Optimization of subcritical water extraction of phenolic antioxidants from pomegranate (Punica granatum L.) peel by response surface methodology

Linlin Yan a, Yungang Cao bc and Guangyao Zheng *a
aInstitute of Chemical Industry of Forest Products CAF, National Engineering and Technology Research Center of Forest Chemical Industry, Key Laboratory of Biomass Energy and Material, Nanjing 210042, Jiangsu Province, China. E-mail: guyazheng@sina.com; Fax: +86 25 85482465; Tel: +86 25 85482465
bBeijing Engineering and Technology Research Center of Food Additives, Beijing Technology & Business University, Beijing 100048, People's Republic of China
cSchool of Food and Biological Engineering, Shaanxi University of Science and Technology, Xi'an 710021, People's Republic of China

Received 14th June 2017 , Accepted 17th July 2017

First published on 18th July 2017

Subcritical water extraction (SWE), a ‘green’ and efficient extraction technology, was applied to extract phenolic antioxidants from pomegranate peel in this study. In single-factor experiments, the effects of extraction temperature (100–220 °C), time (5–80 min) and the water/solid ratio (20–60 mL g−1) on the extraction yields of total phenolics (TP), total flavonoids (TF), and major phenolic compounds (punicalagin, punicalin, and ellagic acid) from pomegranate peel were investigated, as well as their antioxidant capacities. The results showed that the yields of phenolic compounds and antioxidant activities were significantly affected by these parameters. The highest levels of TP (314.65 mg GAE per g), TF (153.66 mg RE per g), and antioxidant activities (DPPH and ABTS radical-scavenging abilities) were observed at 130 °C for 20 min with a solid/water ratio of 50 mL g−1. The highest yield of punicalagin (81.04 mg g−1) from pomegranate peel was also observed under these conditions. However, punicalin and ellagic acid yields both continued to increase with increasing temperature and extension of time. Subsequently, on the basis of the single-factor experiments, three independent variables, temperature (X1, 110–150 °C), time (X2, 10–30 min), and the water/solid ratio (X3, 40–60 mL g−1), were optimized to maximize the total phenolic yields (Y1) and DPPH antioxidant activities (Y2) by a response surface methodology with a three-factor, three-level Box–Behnken design. The optimized conditions for phenolic antioxidant extraction were 126.1 °C, 18.5 min, and a water/solid ratio of 54.8 mL g−1, and the corresponding predicted values of Y1 and Y2 were 323.10 mg GAE per g and 476.81 mg Trolox equivalent antioxidant capacity (TEAC) per g, respectively.


Pomegranate (Punica granatum L.) has been considered a health-beneficial fruit that can reduce the risk of cardiac and cerebrovascular diseases because it possesses numerous nutritional components and bioactive compounds, some with potent antioxidant activities.1–3 As by-products of pomegranate consumption or industrial utilization, pomegranate peels have been used in China in traditional medicine for various diseases, such as diarrhea, metrorrhagia, and stomach ache.4 Pomegranate peels contain abundant phenolic compounds, such as ellagic acid, gallic acid, punicalagin, gallotannin, and ellagitannins,3,5,6 and show much stronger antioxidant activities than pomegranate seeds and juice.3,7–9

Various conventional extraction methods have been developed for extracting bioactive phenolics from pomegranate peels, such as leaching, heating, or ultrasound-assisted extraction with aqueous and organic solvents.8,10 Most of these technologies using aqueous and organic solvents show good extraction yields, but require too much time and some organic solvents are potentially toxic to the environment or human health.11 Thus, there is a need for new extraction technologies with low processing costs, mild operating conditions, short process times, and environmentally friendly solvents.

Subcritical water extraction (SWE) is a ‘green’ extraction method developed recently using water as the solvent.12 By increasing the temperature between 100 and 374 °C and keeping the pressure high enough to maintain the liquid state (below the critical pressure of 22 MPa), the physical properties of water are changed and specifically the dielectric constant (ε) of water becomes similar to that of an organic solvent, like ethanol or methanol.11,13,14 Thus, subcritical water can dissolve many compounds with medium or low polarity which would dissolve readily in organic solvents.15,16 In fact, SWE has been used successfully to extract natural phenolics, flavonoids, and anthocyanins from various foods, vegetables, and plant materials, such as grape skin and canola meal.13,14,17–19.

Considering its extensive application in the extraction of phenolics, SWE has been used as a promising technology for extracting bioactive phenolics from pomegranate seeds.11 However, few efforts have been made regarding the application of SWE to extract phenolics from pomegranate peels. Moreover, little is known about the influences of the vital extraction parameters (e.g., temperature, time, and water/solid ratio) on the phenolic extraction yield, phenolic composition, or antioxidant activities, even though these parameters would be expected to have significant impacts on the chemical profile and bioactivity.11,20–22

Thus, the aims of this study were (1) to assess the possibility of using SWE to extract phenolic antioxidants from pomegranate peels, (2) to determine the impact of extraction parameters on the extraction yields of phenolic compounds and their antioxidant activities, and (3) to optimize the extraction conditions using a response surface methodology (RSM) to maximize the total phenolic yield and antioxidant capacity. The present study provides fundamental information for the application of SWE in industrial pomegranate peel extraction.


