to a vacuum drying oven and dried under vacuum at 120℃ for 4 hours to obtain an activated mesoporous silica support.

[0066] A2. Silanization and Initiation Site Construction The activated support was dispersed in 150 mL of anhydrous toluene (analytical grade, ≥99.5%, commercially available), and 5.5 mL of 3-aminopropyltriethoxysilane (analytical grade, ≥98%, commercially available) was added. The mixture was reacted under nitrogen protection at 20 °C with magnetic stirring (300 rpm) for 0.5 h (low-temperature short-time treatment to construct a thinner functional layer). After cooling to room temperature, the mixture was filtered and washed three times each with toluene and ethanol, and then dried under vacuum to obtain the aminated support. This support was redispersed in 100 mL of anhydrous dichloromethane (analytical grade, ≥99.8%, commercially available), and a mixed solution of 3.5 mL of α-bromoisobutyryl bromide (chemically pure, ≥98%, commercially available) and 7.0 mL of triethylamine (analytical grade, ≥99%, commercially available) was slowly added dropwise over an ice bath for 30 min. After the addition was complete, the mixture was heated to room temperature and reacted for 3 h. The mixture was then filtered, washed, and dried under vacuum to obtain a support with free radical polymerization initiation sites on its surface.

[0067] A3. Polymer grafting Weigh 5.0 g of the above-mentioned initiator-modified support and disperse it in a mixed solvent of 80 mL anhydrous methanol (analytical grade, ≥99.9%, commercially available) and 20 mL deionized water. Add 8.0 g of polyethylene glycol methacrylate (average molecular weight 200, chemically pure, commercially available), 1.8 g of propyne methacrylate (chemically pure, ≥97%, commercially available), 80 mg of 2,2'-bipyridine (analytical grade, ≥99%, commercially available), and 40 mg of cuprous bromide (I) (chemically pure, ≥98%, commercially available). After bubbling the system with nitrogen to remove oxygen for 15 min, react it magnetically at 20 °C (300 rpm) for 0.5 h (low temperature and short reaction time to construct a thinner shell layer). After the reaction is complete, wash with methanol until the filtrate is colorless, and vacuum dry to obtain the intermediate with the grafted organic layer. At this point, the total organic layer mass increase is approximately 2.5 wt%.

[0068] A4. End-base coupling fixed dual-boron module Preparation of azido-functionalized bis-boron module solution: 1.0 g of 1,4-phenyldiboronic acid (pinacol ester) (chemically pure, ≥95%, commercially available) was dissolved in 40 mL of anhydrous dimethylformamide (analytical grade, ≥99.8%, commercially available) after side chain shortening modification. 0.5 g of 3-azidopropylamine (chemically pure, ≥98%, commercially available) and a condensing agent were added, and the reaction yielded a short-chain bis-boron module with azido end groups (the distance between boron to the boron center was reduced to 0.32 nm). 3.0 g of the above grafting intermediate was added to this solution, followed by catalysts: 50 mg of copper sulfate (analytical grade, ≥99%, commercially available) and 100 mg of sodium ascorbate (analytical grade, ≥99%, commercially available). The catalyst dosage was 1 mol% of the molar amount of the azido group (minimum catalyst dosage). The reaction was carried out at 20 °C with magnetic stirring (200 rpm) for 4 h, during which sodium ascorbate was added every 2 h to maintain the Cu(I) concentration. After the reaction, the product was washed successively with dimethylformamide and methanol, and then dried under vacuum to obtain the diboron-functionalized product.

[0069] A5. Single Boron Complementation and Crosslinking Lock-in 2.5 g of the above product was redispersed in 60 mL of methanol, and 0.8 g of phenylboronic acid (pinacol ester) (chemically pure, ≥98%, commercially available) was added to supplement single boron sites (to maintain a low double boron content of 20%). The mixture was stirred at room temperature (150 rpm) for 4 h. Subsequently, 0.3 g of polyethylene glycol methacrylate (average molecular weight 200, chemically pure, commercially available) and 30 mg of potassium persulfate (analytical grade, ≥99%, commercially available) were added as initiator, and the mixture was reacted at 50 °C for 2 h to construct a polyethylene glycol methacrylate brush. Finally, 0.2 g of N,N'-methylenebisacrylamide (chemically pure, ≥99%, commercially available) was added as a crosslinking agent and crosslinked at 65 °C for 1.5 h to achieve immobilization. After washing and drying, a core-shell mesoporous functional adsorbent was obtained, with an organic shell thickness of approximately 3.0 nm, a total organic layer mass increment of 2.0 wt%, and a total boron site density of 1.2 mmol·g. -1 (Dual calibration by titration and XPS atomic fraction), 20% of the sites are boron sites, the boron-to-boron center distance is 0.32 nm, and the site density retention rate is 90%.

[0070] II. Separation Process S1. Feed solution preparation and adsorbent loading Industrial hemp extract (containing o-cannabidiols, commercially available) was dissolved in an 85 vol% alcohol-water mixture of ethanol (food grade, ≥95 vol%, commercially available) and deionized water (lower alcohol content enhances aqueous solubility) to prepare a feed solution. 40 g of the prepared core-shell mesoporous functional adsorbent was packed into a glass fixed bed column with an inner diameter of 2.5 cm and a height of 35 cm, resulting in a bed height of approximately 28 cm.

[0071] S2. Selective adsorption The feed solution is pumped through a peristaltic pump at a liquid hourly space velocity of 0.5 h⁻¹. -1 (Minimum liquid hourly space velocity) The adsorbent ascends through a fixed bed, with the temperature controlled at 30°C (lower temperatures extend residence time). The pH of the adsorption section is adjusted to 7.0 (lower pH limit) using a trace amount of sodium bicarbonate buffer. Under the conditions of low alcohol content, low temperature, and low pH in this embodiment, the clamp-type double boron recognition cavity on the adsorbent surface forms a coordination complex with the ortho-dihydroxy structure of ortho-dihydroxyl-type industrial hemp metabolites through reversible borate ester bonds, achieving selective adsorption. Feeding continues until the target component is detected to have penetrated through the column (UV detection wavelength 280 nm).

[0072] S3. Desorption and staged recovery After stopping the feed, the bed is first flushed for 30 minutes with an alcohol-water mixed solvent of the same composition as the adsorption section to remove unadsorbed components. Desorption is carried out in two stages: First stage: Introduce carbon dioxide (industrial purity, ≥99.5%, commercially available) at a flow rate of 60 mL / min. -1Bubbling into the bottom of the column to slowly lower the pH of the system to 7.0 (only a slight decrease in pH), stopping the bubbling and letting it stand for 10 minutes (the minimum standing time), then collecting the first portion of the desorbed solution.

