Rapid and Selective Extraction and Detection of Baicalin From Scutellariae Radix Using Magnetic Molecularly Imprinted Polymers

ABSTRACT

Baicalin (BA), a key flavonoid isolated from Scutellariae Radix (SR), possesses diverse pharmacological activities encompassing anticancer, anti‐inflammatory, and hepatoprotective effects. In order to achieve rapid separation and enrichment of BA from complex matrix samples, core–shell magnetic molecularly imprinted polymers (MMIPs) were fabricated using BA as the template molecule and Fe₃O₄ nanoparticles as the magnetic core. Subsequently, key preparation parameters, including the ratio of functional monomers to cross‐linking agents, imprinting time, and the dosage of BA were systematically optimized. Under the optimized conditions, the prepared MMIPs exhibited rapid binding kinetics and a remarkable maximum adsorption capacity of 18.21 mg/g, accompanied by an imprinting factor (IF) of 3.919. Furthermore, the developed MMIPs were successfully employed for the selective extraction and determination of BA from SR. Specifically, the average recoveries of BA at three spiked concentrations (5.00, 11.20, and 20.00 µg/mL) were 107.0%, 92.7%, and 105.2%, respectively, with corresponding relative standard deviations (RSDs) of 5.31%, 1.65%, and 3.28% (n = 3). Overall, this work provides a novel and efficient adsorbent for trace flavonoids enrichment from complex herbal samples, offering a promising pretreatment strategy for active ingredient analysis in traditional Chinese medicine research.

The MMIPs exhibited high‐specificity recognition and fast separation. They were employed for the specific extraction and accurate quantification of baicalin from Scutellariae Radix (SR), offering a reliable strategy for precise analysis of active ingredients from natural products.

1. Introduction

Scutellariae Radix (SR), derived from the root of Scutellaria baicalensis Georgi, has been extensively applied in disease prevention and treatment, including cardiovascular diseases [1], central nervous system diseases [2], inflammation [3], and cancer [4]. Baicalin (BA), 5,6‐dihydroxy‐4‐oxo‐2‐phenyl‐4H‐1‐benzopyran‐7‐yl β‐D‐glucopyranosiduronic acid, is recognized as the major active ingredient of SR and demonstrates some therapeutic effects, such as anti‐cancer [5], anti‐inflammatory [6, 7], anti‐microbial [8], hepatoprotective [9, 10]. Given its significant biological efficacy, numerous extraction and separation techniques, including chromatography, two‐dimensional countercurrent chromatography [11], supercritical fluid extraction [12], ultrasound‐assisted deep eutectic solvents [13, 14, 15], and resin‐based separation [16], have been developed to isolate flavonoids from complex matrices. However, due to the trace concentration of BA and the presence of complex interfering substances in SR, these methodologies suffered from low selectivity, long operation time, excessive reagent consumption, and high instrumental costs [17]. Therefore, it is urgent to develop efficient adsorbents for the selective enrichment and rapid isolation of bioactive compounds from SR.

Molecularly imprinted polymers (MIPs) are ideal plastic antibodies with selective recognition and binding abilities for target substances [18]. By virtue of the exceptional selectivity, high stability, low cost, and facile preparation [19], MIPs have attracted considerable attention in the context of extracting and isolating active ingredients from traditional Chinese medicine (TCM) [19, 20, 21, 22]. However, MIPs synthesized via bulk polymerization exhibit severe mass transfer resistance that hinders the template removal. The resulting imprinted cavities are imperfect, leading to compromised selectivity and reduced adsorption capacity [23, 24], thus restricting the practical application of MIPs for the separation of flavonoids from complex natural matrices.

The core–shell structure of surface MIPs, where the imprinted layer is coated onto nanoparticle surfaces, overcomes the difficulties in template elution and exhibits promising applicability for flavonoid separation [25]. For example, Qi et al. prepared MIPs immobilized on macroporous adsorption resins to achieve rapid and efficient separation of rutin from natural plant extracts [21]. Similarly, molecularly imprinted membranes were fabricated by Li et al. using green lignocellulosic nanofibers as the substrate for the selective adsorption of baicalin from S. baicalensis [26]. Although these strategies improve adsorption selectivity and capacity, the separation process still relies on centrifugation, leading to cumbersome and time‐consuming operations. To overcome these limitations, Fe₃O₄ nanoparticles (MNPs), with a large specific surface area and favorable superparamagnetic properties, are widely used as a magnetic core to construct magnetic surface MIPs for the efficient capture of active ingredients from TCM [27, 28]. Nevertheless, few studies have focused on constructing magnetic MIPs for the specific extraction and accurate quantification of baicalin in complex herbal extracts.

