Response Surface Methodology Optimization of Ultrasonic-Assisted Extraction of Acer Truncatum Leaves for Maximal Phenolic Yield and Antioxidant Activity

This study is the first to report the use of response surface methodology to improve phenolic yield and antioxidant activity of Acer truncatum leaves extracts (ATLs) obtained by ultrasonic-assisted extraction. The phenolic composition in ATLs extracted under the optimized conditions were characterized by UPLC-QTOF-MS/MS. Solvent and extraction time were selected based on preliminary experiments, and a four-factors-three-levels central composite design was conducted to optimize solvent concentration (X1), material-to-liquid ratio (X2), ultrasonic temperature (X3) and power (X4) for an optimal total phenol yield (Y1) and DPPH• antioxidant activity (Y2). The results showed that the optimal combination was ethanol:water (v:v) 66.21%, material-to-liquid ratio 1:15.31 g/mL, ultrasonic bath temperature 60 °C, power 267.30 W, and time 30 min with three extractions, giving a maximal total phenol yield of 7593.62 mg gallic acid equivalent/100 g d.w. and a maximal DPPH• antioxidant activity of 74,241.61 μmol Trolox equivalent/100 g d.w. Furthermore, 22 phenolics were first identified in ATL extract obtained under the optimized conditions, indicating that gallates, gallotannins, quercetin, myricetin and chlorogenic acid derivatives were the main phenolic components in ATL. What’s more, a gallotannins pathway existing in ATL from gallic acid to penta-O-galloylglucoside was proposed. All these results provide practical information aiming at full utilization of phenolics in ATL, together with fundamental knowledge for further research.


Introduction
Acer truncatum is a prominent maple (Aceraceae) species widely cultivated in China, Korea and Japan, and is also found in Europe and Northern America [1,2]. In northern China, maple leaves, mainly A. truncatum leaves (ATL), are often used as health drinks and folk medicines for treating coronary artery cirrhosis, cerebrovascular diseases and angina pectoris [3]. Previous investigations of ATL indicated that it possessed various biological functions, such as antioxidant [4], antibacterial [5,6], and antitumor properties [7,8] due to its high contents of tannins, flavonoids, and chlorogenic acid. However, little effort has been made to optimize of methods of extraction of phenolics from ATL. Therefore, in order to ensure the full utilization of ATL, in the current study an ultrasound-assisted extraction (UAE) method was established through response surface methodology (RSM). UAE is well known to display positive effects on the rate of various extraction processes used in the food, pharmaceutical and cosmetic industries. Compared to other conventional methods, UAE offers a net advantage in terms of productivity and selectivity with shorter processing times, enhanced quality, reduced chemical and physical hazards, and it is environmentally friendly and causes less damage to the structural and molecular properties of compounds in plant materials [9,10]. For these reasons, UAE from plant materials has been widely used lately to facilitate extractions of phenols from Chrysanthemum morifolium flower heads [10], Olea europaea fruits [11], Morus nigra must [12], and Curcuma longa rhizomes [13], lipids from rice [14], polysaccharides from pumpkin [15] and mycelial fermentation of Phellinus igniarius [16], proteins from Jatropha curcas seeds [17], lignans from Schisandra chinensis fruits [18] and other value-added compounds from various natural resources. However, to the best of our knowledge, no report is available regarding improvements of phenolic yield by itself, let alone the phenolic yield together with antioxidant activity from ATL by using ultrasonic stimulation. Thus the present study could provide useful information for the industrial manufacture of ATL as a phenolic and antioxidant resource.
Being a valuable tool to investigate the interaction among factors and quantitatively depict the effects of given parameters on their measured responses [19], RSM is a collection of statistical and mathematical methods for developing, improving, and optimizing a process [20]. In comparison with single variable optimization methods, RSM is a time-and cost-effective means of simultaneously evaluating interactions as well as the key experimental parameters [21], thus have been wildly applied to optimize many bioprocesses [17,[22][23][24]. Central composite design (CCD) used in RSM, is superior compared with the classical approach in terms of the comprehensiveness of information gained and the accuracy of the experiments conducted [22,25], thus is applied in the present study to optimize the UAE process.
The LC-MS/MS technique is gaining increasing interest in the characterization of phenolic components due to its high selectivity and sensitivity [26][27][28]. For example, Melguizo-Melguizo et al. [29] analyzed 22 compounds from Artemisia vulgaris leaves, 15 of them were phenolics; and Kolniak-Ostek [30] tentatively identified 65 phenolic components in ten pear cultivars. Therefore, the variety of phenolics in ATL extract, determined with UPLC-QTOF-MS/MS analysis under the optimized extraction condition, was analyzed to investigate the phenolic composition as well as the quality of the extraction.
Overall, the present study was designed to optimize the UAE process with RSM and CCD for the maximum phenolic yield and antioxidant activity from ATL. The phenolic composition of the ATL extract obtained under the optimized UAE conditions was then analyzed with the UPLC-QTOF-MS/MS technique. These results should contribute to the baseline data for industrial manufacture as well as further exploitation of ATL as a phenolic and antioxidant compound source.

