Two reagents were employed for the MAE procedure: HPLC-grade methanol (Fischer Scientifics, Loughborough, UK) and ultra-pure water obtained from a Milli-Q water-purification system (EMD Millipore Corporation, Bedford, MA, USA). Furthermore, a sodium hydroxide solution (NaOH, 1 M) and a hydrochloric acid solution (HCl, 1 M), both from Panreac (Barcelona, Spain) were used to adjust the pH of the extraction solvents. For the analysis and quantification of the flavonols in the extract, different reagents were employed. Milli-Q water, methanol, and formic acid (Merck KGaA, Darmstadt, Germany) were used for the identification of the flavonols by liquid chromatography coupled with mass spectrometry, while Milli-Q water, acetonitrile (Panreac, Barcelona, Spain) and acetic acid (Merck KGaA, Darmstadt, Germany) were employed for the quantification of the flavonols by ultra-high-performance liquid chromatography. Quercetin 3-O-glucoside supplied by Sigma-Aldrich (Steinheim, Germany) was the standard used for the quantification of the flavonols. The reagent DPPH (2,2-diphenyl-1-picrylhydrazyl) radical scavenging (Sigma—Aldrich, San Luis, MO, USA) was employed for the quantification of the antioxidant activity. The standard for the quantification of the antioxidant activity was 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) supplied by Sigma—Aldrich (Steinheim, Germany).
2.3. Microwave-Assisted ExtractionAs mentioned above, a microwave-assisted extraction procedure was employed to obtain onion extracts with a large amount of total flavonols and high antioxidant activity. The MAE equipment specifically employed for this study was a MARS 6 One TouchTM Technology system (1800 W) (CEM Corporation, Matthews, NC, USA). The following protocol was followed for routine sample preparation: 0.2 g of the homogenized onion sample was weighed into 75 mL MARSXpress vessels (CEM Corporation), and then the extraction solvent was added to each vessel. The vessels were securely closed and placed inside the microwave oven with eight other tubes with the same methanol percentage and total volume of solvent. The variables to be considered for the study were adjusted to each experimental condition. The factors to be investigated in this study were within the following ranges: solvent (50:50–100:0% methanol:water), pH (2–7), temperature (50–100 °C) and sample:solvent ratio (0.2:10–0.2:20 g:mL). The initial extraction time remained constant at 5 min, plus an additional cooling period. The onion extracts were centrifuged at 1702× g for 5 min. The precipitate was centrifuged again after re-dissolving in 5 mL of the same extraction solvent. Both supernatants were added to a 25 mL volumetric flask, which was then filled up to the mark with the same solvent. Finally, the flask content was stored at −20 °C until further analysis.
2.4. Identification and Quantification of FlavonolsThe method used to identify flavonoids and subsequently quantify their content in the onion samples had been previously developed by our research group [26]. Before both analyses, the extracts were filtered through a 0.20 µm nylon syringe filter (Membrane Solutions, Dallas, TX, USA), and 3 µL were injected into each extraction system.For the identification of the flavonols in the onion samples, UHPLC coupled with a quadrupole-time-of-flight mass spectrometric system (Q-ToF-MS) (Xevo G2 QToF, Waters Corp., Milford, MA, USA) was employed. The general characteristics of the UHPLC-Q-ToF-MS method were the following: a C18 analytical column (1.7 µm, 2.1 × 100 mm., AQUITY UPLC CSH C18, Waters Corp., Milford, MA, USA), a binary solvent system (phase A, 2% solution of formic acid in water, and phase B, 2% solution of formic acid in methanol) and a flow rate of 0.4 mL min−1. The compounds identified according to their retention time and molecular weight were the following: quercetin 3,7,4′-O-triglucoside (2.873 min, m/z = 787.1421), quercetin 7,4′-O-diglucoside (4.610 min, m/z = 625.1396), quercetin 3,4′-O-diglucoside (5.151 min, m/z = 625.1398), isorhamnetin 3,4′-O-diglucoside (5.311 min, m/z = 639.1559), quercetin 3-O-glucoside (5.202 min, m/z = 463.0886), quercetin 4′-O-glucoside (5.543 min, m/z = 463.0873) and isorhamnetin 4-O’-glucoside (5.614 min, m/z = 477.1040).