In the present work, core–shell magnetic molecularly imprinted polymers (MMIPs) were fabricated with Fe₃O₄ particles as the magnetic cores and baicalin as the target template. The morphological, spectral, crystal structure, and the magnetic properties of the obtained materials were fully investigated by TEM, FT‑IR, XRD, as well as VSM to verify their successful fabrication. The adsorption performances, including binding kinetics, isotherms, selectivity, and reusability were systematically evaluated. Furthermore, the as‐prepared MMIPs were successfully used for the selective extraction and quantification of baicalin from SR, demonstrating good recoveries and reproducibility. Therefore, this study not only addresses the limitations of conventional extraction methods for baicalin but also expands the application of magnetic molecular imprinting technology in the precise analysis of active ingredients from natural products.

2. Materials and Methods

2.1. Materials and Instrumentation

Detailed information on all reagents and experimental instruments is provided in the Supporting Information.

2.2. Synthesis of Fe₃O₄ @SiO₂@MPS MNPs

Solvothermal method was used to synthesize Fe₃O₄ nanoparticles according to the literature [29]. The suspension of FeCl₃·6H₂O (2.76 g), trisodium citrate dihydrate (7.94 g), CH₃COONa (6.15 g), and ethylene glycol (80 mL) were evenly dispersed by ultrasonic treatment for 15 min. Then, the homogeneous mixture was reacted under vigorous stirring and remained for 60 min at 150°C. Subsequently, transfer of the reactant mixture into a 100 mL PTFE reactor was performed, and the reaction was conducted at 200°C for 8 h. After natural cooling, the products were purified via alternating washing with deionized water and ethanol, and were finally dried at 60°C for further use.

To synthesize Fe₃O₄@SiO₂, 0.207 g of Fe₃O₄ MNPs was dispersed into 40 mL of ultrapure water and 160 mL of absolute ethanol. Post‐sonication, NH₃·H₂O (8 mL) and TEOS (1 mL) were introduced to the above suspension, and the system was maintained for 6 h at 60°C. The obtained Fe₃O₄@SiO₂ MNPs were alternately rinsed with anhydrous ethanol and deionized water three times each, followed by drying at 60°C.

To modify the surface of Fe₃O₄@SiO₂ with double bonds, 3 mL of 3‐(trimethoxysilyl)propyl methacrylate (3‐MPS), 200 mg of Fe₃O₄@SiO₂, and 50 mL of anhydrous toluene were mixed. After uniform dispersion, the whole suspension was reacted at 90°C for 12 h with nitrogen protection. The obtained Fe₃O₄@SiO₂@MPS was purified via repeated anhydrous ethanol rinsing and dried, and the final solid was stored under ambient conditions.

2.3. Preparation of MMIPs

The free radical polymerization method was used to prepare MMIPs. Briefly, 10 mg of baicalin and 30 µL of MAA were dissolved in anhydrous ethanol (12 mL) and sonicated for 20 min. The solution was pre‐polymerized at 25°C for 12 h to ensure sufficient interaction between baicalin and MAA. Thereafter, the upper suspension was supplemented with 20 mg Fe₃O₄@SiO₂@MPS, 300 µL of EGDMA cross‐linker and 40 mg of AIBN initiator, and the mixture was stirred at 65°C for 6 h under N₂ protection. The as‐prepared MMIPs were collected and washed with methanol/acetic acid (9:1, v/v) repeatedly until no baicalin was detected by UV–vis (280 nm). As a control, magnetic non‑imprinted polymers (MNIPs) were prepared under the same conditions, without the addition of baicalin.

2.4. Adsorption Experiments

In the kinetic adsorption study, MMIPs or MNIPs (10 mg) were scattered into the baicalin ethanol solution (0.1 mg/mL, 5 mL) with continuous shaking at ambient temperature. After incubation for different times (0, 3, 5, 10, 15, 20, 30, 40, 50, 60, 90, and 120 min), the supernatant was detected after magnetically isolating the nanoparticles, and the corresponding adsorption capacities were calculated using the Formula (1) [30].

$$Q=V\left(C_0-C_e\right)/m$$ (1)

where Q (mg/g) means the adsorption capacity of adsorbent. C ₀ and C_e (mg/mL) are the initial and equilibrium concentrations of the baicalin ethanol solution, respectively. V (mL) refers to sample volume, m (g) is the dosage of sorbent.