Solvent Types
The phenolic yield obtained with different extraction solvents (water, methanol, ethanol, 1:1 (v:v) methanol:water, or 1:1 (v:v) ethanol:water) was determined first to select the best solvent (combination). The results showed that different solvents significantly affected the phenolic yield ( Figure 1A). The phenolic yield with methanol (5112.17 GAE/100 g d.w.) was slightly higher than that with ethanol (5049. 23), and both of them were significantly higher than that with water (602.63). However, when ethanol was combined with water at the ratio of 1:1 (v:v), the phenol extraction was significantly increased to 7262.61. This result was consistent with several previous studies reporting that aqueous organic solvents exhibit a higher extraction efficiency than absolute organic solvents [31][32][33]. Cujic et al. [34] explained this phenomenon by the fact water is responsible for swelling of the plant material while the ethanol is responsible for disrupting the bonding between the solutes and the plant matrix, and this synergistic effect leads to a higher phenolic yield. However, the phenolic yield obtained with 1:1 (v:v) methanol:water was worse (2369.13) compared with that extracted merely with methanol. This might result from its inappropriate polarity for extractions targeting phenols in ATL. Therefore, ethanol but not methanol, and aqueous ethanol, instead of pure ethanol, was selected as the best solvent for phenolic extraction. material while the ethanol is responsible for disrupting the bonding between the solutes and the plant matrix, and this synergistic effect leads to a higher phenolic yield. However, the phenolic yield obtained with 1:1 (v:v) methanol:water was worse (2369.13) compared with that extracted merely with methanol. This might result from its inappropriate polarity for extractions targeting phenols in ATL. Therefore, ethanol but not methanol, and aqueous ethanol, instead of pure ethanol, was selected as the best solvent for phenolic extraction. min, with material to liquid ratio at 1:20 g/mL, material to liquid ratio (C, extracted at 240 W, 50 °C, 30 min, and 70% aqueous ethanol), extraction temperature (D, 240 W, 30 min, 70% aqueous ethanol, and 1:20 g/mL), sonication power (E, 50 °C, 30 min, 70% aqueous ethanol, and 1:20 g/mL), and extraction time (F, 240 W, 50 °C, 70% aqueous ethanol, and 1:20 g/mL) on the yields of total phenols in ATL by single factor test. Extraction was repeated thrice. Values marked by the same letter are not significantly different (p < 0.05).

Solvent Concentration
Aqueous ethanol of different concentrations was used to conduct the single factor analysis under the following uniform conditions for material-to-liquid ratio, extraction temperature, ultrasonic power, and extraction time: 1:20 g/mL, 50 °C, 240 W, and 30 min, respectively. As shown in Figure 1B, phenolic yield increased from 6850.04 mg GAE/100 g d.w. to 7329.02 for ethanol concentrations between 50% and 60%, and to the maximum (7388.07) at 70%. The efficiency then decreased with further increases in the ethanol concentration, and reached the minimum (5026.35) at 100%. Therefore, it can be concluded that aqueous ethanol exhibited the closest polarity to the phenols in ATL at the range Figure 1. The effect of solvent type (A) and solvent concentration (B) extracted at 240 W, 50 • C and 30 min, with material to liquid ratio at 1:20 g/mL, material to liquid ratio (C, extracted at 240 W, 50 • C, 30 min, and 70% aqueous ethanol), extraction temperature (D, 240 W, 30 min, 70% aqueous ethanol, and 1:20 g/mL), sonication power (E, 50 • C, 30 min, 70% aqueous ethanol, and 1:20 g/mL), and extraction time (F, 240 W, 50 • C, 70% aqueous ethanol, and 1:20 g/mL) on the yields of total phenols in ATL by single factor test. Extraction was repeated thrice. Values marked by the same letter are not significantly different (p < 0.05).

Solvent Concentration
Aqueous ethanol of different concentrations was used to conduct the single factor analysis under the following uniform conditions for material-to-liquid ratio, extraction temperature, ultrasonic power, and extraction time: 1:20 g/mL, 50 • C, 240 W, and 30 min, respectively. As shown in Figure 1B, phenolic yield increased from 6850.04 mg GAE/100 g d.w. to 7329.02 for ethanol concentrations between 50% and 60%, and to the maximum (7388.07) at 70%. The efficiency then decreased with further increases in the ethanol concentration, and reached the minimum (5026.35) at 100%. Therefore, it can be concluded that aqueous ethanol exhibited the closest polarity to the phenols in ATL at the range between 60% and 80% as the efficiency reached the maximum value with ethanol concentration at 70%. Thus, the variable range of ethanol concentration used in the RSM experiments was 60%-80%.