[0073] Second stage: Add an alcohol-water solution containing 1.0 wt% (based on solution mass, minimum competitor concentration) of 1,2-propanediol (analytical grade, ≥99.5%, commercially available) to the system at a liquid hourly space velocity (LISH) of 0.8 h⁻¹. -1 The second portion of the desorbed liquid is collected through the bed.

[0074] S4. Recycling After desorption, the bed was flushed sequentially with 95 vol% ethanol-water solution and 85 vol% ethanol-water solution, each flushing for one bed volume. Then, it was purged with nitrogen (industrial purity, ≥99.9%, commercially available) for 30 minutes. The bed was then transferred to a vacuum drying oven at 60°C and dried for 2 hours before the next cycle. In this embodiment, the adsorbent maintained an activity retention rate of ≥90% after 100 cycles, and the boron leaching amount was ≤1 ppm after every 100 cycles.

[0075] III. Product Characteristics The collected eluent was concentrated to obtain enriched industrial hemp metabolites. The main components of the product were o-dihydroxy cannabidiols and their analogues. The product was a pale yellow solid or semi-solid, exhibiting good solubility in an 85-96 vol% ethanol-water system.

[0076] IV. Features and Applicable Scenarios of the Implementation Examples Features: This embodiment utilizes a small-pore SBA-15 carrier (1.5 nm), a thinnest organic shell (3 nm), and the lowest percentage of double boron sites (20%), with a boron-to-boron center distance reaching a lower limit of 0.32 nm. Fixed-bed operation was performed at low alcohol content (85 vol%), low temperature (30 °C), low pH (7.0), and the lowest liquid hourly space velocity (0.5 h⁻¹). -1 The experiment was conducted under specific conditions to fully verify the feasibility of the extreme low-parameter process and demonstrate that effective separation can still be achieved under the condition of low site density in the thin shell layer.

[0077] Applicable scenarios: Suitable for cost-sensitive applications, simplified adsorbent preparation process, minimized organic layer loading, suitable for low feed concentration or occasions where adsorption capacity requirements are not high but circular economy is important.

[0078] Example 4 Example 4 (Optimized Hybrid Configuration) I. Adsorbent Preparation A1. Pretreatment of mesoporous carriers Weigh 10.0g of KIT-6 ordered mesoporous silica (average pore size 2.5nm, specific surface area 720m²).2 ·g -1 Commercially available), placed in a tube furnace, and heated at 5°C / min in an air atmosphere. -1 The temperature was rapidly increased to 550℃, and the template agent was removed by holding the temperature for 6 hours. After cooling to room temperature, the sample was transferred to a vacuum drying oven and dried under vacuum at 120℃ for 4 hours to obtain an activated mesoporous silica support.

[0079] A2. Silanization and Initiation Site Construction The activated support was dispersed in 150 mL of anhydrous toluene (analytical grade, ≥99.5%, commercially available), and 6.8 mL of 3-aminopropyltriethoxysilane (analytical grade, ≥98%, commercially available) was added. The mixture was refluxed at 70 °C for 3 h under nitrogen protection. After cooling to room temperature, the mixture was filtered and washed three times each with toluene and ethanol, and then dried under vacuum to obtain the aminated support. This support was redispersed in 100 mL of anhydrous dichloromethane (analytical grade, ≥99.8%, commercially available), and a mixed solution of 4.3 mL of α-bromoisobutyryl bromide (chemically pure, ≥98%, commercially available) and 8.6 mL of triethylamine (analytical grade, ≥99%, commercially available) was slowly added dropwise over an ice bath for 30 min. After the addition was complete, the mixture was heated to room temperature and reacted for 3 h. The mixture was then filtered, washed, and dried under vacuum to obtain a support with free radical polymerization initiation sites on its surface.

[0080] A3. Polymer grafting Weigh 5.0 g of the above-mentioned initiator-modified support and disperse it in a mixed solvent of 80 mL anhydrous methanol (analytical grade, ≥99.9%, commercially available) and 20 mL deionized water. Add 13.5 g of polyethylene glycol methacrylate (average molecular weight 400, chemically pure, commercially available), 3.0 g of propyne methacrylate (chemically pure, ≥97%, commercially available), 130 mg of 2,2'-bipyridine (analytical grade, ≥99%, commercially available), and 65 mg of cuprous bromide(I) (chemically pure, ≥98%, commercially available). After bubbling the system with nitrogen to remove oxygen for 15 min, react it magnetically at 50 °C (300 rpm) for 1.5 h. After the reaction is complete, wash with methanol until the filtrate is colorless, and dry under vacuum to obtain the intermediate with the grafted organic layer. At this point, the total organic layer mass increase is approximately 6.5 wt%.

[0081] A4. End-base coupling fixed dual-boron module Preparation of azido-functionalized bisboron module solution: 1.6 g of 1,4-phenyldiboronic acid (pinacol ester) (chemically pure, ≥95%, commercially available) was dissolved in 50 mL of anhydrous dimethylformamide (analytical grade, ≥99.8%, commercially available). 0.85 g of 3-azidopropylamine (chemically pure, ≥98%, commercially available) and a condensing agent were added, and the reaction was carried out to obtain bisboron modules with azido-terminal groups (the distance between boron to the boron center is approximately 0.42 nm). 3.0 g of the above grafting intermediate was added to this solution, followed by a catalyst: 100 mg of cuprous iodide (chemically pure, ≥98%, commercially available). The amount of catalyst was 3 mol% of the molar amount of the azido groups. The reaction was carried out at 28 °C with magnetic stirring (200 rpm) for 7 h under nitrogen protection. After the reaction, the product was washed successively with dimethylformamide and methanol, and then dried under vacuum to obtain the bisboron functionalized product.

[0082] A5. Single Boron Complementation and Crosslinking Lock-in 2.5 g of the above product was redispersed in 60 mL of methanol, and 1.3 g of phenylboronic acid (pinacol ester) (chemically pure, ≥98%, commercially available) was added to supplement single boron sites. The mixture was stirred at room temperature (150 rpm) for 4 h. Subsequently, 0.55 g of polyethylene glycol methacrylate (average molecular weight 300, chemically pure, commercially available) and 55 mg of potassium persulfate (analytical grade, ≥99%, commercially available) were added as initiator, and the mixture was reacted at 50 °C for 2 h to construct a polyethylene glycol methacrylate brush. Finally, 0.32 g of N,N'-methylenebisacrylamide (chemically pure, ≥99%, commercially available) was added as a crosslinking agent, and crosslinking was performed at 65 °C for 1.5 h to achieve immobilization. After washing and drying, a core-shell mesoporous functional adsorbent was obtained, with an organic shell thickness of approximately 6.5 nm, a total organic layer mass increment of 7.0 wt%, and a total boron site density of 2.0 mmol·g. -1 (Dual calibration by titration and XPS atomic fraction), 50% of the sites are boron sites, the boron-to-boron center distance is 0.42 nm, and the site density retention rate is 96%.