In the static adsorption experiment, MMIPs or MNIPs (10 mg) were distributed into the ethanol solution with different baicalin concentrations (15, 20, 25, 30, 40, 50, 100, 150, and 200 µg/mL), and the compounds were adsorbed for 1 h. After binding equilibrium, the residual baicalin concentrations were investigated, and the amounts of baicalin bound by the adsorbents were calculated accordingly.

In the selectivity adsorption analyses, two structural analogs (rutin and quercetin) were used to evaluate the specificity of MMIPs to baicalin. Ethanol solution of each structural analog (0.1 mg/mL, 5 mL) was individually blended with MMIPs or MNIPs (10 mg). The mixed suspension was incubated at 25°C for 1 h, followed by HPLC detection of supernatants. Meanwhile, imprinting factor (IF) calculated from Formula (2) was adopted to assess the selectivity of MMIPs [30]:

$$\mathrm{IF}=Q_{\mathrm{MMIPs}}/Q_{\mathrm{MNIPs}}$$ (2)

where Q _MMIPs (mg/g) and Q _MNIPs (mg/g) are the adsorption capacities of MMIPs and MNIPs, respectively.

To assess the reusability, the adsorption–desorption experiments were conducted. The baicalin ethanol solution (0.1 mg/mL, 5 mL) was incubated with MMIPs or MNIPs (10 mg). After reaching equilibrium, nanoparticles were magnetically precipitated and the resulting supernatant was taken for HPLC quantification. Then, the adsorbed baicalin‐MMIPs were transferred into the methanol–acetic acid eluent (9:1, v/v). Five adsorption–desorption cycles were performed, and the corresponding adsorption capacity was calculated.

2.5. Real Sample Detection

Here, 20 g of the dried SR were reflux‐extracted in 200 mL of water for 60 min, then filtered. The filter residue was extracted again with 100 mL of water for 30 min. The combined filtrates were adjusted to pH 1–2 using 4 mol/L hydrochloric acid at 60°C, and then incubated at 80°C for 30 min. After standing for 12 h at 25°C, the precipitate was collected by filtration and washed to neutral pH. The crude baicalin extract was obtained after drying at 40°C, and then stored at 4°C.

Around 30 mg of MMIPs were dispersed in the crude baicalin ethanol solution, and the compounds were reacted for 1 h. The baicalin–MMIPs were rapidly harvested magnetically and subjected to 1 h elution with methanol–acetic acid (9:1, v/v). After eluent removal, the dry residue was redissolved in anhydrous ethanol (1 mL), and then filtered through a 0.45 µm membrane filter for further HPLC analysis.

3. Results and Discussions

3.1. Imprinting Strategy and Optimization of MMIPs

As depicted in Figure 1, a new approach was developed to fabricate MMIPs aimed at improving the separation specificity and efficiency of active ingredients in TCM. The integration of magnetic Fe₃O₄ into MIPs could not only retain the high selectivity and specific recognition capability of MIPs, but also impart the remarkable magnetic responsiveness, which greatly simplified the separation and recovery of adsorbents. Herein, we reported baicalin‐templated MMIPs, which were employed for the specific extraction and fast separation of baicalin from complex samples.

FIGURE 1.

Schematic illustration of the preparation and application of MMIPs.

To determine the optimal parameters, the impacts of the MAA‐to‐EGDMA molar ratio, the imprinting time, and the dosage of BA on the adsorption performance of MMIPs were systematically studied.

The MAA‐to‐EGDMA molar ratios significantly affected the specificity of MMIPs, so it was necessary to investigate the optimal ratio. According to Figure 2A, increasing the EGDMA first enhanced and then reduced the adsorption capacity, with the optimal value of 12.45 mg/g observed at a ratio of 1:4.5. Excessive MAA potentially increased non‐specific binding, whereas insufficient EGDMA might result in formation of fewer imprinting sites. Hence, a ratio of 1:4.5 was selected for the subsequent experiments.

FIGURE 2.

The effects of preparation conditions on the adsorption capacity and imprinting factor of MMIPs to baicalin:MAA:EGDMA (A), imprinting time (B), and the amount of BA (C).

