Antioxidant and antimicrobial activity of oils obtained

G Model JIEC 3590 No. of Pages 10

Journal of Industrial and Engineering Chemistry xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

Antioxidant and antimicrobial activity of oils obtained from a mixture of citrus by-products using a modified supercritical carbon dioxide John Ndayishimiyea , Deok Jum Limb , Byung Soo Chuna,* a b

Department of Food Science and Technology, Pukyong National University 45 Yongso-ro, Namgu, Busan 48513, Republic of Korea Y.G. Company, 365, Shinsolo, Namgu, Busan 608-739, Republic of Korea

A R T I C L E I N F O

Article history: Received 13 February 2017 Received in revised form 21 August 2017 Accepted 22 August 2017 Available online xxx Keywords: Antioxidant activity Antimicrobial activity Mixture Supercritical CO2 extraction Modifier

A B S T R A C T

This study investigated the impact of combining citrus seeds and citrus peels on the bioactivity of the resulting oil obtained using a modified supercritical carbon dioxide. The total phenolic and total flavonoid contents were determined, and the citrus-peel and citrus-seed oils exhibited high content. In addition, antioxidant activity was determined, and the oil extracted using SC-CO2 + ethanol at 200 bar exhibited high IC50 values of 0.52 and 0.53 mg/ml for citrus-peel alone and mixture oils, respectively, for the DPPH assay. Oil from the mixture exhibited high antimicrobial activity, and the oils were more susceptible for gram-positive than gram-negative bacteria. © 2017 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Introduction Consumption of citrus fruits, either as fresh produce or in the juice form, is common due to their dietary benefits and particular flavor. These fruits are broadly grown around the world with a yearly production of about 102 million tons [1]. The production of juice and other products from citrus fruits results in the generation of large amounts of citrus by-products every year. This not only wastes useful materials but also may pose some pollution, disposal, and other related environmental problems due to microbial spoilage [2]. These citrus by-products can be valorized since they contain a wide range of healthy bioactive compounds [3,4]. Citrus peels have a high proportion of natural flavonoids and are among the rich sources of phenolic compounds [5]. In addition, several compounds including flavanone glycosides, polymethoxylated flavones, and flavanones that are unique to citrus have been found to be comparatively rare in other plants [6,7]. It has been reported that the citrus peel extracts demonstrate high antioxidant activity [5] and exert antimicrobial effects against food-borne pathogens [7,8] due to the present quinones, terpenoids, polyphenols, phenolic acids, and tannins [9–11]. Citrus seeds are other by-products of citrus fruit processing. Although many researchers have paid much attention to citrus peels, the importance of citrus seeds has been also studied owing to the present diverse

* Corresponding author. Fax: +82 51 629 5824. E-mail address: [email protected] (B.S. Chun).

compounds including polyphenols, tocopherols, phytosterols, and high amount of unsaturated fatty acids that can be useful for adding value to many products [4,12,13]. The different techniques utilized to obtain extracts from plant matrices include solvent extraction and distillation among others [14]. However, those techniques have some drawbacks such as long extraction time, volatile compound loss, residues of toxic substances, and unsaturated compound degradation due to high temperature [15,16]. The supercritical carbon dioxide (SC-CO2) extraction of natural products has recently drawn attention by many researchers. This type of extraction is not only eco-friendly but also affords the minimum degradation of bioactive compounds (since CO2 has a close-room critical temperature of 31 C), and the prospect of getting solvent-free products [17] has made this a promising technique. In SC-CO2 extraction, the solvating power of SC-CO2 fluid can be increased or decreased by manipulating pressure and/or temperature, resulting in high selectivity. Moreover, the separation of dissolved solutes and SCCO2 could be simply performed via depressurization [18]. However, the limitation of CO2 for extraction of polar compounds due to its non-polar characteristic has been a challenge for extracting polyphenols and other polar compounds. Nonetheless, SC-CO2 polarity can be improved by incorporating modifiers such as ethanol, methanol, and water [14,19]. Therefore, the use of SCCO2 extraction with ethanol as a modifier can not only afford a high bio-potentiality extract but also might help in eliminating or notably decreasing the need for eco unfriendly organic solvents [20].

http://dx.doi.org/10.1016/j.jiec.2017.08.041 1226-086X/© 2017 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: J. Ndayishimiye, et al., Antioxidant and antimicrobial activity of oils obtained from a mixture of citrus byproducts using a modified supercritical carbon dioxide, J. Ind. Eng. Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.08.041