[0083] II. Separation Process S1. Feed solution preparation and adsorbent loading Industrial hemp extract (containing o-dihydroxycannabidiol, commercially available) was dissolved in a 92 vol% alcohol-water mixture of ethanol (food grade, ≥95 vol%) and deionized water to prepare a feed solution. 35 g of the prepared core-shell mesoporous functional adsorbent was packed into a glass fixed bed column with an inner diameter of 2.5 cm and a height of 32 cm, resulting in a bed height of approximately 26 cm.

[0084] S2. Selective adsorption The feed solution is pumped through a peristaltic pump at a liquid hourly space velocity of 2.0 h⁻¹. -1(Maximum liquid hourly space velocity) The adsorbent ascends through a fixed bed, with the temperature controlled at 38°C. The pH of the adsorption section is adjusted to 7.4 using a trace amount of sodium bicarbonate buffer. Under the high liquid hourly space velocity conditions of this embodiment, the dynamic adsorption performance of the adsorbent at high flow rates is verified. The clamp-type double boron recognition cavity on the adsorbent surface forms a coordination complex with the ortho-dihydroxy structure of ortho-dihydroxyl-type industrial hemp metabolites through reversible borate ester bonds, achieving selective adsorption. Feeding continues until the target component is detected to have penetrated through the column (UV detection wavelength 280 nm).

[0085] S3. Desorption and staged recovery After stopping the feed, the bed is first flushed for 30 minutes with an alcohol-water mixed solvent of the same composition as the adsorption section to remove unadsorbed components. Desorption is carried out in two stages: First stage: Introduce carbon dioxide (industrial grade, ≥99.5%, commercially available) at a flow rate of 90 mL / min. -1 Bubbling is introduced into the bottom of the column to slowly lower the pH of the system to 7.1. After stopping the bubbling, the system is allowed to stand for 25 minutes, and the first portion of the desorption solution is collected.

[0086] Second stage: Add an alcohol-water solution containing 2.5 wt% (by solution mass) glycerol carbonate (chemically pure, ≥99%, commercially available) to the system at a liquid hourly space velocity (LISH) of 0.8 h⁻¹. -1 The second portion of the desorbed liquid is collected through the bed.

[0087] S4. Recycling After desorption, the bed was rinsed sequentially with 95 vol% ethanol-water solution and 85 vol% ethanol-water solution, each for one bed volume. Then, it was purged with nitrogen (industrial purity, ≥99.9%, commercially available) for 30 minutes. The bed was then transferred to a vacuum drying oven at 60°C for 2 hours before the next cycle. In this embodiment, the adsorbent maintained an activity retention rate of ≥90% after 180 cycles, and the boron leaching amount was ≤1 ppm after every 100 cycles. A regeneration procedure of ethanol-water rinsing and 60°C vacuum drying for 2 hours was performed every 50 cycles.

[0088] III. Product Characteristics The collected eluent was concentrated to obtain enriched industrial hemp metabolites. The main components of the product were o-dihydroxy cannabidiols and their analogues. The product was a yellow solid with good solubility in an 85-96 vol% ethanol-water system.

[0089] IV. Features and Applicable Scenarios of the Implementation Examples Features: This embodiment employs a medium-to-large pore size KIT-6 support (2.5 nm), a medium-thickness organic shell (6.5 nm), and a medium-to-high proportion of diboron sites (50%), achieving a maximum liquid hourly space velocity (2.0 h⁻¹). -1The system was operated under the following conditions to verify the dynamic performance of the adsorbent under high-throughput production conditions. A silanization temperature of 70℃ and a grafting temperature of 50℃ represent an optimized combination of medium and high temperatures. The end-group coupling was performed using a cuprous iodide catalyst to verify the applicability of different catalytic systems.

[0090] Applicable scenarios: Suitable for large-scale industrial continuous production, high-throughput processing requirements, and applications with high requirements for production efficiency and adsorbent utilization. The periodic regeneration process ensures long-term stable operation.

[0091] Comparative Example 1: Basically the same as Example 1, except that the mesoporous carrier used is MCM-41 (average pore size 2.0 nm, hexagonal ordered mesoporous structure, specific surface area 800 m²). 2 ·g -1 In Example 1, SBA-15 was used, while the amounts of other components and preparation conditions remained unchanged.

[0092] Comparative Example 2: It is basically the same as Example 1, except that the volume fraction of ethanol under the borate bond formation conditions is 80 vol%, and the amounts of other components and preparation conditions remain unchanged.

[0093] Comparative Example 3: It is basically the same as Example 1, except that the pH of the adsorption section is 6.5, and the amount of other components and preparation conditions remain unchanged.

[0094] Comparative Example 4: Basically the same as Example 1, except that the liquid hourly space velocity in the fixed bed is 3.0 h⁻¹. -1 The amounts of other components and preparation conditions remain unchanged.

[0095] Comparative Example 5: Basically the same as Example 1, except that the total boron site density is 0.8 mmol·g. -1 This was achieved by reducing the amount of the double boron module in step A4 to 1.0 g and correspondingly reducing the amount of the single boron filler in step A5 to 0.5 g, while keeping other preparation conditions unchanged.

[0096] Comparative Example 6: Basically the same as Example 1, except that the proportion of double boron sites is 10% (achieved by increasing the amount of phenylboronic acid used for single boron supplementation to 2.0g in step A5 and reducing the amount of double boron module used in step A4 to 0.8g, while other preparation conditions remain unchanged).

[0097] Comparative Example 7: Basically the same as Example 1, except that the polymer grafting temperature in step A3 is 10°C and the grafting time is extended to 3 hours to compensate for the low temperature reaction rate, while other preparation conditions remain unchanged.

[0098] Comparative Example 8: It is basically the same as Example 1, except that the end-group coupling in step A4 is carried out by a thermally promoted reaction instead of Cu(I) catalysis. The reaction is carried out at 80°C without a catalyst for 12 hours, while other preparation conditions remain unchanged.

[0099] Comparative Example 9: It is basically the same as Example 1, except that the desorption stage uses a single-step hydrochloric acid adjustment to pH 4.0 for direct desorption, without a first stage and a second stage, and does not use a diol competing agent. Other operating conditions remain unchanged.