Material-to-Liquid Ratio
Phenolic yields with different material-to-liquid ratios were evaluated in the single factor analysis under the following uniform conditions for aqueous ethanol concentration, extraction temperature, ultrasonic power, and extraction time: 70%, 50 • C, 240 W, and 30 min, respectively. As shown in Figure 1C, phenolic yield increased slightly from 6772.95 mg GAE/100 g d.w. to 6858.25 at ratios between 1:5 and 1:10, and much rapidly to the maximum (7320.00) at 1:20. The efficiency declined at ratios between 1:20 and 1:30, then increased to a second peak at 1:40 and reached the minimum (5086.68) at 1:50. Based on the mass transfer principles occurred in UAE, the driving force for mass transfer is considered to be the concentration gradient between the solid and solvent [35], as well as the partition coefficient of phenolics between the natural matrix and the solvent [36]. These results revealed that the optimal ratio leading to the strongest driving force was 1:20 g/mL for phenolic extraction in ATL. Thus, the variable range of material-to-liquid ratio used in the RSM experiments was determined as 1:15-1:25 g/mL.

Extraction Temperature
To study the effect of extraction temperature on phenolic yield, UAE was implemented at different temperatures under the following uniform conditions for ethanol concentration, material-to-liquid ratio, ultrasonic power, and extraction time: 70%, 1:20 g/mL, 240 W, and 30 min, respectively. As shown in Figure 1D, phenolic yield increased from the minimum (6840.00 mg GAE/100 g d.w.) at 30 • C to the maximum (7376.17) at 50 • C, decreased to 6890.21 at 60 • C, then slightly increased up to 80 • C with no significant differences thereafter between 60 and 80 • C. These results suggest that a relative high temperature increased phenolic extraction efficiency as it increased the number of cavitation nucleus formed [37], leading an enhanced mass transfer and therefore a better access of solvent to cell components [38]. However, this increase declined, namely, brought a negative effect on the efficiency when the temperature exceeded a certain value. This was mainly due to the decreased cavitation intensity under increased temperatures [39,40]. In addition, the accelerated evaporation would also be induced at high temperatures, which led to the decrease of driving force [41]. A temperature effect on yield with the same pattern in a UAE process was previously reported, such as in the cases of phenols from olive fruit [11], oil from pomegranate seed [41], and carotene from citrus peels [42]. Thus, an extraction temperature ranging from 40 to 60 • C was chosen in the RSM process of ATL.

Ultrasonic Power
The effect of ultrasonic power on phenolic yield was determined under the following uniform conditions for ethanol concentration, material-to-liquid ratio, extraction temperature, and extraction time: 70%, 1:20 g/mL, 50 • C, and 30 min, respectively. As shown in Figure 1E, phenolic yield increased from 6693.94 mg GAE/100 g d.w. at 150 W to the maximum (7327.12) at 240 W. The yield decreased to the minimum (6456.36) at 270 W, then increased to 6959.13 at 300 W. There was no significant difference between 150, 180, and 210 W or between 270 and 300 W. According to a report by Hemwimol et al. [43], when ultrasonic waves with larger amplitude travel through an extraction solvent, more bubbles would be created, and more collapse and violent shockwaves and high-speed jets might be generated to disrupt the cell walls. Therefore, the penetration of extraction solvent into cells became stronger, resulting in more release of target components from cells into the mass medium. Consequently, the mass transfer rate was thus enhanced and this led to an appropriate increase in yield. However, the extraction efficiency tended to decrease when the ultrasonic power was higher than 240 W, and this might be caused by degradation of bioactive compounds and overproduced bubbles which could hamper the propagation of ultrasound waves at too high an ultrasonic power [44]. Thus, an ultrasonic power range from 210 to 270 W was chosen in the RSM process of ATL.

Extraction Time
The effect of extraction time on phenolic yield was determined under the following uniform conditions for ethanol concentration, material-to-liquid ratio, extraction temperature, and ultrasonic power: 70%, 1:20 g/mL, 50 • C, and 240 W, respectively. As shown in Figure 1F, a biphasic increase in phenolic yield was observed with the increase of the extraction time. Phenolic yield increased from 6760.65 mg GAE/100 g d.w. to 7421.41 when the sonication time was prolonged from 10 to 30 min, then declined to the lowest value (6411.25) at 60 min, and the yield increased again to the maximum (7458.61) at 120 min. There was no significant difference between phenolic yields at 10 and 30 min. The sonication time was recognized as a significant factor increasing the extraction efficiency of UAE processes, however, the extraction efficiency decreased in some cases due to a prolonged sonication time as it will lower the permeability of solvent into the cell walls due to oversuspended impurities [14]. Moreover, prolonged extraction time may increase the chances of decomposition of phenolics [45] and also potentially increase the loss of solvent by vaporization [46], which can directly affect the loss of mass transfer during extraction. Considering energy-savings, 30 min was chosen to conduct the experimental design tests described below.