2

J. Ndayishimiye et al. / Journal of Industrial and Engineering Chemistry xxx (2017) xxx–xxx

Although many studies have been dedicated towards studying citrus by-products [21,22], to the best of our knowledge, no study on the bioactive compounds and the antioxidant and antimicrobial activities of oils resulting from a combination of citrus peels and citrus seeds either by using neat SC-CO2 or modified SC-CO2 extraction has been conducted. The combination of citrus peels and citrus seeds may cause increase in bioactivity of the resulting oils owing to the synergistic effect of the compounds they contain and may enhance the bioavailability of some active compounds [23,24]. Therefore, this study was designed to study the effect of combining the citrus peels and citrus seeds on the bioactive compounds and the antioxidant and antimicrobial activities of the resulting oils to assess if their bioactivity can make them applicable in many fields. The purpose of the present study was threefold: First, to extract the oils from citrus seeds, citrus peels, and the mixture of citrus seeds and peels using neat SC-CO2 or SC-CO2 with ethanol as a modifier. Second, to determine the total phenolic, total flavonoid, tocopherol, and phytosterol contents of the extracted oils. Third, to study the antioxidant and antimicrobial activities of the extracted oils and assess whether these oils exhibit potential bioactivity, so they can be used for different applications. Materials and methods Chemicals Folin–Ciocalteu’s reagent, gallic acid, quercetin, tocopherol standards (a-, b-, g- and d-tocopherol), sterol standards (brassicasterol, campesterol, stigmasterol, sitosterol and D-avenasterol), 1,1 diphenyl-2-picryl-hydrazyl (DPPH), 2,20 -azinobis-(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS), Tryptone Soy Broth (TSB), and Mueller–Hinton agar (MHA) were purchased

from Sigma–Aldrich (St. Louis, USA). Ethanol, methanol, and 2, 3, 5-triphenyltetrazolium chloride (TTC) were purchased from Samchun Company (Busan, South Korea). Carbon dioxide (99.99%) was obtained from KOSEM Company (Busan, South Korea). All the chemicals and reagents were of HPLC or analytical grade. Sample collection and preparation The citrus fruits (Citrus junos), common name: Yuzu, provenance: Namhae-gu, Gyeongsangnam-do Province (Busan, South Korea) provided given by Y.G., Co. Citrus fruits were cleaned and peeled, and the peels and seeds were collected. The citrus peels were freeze dried (50 C for 4 days) while the citrus seeds were oven dried at 103 C and then crushed and sieved (using a 710-mm metal sieve) to obtain the powder for extraction. Extraction procedure The SC-CO2 extraction diagram used in this study is shown in Fig. 1. For neat SC-CO2 (without ethanol) extraction, the extraction setup and procedures were the same as those reported in our previous work [25]. For modified SC-CO2 (with ethanol) extraction, a second pump was connected to the extraction line, which supplied ethanol (ca. 1 ml/min: flow rate) and then mixed with CO2. The CO2 + ethanol passed through a heat exchanger and then flowed through the sample in the extraction vessel. The ethanol-oil mixture was received in a vial, whereas ethanol-saturated CO2 left through the flow meter. The residual ethanol was removed using a rotary evaporator (Model N-1100, Eyela, Japan). The extraction (either neat or modified SC-CO2) was performed at the temperature of 45 C, the pressure of 200 and 300 bar, the extraction time was 2 h, and the CO2 flow rate was 27 g/min.

Fig. 1. Schematic diagram of SC-CO2 and co-solvent extraction process.




The yield of extracted oil was calculated using the following formula: Yieldð%Þ ¼