Chemical reagents and materials

Folin–Ciocalteu reagent, 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azinobis-3-ethylbenzothiazoline-6-sulfonic acid (ABTS), and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). The standard substances of gallic acid, rutin, punicalagin, punicalin, and ellagic acid, were purchased from Nanjing Spring & Autumn Biological Engineering Co., Ltd. (Nanjing, China). Acetonitrile, methanol, and formic acid for high performance liquid chromatography (HPLC) analysis were purchased from Merck (Darmstadt, Germany). Water was obtained from a Milli-Q apparatus (Millipore, Milford, MA, US). All other analytical reagents were purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China).

Fresh pomegranate fruits of the “Qingpi” cultivar were collected during the 2015 harvest season in Zaozhuang, Shandong province. According to a previous method,23,24 pomegranate peels were separated manually, dried in the shade, and then ground to powder (60 mesh) using a pulverizer, and sealed and stored at −20 °C until analysis.

Conventional organic/water solvent extraction

Conventional organic/water extraction methods, such as heat reflux extraction (HRE), Soxhlet extraction (SE), and ultrasonic-assisted extraction (UAE) methods, using methanol, ethanol, and water solutions were performed for comparison. Dried powder of pomegranate peels (2 g) was extracted using HRE with 100 mL of water, 80% (v/v) methanol, and 80% (v/v) ethanol aqueous solutions at 80 °C in a constant-temperature water bath (three times, 2 h each). For the SE, 2 g of dried powder of pomegranate peels were added into 100 mL of 80% (v/v) methanol and 80% (v/v) ethanol aqueous solutions and extracted at 80 °C in a water bath for 2 h (three times). A UAE device was also used: 2 g dried pomegranate peel powder (M) were extracted using a UAE with 100 mL of water, 80% (v/v) methanol, and 80% (v/v) ethanol aqueous solutions for 1 h (three times) at 80 °C. After the extractions, the slurries were combined, filtered, and adjusted to 500 mL (V) with the corresponding extraction solutions. The resulting extract solutions were stored at −20 °C until analyzed for total phenolics and antioxidant activities.

Subcritical water extraction

An SWE experiment was carried out using an SWE apparatus (Jiangsu Huaan Scientific Research Devices Co., Ltd., Nantong, Jiangsu, China). The dried powder of pomegranate peel (2 g) was loaded into the stainless steel vessel (inner volume, 600 mL) packed in the oven with a heating jacket. The oven was connected with temperature and pressure sensors, and the parameters were shown on digital displays. After closing the stainless steel cap, 100 mL of deionized water was added into the extraction vessel. And, nitrogen had been used to remove dissolved oxygen before extraction.

Parameters such as temperature, time, and water/solid ratio, are considered important in the SWE process;21 however, the influence of pressure was considered to be insignificant in previous reports.13,25,26 Thus, an SWE temperature series (100–220 °C) of phenolic profiles and antioxidant activities of pomegranate peels was first analyzed at a constant pressure of 3.0 MPa, a water/solid ratio of 50 mL g−1, and an extraction time of 10 min. Then, an SWE extraction time series (5–80 min) was analyzed at a constant temperature of 130 °C, a water/solid ratio of 50 mL g−1, and a pressure of 3.0 MPa. Finally, a water/solid ratio (20–60 mL g−1) series of extractions was carried out at 130 °C and 3.0 MPa for 20 min. All samples were extracted twice and the resulting slurries were combined, filtered, adjusted to 500 mL (V) with the methanol solution, and stored at −20 °C until further use.

Determination of total phenolics and total flavonoids

The Folin–Ciocalteu method was used to determine the total phenolic (TP) content.27 Total flavonoids (TF) were determined according to a previous method described by Gong et al. (2015).21 The concentrations (C) of TP and TF in extraction solutions were calculated based on calibration curves of the corresponding reference substances (gallic acid and rutin, respectively). The yields (W) of TP and TF in pomegranate peels were expressed as mg gallic acid equivalents (GAE) per g and mg rutin equivalents (RE) per g DW (dry weight) pomegranate peel powder, respectively. Specifically, the resulting calculation was based on the following equation:
W = CV/M(1)
where W (mg g−1) is the yield of TP or TF, C (mg mL−1) is the concentration of TP or TF in the extraction solution, V (mL) is the volume of the resulting extract solution, and M is the dried weight of pomegranate peel powder.

Determination of antioxidant capacity

Antioxidant capacities were determined by DPPH and ABTS free radical-scavenging assays.27,28 For the DPPH assay, an aliquot of extract solution (0.1 mL) was mixed with freshly prepared DPPH solution (0.25 mL, 1 mM in methanol) and methanol (2 mL), after which the mixture was incubated for 20 min in the dark at room temperature before the absorbance was recorded at 517 nm. For the ABTS assay, the extract solution (0.2 mL) was mixed thoroughly with the ABTS˙+ solution (2.8 mL) and after 6–10 min in the dark at room temperature, the absorbance was read at 734 nm.

The radical-scavenging activity (RSA) was calculated as RSA% = 100(A0AS)/A0, where AS is the absorbance of the extract solution and A0 is the absorbance of a control solution, prepared with no added extract. A calibration curve was prepared using Trolox as a standard, and the DPPH and ABTS free radical-scavenging activity of the extract solution (CL) was calculated as mg Trolox equivalent antioxidant activity (TEAC) per mL. Finally, the extraction efficiency of TEAC (ET) of different methods from pomegranate peels was calculated based on the following equation:

ET = CLV/M(2)
where ET (mg g−1) is the extraction efficiency of TEAC of different methods from pomegranate peels, CL (mg mL−1) is the concentration of TP or TF in the extraction solution, V (mL) is the volume of the resulting extract solution, and M is the dried weight of pomegranate peel powder.