Statistical Analysis and Model Fitting
To optimize the operating parameters, 30 random sequential experiments were performed under the designed UAE conditions based on the ranges of every single factor (independent variable) determined above to study the reciprocal influence of independent variables (i.e., solvent concentration, material-to-liquid ratio, extraction temperature and ultrasonic power) on the two dependent (response) variables (i.e., phenolic yield and its corresponding antioxidant activity).
The experimental values, together with predicted values obtained by their response surface central composite design, are presented in Table 1.
The final quadratic equation obtained in terms of actual factors upon applying multiple regression analysis to the experimental data is given below: When the phenolic extraction efficiencies (Y 1 ) were considered as the response: When the antioxidant capacities (Y 2 ) were considered as the response: where, X 1 , X 2 , X 3 , X 4 , Y 1 , Y 2 are the aqueous ethanol concentration, material-to-liquid ratio, extraction temperature, ultrasonic power, phenolic yield response and antioxidant activity response, respectively. The linear effect of solvent concentration (X 1 ) was found to be significant for both response variables while X 4 was only significant for phenol yield (Y 1 ), and X 2 and X 3 were only significant for antioxidant activity (Y 2 ). It can be concluded that solvent concentration was the vital parameter in both responses. However, the quadratic effect (X 1 2 , X 2 2 , X 3 2 ) was found to produce extremely significant (p < 0.01) positive effect on both Y 1 and Y 2 , but X 4 2 was only significant for Y 2 .

Factors
Total Phenols (mg/100 g) DPPH (µmol/100 g) DPPH Activity/Total Phenols (µmol/mg) ANOVA results for multiple regression analysis and response surface quadratic model of Y 1 and Y 2 were evaluated using the corresponding F and p values ( Table 2). F values of Y 1 and Y 2 were calculated to be 25.95 and 50.34, both leading to a p value < 0.0001, suggesting that both the models were statistically extremely significant. The models' coefficient of determination (R 2 ) were 0.9604 and 0.9792, indicating that more than 96.04% and 97.92% of the response variability is explained, and supporting a good accuracy and ability of the established model within the range limits used [47]. Correlation coefficients of 0.9800 and 0.9898 also indicate a good positive correlation between the actual data and the predicted values obtained using CCD. The F-values of Lack of Fit of Y 1 and Y 2 were 0.6032 and 0.2015, respectively, implying that the Lack of Fit is not significant relative to the pure error, thus the model can be used to predict the phenol yield and corresponding antioxidant activity of ATL. In addition, Adj-R 2 , Pre-R 2 and the coefficient of variation (C.V.) were calculated to check the model's adequacy. Pre-R 2 of Y 1 and Y 2 were 0.8341 and 0.8968, which were in reasonable agreement with their Adj-R 2 of 0.9233 and 0.9597, respectively (Adj-R 2 − Pre-R 2 < 0.2), indicating a high degree of correlation between the measured and predicted data from the regression model [48]. The C.V. expressed the standard deviation as a percentage of the mean, and was found to be 1.5169% (<5.00%) for the phenolic yield, and 2.3491% (<5.00%) for antioxidant activity, implying that the models were reproducible [24]. Adequate precision measures the signal to noise ratio, which is desirable when the value is larger than 4. The ratios of 22.3775 and 28.1801 both referred an adequate signal and illustrated that the models (Y 1 and Y 2 ) were applicative for the present UAE process [49]. These correlation analyses between predicted values and actual data can be used to evaluate the suitability of the response surface model [50]. Thus, the model can be used to predict the phenolic yield and corresponding antioxidant activity under various extraction conditions during UAE process. As shown in Table 1, the levels of phenolic yield ranged from 5783.04 to 7591.30 mg GAE/100 g d.w., and the levels of antioxidant activity ranged from 42,030.84 to 73,849.78 µmol TE/100 g d.w. The highest levels of phenolic yield (7256.52 to 7587.83 mg GAE/100 g d.w.) and antioxidant activity (70,844.41 to 73,849.78 µmol TE /100 g d.w.) were obtained under the center point combinations of 70% ethanol, material-to-liquid ratio of 1:20 g/mL, 240 W, 50 • C, and 30 min. Moreover, the conditions involving a solvent concentration of 66.21%, material-to-liquid ratio of 1:15.31 g/mL, ultrasonic temperature 60 • C, power 267.30 W, and extraction time of 30 min were predicted to provide the highest phenolic yield, together with the highest antioxidant activity according to the fitted models. Ratios of DPPH activity/total phenols were calculated and we found that these ratios ranged between 7.29 and 9.91, which means the phenolic components extracted in all 30 runs were similar. Moreover, the ratios of six repetitions with the condition at the central point (9.60-9.91) were higher than those of the others, further indicating that the antioxidant activity of phenolics extracted under the condition represented by the central point was superior and the results of single factor experiments were reliable. Total phenols of the leaves collected in April optimized with UAE in the present study was lower than those (93.08 mg/g, i.e., 9308 mg/100 g) of fallen leaves reported by Cai et al. [51] prepared with a microwave method. However, according to another study conducted by ourselves [52], Acer truncatum fallen leaves naturally possessed a much higher total phenols and radical scavenging activity than those collected in other seasons such as in April. Considering the difference in sampling seasons mentioned above, the data discrepancy between our two studies might be considered to be within a reasonable range.
To further verify the models obtained from RSM, ATL was extracted under the predicted optimal UAE conditions, and its phenolic yield and antioxidant ability were evaluated and compared to the predicted maximum yield. For operational convenience, the optimal parameters were modified slightly in the verification experiment as follows: solvent concentration 66.20%, material-to-liquid ratio 1:15.30 g/mL, ultrasonic power 270.00 W, temperature 60 • C, and extraction time 30 min. The predicted phenolic yield and antioxidant activity under the optimal conditions were 7589.19 mg GAE/100 g and 74,010 µmol TE/ 100 g, and the experimental values under the optimal conditions were 7579.56 ± 354.44 mg GAE/100 g and 73,585.78 ± 790.74 µmol TE/100 g, respectively. No significant differences were observed between predicted and experimental values (p > 0.05), indicating that the experimental results confirmed the adequate fitness of the predicted model.