[0100] Comparative Example 10: Basically the same as Example 1, except that the crosslinking and locking treatment is omitted in step A5, and only the single boron filling and polyethylene glycol methacrylate brush construction are performed. The crosslinking agent N,N'-methylenebisacrylamide is not added, and other preparation conditions remain unchanged.

[0101] Comparative Example 11: Basically the same as Example 1, except that the thickness of the organic shell layer is 12 nm. This was achieved by increasing the amount of polyethylene glycol methacrylate to 18 g and propynyl methacrylate to 4.0 g and extending the grafting time to 2.5 h in step A3. The total increase in the mass of the organic layer was about 12 wt%, and other preparation conditions remained unchanged.

[0102] Comparative Example 12: Basically the same as Example 1, except that the adsorption section operating temperature is 20°C, while the amounts of other components and operating conditions remain unchanged.

[0103] Comparative Example 13: It is basically the same as Example 1, except that the nitrogen purging step is omitted during the regeneration process. After desorption, it is directly rinsed with alcohol-water solution and then air-dried at room temperature for 24 hours instead of vacuum drying at 60°C for 2 hours. Other operating conditions remain unchanged.

[0104] Comparative Example 14: It is basically the same as Example 1, except that 3-mercaptopropyltrimethoxysilane is used instead of 3-aminopropyltriethoxysilane for silanization in step A2. Subsequently, polymerization initiation sites are introduced through a mercapto-olefin click reaction (specifically, mercapto and allyl methacrylate are reacted under ultraviolet light for 2 hours at room temperature). Other preparation conditions remain unchanged.

[0105] Performance testing: Experiment 1: Determination of Selectivity Coefficient Test Subject: The selective separation ability of a core-shell mesoporous functional adsorbent for the separation of ortho- and non-ortho-diphenol metabolites from industrial hemp extract. Test Objective: To evaluate the selectivity coefficient of the adsorbent, verify the specific recognition ability of the clamp-type double boron recognition cavity for ortho-diphenol structures, and ensure a selectivity coefficient ≥ 100. Test Principle: Based on the specific reversible coordination of boron ester bonds to ortho-diphenol structures, the selectivity coefficient is calculated by comparing the partition coefficients of ortho-diphenol target compounds and non-ortho-diphenol impurities, α = (q1 / c1) / (q2 / c2), where q is the adsorption amount, c is the equilibrium concentration, subscript 1 represents ortho-diphenol target compounds, and subscript 2 represents non-ortho-diphenol impurities. Experimental Method: Prepare a mixed standard solution containing ortho-diphenol cannabidiol (CBD) and non-ortho-diphenol tetrahydrocannabinol (THC) (100 mg / L each). -1A mixture of 50 mg of adsorbent and 50 mL of ethanol-water (90 vol%) was batch-contaminated at 35 °C and pH 7.3 for 1 h until equilibrium was reached. After separating the adsorbent, the equilibrium concentrations of CBD and THC in the solution were determined by HPLC, and the selectivity coefficients were calculated. Chromatographic conditions: C18 column (250 mm × 4.6 mm, 5 μm), mobile phase: methanol-0.1% phosphoric acid water (85:15, v / v), flow rate: 1.0 mL·min -1 The detection wavelength was 228 nm, and the column temperature was 30℃. Standards followed: GB / T 20769 or similar liquid chromatography standards. Key parameters: adsorption temperature 35℃, pH 7.3, contact time 1 h, solid-liquid ratio 1 g·L⁻¹. -1 Data processing: Three parallel measurements were performed, and the average selectivity coefficient and standard deviation were calculated. The criterion was α ≥ 100.

[0106] Experiment 2: Determination of Adsorption-Desorption Cycle Stability and Boron Leaching Amount Test Subjects: Activity retention rate and boron leaching amount of core-shell mesoporous functional adsorbent after 100-200 adsorption-desorption cycles. Test Objectives: To evaluate the cyclic stability of the adsorbent after cross-linking and locking, verify the site chemical integrity and low boron leaching characteristics during long-term operation, and ensure that the activity retention rate is ≥90% after 100-200 cycles and the boron leaching amount is ≤1 ppm after every 100 cycles. Test Principle: The trend of adsorption capacity change is monitored by repeated adsorption-desorption-regeneration cycles. The activity retention rate is defined as the percentage of the adsorption capacity in the Nth cycle relative to the adsorption capacity in the first cycle. Boron leaching amount is determined by ICP-MS to measure the total boron concentration in the desorption and rinsing solutions. Experimental Methods: 2g of adsorbent is packed into a fixed bed column with an inner diameter of 1cm, using the process conditions of Example 1 (feed concentration 150 mg·L⁻¹). -1 CBD, liquid hourly space velocity 1.2h -1 The adsorption was continuously cyclic at 35℃ and pH 7.3. Each cycle included adsorption to the breakthrough point, CO2 acidification and fractional desorption, diol competitor desorption, alcohol-water rinsing, nitrogen purging, and regeneration by vacuum drying at 60℃ for 2 hours. Every 50 cycles, samples were taken to determine the adsorption capacity, and the desorbed and rinsed solutions were combined and the boron concentration was determined by ICP-MS. ICP-MS test conditions: RF power 1550W, carrier gas flow rate 1.0 L / min. -1 Detection of boron isotopes 11B. Quantitative analysis was performed using a standard curve method (boron standard solution concentration range 0.1-100 ppb). Standards followed: Adsorption capacity was determined according to GB / T32669; boron element determination followed GB 5009.275 or similar ICP-MS standard methods. Key parameters: 150 cycles, sampling frequency every 50 cycles, ICP-MS detection limit <0.1 ppb. Data processing: Adsorption capacity-cycle count curves were plotted, and the activity retention rate at the 150th cycle and the boron leaching amount per 100 cycles were calculated. Three samples were measured in parallel.