Identification and quantification of phenolic compounds

Major individual phenolic compounds were identified and quantified by ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) and HPLC according to the method previously described by Yan et al. (2017).24 UPLC-diode array detection/electrospray ionization-mass spectrometry (UPLC-DAD/ESI-MSn) analyses were carried out using an Agilent 1260 Infinity II LC system coupled with a QTRAP® 5500 system (AB, USA). Identification of the phenolic compounds (punicalagin, punicalin, and ellagic acid) in pomegranate peel was performed by comparison of retention times, UV-vis absorption spectra, and ESI-MS spectra with reference standards and literature data. The concentration of each phenolic compound (CP, mg mL−1) in the extraction solution was calculated based on the calibration curve of each reference substance, and the yields of phenolic compounds (YP, mg g−1) from pomegranate peel were calculated as follows:
YP = CPV/M(3)
where YP (mg g−1) is the extraction yield of each phenolic compound from pomegranate peels, CP (mg mL−1) is the concentration of each phenolic compound in the extraction solution, V (mL) is the volume of the resulting extract solution, and M is the dried weight of pomegranate peel powder.

Experimental design and analysis using RSM

On the basis of single-factor experiments, three independent variables (X1, temperature; X2, time; X3, water/solid ratio) of the SWE process were optimized to maximize the TP yield (Y1, mg GAE per g) and DPPH TEAC extraction efficiency (Y2, mg TEAC per g) from pomegranate peels by RSM with a three-factor and three-level Box–Behnken design (BBD). In this study, X1X3 were the three independent variables coded at three levels, −1 (lower limit), 0 (central point), and +1 (upper limit), as shown in Table 2. The BBD was carried out in 17 random-order experiments including five replicates at the central point (Table 2).

The response variables of Y1 and Y2 related to X1, X2, and X3 were fitted to the following second-order polynomial model equation:

Yn = α0 + α1X1 + α2X2 + α3X3 + α11X12 + α22X22 + α33X32 + α12X1X2 + α13X1X3 + α23X2X3(4)
where Yn (Y1 or Y2) is the predicted responses for X1X3; α0 is a constant coefficient; α1, α2, and α3 are linear coefficients; α11, α22, and α33 are quadratic coefficients; and α12, α13, and α23 are interaction coefficients. The experimental design and the multiple linear regression analysis were conducted using the Design-expert software (ver. 8.0.6; Stat-Ease Inc., Minneapolis, USA). The results were tested statistically by analysis of variance (ANOVA) with a 0.05 significance level. The adequacy of the models was evaluated by the coefficient of determination (R2), coefficient of variance (C.V.%), and p values for the model and lack of fit testing.

Statistical analysis

Data are expressed as means ± standard deviation (SD) of at least triplicate analyses per sample. Data were evaluated by ANOVA and regression analysis using the SPSS software (ver. 20.0; SPSS, Inc., Chicago, USA). Differences among samples were analyzed using Duncan's multiple range tests (p < 0.05).

Results and discussion

Comparison of the SWE and conventional extraction methods

To evaluate the feasibility of the SWE method, we compared several conventional methods (HRE, SE, and UAE) with SWE in terms of extraction efficiencies of TP, TF, and individual phenolic compounds, and the antioxidant properties of pomegranate peel. The results in Table 1 show that the extraction method significantly affected the yields of phenolic compounds and their antioxidant activities. Of the three conventional methods, methanol solution was the best solvent to extract TP and TF from pomegranate peels, followed by ethanol, while water showed the lowest yields. However, for the SWE method, water at 120 °C as the extraction solvent showed higher extraction efficiencies of TP and TF than HRE, SE, UAE, and SWE at 100 °C, and was even higher than with methanol and ethanol solutions by UAE. Similar to previous studies, the extraction efficiency of the SWE method with an increased extraction temperature reached a level similar to that with an organic solvent.15
Table 1 Comparison of the yields of phenolic compounds and antioxidant activities of pomegranate peel using different extraction methodsa
Extracts TP (mg GAE per g) TF (mg GAE per g) DPPH (mg TEAC per g) ABTS (mg TEAC per g) Punicalagin (mg g−1) Punicalin (mg g−1) Ellagic acid (mg g−1)
a TP, total phenolics; TF, total flavonoids; TEAC, Trolox equivalent antioxidant capacity, each value in the table represents the mean ± standard deviation (n = 3), means within each column with different letters are statistical significantly different (p < 0.05).
Hot reflux extraction (2 h × 3 times)
Methanol 322.04 ± 5.02a 133.81 ± 1.97b 436.10 ± 3.84b 543.96 ± 4.34a 85.02 ± 2.24a 4.55 ± 0.24d 14.55 ± 0.12b
Ethanol 305.28 ± 3.89b 119.18 ± 2.61def 426.30 ± 2.24c 492.67 ± 13.59cd 69.47 ± 3.69b 3.55 ± 0.17e 13.22 ± 0.07c
Water 274.75 ± 4.65e 117.69 ± 4.10ef 382.94 ± 5.83gh 456.23 ± 6.66e 17.24 ± 1.45g 3.65 ± 0.12e 7.67 ± 0.41h
[thin space (1/6-em)]
Soxhlet extraction (6 h)
Methanol 301.29 ± 3.89b 125.38 ± 1.55c 429.12 ± 1.36bc 523.71 ± 5.11b 31.85 ± 2.53e 3.29 ± 0.21ef 11.36 ± 0.35d
Ethanol 278.74 ± 5.49de 114.71 ± 5.68f 411.15 ± 3.89d 491.77 ± 5.88cd 22.39 ± 0.46f 2.36 ± 0.07g 10.52 ± 0.25ef
[thin space (1/6-em)]
Ultrasound assisted extraction (1 h × 3 times)
Methanol 285.53 ± 3.51cd 123.39 ± 1.49cde 399.27 ± 8.43e 487.72 ± 3.57d 65.74 ± 2.04c 3.62 ± 0.10e 10.69 ± 0.26e
Ethanol 280.14 ± 5.43de 115.95 ± 4.53f 395.56 ± 3.71ef 446.78 ± 6.93e 53.72 ± 1.38d 3.24 ± 0.07ef 8.67 ± 0.41g
Water 275.75 ± 4.19e 101.32 ± 1.14g 379.52 ± 3.84h 414.39 ± 4.74f 24.12 ± 1.67f 2.83 ± 0.07fg 7.87 ± 0.27h
[thin space (1/6-em)]
Subcritical water extraction (10 min × 2 times)
100 °C 266.17 ± 2.16f 106.52 ± 5.23g 389.35 ± 4.24fg 447.59 ± 10.17e 8.85 ± 0.21h 1.62 ± 0.02d 4.62 ± 0.39i
120 °C 290.91 ± 4.07c 124.39 ± 1.87cd 424.26 ± 3.25c 499.96 ± 6.62c 51.83 ± 1.92d 11.65 ± 0.62c 10.03 ± 0.65f
[thin space (1/6-em)]
Optimal condition of subcritical water extraction (126.1 °C, 18.5 min, and 54.8 mL g −1 )
327.32 ± 5.71a 158.12 ± 7.07a 468.24 ± 7.52a 537.21 ± 6.79a 84.69 ± 2.27a 17.53 ± 0.57a 16.82 ± 0.96a