Effect of Interactions Among Variables on Phenolic Yield and Antioxidant Activity in ATL
To visualize the interactions of two operational parameters on extraction efficiency, the responses were generated as planar contour plots (Figure 2). Two variables unshown in the Figures were kept constant at their respective central experimental values and the other two variables presented on the two horizontal axis varied within their experimental ranges in order to understand their main and interactive effects on the dependent variables. Figure 2a-f show the results of interactive influence of solvent concentration, material-to-liquid ratio, extraction temperature, and ultrasonic power on phenolic yield, while Figure 2g-h exhibit the impact of these variables on antioxidant activity. Figure 2a-c, g-i show the phenols and antioxidant activities as responses of aqueous ethanol concentration (X 1 ) and the other factors (X 2 , X 3 , X 4 ). The extraction efficiencies first increased then decreased with increase of aqueous ethanol concentration from 50% to 90%, material-to-liquid ratio from 1:10 to 1:30 g/mL, ultrasonic power from 180 to 300 W, and extraction temperature from 30 to 70 • C.
The interaction of solvent concentration and material-to-liquid ratio (X 1 X 2 ) showed an extremely significant positive effect (p < 0.001) on both responses (see Table 2). At lower ethanol concentration with increasing material-to-liquid ratio, total phenols and antioxidant activity kept generally mild. However, the total phenols with antioxidant activity first increased, then decreased at lower material-to-liquid ratio with increasing ethanol concentration (Figure 2a,g).
However, the total phenols with antioxidant activity first increased, then decreased at lower materialto-liquid ratio with increasing ethanol concentration (Figure 2a,g). The possible interaction mechanism between solvent concentration and material-to-liquid ratio might be interpreted as that the positive interaction caused by appropriate solvent concentration and material-to-liquid ratio affected the polarity and viscosity of aqueous ethanol and solubility of target phenolic compounds in the extraction solvent, thus influenced the yield of phenols.
As shown in Figure 2f,l, the interaction between sonication power and temperature (X3X4) also showed a significant positive effect (p < 0.05) on total phenols and antioxidant activity (Figure 2f). The possible interaction mechanism between temperature and power may be the change of cavitation threshold affected by changing temperature, which is responsible for acoustic cavitation and also results in the formation of a cavitational nucleus. The influence of relatively greater forces ruptures and erupts the formed cavitational nucleus and disrupts the cell tissues during extraction, which in turn enhances the mass transfer rate [38].
With respect to antioxidant activity (Y2), the interaction effect of solvent concentration and The possible interaction mechanism between solvent concentration and material-to-liquid ratio might be interpreted as that the positive interaction caused by appropriate solvent concentration and material-to-liquid ratio affected the polarity and viscosity of aqueous ethanol and solubility of target phenolic compounds in the extraction solvent, thus influenced the yield of phenols.
As shown in Figure 2f,l, the interaction between sonication power and temperature (X 3 X 4 ) also showed a significant positive effect (p < 0.05) on total phenols and antioxidant activity (Figure 2f). The possible interaction mechanism between temperature and power may be the change of cavitation threshold affected by changing temperature, which is responsible for acoustic cavitation and also results in the formation of a cavitational nucleus. The influence of relatively greater forces ruptures and erupts the formed cavitational nucleus and disrupts the cell tissues during extraction, which in turn enhances the mass transfer rate [38].
With respect to antioxidant activity (Y 2 ), the interaction effect of solvent concentration and sonication power (X 1 X 4 ) was also found to be significant (p < 0.05). According to the study of Hemwimol et al. [43], only a small fraction of the electric energy from the ultrasound actually entered the extraction solvent in the ultrasonic bath system, and most of it was absorbed by the water in the bath. Under this circumstances, the rise of the solvent concentration might increase the utilization efficiency of the limited electric energy, thus playing a vital role in the improvement of the extraction yield. Table 3 shows a list of 29 phenolic compounds identified in ATL prepared under the optimal extraction conditions obtained above through UPLC-QTOF-MS/MS experiments, along with their retention times (RT), experimental m/z, calculated m/z, error values (ppm), molecular formula and MS/MS fragments. As signals in negative mode was stronger than those in the positive mode, data collected in negative mode was thus chosen to conduct the identification. Mass spectra in negative ion mode ( Figure 3A), together with MS ( Figure 3B) and MS/MS ( Figure 3C) fragments of the peak with retention time at 26.51 min were presented as follows. bath. Under this circumstances, the rise of the solvent concentration might increase the utilization efficiency of the limited electric energy, thus playing a vital role in the improvement of the extraction yield. Table 3 shows a list of 29 phenolic compounds identified in ATL prepared under the optimal extraction conditions obtained above through UPLC-QTOF-MS/MS experiments, along with their retention times (RT), experimental m/z, calculated m/z, error values (ppm), molecular formula and MS/MS fragments. As signals in negative mode was stronger than those in the positive mode, data collected in negative mode was thus chosen to conduct the identification. Mass spectra in negative ion mode ( Figure 3A), together with MS ( Figure 3B) and MS/MS ( Figure 3C) fragments of the peak with retention time at 26.51 min were presented as follows.