[0107] Experiment 3: Adsorption Kinetics and Isotherm Measurement Test Subject: Adsorption kinetics and adsorption isotherms of core-shell mesoporous functional adsorbents for o-biphenol-type CBD. Test Objective: To evaluate the mass transfer rate and equilibrium adsorption capacity of the adsorption process, elucidate the diffusion-controlling step, and verify the influence of organic shell thickness (3-8 nm range) on the accessibility of the recognition cavity. Test Principle: Adsorption kinetics describes the change in adsorption amount over time; a pseudo-first-order or pseudo-second-order kinetic model is fitted to determine the rate constant. The adsorption isotherm describes the relationship between equilibrium adsorption amount and equilibrium concentration; a Langmuir or Freundlich model is fitted to determine the maximum adsorption capacity and adsorption constant. Experimental Method: Kinetic Experiment: An initial CBD concentration of 100 mg·L⁻¹ was prepared. -1 A 90 vol% ethanol-water solution was prepared. 50 mg of adsorbent was batch-contaminated with 50 mL of the solution at 35 °C, pH 7.3, and magnetic stirring at 300 rpm. Samples (0.5 mL each time, immediately filtered) were taken at 5, 10, 15, 30, 60, 120, and 180 min. The CBD concentration was determined by HPLC, and the adsorption capacity q at each time point was calculated. t =(c0-c t V / m. Isotherm experiment: CBD initial concentration gradients of 20, 50, 100, 200, 400, and 600 mg·L⁻¹ were prepared. -1 For a series of solutions, 50 mg of adsorbent was batch-contaminated with 50 mL of solutions of various concentrations at 35 °C and pH 7.3 for 2 h until equilibrium was reached. The equilibrium concentration c_e was determined by HPLC, and the equilibrium adsorption capacity q was calculated. e The pseudo-second-order kinetic equation and Langmuir isotherm equation were fitted using Origin software. Standards followed: Adsorption isotherm determination principle in GB / T 21650.3; kinetic and isotherm models referenced the IUPAC Adsorption Data Processing Guidelines. Key parameters: temperature 35℃, pH 7.3, stirring rate 300 rpm, solid-liquid ratio 1 g·L⁻¹. -1 Data processing: Plotting q t-t Curve and q e -c eCurves were used to calculate the pseudo-second-order kinetic rate constant k2 and the Langmuir maximum adsorption capacity q. max With adsorption constant K L Evaluation of R 2 A value >0.95 indicates a good fit; the average value is taken from three parallel measurements.

[0108] Experiment 4: Pore structure characterization and BET specific surface area determination Test Subjects: Pore size distribution, specific surface area, and pore volume of core-shell mesoporous functional adsorbents. Test Objectives: To characterize the retention of the mesoporous structure after organic shell grafting, verify the average pore size range of 1.5-3.0 nm and specific surface area, and evaluate the effect of shell thickness (3-8 nm) on pore blockage. Test Principle: The nitrogen adsorption isotherm of the adsorbent was determined at 77 K using the nitrogen adsorption-desorption method. The specific surface area was calculated according to the BET equation, and the pore size distribution was calculated from the desorption branch using the BJH model. The pore volume was converted from the adsorption amount at P / P0=0.99. Experimental Methods: Approximately 100 mg of sample was placed in a sample tube and tested using a fully automated specific surface area and porosity analyzer (such as Micromeritics ASAP 2020 or a similar instrument). Sample Pretreatment: Degassing was performed at 200℃ under vacuum for 4 h to remove surface adsorbed water and residual organic solvents. Test conditions: Liquid nitrogen bath temperature 77K, relative pressure P / P0 range 0.01-0.99, equilibrium time ≥10s per point. BET specific surface area was calculated using a linear range of P / P0 = 0.05-0.30, and BJH pore size distribution was calculated using the desorption branch, with the average pore size defined as 4V / S (V is pore volume, S is BET specific surface area). The pore structure parameters of the original mesoporous silica support (after template removal) and the functionalized adsorbent were compared and determined. Standards: GB / T 21650.2 and GB / T 21650.3. Key parameters: Degassing temperature 200℃, time 4h, test temperature 77K, BET calculation range P / P0 = 0.05-0.30, BJH model corrected using Harkins-Jura thickness curve. Data processing: Output BET specific surface area (m²). 2 ·g -1 BJH cumulative pore volume (cm³) 3 ·g -1 The average pore size (nm) and pore size distribution curves were compared to the percentage changes in parameters of the original support and the functionalized adsorbent. The average value of two parallel measurements was taken.

[0109] Experiment 5: Desorption efficiency and recovery rate determination Test Subject: Desorption efficiency and total recovery rate of adsorbed catechol metabolites by a staged desorption process (CO2 acidification + diol competing agent). Test Objective: To evaluate the completeness of desorption under mild regeneration conditions, verify the synergistic desorption effect of the first stage (pH reduced to 7.0-7.2) and the second stage (1-3 wt% diol competing agent), and ensure a recovery rate ≥95%. Test Principle: Desorption efficiency is defined as the percentage of the target substance mass desorbed relative to the total adsorbed mass. By collecting the eluent stepwise and measuring the desorption amount at each stage, the necessity of staged desorption and the contribution of the diol competing agent are evaluated. Experimental Method: 1 g of adsorbent that has completed adsorption (pre-treated at 100 mg / L...) was used. -1 (Adsorption to saturation in CBD solution), proceed with the staged desorption procedure of Example 1. First stage: Introduce CO2 (flow rate 80 mL / min). -1 First stage: Adjust the pH of 50 mL of alcohol-water solution from 7.3 to 7.1, let it stand for 20 min, separate the solution, and determine the CBD concentration and volume using HPLC to calculate the desorption amount D1. Second stage: Add 50 mL of alcohol-water solution containing 2 wt% 1,2-propanediol to the adsorbent, stir at 35℃ for 45 min, separate the solution, and determine the desorption amount D2 using HPLC. Calculate the desorption efficiency of the first stage η1 = D1 / Q_total × 100%, the desorption efficiency of the second stage η2 = D2 / Q_total × 100%, and the total desorption efficiency η_total = (D1 + D2) / Q_total × 100%. A comparative experiment was used: single-step hydrochloric acid adjustment to pH 4.0 desorption was used as a control. Recovery rate determination: Simulate continuous operation of a fixed bed, collect all the desorbed liquid, concentrate it, weigh it, and determine the CBD mass. Compare the mass with the total feed to calculate the recovery rate. Standard basis: Refer to the adsorbent performance evaluation method in GB / T 32669. Key parameter: CO2 flow rate 80 mL·min -1 pH control accuracy ±0.1, standing time 20 min, competitor concentration 2 wt%, stirring temperature 35℃. Data processing: Calculate the desorption efficiency (%), total desorption efficiency (%), and recovery rate (%) for each stage, plot the cumulative desorption curve, compare the effect of staged desorption and single-step desorption, perform three parallel measurements, and the judgment criterion is a recovery rate ≥95%.