After the flavonols in the red onion extracts had been identified, an UHPLC coupled with a photodiode array (PDA) detector (ACQUITY UPLC® H-Class, Waters Corporation, Milford, MA, USA) was used for their separation and quantification. The general characteristics of the UHPLC-PDA method were the following: a C18 analytical column (1.7 µm, 2.1 × 50 mm., UPLC® BEH C18, Waters Corporation, Milford, MA, USA), a binary solvent system (phase A, 2% solution of acetic acid in water, and phase B, 2% solution of acetic acid in acetonitrile), a flow rate of 0.6 mL min−1 and a column at 55 °C. The gradient employed, which had been previously optimized based on a multi-response (MRO) Box–Behnken (BBD) experimental design, allowed the separation and quantification of the seven flavonols that have been previously mentioned in less than 2.7 min [26]. The UHPLC chromatogram obtained can be seen in Figure 1. The flavonols were quantified by means of Empower3 Chromatography Data Software (Waters Corporation, Milford, MA, USA) with the PDA set at 360 nm. Specifically, a calibration curve of quercetin 3-O-glucoside (y = 13,884.19× − 19,765.16; R2 = 0.9991; 0.5–100 mg L−1) was constructed. The curve was built by injecting the quercetin 3-O-glucoside standard at different known concentrations (mg L−1) and measuring the area of the peak obtained (µV sec) (plot of peak areas versus concentrations). Since the other flavonols have similar chemical structures, and therefore similar absorbance, they were quantified based on this curve, but taking into account the molecular mass ratio of each compound. The sum of the extracted amounts of each of the flavonols identified has been expressed in this work as total flavonols (TF) and constitutes one of the response variables to be optimized. 2.5. Determining the Antioxidant ActivityThe antioxidant capacity (electron donation) of the extracts obtained was measured as the loss of purple color of the DPPH radical solution. The molecules of α-diphenyl-β-picrylhydrazyl (DPPH; C18H12N5O6) change their optical absorption at 515 nm, and a pale-yellow color appears when the odd electron in its nitrogen atoms is reduced by the presence of antioxidant substances. For the routine determination of the antioxidant capacity of the samples, the following protocol was carried out: 100 μL of onion extract, previously filtered through a 0.45 µm nylon syringe filter (Membrane Solutions, Dallas, TX, USA), was mixed with 900 μL of 0.1 M Tris-HCl buffer (pH 7.4) and 2 mL of the DPPH solution (6 × 10−5 mol L−1 DPPH in methanol). The mixture was incubated (40 min) in the absence of light at room temperature and spectrophotometrically analyzed at 515 nm. This protocol was previously designed by Brand-Williams et al. [27]. The results were expressed as mg of Trolox equivalents (TE) per g of dry-weight sample (mg TE g−1 DW) using Trolox as standard solvent (y = 88.94 x + 0.7478; R2 = 0.9959; 0–1.4 10−3 mol L−1). Both supernatants were added to a 25 mL volumetric flask, which was then filled up using the same solvent. 2.6. MAE Optimization through Experimental DesignIn this study, four factors or independent variables (X1: %methanol in water, X2: pH, X3: temperature and X4: ratio between sample and solvent) and two responses, or dependent variables (YTF: extraction of total flavonols and YDPPH: antioxidant activity of the extracts) have been investigated. Each of the response variables were first independently studied by applying BBD together with RSM. The purpose of using this spherical response surface methodology was to find out how MAE experimental parameters affect the extraction of flavonols, on the one hand, and the antioxidant activity of the extracts, on the other. Furthermore, one of the advantages of a Box–Behnken design is that the factors are in no case either all high or all low at the same time, i.e., no extreme combinations occur [28]. Subsequently, a MRO (Multi-Response Optimization) with the desirability function was simultaneously applied to both variables. MRO has the advantage that the two response variables can be adjusted to produce good yields from a single extraction. This is a very attractive feature from an economic point of view, since it represents a considerable saving with respect to the amount of solvent and length of time used.For the individual optimization, three ranges were established for each factor, as follows: a lower level (−1), an intermediate level (0) and an upper level (1). The levels chosen for this study of onion matrices were selected considering the previous experiences of our research group, and can be seen in Table 1. On the other hand, because of the thermability of the bioactive compounds of interest, temperature was examined by means of an independent stability study, as described in Section 3.1. An analysis of variance (ANOVA) was applied to the results from this stability study and Tukey’s honest significant difference test (Tukey’s HSD) was used to determine the differences between the sample means at 95% confidence level.According to the four factors and one response variable, a BBD design was produced consisting of 27 experimental points with 3 repetitions at the center point, which was randomly conducted. The whole experimental design matrix used can be seen in Table 2.The RSM was applied to the results of the BBD to obtain a regression model and its corresponding second-order polynomial equation (Equation (1)), which allows the correlation of the independent variables with the response one. Furthermore, the suitability of the elaborated model was determined by ANOVA. All these statistical studies were performed by means of the software application Statgraphics Centurion version XVI (Warrenton, VA, USA).Y=β0+∑i=1kβiXi+βii Xi2+∑i∑i=1kβijXiXj+r
(1)
where Y represents the responses; β0 the model constant; βi the coefficient for each main effect; βij the coefficient corresponding to the interactions between factor i and factor j; βii the coefficient of the quadratic factors that represent the curvature of the surface; X each one of the factors studied; and r the residual value (random error).In addition to conducting the individual optimization of each response, the desirability function (D) was used to perform the MRO. The objective of MRO is to maximize D, which is defined as the geometric average of each response variable (Yi) individual desirability function (di). Thus, di is obtained by converting the predicted values from each response surface into a scale-free value within the range 0–1. In order to compare the results obtained from each individual optimization and from the MRO, a two-tailed t-test was used while assuming normal distribution and the same variance. These procedures were also carried out by means of the software application Statgraphics Centurion version XVI (Warrenton, VA, USA).
3. Results and Discussion 3.1. Stability StudyAs mentioned above, the ranges of the independent variables were chosen based on previous experiences of our research group on the same natural matrix [12,28,29]. However, to determine the temperature ranges to be investigated, an additional study on the preservation of the antioxidant capacity of the extracts and its flavonol content when subjected to different temperatures was carried out. Although high temperatures generally promote large bioactive compound recoveries in solid–liquid extraction procedures, they may also promote the oxidative degradation of the compounds of interest [30]. Because of these contrary effects, the whole operational temperature range needs to be evaluated. For this purpose, first of all, a control extract was elaborated and subjected to intermediate extraction conditions (50:50 MeOH:H2O extraction solvent, 0.2:15 g:mL ratio and 20 min extraction time) but without any external temperature. Then, different samples from this control extract (15 mL to keep the same ratio) were subjected to different temperatures (10, 50, 75, 100, 125 and 150 °C) and the same intermediate conditions. The data corresponding to these extracts are shown in Figure 2a,b.Based on the results obtained, it can be concluded that intermediate temperatures do not cause the degradation of the compounds of interest. However, when a certain temperature is reached, 125 °C for flavonols and 150 °C for antioxidant activity, there is a falling trend (according to Tukey’s test, they belong to different groups). This is probably due to the degradation of not only the flavonols, but also the rest of the compounds with antioxidant activity that can be found in the extracts. Based on these results and for the response variables to remain within the optimal region, it was decided to work during the Box–Behnken design in a temperature range between 50 and 100 °C.
3.2. Optimization of the MAE Method for Total FlavonolsWith the BBD matrix finished, the RSM and ANOVA were applied to the data from the experimental extractions. Based on the ANOVA results (Table 3), it could be concluded that the analysis explains 96.77% of the total variability (R2 96.77% and R2 adjusted 93.01%) and that the model fits well, as the lack of fit test presented a p-value (0.33) greater than 0.05.To confirm the validity of the model, the ANOVA results included the coefficients for the parameters of the quadratic polynomial equation and their significance (p-values). Based on these significances, it is possible to know which parameters (factors and/or interactions between factors) have a relevant influence on the response (total flavonols extraction). The parameters with p-values lower than 0.05 were considered to have a relevant influence on the response at a 95% significance level.