weight of extracted oil 100 weight of sample

Determination of total phenolic content The total phenolic content (TPC) was determined using the Folin–Ciocalteu method. The reaction mixture was composed of 0.5 ml of the diluted extracted oil (5 mg/ml in ethanol), 0.5 ml of 1 N Folin–Ciocalteu’s reagent, and 5 ml of distilled water. 1 ml of sodium carbonate solution (20%) was added to this mixture, vigorously shaken, and stayed for 1 h for reaction. The absorbance was read at 725 nm using UV–vis spectrophotometer (UVmini1240, Shimadzu, Japan). The gallic acid standard curve was used for quantifying TPC and was expressed as mg gallic acid equivalent (GAE)/g of extracted oil. Determination of total flavonoid content The method developed previously by Meda et al. [26] was used for determining total flavonoid content (TFC) with some modifications. In brief, a mixture comprising 0.5 ml of diluted extracted oil (2.5 mg/ml in ethanol), 0.5 ml of methanol, 50 ml of AlCl3 (10%), 50 ml of 1 M potassium acetate, and 1.4 ml of distilled water was allowed to stand for 30 min at ambient temperature. Subsequently, the absorbance was read at 415 nm using the same spectrophotometer used for TPC. The TFC quantification was based on the quercetin standard curve and expressed as mg quercetin equivalent (QE)/g of extracted oil.

3

standard curve that was used to identify and quantify the phytosterols in the extracted oils. Antioxidant activity determination DPPH radical scavenging assay The DPPH assay of the extracted oils was performed in accordance with a previously reported method [27] with some modifications. 1.5 ml of oil with different concentrations of ethanol was added to 1.5 ml of 0.1 mM ethanol DPPH radical solution. The absorbance was read at 517 nm using a UV–vis spectrophotometer (UVmini-1240, Shimadzu, Japan) after 30 min of incubation at ambient temperature (in the dark). Ethanol was used as a control. An inhibition percentage against concentration curve was plotted, and the concentration of oil required for 50% inhibition was determined and expressed as IC50 values. ABTS+ radical scavenging assay The ABTS+ radical scavenging assay of extracted oils was performed by a method reported previously by Re et al. [28] with some modifications. In brief, the pre-formed ABTS+ radical was obtained by reacting 2.45 mM of potassium persulfate with ABTS solution (7 mM) for 14 h (in the dark at room temperature). Then, the solution was diluted with ethanol to get an absorbance of 0.7 0.2 at 734 nm. The extracted oils were dissolved in ethanol at different concentrations, and a 1-ml aliquot of each concentration was added to 3 ml of already prepared ABTS+ radical solution and shaken vigorously. After 1 h in the dark, the absorbance was measured at 734 nm using the UV–vis spectrophotometer used earlier in the study, and the IC50 values were calculated. Antimicrobial activity

Tocopherol content The tocopherol content of extracted oils was determined via normal-phase HPLC using a Hitachi chromatography system. Approximately 20 mg of extracted oil were dissolved in 1 ml HPLC grade hexane and filtered through 0.45-mm filters, and 20 ml of the sample was injected directly into a column (5 mm, 4.6 150 mm2) (Agilent Technologies, Hewlett-Packard, CA, USA). The mobile phase was hexane/isopropanol (99.5:0.5 v/v), the flow rate was 1.0 ml/min, and the detection wavelength was 294 nm employing a Hitachi L-2420 UV–vis detector. The individual tocopherol standards were diluted with hexane (HPLC grade) at different concentrations to construct the external standard curve that was used to identify and quantify the tocopherols in the extracted oils.

Test microorganisms Four pathogenic bacteria were used to test the antimicrobial activity of extracted oils. These bacteria include two Gramnegative bacteria: Salmonella typhimurium KCCM 11862 and Escherichia coli ATCC 25922; two Gram-positive bacteria: Staphylococcus aureus KCCM 11335 and Bacillus cereus ATCC 13061.

Phytosterol content

Disk diffusion assay The preliminary screening of antimicrobial activity was performed using the disc diffusion method. MHA media was prepared and then poured into plates. The plates were inoculated and spread with 100 ml of bacterial suspensions (1 107 CFU/ml). Then, filter paper discs (6 mm Ø) were put on the surface of plates and impregnated with 15 ml of extracted oils. After staying for 2 h (4 C), all plates were incubated at 37 C (overnight), and the diameter of the zone of inhibition (in mm) was measured.