To further clarify the differences in phenolic components, we identified major phenolic compounds (punicalagin, punicalin, and ellagic acid) in pomegranate peel, obtained using the different extraction methods, by comparing retention times, UV-vis absorption spectra, and ESI-MS spectra with reference standards (Fig. S1), and by quantifying the compounds using a calibration curve for each reference substance according to our previous methods.24 It can be seen from Table 1 that the yields of punicalagin, punicalin, and ellagic acid in SWE extracts at 120 °C were higher than other water extracts by conventional methods and SWE at 100 °C, and reached similar extraction efficiencies as organic solvents (methanol or ethanol) with conventional methods. Specifically, there was no difference between the punicalagin yields in SWE at 120 °C (51.83 mg g−1) and UAE with ethanol (53.72 mg g−1). Moreover, the punicalagin yield in SWE at 120 °C was higher than that in SE with methanol or ethanol (31.85 and 22.39 mg g−1, respectively).

By comparing the results of antioxidant activities (Table 1), we found that the methanol extract of HRE showed the highest antioxidant activities, followed by SE with methanol, and HRE with ethanol. Water extracts of conventional extraction methods demonstrated poor free radical-scavenging activities. However, the SWE extracts at 120 °C showed higher antioxidant activities than the ethanol extracts obtained by SE and UAE, which may be due to the higher phenolic contents of SWE at 120 °C.18

In summary, with respect of extraction time (6 h for HRE and SE, 3 h for UAE, and 20 min for SWE) and extraction solution toxicity (methanol, ethanol, and water), SWE showed a more efficient process and other advantages (time-saving, energy-saving, and less toxicity) than conventional extraction methods for the extraction of phenolics from pomegranate peels. Moreover, it is important to optimize SWE parameters, such as extraction temperature, time, and the water/solid ratio, to achieve the maximum extraction efficiency and best antioxidant activity; these parameters are considered important factors in extraction yields in the SWE process.20,21,26

Effects of extraction temperature

It can be seen from Fig. 1a that the yields of TP and TF reached the highest levels (300.09 mg GAE per g and 133.07 mg RE per g) as the temperature increased up to 130 °C, and then decreased rapidly as the temperature continued to increase (130 to 220 °C). As reported in previous studies, subcritical water can dissolve most phenolic compounds at relatively high temperatures, to the same extent as the organic solvents.21 This can be attributed to the ε of subcritical water, which decreases significantly from 61 to 31 with a temperature increase from 80 to 220 °C, which is close to the ε values of methanol (33) and ethanol (25).11,29 However, if heated too much, the phenolic compounds become unstable, probably suffering from degradation or reactions with other compounds.11,21,22
image file: c7ay01475a-f1.tif
Fig. 1 Effects of extraction temperature in SWE on phenolics extraction yields and antioxidant activities. (a) Yields of total phenolics (TP) and total flavonoids (TF) in pomegranate peel. (b) Yields of punicalagin, punicalin, and ellagic acid. (c) ABTS and DPPH antioxidant activities of pomegranate peel.