In the case of the total flavonols, the linear variables percentage of methanol in the solvent (X1) and pH of the solvent (X2) showed p-values 1X2)) also proved to have some influence on the response variable (p-value lower than 0.05). Regarding the quadratic interaction of the percentage of methanol (X12), it also showed a relevant effect on the response (p-value 10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36] and with those obtained in a previous study published by our research group on the application of UAE to onions [12], which indicate that the characteristics of the solvent used (pH and polarity) has a major influence on the extraction of these bioactive compounds from onion. Regarding the extraction of compounds such as the flavonoid glycosides that can be found in onion matrices, certain polar solvents, such as methanol or ethanol, have been used, since the compounds of interest are more soluble in this type of solvent [31]. However, since different flavonols may present varying polarity, no standardized ideal solvent composition can be applied to the extraction of every flavonol. Thus, the ideal methanol–water ratio to be used as the extraction solvent is generally, as already obtained in this study, one of the major factors that affect the extraction. In addition, polar solvents also contribute to improving heat transfer [36], as they can absorb microwave energy thanks to their high dielectric constant and dielectric loss [9]. In this regard, and for adequate microwave heating efficiency, solvent mixtures should be adjusted according to the dielectric constant that most favors the target extraction from each specific sample [37]. Finally, acidified solvents also promote bioactive compound extractions [38]. Acid solvents promote the breakdown of cell membranes, thus enhancing the release and solubilization of certain compounds [39].A Pareto chart (Figure 3) was used for a more evident representation, where the horizontal bars corresponding to the relevant influencing factors or interactions cross the vertical reference line. In addition, the different colors indicate whether each factor has a positive or negative influence on the response variable. The percentage of methanol showed a positive effect on the response variable (b1 = 1.232), which indicates that by increasing the percentage of methanol in the solvent, the extraction of flavonols from red onion is enhanced. On the other hand, pH exhibited a negative effect (b2 = −0.2212), which indicates that flavonols are better extracted when acidified solvents are used.Finally, based on the coefficients of the effects from the independent variables and their interactions (Table 3), a second-order mathematical equation can be built that predicts the total flavonols extracted (YTF, response variable) as a function of the independent variables (Equation (2)):YTF (mg g−1) = 6.468 + 1.232·X1 − 0.2212·X2 − 0.1450·X3 − 0.1176·X4 − 1.233·X12 − 0.5124·X1X2 − 0.006252·X1X3 + 0.03175·X1X4 + 0.003131·X22 + 0.04233·X2X3 − 0.09485·X2X4 − 0.08731·X32 − 0.1162·X3X4 − 0.1276·X42,
(2)
A two-dimensional (2D) contour plot (Figure 4) using the fitted model was built to graphically represent the main effect from the interactions of the most influential parameters: effects from the methanol–pH interaction on the total flavonols recovery. 3.3. Optimization of the MAE Method for Antioxidant ActivityIn the case of the antioxidant activity, the RSM and ANOVA were also applied to the experimental matrix. The results obtained are shown in Table 4 and in the Pareto chart (Figure 5). According to these ANOVA results, the lack-of-fit test showed a p-value 40]. Furthermore, it could be confirmed that the analysis explains 90.20% of the total variability (R2 90.20% and R2 adjusted 78.77%).With regard to the flavonols, the ANOVA and the Pareto chart allowed us to determine which factors and interactions between them had a more relevant (either positive or negative) influence on the antioxidant capacity of onions. In this case, the linear variables percentage of methanol in the solvent (X1), pH of the solvent (X2), and temperature (X3) showed p-values 2X4) also showed a relevant influence on the response variable, with a p-value (0.047) lower than 0.05. Finally, the quadratic interaction percentage methanol–percentage methanol (X12) also revealed a relevant effect on the response (p-value 2 = −0.66), which results in greater antioxidant activity of the onion extract when lower solvent pH is used. As already mentioned, acid solvents facilitate the extraction of bioactive compounds through the rupture of cell membranes. On the other hand, the effect of the percentage of methanol on this response was contrary to that obtained for the flavonols. In the case of antioxidant capacity, decreases in the percentage of methanol favor the antioxidant capacity of the extracts (b1 = −0.64). This contrary trend is probably because flavonols are not the only compounds that determine the antioxidant capacity of the extracts. Thus, there are other bioactive compounds, such as anthocyanins, that also promote said antioxidant capacity and that can be better extracted using solvents of different characteristics. Finally, in this case, the temperature also affected the antioxidant capacity of the extracts. Specifically, temperature had a positive influence on this variable response (b3 = 0.45), which means that higher temperatures produced extracts with a greater antioxidant capacity. Like pH, higher temperatures favor the extraction of bioactive compounds, which results in extracts with greater antioxidant capacity. This is explained by the higher solubility of the target compounds in a warmer solvent, which achieves an improved mass transfer of the molecules of interest and a more efficient extraction [41].Finally, the second-order mathematical equation (Equation (3)) intended to predict the antioxidant activity of the extracts (YDPPH, dependent variable) as a function of the independent variables was built:
The fitted model was used to elaborate two-dimensional (2D) contour plots as a graphical representation of the main effects from the most influential parameters and their interactions. Figure 6 shows the effect from methanol–pH, methanol–temperature and pH–temperature interactions on the antioxidant activity of the extracts. 3.4. Optimal Conditions for the Individual MethodIn addition to all the above-mentioned factors, the RSM should allow us to determine the values of the independent factors to obtain the best possible values of the response variables, i.e., the maximum extraction of flavonols, on the one hand, and the greatest antioxidant capacity of the extracts, on the other (Table 5).YDPPH (mg g−1) = 8.7372 − 0.6405·X1 − 0.6663·X2 + 0.4522·X3 − 0.2948·X4 − 1.6590·X12 − 0.3184·X1X2 + 0.1641·X1X3 + 0.07450·X1X4 + 0.1587·X22 + 0.03630·X2X3 − 0.6355·X2X4 − 0.06580·X32 + 0.02100·X3X4 − 0.2766·X42,
(3)
Given that, as acid solvents achieve greater yields and with higher antioxidant activity, pH 2 was established as the optimum value. However, regarding the percentage of methanol, the response variables presented opposite trends. Thus, although the optimal percentage of methanol for the extraction of flavonols was as high as 93.8%, with respect to the antioxidant activity of the extracts, the best results were obtained using 74.2% of methanol. This evidences that there are other compounds, apart from flavonols, that also provide antioxidant activity to onion extracts. Temperature was another independent variable that presented opposite trends so that the greatest flavonol yields were obtained at the lowest value of the range studied (50 °C), while the highest antioxidant activity was registered at the upper end of the temperature range (99.9 °C).
3.5. Extraction Time, Repeatability and Applicability to Different Onion VarietiesAfter each optimal factor had been established, the influence from different extraction times on each optimized MAE method was determined. For this purpose, each optimized extraction method was tested in triplicate for 2, 5, 10, 15, 20, 25 and 30 min extraction times. The test data can be seen in Figure 7.From Figure 7a, it can be concluded that 5 min allows the largest yields of total flavonols (7.557 ± 0.3261 mg g−1). A shorter time (2 min) produces smaller yields, since the extraction process probably cannot be completed in such a short time. Longer extraction times do not show statically significant differences according to Tukey´s test. So, for rather similar yields, the best choice would be the one which demands the least time and energy.It can also be concluded from Figure 7b that 10 min produces the onion extracts with the highest antioxidant activity (12.08 ± 0.03788 mg TE g−1); while, as in the case of total flavonols, no significant differences can be noticed as the time taken grows longer, according to Tukey´s test.Once the optimum extraction times had been established, the precision of each method was studied in terms of repeatability and intermediate precision. Both aspects were evaluated by analyzing the coefficients of variance (CV): 30 experiments were conducted on three consecutive days to determine their intermediate precision, and 10 experiments were conducted on a single day to determine their repeatability. If both CVs are less than 5%, the method can be considered precise according to the AOAC for this type of assessment [42]. The MAE method for total flavonols showed the following CVs for repeatability and intermediate precision, respectively: 2.051% and 3.012%. The MAE method for antioxidant activity showed the following CVs for repeatability and intermediate precision, respectively: 3.980% and 4.561%. So, both methods can be considered precise.Finally, both optimized and validated MAE methods were applied to several onion varieties to verify their efficacy in onions with different chemical compositions. Thus, 15 types of onion of different color, origin and variety were studied. The results regarding individual flavonols, as well as TFs, correspond to the optimized MAE method for the extraction of TF, and the results regarding antioxidant activity correspond to the optimized MAE method for maximum antioxidant activity. The results obtained are shown in Table 6.It can be concluded that both MAE methods, for the extraction of total flavonols and to obtain onion extracts with high antioxidant activity, could be successfully applied to many onion varieties, which confirms their applicability to the production of extracts from onions with different chemical compositions. Such differences in chemical composition can be observed, for example, in the purple varieties, which contain, apart from flavonols, anthocyanins (a type of compounds that are only found in red/purple vegetables) in their matrices, and their extracts exhibit greater antioxidant activity.