The extracted oils were first saponified and then analyzed for phytosterol content. Approximately 1 g of extracted oil and 100 ml 1 M ethanolic KOH were put in an Erlenmeyer flask and heated at 60 C for 45 min. The mixture was put in a 250-ml separatory funnel with 100 ml of hexane, shaken and washed with 50 ml of distilled water to remove the hydro-soluble components. The hexane layer was dried over anhydrous sodium sulfate and evaporated completely in a rotary evaporator at 40 C. Prior to analysis, the residue was dissolved in 3 ml of hexane, filtered through 0.45-mm filters, and then 20 ml of the sample was injected into the same HPLC column. The mobile phase was hexane/ethanol (70:30 v/v) at a flow rate of 1.0 ml/min, and the detection wavelength was 205 nm employing the same UV–vis detector used for tocopherol content. The individual sterols standards were diluted at different concentrations to construct the external

Determination of the minimum inhibitory concentration Minimum inhibitory concentration (MIC) was determined using a 96-flat well microtiter broth dilution method in 0.15% agar amended TSB as recommended by Mann and Markham [29]. Stock solutions and serial dilutions of the extracted oils were prepared in TSB + 0.15% agar. 100 ml of each dilution was placed into rows of wells in microtiter plates (96 250-ml wells). An equal volume of inoculum was dispensed into the proper wells and mixed with the growth medium, and then the plates were incubated at 37 C with the covers on. After 24 h of incubation, 40 ml of TTC (3 mg/ml) prepared in distilled water was added to each well and incubation was performed again for 30 min. The change to red color was an indication of the biological activity of bacteria. The MICs were recorded to the wells, which showed no change of color of TTC.



4


Fig. 2. Yield of the extracted oils at different conditions. Values are presented as mean standard deviation (n = 3). Different letters on the histogram imply the significant difference (P < 0.05).

The results were reported as the mean standard deviation of 3 replicates. The analysis of variance was performed to compare the results using SPSS for Windows (version 20.0.0, SPSS Inc.).

the yield (p < 0.05) at 300 bar, the yield did not appear to increase significantly (p > 0.05) by addition of ethanol at 200 bar. This might be attributed to the fact that the addition ethanol could not help in increasing the density of CO2 to the level at which it could extract much denser compounds compared with that possible at 300 bar.

Results and discussion

Total phenolic and total flavonoid contents

Yield of extracted oils

The TPC of the extracted oils is presented in Table 1. The oil from citrus peels showed significantly (p < 0.05) higher TPC than the others. The TPC was in the following order: citrus-peel oil > mixture oil > citrus-seed oil. The highest TPC (43.64 mg GA/g) was found in the oil from citrus peels extracted using SC-CO2 + ethanol at 200 bar, while the oil of citrus seeds under the same conditions exhibited TPC of 29.18 mg GA/g only. The lowest TPC (12.17 mg GA/ g) was exhibited by the oil from citrus seeds extracted using neat SC-CO2 at 300 bar, whereas the citrus-peel oil under the same conditions exhibited a TPC of 22.40 mg GA/g. Generally, phenolic compounds might work as protective agents against UV lights, predators, and pathogens in fruits and vegetables [32]. So, it is plausible that the oil from citrus peels exhibited a higher proportion of TPC because the peels might have a higher proportion of these compounds because they are the external covering of the fruit, which is a more likely site of synthesis of phenolic compounds [33]. These results are in agreement with those reported by previous researchers [34] who found higher levels of phenolic compounds in citrus peel than in the other parts. Regarding the addition of ethanol as a modifier, the oils extracted using SC-CO2 + ethanol showed significantly (p < 0.05) higher TPC than those extracted using neat SC-CO2. The TPC ranged between 12.17 and 27.94 mg GA/g for the oils extracted by neat SCCO2, whereas the TPC range was between 27.31 to 43.64 mg GA/g when SC-CO2 + ethanol was used. Therefore, by adding ethanol, the TPC increased regardless of the sample type. This increase in TPC, resulting from ethanol addition, might be explained by considering the polarity of ethanol. SC-CO2 is a non-polar solvent; even at high density, the capability of SC-CO2 to dissolve polar compounds is limited. The addition of a modifier (ethanol in this case) to SC-CO2 could ameliorate the extraction efficiency of polar compounds by increasing their solubility [35]. Since phenolics are polar in nature, it can be deduced that adding ethanol might have enormous impact on the extraction of phenolic compounds by accelerating the desorption process. It may exert its effect by competing with the phenolics for their active binding site, which may cause release of those phenolic compounds and therefore increased TPC [7,36]. In addition, there was a significant difference (p < 0.05) among oils extracted under different conditions. The TPC exhibited a range of 15.32–43.64 mg GA/g for the extracted oils (either neat SC-CO2

