In recent years, green extraction of bioactive compounds from herbal medicines has generated widespread interest. Deep eutectic solvents (DES) have widely replaced traditional organic solvents in the extraction process. In this study, the efficiencies of eight tailor-made DESs in extracting flavonoids from Acanthopanax senticosus (AS) were compared. Response surface methodology (RSM) was employed to optimize the influencing parameters including ultrasonic power, HBA-HBD ratio, water content, solid-liquid ratio, extraction temperature and extraction time. DES composed of glycerol and levulinic acid at a 1:1 ratio was established as the most suitable extraction medium. Optimal conditions were ultrasonic power of 500W, water content of 28%, solid-liquid ratio of 1:18 g·mL^-1, extraction temperature of 55℃ and extraction time of 73 minutes. The extraction yield of AS total flavonoids reached 23.928±0.071 mg·g^-1, which was 86.3% and 43.8% higher compared with traditional solvent soak and ethanol reflux extraction methods, respectively. Macroporous resin (D-101, HPD-600, S-8 and AB-8) was used to recover flavonoids from extracts. The AB-8 resin showed higher adsorption/desorption performance, with a recovery rate of total flavonoids of up to 71.56±0.256%. In addition, DES solvent could be efficiently recovered through this process and reused. In summary, ultrasonic-assisted DES combined with the macroporous resin enrichment method is exceptionally effective in extracting flavonoids from AS and provides a promising environmentally friendly and recyclable strategy for flavonoid extraction from natural plant sources.

1. Preparation of DES

HBA and HBD reagents were mixed at specific molar ratios, heated and stirred in a water bath at 80°C until completely dissolved. Since the majority of flavonoids are acidic and easily soluble in polar solvents, choline chloride (ChCl), levulinic acid (La), Lactic acid (Lac), Malic acid (Ma), and Citric acid (Ca), which are environmentally friendly and biodegradable, were selected as HBA and ethylene glycol (Eg), La, 1,4-butanediol (Buta), and glycerol (Gly) as HBD components. According to the self-made combination method and molar ratio, eight different DES types were prepared, all of which were uniform and stable transparent liquids.

2. Impact of different DES types and auxiliary extraction methods

DES (7 mL) configured according to the corresponding molar ratio was added to a 20 mL centrifuge tube, followed by 3 mL water (DES water content was 30%). The solution was heated and thoroughly mixed in a water bath at 80℃ to prepare DES extraction solvent with a solid-liquid ratio of 1:20 g·mL^-1. Extraction rates under mechanical shaking with constant temperature oscillator-assisted extraction (200, 225 or 250 r·min^-1, 40 min, 40℃) and ultrasonic-assisted extraction (500W, 40 min, 40℃) were compared.

3. Single factor effects

The effects of six single factors, specifically, HBA:HBD molar ratio (2:1, 1.5:1, 1:1, 1:1.5, and 1:2), DES water content (20%, 30%, 40%, 50%, and 60%), solid–liquid ratio of AS powder to DES solvent volume (1:10, 1:20, 1:30, 1:40, and 1:50 g·mL^-1), extraction temperature (30, 40, 50, 60, and 70℃) and extraction time (40, 50, 60, 70, and 80 min), on extraction rates of AS total flavonoids were evaluated.

4. Determination of total flavonoids

AS powder (0.5 g) was accurately weighed and extracted with the corresponding extraction solvent. Following shaking, 1.5 mL was centrifuged at 8000 rpm·min^-1 for 10 minutes. A 1 mL aliquot of the experimental extract was accurately measured and placed in a 25 mL volumetric flask, 1 mL of 5% (v/v) NaNO₂ solution added, and shaken for 6 minutes. After addition of 1 mL of 10% (v/v) Al(NO₃)₃ solution to the bottle, the mixture was re-shaken and incubated for 6 minutes. Finally, 9 mL of 4% NaOH solution was added, diluted with 70% ethanol, and shaken for 15 minutes. Absorbance was measured at 510 nm to calculate the total flavonoid content.

5. Optimization of RSM conditions for extraction

BBD was conducted by selecting the optimal levels of each of the four single factors (DES water content, solid-liquid ratio, extraction temperature, and extraction time). On the basis of the experimental data, three levels of response surface experiments were conducted for each factor and the extraction rate of AS total flavonoid used as an evaluation index to analyze optimal combinations of the four factors.

6. Macroporous resin pretreatment

Macroporous resin pretreatment was performed according to a previous report, with appropriate adjustments to the operating method. Specifically, macroporous resin was soaked with absolute ethanol for 24 hours and washed with deionized water until the smell of alcohol was eliminated. The resin was further soaked in five times the volume of 5% (g/v) NaOH for 6 hours and washed with deionized water to ensure neutrality, followed by soaking in five times the volume of 5% (v/v) HCL for 4–5 hours and re-washing until neutral pH. For storage, the resin was soaked in absolute ethanol and washed again until the alcohol smell was completely removed before use.

7. Screening for the optimal macroporous resin and analytical solvent

Columns (16 mm×40 cm) were wet-packed with the four types of four pretreated macroporous resin. Flavonoid crude extract solution (25 mL; 1.285 mg·mL^-1) was mixed with DES/8 solvent (0.161 mg·mL^-1) as the sample loading solution, and the column loaded at a flow rate of 1 mL·min^-1 to initiate dynamic adsorption. The concentration of total flavonoids in the final effluent of each resin was determined. The column was eluted with deionized water at the same speed until the upper layer of the macroporous resin was colorless, followed by an equal volume of 50% ethanol at a flow rate of 1 mL·min^-1 for dynamic desorption. In order to ensure that the adsorption effect of the resin and the resolution effect of the elution solvent are both optimal, the collected effluent was passed through the column twice and the flavonoid content of each collected solution calculated. All experiments were repeated three times in parallel. The adsorption rate (A, %), resolution rate (B, %) and total flavonoid recovery rate (W1, %) of various resins were calculated.

8. Reusability of DES

Following screening of the optimal recovery conditions, the operation was repeated according to the above steps, with extraction of total AS flavonoids from the recovered DES solvent under the optimized conditions. All experiments were repeated three times in parallel according to the above steps and the total flavonoid yield calculated each time to establish the reusability of DES.