[0110] Experiment 6: Site density retention rate and XPS characterization Test Subjects: Boron site density and surface chemical state of the adsorbent before, during, and after cross-linking and locking, and after repeated use. Test Objectives: To verify the effect of cross-linking and locking treatment on site density retention (retention rate 90-100%), and the impact of repeated use on the surface boron chemical state. Test Principles: Boron site density was calibrated using both acid-base titration and X-ray photoelectron spectroscopy (XPS). The titration method quantifies the total boron content by the neutralization reaction of borate ions with a standard acid under alkaline conditions. XPS quantifies the surface boron atom fraction by detecting the area integral of the B 1s electron binding energy peak. Site density retention rate is defined as the percentage of site density after cross-linking and locking / after N cycles relative to the site density before cross-linking. Experimental Methods: Titration Method: Weigh 50 mg of sample, add 10 mL of 0.1 M NaOH solution (containing 20 wt% mannitol to stabilize borate ions), ultrasonically disperse for 30 min to completely release borate ions, titrate to pH 7.0 with 0.05 M hydrochloric acid standard solution, record the volume consumed V (mL), and calculate the boron site density C. B = (0.05 × V - blank consumption volume) / sample mass (mmol·g) -1 XPS test: Samples were tested in an ultra-high vacuum chamber (<10). -7 Excitation was performed using an Al Kα X-ray source (1486.6 eV) at 150 W. A full spectrum was obtained at 280 eV, and high-resolution spectra for B 1s, C 1s, O 1s, and Si 2p were obtained at 30 eV. The binding energy of the B 1s peak was approximately 192 eV (boron in the boron ester bond). Shirley background subtraction and Gaussian-Lorentz mixture function fitting were used. The samples measured included: ① samples before crosslinking and locking, ② samples after crosslinking and locking, and ③ samples after 150 cycles. Standards followed: Titration method referenced GB / T 14506.7; XPS testing referenced ISO 18115 and the XPS general specifications in GB / T 16594. Key parameters: titration endpoint pH 7.0, NaOH concentration 0.1 M, mannitol concentration 20 wt%; XPS vacuum degree <10. -7 Pa, X-ray source Al Kα, pass energy 30 eV (high resolution), scan step size 0.05 eV. Data processing: Calculate the boron site density (mmol·g) for each sample. -1 The retention rate was calculated as (density after crosslinking or cycling / density before crosslinking) × 100%, with a judgment criterion of 90-100%. XPS outputs the B 1s peak position, half-peak width, and atomic fraction, and a B 1s high-resolution spectrum comparison diagram is plotted. Two samples are measured in parallel.

[0111] The results of single-factor experiments fully demonstrate the rationality, reliability, and effectiveness of the technical solution of this invention. For example... Figure 1 As shown, the effect of total boron site density on the selectivity coefficient and adsorption capacity exhibits a distinct single-peak response curve, with the highest peak at 1.9 mmol·g⁻¹. -1Optimal performance was achieved at a selectivity coefficient of 130 and an adsorption capacity of 188 mg·g. -1 Less than 1.2 mmol·g -1 Insufficient timing points lead to performance degradation, exceeding 2.4 mmol·g. -1 Increased spatial steric hindrance leads to increased mass transfer resistance. Figure 2 This indicates that the performance is optimal when the proportion of diboron sites is 43%, with a selectivity coefficient of 132 and an adsorption capacity of 186 mg·g. -1 When the ratio is below 20%, the selectivity of the dual-boron recognition cavity is insufficient and decreases; when it is above 60%, the total capacity decreases due to the limitation of single-boron supplementation. This confirms the importance of balancing the ratio of dual-boron clamp-type recognition cavity to single-boron sites. Figure 3 The results showed that the optimal separation effect was achieved when the organic shell thickness was 5.8 nm, with a selectivity coefficient of 129 and an adsorption capacity of 187 mg·g⁻¹. -1 Below 3.0 nm, the crosslinking and locking are insufficient and the stability is poor. Above 8.0 nm, the mass transfer resistance increases and the kinetic performance decreases, revealing the synergistic optimization mechanism of site protection and mass transfer efficiency. Figure 4 The performance is optimal when the volume fraction of ethanol is 90.5 vol%, with a selectivity coefficient of 130 and an adsorption capacity of 187 mg·g. -1 When the water content is below 85 vol%, the excessive water content inhibits the formation of borate ester bonds, and when it is above 96 vol%, the polarity is too weak and the solubility of the target substance decreases, indicating that the alcohol-water ratio plays a key regulatory role in the formation environment of borate ester bonds. Figure 5 The results show that the separation performance of the adsorption section reaches its peak at pH 7.35, with a selectivity coefficient of 130 and an adsorption capacity of 187 mg·g⁻¹. -1 Below pH 7.0, insufficient borate ion concentration reduces coordination efficiency; above pH 7.6, hydrolysis competition intensifies, decreasing the stability of borate ester bonds. This confirms the necessity of precise pH control over borate-borate equilibrium and coordination reactions. The experimental design employed a wide-range gradient scan, with sufficient sampling points both inside and outside the optimal range. Data continuity was good, and trends were evident. Boundary validation experiments showed a significant performance decrease when parameters deviated from the optimal range, demonstrating the repeatability of the experiment and the reliability of the data. This provides ample parameter margin and controllability for industrial applications.

[0112] Embodiment 1 of the present invention Figure 6 The FTIR results showed that, compared with the control sample, the target sample was 1715–1735 cm⁻¹ -1 The significantly enhanced C=O stretching vibration and narrowed peak at 1250–1280 cm⁻¹ indicate a more ordered ester / carbonyl environment; -1 With 1020–1060 cm -1 The C–O and B–O–C characteristic bands in the region show synchronous enhancement accompanied by a slight blue shift, suggesting the formation of stable borate ester bonds; 3200–3500 cm⁻¹-1 The free –OH component in the broad peak was significantly weakened, while the hydrogen-bonded –OH signal shifted slightly downward, indicating that reactive –OH was effectively consumed and the intermolecular hydrogen bond network was reconstructed; simultaneously, the 800–900 cm⁻¹ peak... -1 The skeletal vibrations and Si–O-related fingerprint bands remained stable, eliminating interference from substrate structural changes. These characteristics collectively demonstrate the successful introduction of functional groups and their solidification in the form of borate esters / ester bonds, resulting in more stable interfacial chemistry. This provides a rational molecular-level basis for improving subsequent cycle stability and surface chemical resistance. Figure 7 The adsorption capacity remained ≥90% after 1–150 cycles, and the boron leaching level remained low throughout. The chemical bond stability evidence from FTIR was consistent with the macroscopic performance trend, which fully demonstrates that the dual locking mechanism of "polyethylene glycol methacrylate brush + N,N'-methylenebisacrylamide" can effectively suppress site loss and support the rationality of the scheme with ultra-high cycling stability and extremely low boron leaching.