Statistical analysis

The yield results are provided in Fig. 2. The yield of extracted oils was significantly different (p < 0.05) depending on the extraction conditions and the used raw materials (namely citrus seeds, citrus peels, and mixture). The yield of oil extracted from citrus seeds, citrus peels, and mixture using neat SC-CO2 at 200 bar was 15.45%, 1.57%, and 8.22%, respectively. The yield of oil extracted from citrus seeds, citrus peels, and mixture using neat SC-CO2 at 300 bar was 22.12%, 1.87%, and 11.97%, respectively. On the other hand, using SC-CO2 + ethanol for extraction at 200 bar yielded 15.67%, 1.66%, and 8.39% for oil from citrus seeds, citrus peels, and mixture, respectively. Using SC-CO2 + ethanol for extraction at 300 bar yielded 23.04%, 1.91%, and 12.73% for citrus seeds, citrus peels, and mixture, respectively. These results indicate as pressure increased, the extraction yield increased significantly (p < 0.05) for all samples. This increase in yield, resulting increase in pressure, might be justified by the fact that pressure is one of the main driving parameters for SC-CO2 extraction [18]. Variation in pressure causes variation in CO2 density, which thereby affects the solubility of analytes in CO2. As the density of CO2 increases (from 18.46 mol/l at 200 bar to 20.23 mol/l at 300 bar at a constant temperature of 45 C), the distance between molecules decreases. Hence, the analytes and CO2 interact, resulting in increased solubility of analytes in CO2, which in turn increases the yield [30]. Furthermore, this increase in yield caused by an increase in pressure might occur because, at lower pressure, the selectivity of CO2 is high due to lower density. When the density increases (with increasing pressure at a constant temperature), it causes even the solubility of more dense compounds which consequently increases the yield. Similarly, previous studies [15] reported that during the extraction of oil from coriander seeds using SC-CO2, the fraction of the non-volatile part was markedly increased when the pressure increased from 100 to 350 bar, with a clear reduction in the volatile fraction, which therefore increased the total yield. Furthermore, it has been reported that the addition of ethanol (as a modifier) enhances the solvent power of SC-CO2 and promotes sample matrix swelling. This increases the surface area and the inner volume for contact with SC-CO2 [31], which might increase the yield. Even though the addition of ethanol increased



37.28 0.8de 4.01 0.01c Values are presented as mean standard deviation (n = 3). Values with different letter in the same row are significantly different (P < 0.05).

Citrus peel

40.71 1e 3.67 0.1bc 27.31 0.2cd 4.46 0.06cd

Citrus seed Mixture

40.94 0.3e 5.06 0.4de 43.64 0.5ef 3.32 0.01b

Citrus peel Citrus seed

29.18 0.17d 5.65 0.02e 20.16 0.02b 5.14 0.1de 24.1 0.1c 4.95 0.2d

12.17 0.3a 6.75 0.05f

Citrus seed

27.94 0.1cd 2.71 0.01ab 15.32 0.02ab 5.17 0.01de TPC TFC

Mixture Citrus peel Citrus seed

Citrus peel

Mixture

SC-CO2 + ethanol, 200 bar, 45 C SC-CO2, 300 bar, 45 C SC-CO2, 200 bar, 45 C

Table 1 Total phenolic content (mg GAE/g of oil) and total flavonoid content (mg QE/g of oil) of extracted oils.