Gallates and Gallotannins Derivatives
Compound 2 (R t = 4.03 min) was identified as theogallin (a galloylquinic acid) with a [M − H] − ion at m/z 343.0654. It showed a major fragment at m/z 191 (quinic acid) due to the loss of a galloyl moiety (152 amu) and another fragment at m/z 169 (gallic acid) due to the loss of a part of the quinic acid moiety (174 amu). Compound 3 (4.29 min) gave a [M − H] − ion at 169.0143 and a major fragment ion at m/z 125 caused by the loss of a -CO 2 group, and was identified as gallic acid. Compound 4, with a [M − H] − ion at m/z 183.0323 and a major fragment ion at m/z 124 due to the loss of -CO 2 CH 3 (59 amu) was assigned as methyl gallate. The parent ions and MS/MS profiles of these three gallates accorded with the data reported in the literature [53].
Moreover, the gallic acid ion at m/z 169 occurred not only in compounds 2 and 3, but also in compounds 16, 21, 22, and 26, indicating that all these compounds share the gallic acid moiety.  [54]. Compound 26 (56.11 min) with a deprotonated ion at m/z 1091.12 was identified as hexa-O-galloyl-glucoside. It fragmented at m/z 939 and 769 following the mode described above. The parent ions and fragmentation mode of these gallotannins were reported in the literature [55].
Accordingly, we hypothesized that a gallotannins pathway exists in ATL, leading to the biosynthesis of tetra-O-galloylglucoside, penta-O-galloylglucoside and hexa-O-galloylglucoside from gallic acid. Grundhöfer et al. [56] elucidated a pathway forming complex gallotannins in Rhus typhina through enzyme studies. They reported the detailed conversion process from gallic acid to β-glucogallin, digalloyglucose, trigalloyglucose, tetragalloyglucose, pentagalloylglucose, and hexagalloyglucose. As most of galloylglucoses mentioned above were also found in ATL, we conjecturethat the gallotannins pathway existing in ATL might be quite similar with the one in Rhus typhina. In addition, 1,2,3,4,6-pentakis-O-galloyl-β-D-glucose was the common precursor of two subclasses of hydrolyzable tannins, the gallotannins and the related ellagitannins [53]. Thus the pathway from gallic acid to penta-O-galloylglucoside might be the tip of the iceberg. Unfortunately, gallotannins with a higher degree of polymerization than that of hexagalloyglucose were not found in the present study.