3.6. Comparison against Other Extraction MethodsAs previously mentioned, other authors have reported on different extraction methods to obtain flavonol-rich extracts, as well as extracts with high antioxidant activity. Our research group carried out a study on the development of a UAE method for these goals. As a summary of the results obtained in this study, our data have been compared against those reported by other studies, so that the efficiency of MAE can be appreciated.
It should first be pointed out that many of these studies employed traditional techniques for onion extraction purposes. Ren F. et al. [17] carried out a homogenization and shaking of the onion bulbs overnight (Red Baron and Hyskin varieties). By employing these extraction techniques, they reported 5.61 ± 0.27 mg g−1 and 0.90 ± 1.42 mg g−1 of total flavonols (sum of the individual flavonols analyzed by HPLC) in each variety, respectively. Regarding antioxidant activity, they reported 0.44 ± 0.01 mg g−1 and 0.01 ± 0.00 mg g−1 in each variety employing DPPH Antioxidant Power Assay. Majid I. et Nanda V. [18] centrifuged different varieties of onion bulbs for 15 min. They identified five individual flavonols by HPLC and reported the following total flavonols in an assortment of onion varieties: 212.37 ± 2.76, 199.10 ± 2.14, 279.50 ± 5.3, 302.80 ± 2.66, 359.80 ± 5.52, 440.20 ± 7.94, 149.50 ± 3.50 and 147.60 ± 2.45 mg kg−1. Traditional extraction techniques required long extraction times (in many cases, up to 24 h) but did not achieve large recoveries.Apart from the traditional techniques, some articles found in the bibliography also investigated microwave-assisted extraction applied to onion matrices. In this regard, even though Zill-e-Huma et al. [8,21], Kumar B. et al. [9] and Soltoft M. et al. [22] employed MAE, they did not carry out any optimization of the extraction factors by DOE with RSM. They simply optimized some of the factors individually. According to the HPLC analyses carried out by Zill-e-Huma et al., flavonols were obtained at amounts between 3.30 mg g−1 and 0.04 mg g−1, depending on the onion variety studied. Kumar B. et al., reported 2.09 mg g−1 obtained from the onions’ edible outer layers and apical trimmings of red onion, and Soltoft M. et al., reported 0.46 mg g−1 from onion bulbs. Even though using an efficient extraction technique such as MAE is very important, correct optimization of the extraction factor is also vital. When the most suitable conditions for each specific purpose are applied, it is possible to maximize the recovery of the compounds of interest, minimize the extracted adjunct compounds and limit the degradation or alteration of the natural state of the extracts.It should also be mentioned that three articles have been found in the bibliography that use MAE in combination with an experimental design. Thus, Jin E.Y. et al. [20] carried out a Central Composite Design (CCD) with RSM. These authors specifically studied how the percentage of ethanol and time affect the extraction of quercetin, as determined by HPLC. These studies reported extractions of 4.84 ± 0.14 mg g−1 quercetin from onion skin. Dairi S. et al. [23] and Pal, C.B. et Jadeja, G.C. [24], similarly to the method employed in this work, used a Box–Behnken design. However, these authors used the total phenolic compounds as the response variable determined by the Folin–Ciocalteu method. So, the MAE method developed in this work achieved yields comparable to or even greater than those obtained by the above-mentioned authors. This is probably attributable to the greater number of factors that have been considered for the DOE and to the chromatographic analysis used.Finally, it is rather interesting to compare the results obtained in this study against our previous work, where UAE was used as the extraction method [12]. The red onion samples employed were alike, and the extractions and analyses were carried out at the same moment, employing the two optimized developed methods: UAE and MAE, respectively. On the one hand, the UAE method that had been developed for total flavonols could extract 8.193 ± 0.2500 mg g−1 flavonols from red onion bulbs, employing a solvent at 79% methanol content, pH 2, 60 °C temperature and 10 min extraction time. On the other hand, the onion extracts obtained through the UAE method could only reach antioxidant activity of 11.45 ± 0.05592 mg g−1 of TE from the red onion bulbs by employing a solvent at 77% methanol, pH 2, 59 °C temperature and 10 min of extraction time. First, it should be highlighted that the optimized MAE method used greater percentages of methanol stretching close to 100% for flavonols’ maximum yields. This fact had already been observed when UAE and MAE were compared against each other about total phenolic compounds (TPC) or total anthocyanins (TA) yields [28,29]. This coincidence led to the conclusion that, regardless of the bioactive compound of interest, MAE applied to onions is more efficient when using high methanol percentages. This is probably explained by, as previously mentioned, the dielectric constant of methanol, which enhances the solvent-heating efficiency of the microwaves. It should also be borne in mind that MAE is more efficient when used to extract compounds with a greater affinity for methanol than to water (according to the rule of equal polarity between solvent and compounds to be extracted that has been previously explained). Secondly, and based on the specific results obtained, it can be concluded that UAE achieves greater flavonol yields than MAE. This is mainly due to the two major compounds in onions: flavonol 3. quercetin 3,4′-O-diglucoside and flavonol 6. quercetin 4′-O-glucoside, which are more easily extracted by UAE (3.517 ± 0.04120 and 3.661 ± 0.04810 mg g−1, respectively) than by MAE (2.932 ± 0.03603 and 3.053 ± 0.02494 mg g−1, respectively), probably due to the greater amount of polar solvent used for UAE. However, regarding antioxidant activity, this is higher in the extracts obtained by MAE. This is explained by the larger amounts of CFT and TA extracted by MAE when compared against those obtained by UAE. Since bioactive compounds with electron donation capacity are more efficiently extracted by MAE rather than by UAE, MAE extracts present a greater content of these compounds and, logically, superior antioxidant activity. 3.7. Multi-Response OptimizationIn addition to the optimization of each individual method, a multi-response optimization was also conducted. Multi-response optimization allows the simultaneous optimization of the two methods according to all the factors considered. MRO provides optimal results for both response variables with a single extraction, which represents a considerable saving in costs and time. The optimum conditions established were the following: 84.9% methanol in water as the extraction solvent, with a pH of 2.5, 50 °C extraction temperature, and 0.2:10 g:mL sample–solvent ratio. By applying these extraction conditions to the same red-onion sample for 10 min, onion extracts with the following characteristics were obtained: 7.63 ± 0.00 mg g−1 total flavonols and 12.05 ± 0.24 mg Trolox equivalent g−1. Assuming that the values follow a normal distribution and that there is no difference between the variances, a two-tailed t-test was applied to compare these results with those obtained by applying the individual methods using the same extraction time. Regarding TFs, it can be stated that there is a statistically significant difference between the means of the two variables (7.72 ± 0.02 mg g−1 of the individual optimization vs. 7.63 ± 0.00 mg g−1 of the multi-response optimization) because the p-value of the F-test is less than 0.05. Regarding the antioxidant activity, the result obtained by applying the multi-response optimization (12.08 ± 0.24 mg Trolox equivalent g−1) does not differ from the one obtained when using the individually optimized method (12.05 ± 0.24 mg Trolox equivalent g−1) and according to the F-test (the p-value was 0.89). These results confirm that the multi-response optimization method would be suitable for their application in quality control analytical laboratories, where time and costs must be minimized [43].
留言 (0)