22.4 0.04bc 2.2 0.02a

SC-CO2 + ethanol, 300 bar, 45 C

Mixture


5

or modified) at 200 bar, while the TPC fell in the range of 12.17–40.71 mg GA/g for the extracted oils (either neat or modified) at 300 bar. This decrease in TPC by an increase in pressure might be understood by considering their density and solubility of phenolics. Usually, the solubility of phenolics increases with increasing pressure [37,38], resulting in an increase in phenolics release from the plant matrix [39]. However, even though the release of phenolic compounds increases with increasing pressure during the extraction process, the overall TPC of the extracted oils decreases because the extraction yield also increases simultaneously (Fig. 2). Therefore, our results indicate that pressure can have a greater effect on the solubility of other heavier molecules than on the solubility of phenolic compounds, which might increase the overall extraction yield but decrease the phenolic concentration in the extracted oils. As discussed above, an increase in pressure increases the solubility of a wide range of molecules, so this affects the concentration of TPC in the extracted oil adversely because it acts like diluting effect, decreasing the TPC in extracted oils. This behavior was also reported by previous researchers who demonstrated that amount of phospholipids and paraffinic compounds (i.e., waxes) increased with increasing pressure during SC-CO2 extraction [40,41]. Our results are similar to those published by Lou et al. [42] (5.71–30.00 mg GA/g) for kumquat extracts and Zhang et al. [43] (29.38–51.14 mg GA/g) for Chinese wild mandarin peel extract. The range of variation in our results is higher than those reported by Karimi et al. [44] (3.93–4.83 mg GAE/G) for the phenolics of Citrus aurantium peels and Chen et al. [45] (15.595–18.950 mg GA/g) for TPC of ethanol extracts of Citrus reticulata Blanco cv. Ougan fruits peels. But our results show lower values than those reported by Ghasemi et al. [46] (104.2–172.1 mg GAE/g DW) for the extracts of C. reticulate Blanco fruits peels and Goulas and Manganaris [47] (112.2–196.2 mg GAE/G) for citrus fruits pulps and peels grown in Cyprus. The TFC results for the extracted oils are presented in Table 1. The TFC ranged between 2.20 and 6.75 mg QE/g for all extracted oils. In general, unlike the case of TPC, the citrus-seed oils showed significantly (p < 0.05) higher TFC than the others. The order of TFC was as follow: citrus-seed oil > mixture oil > citrus-peel oil. The highest TFC was exhibited by the oil extracted from citrus seeds using neat SC-CO2 at 300 bar (6.75 mg QE/g), while the lowest TFC was found under the same conditions for the oil extracted from citrus peels (2.2 mg QE/g). The TFC of oils from the mixture varied from 4.01 to 5.14 mg QE/g under all conditions. The extraction conditions appeared to influence the TFC. The increase in pressure seemed to increase the TFC for oil from citrus seeds and mixture while decreasing the TFC for citrus peels when neat SC-CO2 was used. However, the TFC exhibited an increase with increasing pressure for oil from citrus peels while decreasing the TFC for citrus seeds and mixture when SC-CO2 + ethanol was used. For instance, TFC increased from 5.17 and 4.95 mg QE/g at 200 bar to 6.75 and 5.14 mg QE/g at 300 bar, for citrus-seed oil and mixture oil, respectively, when neat SC-CO2 was used, whereas TFC slightly reduced from 2.71 to 2.20 mg QE/g for citrus-peel oil under the same conditions. Conversely, TFC decreased from 5.65 and 5.06 mg QE/g at 200 bar to 4.46 and 4.01 mg QE/g at 300 bar for citrus seed oil and mixture oil, respectively, when SC-CO2 + ethanol was used, while the oil from citrus peels showed increased TFC under the same conditions. Moreover, similar to the case of TPC, adding ethanol yielded a significant increase in TFC (p < 0.05). For example, TFC increased from 5.17, 2.71, and 4.95 mg QE/g for oils extracted using neat SCCO2 at 200 bar to 5.65, 3.32, and 5.06 mg QE/g for those extracted using SC-CO2 + ethanol from citrus seeds, citrus peels, and mixture, respectively, at the same pressure. This might again be attributed to the polar nature of ethanol, which possibly helps in extracting