Flavonoids
Compound 29 (71.72 min) with a parent ion at m/z 301.0354 was identified as quercetin by comparison with the MS/MS data of a previous study [57]. Moreover, the quercetin moiety as a daughter ion at m/z 301 also existed in compounds 10, 15, 19, 20, 23-25, and 28, from which the abundance of quercetin derivatives in ATL could be inferred. Compound 10 (16.08 min) was assigned as quercetin 3-rhamninoside as its molecular ion was at m/z 755.2031 [58]. It fragmented at m/z 301 (quercetin) as its aglycone ion, and fragmented at m/z 609 due to the loss of a rhamnosyl (146 amu), and fragmented at m/z 463 due to another loss of a rhamnosyl. Compound 15 (19.38 min) was identified as quercetin-3,7-O-alpha-L-dirhamnopyranoside with a [M − H] − ion at 593.1513 and a daughter ion at 447 caused by the loss of a rhamnosyl (146 amu) [59]. According to Lin and Harnly [60], the glycosylated quercetins with a monosaccharide at the same position elute from C 18 columns in the following order: galactoside, glucoside, xyloside, arabinopyranoside, arabinofuranoside, rhamnoside and glucuronide, and this order was further confirmed by Keinänen and Julkunen-Tiitto [61].  [57], 20 [57], 25 [53], and 28 [59] were verified.
Compounds 13 and 14 shared a parent ion at m/z 479, and their common daughter ion at m/z 316 was assigned as a myricetin moiety. According to Heras et al. [28], compounds 13 (18.65 min) and 14 (19.01 min) were identified as myricetin-O-hexoside I and myricetin-O-hexoside II, respectively. Besides compounds 13 and 14, compounds 17 and 18 also shared the myricetin moiety ion at m/z 316, thus they were considered myricetin derivatives. Compound 17 (21.95 min) with molecular ion at m/z 449.0727 was identified as myricetin-arabinoside/xylopyranoside as its MS/MS profile corresponded with that of the literature [27]. Compound 18 (23.10 min) was assigned as myricetrin-3-O-rhamnoside (myricitrin) with a [M − H] − ion at m/z 463.0878 and an aglycone ion at m/z 316 [62].
Compound 11 (16.39 min) was identified as catechin, as its molecular ion was observed at m/z 289.0717. It fragmented at m/z 245 and 211, corresponding with its MS/MS data in the literature [26]. Compound 27 (60.30 min) possessed a deprotonated ion at m/z 431.0974, and was identified as kaempferol-3-O-rhamnoside [59]. Its fragment ion at 285 was assigned as the kaempherol moiety.

Chlorogenic Acid Derivatives
Compound 1 (2.35 min) with deprotonated ion at m/z 191.0564 was identified as quinic acid by comparing its MS/MS data with the literature [29]. Quinic acid is a cyclic polyol, instead of a phenol itself, however, it is an important part of caffeoylquinic acid derivatives. Compounds 4 (6.91 min) and 5 (7.85 min) were found to share a deprotonated ion at m/z 353, indicating the presence of monocaffeoylquinic acid isomers. According to Ncube et al. [63] and Melguizo-Melguizo et al. [29], the common base peak in their MS/MS spectra at m/z 191 illustrated the presence of a quinic acid moiety, and the fragment ion at m/z 179 occurring in compound 4 indicated that caffeoyl group is linked to the 3-OH position of quinic acid. Therefore, compound 4 was identified as 3-O-caffeoylquinic acid, while compound 5 was identified as 5-O-caffeoylquinic acid due to the different fragmentation pattern.
Compounds 7 (12.64 min) and 8 (12.93 min) were identified as p-coumaroylquinic acid isomers as they shared a deprotonated ion at m/z 337. According to Ncube et al. [63], transand cis-5-p-coumaroylquinic acids possess a base peak at m/z 191, while transand cis-4-p-coumaroylquinic acid possess one at m/z 173. As they shared a base peak at m/z 191 in the present study, and considering their reported eluting order, compounds 7 and 8 were identified as transand cis-5-p-coumaroylquinic acids, respectively.

Other Phenolic Compounds
Compound 6 (11.05 min) with a molecular ion at m/z 285.0612 was identified as uralenneoside as it fragmented at m/z 153 due to the loss of a hydroxybenzoic acid unit and at m/z 109 because of another loss of a CO 2 group from the carboxylic acid moiety, and its MS/MS data were accordance with the fragmentation pattern reported by Yu et al. [64]. Compound 12 (16.91 min), with a deprotonated ion at m/z 863.1819, was identified as cinnamtannin B1 as it consisted of three epicatechin units, which was reflected by the daughter ion at m/z 289 [65].
Twenty Acer truncatum Bunge trees, whose tree-age (around 15 years old) and growth environment were approximatively identical, were randomly selected in Bajiajiaoye Park, Haidian District (N 40 • 00 57.21 ; E 116 • 19 43.36 with altitude 39-41 m) and authenticated by associate professor Zhonghua Liu, Beijing Forestry University, Beijing, China. A. truncatum leaves (ATL) were uniformly collected from the selected trees in 20 April 2015 and taken back to the laboratory immediately. The leaves were cleaned with distilled water and air-dried, the dried leaves were ground and passed through a 250 × 250-µm sieve, and the powder was stored at −20 • C in a refrigerator for extraction.

Preliminary Experiments
To test the impact of solvent type on phenolic yield, 1.000 g leaf powder was mixed thoroughly with different solvents, including water, methanol, ethanol, methanol: water (1:1, v:v) or ethanol:water (1:1, v:v) at a material-to-liquid ratio (g/mL) of 1:20 in a plastic centrifuge tube. The tube was then immersed into a tunable ultrasonic cleaning bath (KQ-300DE type, Kunshan Ultrasonic Instrument Co., Ltd., Kunshan, China) with the liquid level in the tube kept lower than that of the cleaner tank, and extracted under ultrasonic conditions at 50 • C, 240 W and 30 min. The sample mixture was then centrifuged for 10 min at 6000 rpm and the supernatant was collected. The resulting residue were repeated for extraction twice more with the same volume of solvent under the specific material-to-liquid ratio, and all the supernatants were combined, filtered, and diluted to a final volume of 60 mL. The resulting solutions were analyzed for total phenols. The optimal solvent with the highest total phenol content was selected for the following experiments.