We analyzed the contents of punicalagin, punicalin, and ellagic acid by HPLC. As shown in Fig. 1b, the yields of punicalagin, punicalin, and ellagic acid increased with increasing extraction temperature (from 100 to 130 °C). At 130 °C, the punicalagin yield reached the highest level of 66.77 mg g−1 DW. As the temperature rose from 140 to 200 °C, the punicalagin yield decreased sharply, from 15.08 to 0.89 mg g−1 DW, and almost no punicalagin was detected at 220 °C. The temperature series of extraction yields for punicalin and ellagic acid were significantly different from that of punicalagin. Specifically, when the temperature increased from 140 to 220 °C, the yields of both punicalin and ellagic acid decreased slightly at first and then increased to maximum values (28.52 and 37.21 mg g−1, respectively) at 220 °C. This was possibly due to degradation of punicalagin at higher temperature, resulting in the release of more punicalin and ellagic acid from punicalagin or other ellagitannins at higher temperatures.30 Ellagic acid is more thermodynamically stable.31 For SWE, the extraction efficiencies of the various compounds depended largely on the extraction temperature.20 The yields of TP, TF, and three phenolic compounds (punicalagin, punicalin, and ellagic acid) were affected significantly by temperature in SWE, and 130 °C was suitable for the extraction of punicalagin, total phenolics, and total flavonoids, simultaneously.

As shown in Fig. 1c, the TEAC values of DPPH and ABTS radical-scavenging activities were affected significantly by the extraction temperature in the SWE process, and the same trends as the TP yields were observed (Fig. 1a). From 100 to 130 °C, the values increased, which may be attributable to the increase in TP yields with the decrease in ε. However, the values of antioxidant activities decreased rapidly from 130 to 220 °C, which may be due to the degradation of some phenolic compounds at higher temperature during SWE. The high correlation coefficients indicated that the DPPH and ABTS radical-scavenging activities were positively correlated with TP yields (Table S1), consistent with previous studies.11,18

Effects of extraction time

Extraction time is a key factor for bioactive compounds to reach an equilibrium concentration in the extraction process, and it may affect not only the extraction efficiency but also the bioactivity of the extracts.21,29,32 We assessed the impacts of extraction time ranging from 5 to 80 min on phenolic yields (Fig. 2a and b) and antioxidant activities (Fig. 2c) at 130 °C with a constant water/solid ratio of 50 mL g−1 and a pressure of 3.0 MPa. The yields of TP and TF increased significantly from 5 to 20 min, and reached the highest level (314.65 mg GAE per g and 153.66 mg RE per g, respectively) at 20 min. However, with extended extraction times (20 to 80 min), the yields of TP and TF decreased sharply. Similar to the trend in TP and TF, the yield of punicalagin increased significantly from 5 to 20 min and reached the highest value (81.04 mg g−1) at 20 min, but then decreased sharply from 20 to 80 min. In comparison, the yields of punicalin and ellagic acid increased at first and reached the highest levels at 60 and 40 min (24.16 mg g−1 and 26.18 mg g−1), respectively. Then, their yields decreased significantly. These data indicated that extraction time extension will not necessarily result in higher extraction efficiencies, and too long an extraction time probably leads to degradation or inter-reaction of some compounds.11,21,30,33
image file: c7ay01475a-f2.tif
Fig. 2 Effects of extraction time in SWE on phenolics extraction yields and antioxidant activities. (a) Yields of total phenolics (TP) and total flavonoids (TF) from the pomegranate peel. (b) Yields of punicalagin, punicalin, and ellagic acid. (c) Antioxidant activities of pomegranate peel.

As shown in Fig. 2c, the TEAC values of DPPH and ABTS radical-scavenging activities were also significantly affected by SWE extraction time. From 5 to 20 min, the antioxidant activities increased significantly with extended time. However, further SWE time extension (>20 min) resulted in decreased antioxidant activities. The SWE time series of antioxidant activity trend was similar to those of TP, TF, and punicalagin, and good correlations among them were observed (Table S1), consistent with previous studies.34 Considering the best performance of TP, TF, and punicalagin yield and antioxidant activities, 20 min was considered the optimal time.

Effects of water/solid ratio

Fig. 3 shows that the influence of the water/solid ratio on the yields of TP, TF, and three phenolic compounds (punicalagin, punicalin, and ellagic acid) was significant. The extraction yields of TP, TF, and the three phenolic compounds improved with an increase in the water/solid ratio (from 20 to 50 mL g−1), and reached the highest levels at 50 mL g−1, then remained relatively stable at 60 mL g−1. Similarly, He et al. (2012)11 found that the TP yields of pomegranate seeds increased with an increased water/solid ratio. Qu et al. (2014)35 also reported that the water/material ratio had significantly positive influences on extraction yields and proanthocyanidins from pomegranate peel. A possible explanation for this is that the higher water/solid ratio resulted in faster diffusion of the phenolic compounds from the materials into the extraction solution.35 Taking the cost and evaporation time into account, 50 mL g−1 was considered the preferred choice.
image file: c7ay01475a-f3.tif
Fig. 3 Effects of water/solid ratio of SWE on phenolics extraction yields and antioxidant activities. (a) Yields of total phenolics (TP) and total flavonoids (TF) from pomegranate peel. (b) Yields of punicalagin, punicalin, and ellagic acid. (c) Antioxidant activities of pomegranate peel.