[0113] The performance of the examples and comparative examples is summarized in Table 1. The performance comparison between the examples and comparative examples fully demonstrates the synergistic contribution and rationality of each key element in the technical solution of the present invention to the separation performance. Comparative Example 1 uses MCM-41 carrier instead of SBA-15, although the specific surface area is higher (520m²). 2 ·g -1 However, poor pore connectivity and thin pore walls led to insufficient hydrothermal stability, with the selectivity coefficient dropping to 98 and the activity retention rate only reaching 85% after 150 cycles. Boron leaching increased to 1.2 ppm, indicating that the pore structure and wall thickness of the support are crucial for long-term stability. The thick walls and three-dimensional interconnected pores of SBA-15 can better support the functional layer load and cycling stability. In Comparative Example 2, reducing the ethanol volume fraction to 80 vol% and increasing the water content (which inhibited borate ester bond formation and promoted borate hydrolysis) resulted in a sharp drop in selectivity to 75 and an adsorption capacity to 142 mg / g. -1 The recovery rate decreased to 91.5% and the boron leaching amount increased to 1.5 ppm, proving that the alcohol-water ratio needs to be precisely controlled within the range of 85-96 vol% to balance the borate ester bond formation environment and the solubility of the target substance. Too low an alcohol content disrupts the hydrophobic microenvironment of the coordination reaction. In Comparative Example 3, the pH of the adsorption section was lowered to 6.5. Insufficient borate ion concentration significantly reduced the coordination efficiency, the selectivity coefficient decreased to 68, and the adsorption capacity decreased to 135 mg·g⁻¹. -1 The cycling stability also deteriorated to 78% activity retention, indicating that the pH needs to be controlled within the range of 7.0-7.6 to maintain the borate-borate balance and ensure the efficient formation and stability of reversible borate ester bonds. Comparative Example 4 increased the liquid hourly space velocity to 3.0 h⁻¹. -1 Insufficient contact time prevented mass transfer from reaching equilibrium, causing the selectivity to drop to 92 and the adsorption capacity to plummet to 125 mg·g⁻¹. -1Although the processing efficiency is improved, the separation effect deteriorates, proving that the liquid hourly space velocity needs to be between 0.5 and 2.0 h⁻¹. -1 Optimization within the specified range is necessary to balance mass transfer adequacy and economic efficiency; excessively high space velocities would sacrifice equilibrium adsorption capacity, which would be counterproductive. Comparative Example 5 reduced the total boron site density to 0.8 mmol·g⁻¹. -1 Insufficient adsorption sites severely limited recognition ability, reducing the selectivity coefficient to 62 and the adsorption capacity to only 98 mg·g⁻¹. -1 The cycling stability decreased to 75% and the boron leaching amount increased to 2.2 ppm, indicating that the total boron site density needs to be maintained at 1.2-2.4 mmol·g. -1 The range is designed to provide a sufficient number of recognition cavities; excessively low site density cannot meet the requirements for efficient separation. Comparative Example 6 reduced the proportion of diboron sites to 10%, and the severe insufficiency of the diboron clamp-type recognition cavities caused the selectivity coefficient to drop to 55. Although the number of single-boron sites increased, the non-specific adsorption of non-ortho-bisphenol type impurities increased, with an adsorption capacity of 120 mg·g⁻¹. -1 However, selectivity deteriorated, with boron leaching increasing to 2.5 ppm, confirming that the proportion of diboron sites needs to be in the range of 20-60% to ensure the specific recognition advantage of the ortho-biphenol structure. Too low a proportion of diboron sites results in the loss of the core function of clamp-like synergistic recognition. In Comparative Example 7, the polymer grafting temperature was lowered to 10℃. The slow reaction rate and short grafted chains at low temperatures led to an uneven shell, reducing the selectivity coefficient to 85 and the adsorption capacity to 138 mg·g⁻¹. -1 The cycle stability decreased to 80% and the boron leaching amount was 1.6 ppm, indicating that the grafting temperature needs to be in the range of 20-60℃ to ensure the polymerization reaction activity and shell quality. Too low a temperature makes it difficult to construct a complete protective layer. Comparative Example 8 used a thermally promoted reaction instead of Cu(I) catalysis for end-group coupling. Under catalyst-free conditions at 80℃ for 12 hours, the efficiency was low and side reactions increased, the selectivity coefficient decreased to 78, and the adsorption capacity was 145 mg·g⁻¹. -1 The cycle stability was 81%, demonstrating that Cu(I)-catalyzed click chemistry is crucial for the efficient and specific coupling of the double boron module. While thermally promoted reactions are feasible, their efficiency and selectivity are inferior to the catalytic system. Comparative Example 9 used a single-step hydrochloric acid adjustment to pH 4.0 for direct desorption. Although desorption was rapid under strong acid conditions, it severely damaged the boron ester bonds and caused site hydrolysis and loss. Although the recovery rate was 90.8%, the activity retention rate dropped sharply to 68% after 150 cycles, and the boron leaching amount was as high as 3.5 ppm. This indicates that staged desorption (mild acidification with CO2 + diol competing agent) is indispensable for protecting the chemical integrity of the sites and achieving high cycle stability. Sacrificing long-term stability for short-term efficiency with strong acid desorption is not advisable. Comparative Example 10 omitted the cross-linking and locking step, relying solely on physical adsorption and hydrogen bonding to fix the sites. Although the initial performance was acceptable (selectivity coefficient 95, adsorption capacity 165 mg·g), the results were still satisfactory. -1However, after 150 cycles, the activity retention rate plummeted to 58%, and the boron leaching amount reached as high as 8.2 ppm. This demonstrates that N,N'-methylenebisacrylamide crosslinking and locking are absolutely necessary to inhibit site loss and maintain ultra-high cycling stability (≥90%) and extremely low boron leaching (≤1 ppm). Uncrosslinked adsorbents lose practical value due to significant site detachment during cycling. In Comparative Example 11, the organic shell thickness was increased to 12 nm. The excessively thick shell significantly increased mass transfer resistance, reduced the selectivity coefficient to 82, and decreased the adsorption capacity to 132 mg·g⁻¹. -1 BET specific surface area decreased to 320m² 2 ·g -1 This indicates severe pore blockage, with a cycle stability of 86%, demonstrating that the shell thickness needs to be controlled within the 3-8 nm range to balance site protection and mass transfer efficiency. An excessively thick shell hinders the diffusion of the target analyte into the recognition cavity. In Comparative Example 12, the adsorption section temperature was lowered to 20 °C. While lower temperature is beneficial for equilibrium adsorption, the kinetic rate decreased significantly, the selectivity coefficient dropped to 70, and the adsorption capacity was 128 mg·g⁻¹. -1 The cycle stability was 83%, indicating that the temperature needs to be in the range of 30-40℃ to balance thermodynamic equilibrium and kinetic rate. Too low a temperature prolongs the time required to reach equilibrium and reduces the efficiency of borate ester bond formation. Comparative Example 13 omitted the nitrogen purging step and replaced it with room temperature air drying for 24 hours. Insufficient regeneration led to the accumulation of residual solvent and desorbent. After 150 cycles, the activity retention rate dropped to 62%, and the boron leaching amount increased to 4.8 ppm. Although the initial performance was acceptable, long-term stability deteriorated. This demonstrates that a regeneration procedure of 60℃ vacuum drying for 2 hours combined with nitrogen purging for 30 minutes is crucial for thoroughly removing residual substances and restoring adsorbent activity. Simplifying the regeneration steps led to a rapid decline in cycle performance. Comparative Example 14 used a mercapto-alkene click reaction to replace the amino-bromine system to introduce initiation sites. Mercapto oxidation showed high sensitivity, but UV initiation uniformity was poor, resulting in a selectivity coefficient of 72 and an adsorption capacity of 140 mg·g⁻¹. -1 The cycling stability was 76% and the boron leaching amount was 2.0 ppm, indicating that the ATRP initiation system constructed by 3-aminopropyltriethoxysilane and α-bromoisobutyryl bromide is more stable and controllable. Although the thiol system is feasible, its stability and reproducibility are not as good as the amino route. Examples 1-4 all achieved selectivity coefficients ≥100 and adsorption capacities of 155-192 mg·g. -1 Recovery rate ≥95%, activity retention rate ≥90% after 150 cycles and boron leaching ≤1ppm, demonstrating that the SBA-15 / KIT-6 ordered mesoporous support has a boron leaching rate of 1.2-2.4mmol·g. -1 Total boron site density, 20-60% diboron site ratio, 3-8 nm organic shell thickness, cross-linking and locking treatment, 85-96 vol% alcohol content, 7.0-7.6 adsorption pH, 0.5-2.0 h -1Key technical elements such as liquid hourly space velocity and staged desorption process are indispensable and need to be optimized in a coordinated manner. Any deviation of any element from the optimal range will lead to significant performance degradation. This fully verifies the overall rationality of the technical solution of this invention, the synergistic necessity of each component, and the accuracy of the parameter window, providing a reliable technical route for the highly selective and stable separation of o-diphenol metabolites from industrial hemp.