To some extent, the polymerization time influenced the thickness of the imprinting layer. According to Figure 2B, extending the imprinting time initially enhanced the adsorption capacity of MMIPs, but the capacity subsequently declined. Excessive imprinting time could result in a thicker imprinting layer. Moreover, the template molecules were embedded and could not be eluted, leading to insufficient imprinting sites and reduced adsorption capacity. Conversely, insufficient imprinting time led to a compromised specificity of the imprinted layer. Thus, the optimal imprinting time was determined to be 6 h and was used for the preparation of MMIPs.

The dosage of template molecule directly affected the number of imprinted cavities, so it was necessary to investigate the addition dosage of BA. The amount of MAA was fixed at 30 µL. As presented in Figure 2C, with the gradually increasing amount of BA, the adsorption capacities of MMIPs first increased and then decreased, reaching the maximum adsorption capacity (14.46 mg/g) and IF (3.919) at 10 mg of BA. Because less BA led to the formation of non‐specific binding sites and reduced the selectivity. On the other hand, excessive BA resulted in low imprinting efficiency. However, the adsorption capacities of MNIPs showed little changes, mainly because no template molecules were used in its preparation.

3.2. Characterization

As shown in Figure 3, the morphologies of the Fe₃O₄ (A), Fe₃O₄@SiO₂ (B), MMIPs (C), and MNIPs (D) were described as uniform spherical shapes by the TEM images. The size of the Fe₃O₄@SiO₂ increased from 150 to 200 nm, as a thin and smooth layer of SiO₂ was uniformly deposited on the surface of Fe₃O₄, and then, when compared with the Fe₃O₄@SiO₂, MMIPs and MNIPs appeared rough surfaces and particle sizes increased to 250 nm. All these morphological changes proved that successful formation of both the silica layer and the molecularly imprinted layer.

FIGURE 3.

(A) TEM of Fe₃O₄; (B) TEM of Fe₃O₄@SiO₂; (C) TEM of MMIPs; (D) TEM of MNIPs; (E) FT‐IR spectra of Fe₃O₄ (a), Fe₃O₄@SiO₂ (b), Fe₃O₄@SiO₂@MPS (c), and MMIPs (d); (F) XRD patterns of Fe₃O₄ (a) and MMIPs (b).

The FT‐IR spectra of Fe₃O₄ (a), Fe₃O₄@SiO₂ (b), Fe₃O₄@SiO₂@MPS (c), and MMIPs (d) were determined to characterize their structure and functional groups. In Figure 3E, the absorption band at approximately 588 cm⁻¹ was attributed to the Fe─O, which indicated the successful synthesis of Fe₃O₄ [31]. A set of prominent absorption peaks appeared around 790, 948, and 1082 cm⁻¹, arising from Si─O, Si─O─H, and Si─O─Si stretching, respectively [32], hence confirming the efficient coating of silica layer. The clear observation of C═C bonds at 1618 cm⁻¹ indicated effective modification with 3‐MPS [33]. The occurrence of peak at 1724 cm⁻¹, assigned to the O─C═O of the EGDMA, revealed that the MMIPs were successfully aggregated [34].

The XRD patterns of Fe₃O₄ (a) and MMIPs (b) were employed to validate the crystal structure. As presented in Figure 3F, the diffraction peaks at 2θ = 30.1°, 35.4°, 43.9°, 53.7°, 57.3°, and 62.8°, which were in good agreement with the (220, 311, 400, 422, 511, and 440) of Fe₃O₄ crystal in JCPDS No. 65–3107, respectively, were possessed by both materials. These observations demonstrated the successful synthesis of Fe₃O₄ and confirmed the preservation of its crystalline integrity during the subsequent surface modification and imprinting procedures [35].

The magnetization curves of Fe₃O₄ (a), Fe₃O₄@SiO₂ (b), and MMIPs (c) were characterized by VSM to evaluate the magnetic properties. As shown in Figure S1, compared with the Fe₃O₄, the saturation magnetization (Ms) values of Fe₃O₄@SiO₂ and MMIPs showed a stepwise decrease, from 65.17 emu/g for Fe₃O₄ to 52.9 emu/g for Fe₃O₄@SiO₂ and 36.48 emu/g for MMIPs, which was mainly attributed to the coating of silica shells and the subsequent imprinting layer on the surface of Fe₃O₄ [37]. Nevertheless, the Ms of MMIPs still allowed to complete magnetic collection within 11 s. Thus, all the fabricated materials exhibited excellent superparamagnetic properties and allowed the rapid magnetic separation.