Model fitting and response surface analysis

We applied RSM to optimize the extraction parameters for subcritical water extraction. The experimental and predicted values of RSM are listed in Table 2. The following multiple quadratic eqn (5) and (6) were used to fit the function relationship between actual factors and response values:
Y1 = −1472.52787 + 23.98205X1 + 7.707055X2 + 7.23550X3 + 0.020012X1X2 − 1.98750 × 10−3X1X3 − 5.70000 × 10−3X2X3 − 0.096047X12 − 0.24907X22 − 0.056620X32(5)
Y2 = −2501.76063 + 36.81450X1 + 8.43650X2 + 21.15650X3 + 0.055100X1X2 − 5.80000 × 10−3X1X3 − 5.00000 × 10−3X2X3 − 0.14877X12 − 0.40727X22 − 0.18568X32(6)
where Y1 and Y2 were the predicted response of TP yields and DPPH antioxidant activity, respectively, and X1, X2, and X3 were the extraction temperature, extraction time, and water/solid ratio, respectively.
Table 2 Three-factor, three-level Box–Behnken experimental design and results of the response surface methodology
Run Code levels Independent variables Response Y1a (mg GAE per g) Response Y2b (mg TEAC per g)
X 1 X 1 X 1 X 1 (°C) X 2 (min) X 3 (mL g−1) Experiment Predicted Experiment Predicted
a TP yield (mg GAE per g). b DPPH antioxidant activity (mg TEAC per g).
1 0 0 0 130 20 50 315.50 316.91 473.25 469.95
2 0 −1 1 130 10 60 304.42 304.68 441.41 437.92
3 0 0 0 130 20 50 317.52 316.91 460.40 469.95
4 1 0 −1 150 20 40 248.98 247.28 360.55 354.62
5 0 0 0 130 20 50 312.48 316.91 463.10 469.95
6 1 1 0 150 30 50 235.94 237.90 347.81 350.25
7 0 1 1 130 30 60 293.33 292.02 422.40 418.09
8 0 1 −1 130 30 40 269.41 269.15 380.90 384.39
9 −1 −1 0 110 10 50 279.22 277.26 413.66 411.22
10 −1 0 1 110 20 60 297.46 299.16 425.52 431.45
11 0 0 0 130 20 50 318.53 316.91 478.98 469.95
12 −1 1 0 110 30 50 258.14 257.74 371.97 370.35
13 1 −1 0 150 10 50 241.01 241.41 345.42 347.04
14 −1 0 −1 110 20 40 273.68 274.34 396.30 394.43
15 1 0 1 150 20 60 271.17 270.51 385.13 387.00
16 0 −1 −1 130 10 40 278.22 279.52 397.91 402.22
17 0 0 0 130 20 50 320.54 316.91 474.02 469.95

The reliability and suitability of the models were evaluated by the ANOVA results and regression coefficients.36 The ANOVA results and regression coefficients of the model for two responses are presented in Table 3. The model F values of Y1 and Y2 were 176.29 and 63.00, respectively, and the corresponding p values were both <0.001, indicating that the models were statistically significant. The lack of fit values of Y1 and Y2 were 0.6283 and 0.5377, respectively, indicating that the lack of fit for the two responses was insignificant (p > 0.05) relative to the pure error. The R2 value is defined as the regression of the sum of squares in proportion to the total sum of squares, which illustrates the adequacy of a model.37 The R2 values of the models were 0.9956 and 0.9878 (close to 1), respectively, indicating good accuracy and ability for the established models.37 The pre-R2 values of Y1 and Y2 were 0.9726 and 0.9128, which were in reasonable agreement with their adj-R2 values of 0.9900 and 0.9721, respectively, indicating that the predicted data from the regression model were highly correlated with the experimental results.38 The adeq. precision measures the signal to noise ratio, and a ratio more than 4 is desirable. The adeq. precision statistic measures the signal to noise ratio; a ratio >4 is desirable. In this study, the ratios of 36.452 and 21.115 both indicated an adequate signal and illustrated that the models (Y1 and Y2) were appropriate for the subcritical water extraction process.39 Additionally, the C.V.% expressed the dispersion degree of the data; it was found to be 0.99 and 1.83 (<10%) for Y1 and Y2, respectively, indicating that the two models were reproducible.40 Thus, the models appropriately fitted the experimental data, and can be used to predict the TP yields and the corresponding DPPH values for pomegranate peel.37

Table 3 Analysis of variance (ANOVA) for the total phenolic yield (Y1) and DPPH antioxidant activity (Y2)
Responses Sources Sum of squares df F-Value p-Value
a Significant difference (p < 0.05). b Highly significant difference (p < 0.01).
Y 1 Model 12[thin space (1/6-em)]673.65 9 176.29 <0.0001b
X 1 4842.00 1 606.18 <0.0001b
X 2 241.07 1 30.18 0.0009b
X 3 147.73 1 18.49 0.0036b
X1X2 64.08 1 8.02 0.0253
X1X3 0.63 1 0.079 0.7866
X2X3 1.30 1 0.16 0.6987
X 12 6218.22 1 778.47 <0.0001b
X 22 2612.04 1 327.01 <0.0001b
X 32 134.98 1 16.90 0.0045b
Residual 55.91 7
Lack of fit 18.13 3 0.64 0.6283
Pure error 37.79 4
R 2, 0.9956; adj-R2, 0.9900; pred R2, 0.9726
Adeq. precision 36.452
C.V.% 0.99
Y 2 Model 32[thin space (1/6-em)]659.62 9 63.00 <0.0001b
X 1 11[thin space (1/6-em)]410.11 1 198.10 <0.0001b
X 2 343.22 1 5.96 0.0447a
X 3 1263.06 1 21.93 0.0023b
X1X2 485.76 1 8.43 0.0229
X1X3 5.38 1 0.093 0.7687
X 2 X 3 1.00 1 0.017 0.8989
X 12 14[thin space (1/6-em)]910.07 1 258.86 <0.0001b
X 22 6984.12 1 121.25 <0.0001b
X 32 1451.59 1 25.20 0.0015b
Residual 403.19 7
Lack of fit 156.07 3 0.84 0.5377
Pure error 247.12 4
R 2, 0.9878; adj-R2, 0.9721; pred R2, 0.9128
Adeq. precision 21.115
C.V.% 1.83

