*2.3. Optimization of Extraction Conditions for Flavonoids by RSM*

Four factors were considered for optimizing the extraction conditions of the DES prepared using GA and CC: temperature, extraction time, DES content, and stirring speed. According to the results of the single-factor experiment (Figure S2), the four factors were subjected to the Box–Behnken design (BBD) analysis. The three levels of extraction temperature were 30, 60, and 90 ◦C, while those of the extraction time were 1, 24.5, and 48 h. The three levels of DES content were 10, 50, and 90%, and those of the stirring speed were 450, 850, and 1250 rpm. The design parameters were determined using the Design-Expert software and the results are listed in Table 2. The Design-Expert software was used to simulate the results of the BBD analysis. The obtained quadratic polynomial regression equation for the extraction efficiency Y, as the sum of the contents of rutin, hyperoside, and isoquercitrin, versus the temperature (A), extraction time (B), DES content (C), and stirring speed (D) was as follows:

*Y (content of flavonoids) =* 147.43 − 16.44*A* − 10.20*B* − 28.63*C* + 25.72*D* − 32.66*AB* + 1.18*AC* − 15.59*AD* + 10.62*BC* − 0.6583*BD* − 9.24*CD* + 20.68*A*<sup>2</sup> + 10.77*B*<sup>2</sup> − 48.72*C*<sup>2</sup> + 13.77*D*<sup>2</sup>


**Table 2.** Design and experimental response values based on BBD analysis.

The results of the variance analysis of the above equation are listed in Table 3. According to the analysis of variance (ANOVA), the *p*-value (0.0441) of the four-factor global model was significant. The model F-value of 2.57 implied that the model was significant. The quadratic coefficients (A2, B2, C2, and D2) also had a significant effect, while the linear coefficients (A, B, C, and D) and terms representing the interaction between two factors had a negligible effect. Therefore, the above model can be employed to predict the extraction efficiency of the DES toward PUL. In the DES extraction, the content of DES was found to be the variable with the highest influence on the content of flavonoids. It also affected the order of the stirring speed, extraction temperature, and extraction time. The results confirmed that all the interactions, except that between the extraction temperature and extraction time, were insignificant. Figure 3 depicts a three-dimensional (3D) response surface curve based on the regression polynomial. It represents the range of three independent variables, while the fourth variable is held at level 0. As can be observed in Figure 3A, the extraction efficiency increased as the extraction temperature and time increased. On the other hand, it tended to increase initially and then decrease, depending on the DES content. The maximum extraction efficiency was achieved at a DES content of 30.85%. At low stirring speeds, the extraction efficiency was found to increase with temperature. However, when the stirring speed is fast, the extraction efficiency decreases as the temperature increases, and then increases again when the temperature reaches 60 ◦C. The optimal extraction conditions predicted from the three-dimensional response surface curves were as follows: extraction temperature: 30◦C, extraction time: 1 h, DES content: 30.85%, and stirring speed: 1113.1 rpm. The optimized conditions corresponded to a predicted flavonoid content of 217.515 μg/mL. Three confirmatory tests, performed under the predicted conditions, showed high significance, with error rates within 5%, and flavonoid yields of 223.37, 219.10, and 227.11 μg/mL.


**Table 3.** ANOVA statistical results.

**Figure 3.** Response surface in Box–Behnken experimental design for the four factors influencing the extraction of flavonoids. (**A**) Effect of temperature and extraction time; (**B**) effect of temperature and DES content; (**C**) effect of temperature and stirring speed; (**D**) effect of extraction time and DES content; (**E**) effect of extraction time and stirring speed; and (**F**) effect of DES content and stirring speed.

#### *2.4. Evaluation of Antioxidant Activity*

The antioxidant activity of the flavonoids extracted from the PUL under the optimum conditions was evaluated at various concentrations. The flavonoid concentration corresponding to 50% free radical scavenging activity was calculated based on L-ascorbic acid, which constituted the positive control group. As can be observed in Figure 4, the 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical scavenging activity increases as the concentration of the extract is increased. However, it does not increase when the maximum concentration is reached. The concentrations of L-ascorbic acid (positive group) and the eluted extract corresponding to 50% DPPH radical scavenging activity were 59.54 and 299.68 ppm, respectively. The antioxidant activity of the active component of the PUL extracted using the DES was approximately 1/6 that of L-ascorbic acid. It was, thus, confirmed that the active components of the PUL extracted using the DES had potential utility as antioxidants.

**Figure 4.** DPPH radical scavenging activity of *PUL* extract under optimal extraction conditions.