[0114] Table 1 Performance summary of examples and comparative examples

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that any equivalent structural transformations made under the concept of the present invention and using the contents of the specification and drawings of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for separating industrial hemp metabolites based on a reversible borate ester reaction, characterized in that, Including the following steps: S1. Provide a core-shell mesoporous functional adsorbent, the adsorbent comprising a mesoporous silica core and an organic shell, the organic shell comprising a clamp-type dual boron recognition cavity, the boron-to-boron center distance being 0.32-0.50 nm; S2. The feed solution containing industrial hemp metabolites is contacted with the adsorbent in an alcohol-water mixed solvent with an ethanol volume fraction of 85-96 vol% at a temperature of 30-40°C, and selective adsorption is carried out under the condition of reversible formation of borate ester bonds. S3. Reduce the acidity of the system and / or introduce a diol competitor to break the borate ester bond and desorb the metabolite; S4. Collect the desorption liquid and rinse the adsorbent with alcohol and water and dry it to achieve recycling; During fixed-bed operation, the liquid hourly space velocity is 0.5-2.0 h⁻¹. -1 Alternatively, the batch contact time is 30-60 minutes; The feed solution is a solution of industrial hemp extract in ethanol and water, and the industrial hemp metabolites are selected from cannabidiols of the ortho- and tho-diol type; and the operating pH of the adsorption section is 7.0-7.6; The total boron site density, calculated as boron element, is 1.2–2.4 mmol·g. -1 The proportion of 20-60% of the diboron sites is high; and the relative site density retention rate is 90-100% after cross-linking and locking treatment. The mesoporous silica core is SBA-15 or KIT-6 ordered mesoporous silica with an average pore size of 1.5-3.0 nm; and the organic shell thickness is 3-8 nm or characterized by a total organic layer mass increment of 2-8 wt%. The method for preparing the adsorbent includes the following steps: A1. Detemplating and drying mesoporous silica carriers with an average pore size of 1.5-3.0 nm; A2. An amino site is introduced by silanization with 3-aminopropyltriethoxysilane, and the amino site is converted into a free radical polymerization initiation site by α-bromoisobutyryl bromide; A3. Grafting an alcohol-loving polymer with an alkyne-containing functional segment via free radical polymerization; A4. To enable end-group coupling between the clamp-type double boron module with azide end groups and the alkyne-containing functional segment; A5. After performing single-boron site replacement and constructing with polyethylene glycol methacrylate brushes, crosslinking and locking were implemented. The final total boron site density was 1.2–2.4 mmol·g. -1 The proportion of double boron sites is 20-60%; The silanization temperature is 20-80℃ and the time is 0.5-4h; the grafting temperature of the alkyne-containing functional segment is 20-60℃ and the time is 0.5-2h; the end-group coupling temperature is 20-35℃ and the time is 4-8h; and the alcohol-loving segment polymer is selected from polyethylene glycol methacrylate with an average molecular weight of 200-600, and the alkyne-containing functional segment polymer is formed from propynyl methacrylate monomer.

2. The method according to claim 1, characterized in that, The desorption fractionation includes a first stage of lowering the pH of the system to 7.0-7.2 with carbon dioxide, and a second stage of further lowering the pH of the system to 6.0-6.5 or adding 1-3 wt% of a diol competitor, namely 1,2-propanediol or glycerol carbonate, to the solution.

3. The method according to claim 1, characterized in that, The method performance meets the following requirements: selectivity coefficient ≥100, recovery rate ≥95%, activity retention rate ≥90% after 100-200 cycles, and boron leaching amount ≤1ppm after every 100 cycles.

4. The method according to claim 1, characterized in that, The clamp-type dual-boron recognition cavity is formed by grafting a dual-boron module with end-group functionalization. The dual-boron module is derived from 1,4-phenylenediboric acid or 1,8-naphthalenediboric acid and is fixed to the alkyne-containing functional segment of the organic shell by end-group coupling. The organic shell contains a polyethylene glycol methacrylate brush.

Images