The relationships between the response variables (Y1 and Y2) and the independent variables (X1X3) were visualized by the two-dimensional contour plots of response surfaces (Fig. 4a–f). Y1 and Y2 varied from 241.01 to 320.54 mg GAE per g and 345.42 to 478.98 mg TEAC per g, respectively (Table 2). Both Y1 and Y2 increased initially by increasing temperature and time, but then decreased significantly. Similar to previous reports, this is probably explained by a higher temperature within a range increasing the solubility of the polyphenols, while too high a temperature or a prolonged time heating might result in the degradation of polyphenols during SWE.11,21,22 The interaction between temperature (X1) and time (X2) was significant for the two responses, whereas the interactions of X1X3 and X2X3 were not. Moreover, according to the p values of the regression coefficients (Table 3), X1, X2, X3, X1X2, X12, X22, and X32 were significant model terms (p < 0.05) for both response variables (Y1 and Y2), while the interactions of X1X3 and X2X3 were not significant (p > 0.05).

image file: c7ay01475a-f4.tif
Fig. 4 Contour plots of total phenolic yield (Y1) and DPPH anti-oxidant activity (Y2) from pomegranate peel as a function of extraction temperature (X1), time (X2), and water/solid ratio (X3).

As predicted by the model, the optimal extraction conditions for the maximal values of TP yield (Y1, 323.10 mg GAE per g) and DPPH antioxidant activity (Y2, 476.81 mg TEAC per g) were both obtained at 126.1 °C, 18.5 min, and 54.8 mL g−1. We performed a verification experiment under these conditions, and the mean values of the experiment data for TP yield and DPPH antioxidant activity were 327.32 ± 5.71 mg GAE per g and 468.24 ± 7.52 mg TEAC per g, respectively. No significant difference was observed between the experiment and predicted values. Additionally, the yields of three phenolic compounds (punicalagin, punicalin, and ellagic acid) under these optimized conditions were 84.69 ± 2.27, 17.53 ± 0.57, and 16.82 ± 0.96 mg g−1 DW pomegranate peel, respectively, and higher than the extract efficiencies with water and ethanol solvents in the conventional methods tested (HRE, SE, and UAE).


In this study, SWE was successfully used to extract phenolic antioxidants from pomegranate peel, and showed higher extraction efficiency than conventional extraction methods. The extraction parameters of SWE, including temperature, time, and the water/solid ratio were optimized using RSM with a three-factor, three-level Box–Behnken design. Under the optimal extraction conditions (126.1 °C, 18.5 min, and 54.8 mL g−1), both the total phenolics and DPPH antioxidant activity showed predicted maximum values (323.10 mg GAE per g and 476.81 mg TEAC per g, respectively) that were very close to the experimental values (327.32 mg GAE per g and 468.24 mg TEAC per g, respectively). Thus, subcritical water extraction is not only an environmentally friendly process, but also a highly efficient method for the extraction of bioactive phenolic compounds from pomegranate peel.

Conflict of interest

The authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.

Ethical approval

All applicable international, national, and/or institutional guidelines for the care and use of animals were followed. This article does not contain any studies with animals or human participants performed by any of the authors.


This study was supported by the Program of Innovative Research Team in Institute of Chemical Industry of Forest Products, CAF (LHSXKQ4), the fund of the Beijing Engineering and Technology Research Center of Food Additives, Beijing Technology & Business University (BTBU), and the Prospective Joint Research Project of Production-study-research Cooperation in Jiangsu Province, China (BY2011116).