3.3. Adsorption Performance

Binding kinetics tests were conducted to assess the binding efficiency of MMIPs. Figure 4A revealed that the adsorption capacity of MMIPs rose sharply before 20 min and then gradually stabilized, with a maximum equilibrium adsorption capacity of 18.21 mg/g. Meanwhile, MNIPs obtained adsorption profile with the same trend, but their capacity (10.98 mg/g) remained lower than that of MMIPs. These findings suggested that the presence of abundant imprinted cavities on the MMIPs surface could specifically recognize the template. In addition, the adsorption rate steadily decreased due to the progressive saturation of the imprinted cavities, eventually achieving adsorption equilibrium.

FIGURE 4.

(A) Adsorption kinetics curves; (B) adsorption equilibrium curves; kinetic fitting results of MMIPs and MNIPs via the pseudo‐first‐order (C) and pseudo‐second‐order (D) kinetic model; the adsorption isotherm fitting of MMIPs and MNIPs by the Langmuir (E) and Freundlich (F) models.

Furthermore, the kinetic mechanism was evaluated by the pseudo‐first‐order kinetic model (3) and pseudo‐second‐order kinetic model (4) [36].

$$Q_t=Q_e\left(1-\exp \left(-k_{1}t\right)\right)$$ (3)

$$t/Q_t=1/(k_2{Q_e}^2)+t/Q_e$$ (4)

In these models, Q_e (mg/g) refers to the equilibrium adsorption amount of baicalin, while Q_t (mg/g) is the adsorption amount at time t (min). Meanwhile, the terms k ₁ (1/min) and k ₂ (g/[mg·min]) refer to the rate constants of the pseudo‐first‐order and pseudo‐second‐order kinetic model, respectively.

According to Figure 4C,D and Table S1, the pseudo‐second‐order model presented superior fit to kinetic data of MMIPs and MNIPs nanoparticles, as demonstrated by the greater correlation coefficients (r ²). Abundant imprinted binding sites contributed to the fast kinetic adsorption performance of the nanoparticles.

In characterizing the adsorption capacity of MMIPs, equilibrium experiments were conducted. As depicted in Figure 4B, the adsorption capacities rose sharply along with rising initial baicalin concentration, and gradually reached saturation when the concentration exceeded 0.1 mg/mL. Benefiting from the abundant and tailor‐made recognition cavities, MMIPs could effectively capture the template via robust physicochemical forces. In contrast, adsorption of MNIPs was restricted to feeble nonspecific affinity. These data suggested that MMIPs had obviously superior adsorption capacities to MNIPs under identical conditions.

As displayed in Figure 4E,F, the equilibrium data were evaluated with the Langmuir (5) and Freundlich (6) isotherm models [37] to further appraise the adsorption properties of nano‐adsorbents. In Table S2, the calculated correlation coefficients (r ²) for the Langmuir model were 0.9878 for MMIPs and 0.9910 for MNIPs, both higher than those for the Freundlich model. In addition, the adsorption capacity (13.02 mg/g) of the MMIPs obtained from the Langmuir model was in good agreement with the experimentally measured value (12.90 mg/g). These findings consistently demonstrated that the Langmuir isotherm model provided a better description of the adsorption behavior, implying monolayer adsorption of the synthesized MMIPs [35].

$$Q_e=\left(k_L{Q}_{\max }{C}_e\right)/\left(1+C_e{K}_L\right)$$ (5)

$$\log Q_e=n\log C_e+\log K_F$$ (6)

Specifically, C_e (mg/mL): The equilibrium baicalin concentration in adsorption system; Q_e (mg/g): Actual adsorption capacity of MMIPs or MNIPs toward baicalin; Q _max (mg/g): Theoretical maximum adsorption capacity of adsorbing material; K_L (mL/mg): Characteristic constant of Langmuir; K_F (mg/g): Freundlich constant; n: Heterogeneity index.

3.4. Selective and Reusability Adsorption

Quercetin and rutin possessed the same C6–C3–C6 flavonoid skeleton and multiple phenolic hydroxyl groups as baicalin (Figure S2). In addition, both rutin and baicalin possessed glycosidic moieties. Given that all three compounds could form hydrogen bonds and π–π interactions with MAA, they were selected as structural analogs of baicalin to evaluate the binding selectivity of the MMIPs. As revealed in Figure 5A, baicalin was adsorbed most effectively by the MMIPs compared to other substrates, due to the good spatial match between the imprinted sites and baicalin. Conversely, owing to the lack of specific cavities in the MNIPs, there was no significant difference for the adsorption of all substrates. These findings demonstrated that these obtained MMIPs had excellent specificity and selectivity.