RSM Experiment
After the single factor tests, RSM with CCD was applied to estimate the effect of independent variables (i.e., X 1 , extraction temperature; X 2 , ultrasonic power; X 3 , solvent concentration; X 4 , liquid-to-material ratio) and their interactions on UAE of phenolic yield (Y 1 ) and antioxidant activity (Y 2 ). Based on preliminary single factor analysis and literature data, levels of independent parameters were selected and coded at five levels according to Equation (1): where x i and X i are the coded and actual values of an independent variable, respectively. X 0 is the actual value on the center point of X i , and ∆X i is the value of the step change. The design values of independent variables and their coded values are represented in Table 4. In the present study, CDD conducted 30 experimental points including six replicates at the central point, sixteen factorial points and eight axial points for a full factorial design to study the effect of independent variables on the response. The experiments were randomized and the response values in each trial were analyzed using Design-Expert (Version 8.0.6, Stat-Ease Inc., Minneapolis, MN, USA) and fitted to a second-order polynomial regression model expressing mathematical relationship between independent variables (X 1 , X 2 , X 3 and X 4 ) and responses (Y 1 and Y 2 ): where, Y (Y 1 or Y 2 ) is the predicted response, X i and X j are diverse input variables that influence the response variable Y, β 0 is the constant coefficient, β i is the linear coefficient, β ij is the interaction coefficient of two factors (X i and X j ), and β ii is the quadratic coefficient of one factor (X i 2 ). The obtained model was verified by comparing the phenolic yield and antioxidant ability of the ATL extract obtained under the optimal UAE conditions to those predicted by the models. Furthermore, the resulting solution was analysed by UPLC-QTOF-MS/MS to determine its phenolic composition.

Measurement of Total Phenols
Total phenols were determined according to the Folin-Ciocalteu method [68] with slight modifications. In brief, 40 µL of 25% Folin-Ciocalteu solution was added to designated wells of a 96-well microplate, followed by addition of 20 µL standard (10-400 mg/L gallic acid dissolved and diluted with 50% ethanol, R 2 = 0.999), ATL extract solution or blank (MilliQ water). After blending, 140 µL of 700 mM Na 2 CO 3 solution was added to each well, and the plate was shaken for 5 min at 250 rpm and incubated in dark at 40 • C for 30 min, followed by absorbance measurement at 765 nm with a microplate reader (xMark™ Microplate Absorbance Spectrophotometer, Bio-Rad, Hercules, CA, USA). Results were expressed as mg gallic acid equivalent (GAE)/100 g d.w. of ATL powder.

Determination of DPPH• Scavenging Activity
Assays of DPPH• scavenging activity were performed using the method described by Alañón et al. [69] with slight modifications. In brief, 10 µL standard (10-400 mg/L Trolox dissolved and diluted with 50% ethanol, R 2 = 0.998), ATL extract solution or blank was added to a designated well of a 96-well microplate. Subsequently, 40 µL of 1 mM freshly prepared DPPH solution was added to each well followed by addition of 190 µL methanol, and the plate was then stood in an orbital shaker setting at 200 rpm for 1 min. After 30 min incubation at room temperature in dark, absorbance was recorded at 517 nm using the microplate reader. Results were expressed as µmol Trolox equivalent (TE)/100 g d.w. of ATL powder.

Statistical Analysis
Experiment design was performed using Design-Expert. All experimental results obtained were expressed as means ± SD, and data were analyzed by analysis of variance (p < 0.05) using SPSS software (Version 22.0, SPSS Inc., Chicago, IL, USA). All analyses were performed in triplicate.

Conclusions
The current study is the first report on the effect of ultrasonic stimulation on the phenolic yield and antioxidant activity of ATL. A quadratic model fitted well both responses, i.e., total phenol yield and antioxidant activity. Operating parameters were optimized using single factor experiments and a response surface design. Optimal conditions were found to be percentage of aqueous ethanol 66.21%, material-to-liquid ratio 1:15.25 g/mL, extraction temperature 60 • C; ultrasonic power 270 W, and ultrasound time 30 min, which gave a maximum total phenol yield of 7593.62 mg GAE/100 g d.w. and antioxidant activity of 74,241.61 µmol TE/100 g d.w. The verified results confirmed the adequate fit of the predicted model. Furthermore, 22 phenolic compounds were discovered in the present study with UPLC-QTOF-MS/MS analysis, which indicated the superiority of the extraction conditions and provided new insights into the phenolic components of ATL. All this information should be useful for both researchers and consumers, and contribute to the better utilization of ATL as a source of phenolic compounds with high antioxidant activity.