Notes and references

  1. N. P. Seeram, L. S. Adams, S. M. Henning, Y. Niu, Y. Zhang, M. G. Nair and D. Heber, J. Nutr. Biochem., 2005, 16, 360–367 CrossRef CAS PubMed.
  2. E. P. Lansky and R. A. Newman, J. Ethnopharmacol., 2007, 109, 117–206 CrossRef PubMed.
  3. U. A. Fischer, R. Carle and D. R. Kammerer, Food Chem., 2001, 127, 807–821 CrossRef PubMed.
  4. J. A. T. da Silva, T. S. Rana, D. Narzary, N. Vermae, D. T. Meshramf and S. A. Ranadeg, Sci. Hortic., 2013, 160, 85–107 CrossRef.
  5. J. Lu, K. Ding and Q. A. Yuan, Chromatographia, 2008, 68, 303–306 CAS.
  6. A. Bustamante, A. Hinojosa, P. Robert and V. Escalona, Int. J. Food Sci. Technol., 2017, 35, 85–107 Search PubMed.
  7. R. P. Singh, K. N. Chidambara Murthy and G. K. Jayaprakasha, J. Agric. Food Chem., 2002, 50, 81–86 CrossRef CAS PubMed.
  8. Y. Li, C. Guo, J. Yang, J. Wei, J. Xu and S. A. Cheng, Food Chem., 2006, 96, 254–260 CrossRef CAS.
  9. W. Elfalleh, H. Hannachi, N. Tlili, Y. Yahia, N. Nasri and A. Ferchichi, J. Med. Plants Res., 2012, 6, 4724–4730 CAS.
  10. Z. Pan, W. Qu, H. Ma, G. G. Atungulu and T. H. McHugh, Ultrason. Sonochem., 2011, 18, 1249–1257 CrossRef CAS PubMed.
  11. L. He, X. Zhang, H. Xu, C. Xu, F. Yuan, Ž. Knez, Z. Novak and Y. Gao, Food Bioprod. Process., 2012, 90, 215–223 CrossRef CAS.
  12. M. Herrero, A. Cifuentes and E. Ibañez, Food Chem., 2006, 98, 136–148 CrossRef CAS.
  13. L. Ramos, E. M. Kristenson and U. T. Brinkman, J. Chromatogr. A, 2002, 975, 3–129 CrossRef CAS PubMed.
  14. R. Shaddel, A. Maskooki, M. H. Haddad-Khodaparast, S. Azadmard-Damirchi, M. Mohamadi and B. Fathi-Achachlouei, Food Sci. Biotechnol., 2014, 23, 1459–1468 CrossRef CAS.
  15. M. Plaza and C. Turner, TrAC, Trends Anal. Chem., 2015, 71, 39–54 CrossRef CAS.
  16. A. D. Sousa, A. I. V. Maia, T. H. S. Rodrigues, K. M. Canuto, P. R. V. Ribeiro, R. D. C. A. Pereira, R. F. Vieira and E. S. de Brito, Ind. Crops Prod., 2016, 79, 91–103 CrossRef CAS.
  17. Z. Ju and L. R. Howard, J. Food Sci., 2005, 70, S270–S276 CrossRef CAS.
  18. M. Hassas-Roudsari, P. R. Chang, R. B. Pegg and R. T. Tyler, Food Chem., 2009, 114, 717–726 CrossRef CAS.
  19. M. J. Ko, C. I. Cheigh, S. W. Cho and M. S. Chung, J. Food Eng., 2011, 102, 327–333 CrossRef CAS.
  20. P. Rangsriwong, N. Rangkadilok, J. Satayavivad, M. Goto and A. Shotipruk, Sep. Purif. Technol., 2009, 66, 51–56 CrossRef CAS.
  21. Y. Gong, X. Zhang, L. He, Q. Yan, F. Yuan and Y. Gao, J. Food Sci. Technol., 2015, 52, 1534–1542 CrossRef CAS PubMed.
  22. R. Fan, J. Xiang, N. Li, X. Jiang and Y. Gao, Sep. Sci. Technol., 2016, 51, 609–621 CrossRef CAS.
  23. M. Gundogdu and H. Yilmaz, Sci. Hortic., 2012, 143, 38–42 CrossRef CAS.
  24. L. Yan, X. Zhou, L. Shi, D. Shalimu, C. Ma and Y. Liu, Int. J. Food Prop., 2017 DOI:10.1080/10942912.2017.1289960.
  25. A. Mustafa and C. Turner, Anal. Chim. Acta, 2011, 703, 8–18 CrossRef CAS PubMed.
  26. B. Pavlić, S. Vidović, J. Vladić, R. Radosavljević, M. Cindrić and Z. Zeković, J. Supercrit. Fluids, 2016, 116, 36–45 CrossRef.
  27. C. Wang, L. Shi, L. Fan, Y. Ding, S. Zhao, Y. Liu and C. Ma, Ind. Crops Prod., 2013, 42, 587–594 CrossRef CAS.
  28. S. Zhao, J. Liu, S. Chen, L. Shi, Y. Liu and C. Ma, Molecules, 2011, 16, 8590–8600 CrossRef CAS PubMed.
  29. R. Fan, N. Li, H. Xu, J. Xiang, L. Wang and Y. Gao, Food Chem., 2016b, 190, 912–921 Search PubMed.
  30. J. M. Landete, Food Res. Int., 2011, 44, 1150–1160 CrossRef CAS.
  31. L. Xiang, D. Xing, F. Lei, W. Wang, L. Xu, L. Nie and L. Du, Tsinghua Sci. Technol., 2008, 13, 460–465 CrossRef CAS.
  32. G. Spigno, L. Tramelli and D. M. De Faveri, J. Food Eng., 2007, 81, 200–208 CrossRef CAS.
  33. Y. Yang and F. Hildebrand, Anal. Chim. Acta, 2006, 555, 364–369 CrossRef CAS.
  34. W. Qu, A. P. Breksa III, Z. Pan and H. Ma, Food Chem., 2012, 132, 1585–1591 CrossRef CAS.
  35. W. Qu, S. Shi, P. Li, Z. Pan and C. Venkitasamy, Int. J. Food Eng., 2014, 10, 683–695 CAS.
  36. M. Fattahi and R. Rahimi, Food Anal. Methods, 2016, 9, 2321–2334 CrossRef.
  37. M. Hayta and E. M. İşçimen, LWT-Food Sci. Technol., 2017, 77, 208–216 CrossRef CAS.
  38. L. Yang, P. Yin, H. Fan, Q. Xue, K. Li, X. Li, L. Sun and Y. Liu, Molecules, 2017, 22, 232 CrossRef PubMed.
  39. X. Jin and Y. Ning, Carbohydr. Polym., 2013, 96, 411–416 CrossRef CAS PubMed.
  40. S. Filip, B. Pavlić, S. Vidović, J. Vladić and Z. Zeković, Food Anal. Methods, 2017, 10, 2270–2280 CrossRef.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c7ay01475a

This journal is © The Royal Society of Chemistry 2017