FIGURE 5.

Adsorption performances MMIPs and MNIPs: Selectivity adsorption (A) and reusability (B).

Moreover, the adsorption–elution–recombination cycle experiments were conducted to study the reusability of MMIPs. As shown in Figure 5B, after five adsorption–elution–recombination cycles, the bonding properties of MMIPs remained relatively stable, meanwhile the relative standard deviation (RSD) was below 5.280%. These statistics confirmed that the MMIPs possessed the superb reusability.

3.5. Practical Application

Systematic validation including linear range, limit of detection (LOD) and limit of quantitation (LOQ) were implemented to confirm the feasibility of the developed method. A series of baicalin standard solutions (0.4000–110.0 µg/mL) were quantified for standard curve fitting, yielding the regression formula: y = 858.4C _BA − 145.5 (r² = 0.9993). Furthermore, on the basis of S/N = 3 and S/N = 10 rules, the resultant LOD was 0.1000 µg/mL while the LOQ was 0.4000 µg/mL.

A series of recovery experiments using the standard addition method were performed to validate the reliability and accuracy of the established MMIPs extraction procedure, revealed in Table S3. Recoveries were determined by spiking different concentrations of baicalin into sample solutions, and the values ranged from 92.7% to 107.0%, with RSD values ranging from 1.65% to 5.31%. Therefore, the obtained MMIPs could achieve accurate and selective detection of baicalin in complex matrices, demonstrating favorable practicability and availability.

MMIPs (30 mg) were used for baicalin extraction from the crude extract of SR. As displayed in Figure 6, the baicalin concentrations of the crude extract (a), the baicalin–MMIPs eluate (b), and the baicalin‐MMIPs treated supernatant (c) were 30.36, 25.72, and 4.32 µg/mL, respectively. The baicalin peak in the baicalin‐MMIPs treated supernatant (c) declined obviously, which indicated effective adsorption by the MMIPs. After subsequent elution, the baicalin content in the eluate (b) was comparable to that of the crude extract (a), with a loss during the process. It was confirmed that the obtained MMIPs could selectively capture baicalin and served as an efficient adsorbent for rapid enrichment of target analytes in complex samples.

FIGURE 6.

The HPLC chromatograms of the crude extract of the Scutellariae Radix. The chromatogram of the initial sample of the aqueous extract of the Scutellariae Radix (a), the chromatogram of baicalin obtained after elution from MMIPs (b), and the chromatogram of the extract after being adsorbed by MMIPs (c).

4. Conclusions

MMIPs designed for the extraction and enrichment of baicalin from SR were synthesized successfully and subjected to morphology, spectral properties, crystal structure, and the magnetic properties by TEM, FT‑IR, XRD, as well as VSM. Binding properties including static and dynamic adsorption, selectivity, and reusability tests were performed to assess the recognition and binding performance of MMIPs. These experiments showed that MMIPs possessed fast adsorption rates (20 min), outstanding adsorption capacity (18.21 mg/g), good IF (3.919) and excellent selectivity for baicalin. Moreover, the MMIPs served as dispersive solid‐phase extraction sorbents for purification of baicalin from crude baicalin extract, and the purified product exhibited high quality with minimized matrix interference and satisfactory baicalin recoveries (92.7%–107.0%, RSDs ≤ 5.31%, n = 3). In conclusion, these results indicated that the obtained MMIPs enable an effective and practical approach for the separation and purification of active components, offering a rapid and environmentally friendly alternative for the isolation of target compounds from TCM.

Author Contributions

Jing Li: conceptualization, data curation, investigation, project administration, methodology, writing – original draft. Yihan Sun: investigation, data curation, validation. Jialin Guo: investigation, validation. Wei Zhao: investigation, validation. Xia Liu: funding acquisition, supervision, writing – review and editing, project administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: jssc70507‐sup‐0001‐SuppMat.pdf.

Li J., Sun Y., Guo J., Zhao W., and Liu X., “Rapid and Selective Extraction and Detection of Baicalin From Scutellariae Radix Using Magnetic Molecularly Imprinted Polymers.” Journal of Separation Science 49, no. 8 (2026): e70507. 10.1002/jssc.70507

Data Availability Statement

The data that supports the findings of this study are available in the Supporting Information of this article.