6


more polar flavonoids, possibly elevating the TFC for oils extracted using SC-CO2 + ethanol. These data are partially in accordance with those reported by Chen et al. [45] (0.30–31.1 mg QE/g DW) and Ghasemi et al. [46] (4.67–5.79 mg rutin equivalents/g DW). However, our results showed lower values than those of Zhang et al. [43] (29.38–51.14 mg GA/g) for Chinese wild mandarin peel extract but higher than those reported by Goulas and Manganaris [47] (1.27–2.28 mg rutin/g) for citrus fruits pulps and peels grown in Cyprus. These discrepancies (either for TPC or TFC) between our results and those reported earlier can be attributed to many factors, including the raw materials, extraction methods, pedoclimatic conditions, genotype, variety, and sample matrix preparation, that might alter the TPC and TFC of the resulting extracts. Tocopherol and phytosterol content Table 2 lists the tocopherol content for the extracted oils. The increase of pressure resulted in a significant decrease (p < 0.05) of tocopherol content, and ethanol addition did not lead to a remarkable difference (p > 0.05). The oil of citrus seeds exhibited the highest tocopherol content, followed by mixture oils and citrus-peel oils (citrus-seed oils > mixture oils > citrus-peel oils). Among citrus-seed oils, the highest total tocopherol content was 33.8 mg/100 g recorded at 200 bar (neat SC-CO2), while the lowest was 22.5 mg/100 g recorded at 300 bar (SC-CO2 + ethanol). Similarly, the highest tocopherol content (16.76 mg/100 g) for the oil from a mixture was recorded at 200 bar (neat SC-CO2), while the lowest (8.49 mg/100 g) was recorded at 300 bar (SC-CO2 + ethanol). However, citrus-peel oils exhibited almost no significant difference in tocopherol content under different conditions (p > 0.05). This decrease in tocopherol content due to increasing pressure could result from an increase in SC-CO2 density, which might increase the solubility of a wide range of heavier molecules, which causes dilution of the tocopherol content of the extracted oil and hence a decrease in tocopherol concentration. Similar to our results, Illés et al. [15] reported that the amount of tocopherols in coriander seed oil extracted under mild conditions (100–200 bar and 25 C) was about two times higher than those extracted at 250–350 bar and 35 C. Among the individual tocopherols, a-tocopherol was predominant in all the extracted oils ranging from 16.7 to 24.93 mg/100 g, 5.88 to 13.50 mg/100 g, and 0.56 to 1.01 mg/ 100 g for citrus seeds, mixture, and citrus peels, respectively. It was followed by g-tocopherol, b-tocopherol, and d-tocopherol, which was lowest in amount. Our results are slightly higher than those of Matthaus and Özcan [4], who reported a range of 1.70–20.5 mg/ 100 g for oils of citrus seeds from Vietnam and Turkey. The phytosterol content of the extracted oils in this study is presented in Table 3. The phytosterol content ranged from 245.91 to 367.76 mg/100 g and 90.96 to 140.44 mg/100 g for citrus seeds and mixture, respectively. The phytosterols viz. sitosterol, campesterol, avenasterol, stigmasterol, and brassicasterol were identified, and sitosterol was found to be, by far, the predominant phytosterol in all oils with the range of 66.52–276.08 mg/100 g followed by campesterol (ranging from 14.75 to 49.16 mg/100 g). The third one was avenasterol, except for the oils extracted by SCCO2 + ethanol at 300 bar, where stigmasterol turned out to be the third one. Brassicasterol was not detected in almost all extracted oils. According to our results, the phytosterol content appeared to be affected significantly (p < 0.05) by both pressure and the addition of ethanol. For extracted oils, the phytosterol content increased significantly (p < 0.05) with increasing pressure. This increase of phytosterol content, resulting from an increase in pressure, might be ascribed to the fact that a considerable amount of phytosterols is expected to be attached deep inside the seed tissues, and when the pressure is increased, the SC-CO2 reaches inside, thus successfully extracting more phytosterols. This was

proven by previous studies [48], where the optimization of SC-CO2 extraction of phytosterol-enriched oil from Kalahari melon seeds was studied and an increase in phytosterol content was noted with increasing pressure. This pattern for SC-CO2 extraction of increasing phytosterol content was also reported by other researchers [49], who found SC-CO2 extracted phytosterols better than petroleum ether, which can be ascribed to its ability to penetrate the inner part of the material being extracted. Correlating this with our results, it is plausible that an increase of pressure might cause an increase in penetration by SC-CO2 (due to density increase), which in turn might result in a high release rate of phytosterols. Regarding the influence of ethanol on the extraction of phytosterols, it was shown that phytosterol content increased when ethanol was added. Adding ethanol (possessing OH group) increases the polarity of CO2, hence facilitating the extraction of phytosterols. Within the same line of explanations, Nyam et al. [48] found that the addition of ethanol (flow rate of about 2 ml/min) elevated phytosterol content. As far as the individual sterols are concerned, our results concur with those of previous studies by Lazos and Servos [50] and Matthaus and Özcan [4], who reported that the composition of phytosterols in citrus-seed oil is dominated by sitosterol, which accounted for about 70% or even more of the total phytosterols. Antioxidant activity The results of antioxidant activity of extracted oils expressed as IC50 values are presented in Fig. 3. A low IC50 value represents strong antioxidant activity. The IC50 values differed significantly (p < 0.05) depending on the extraction conditions and/or raw materials. The DPPH values of extracted oils are presented in Fig. 3(a), ranging from 2.73 to 0.52 mg/ml for all extracted oils. The lowest IC50 value (strong antioxidant activity) was recorded for citrus-peel and mixture oils extracted using SC-CO2 + ethanol at 200 bar with 0.52 and 0.53 mg/ml, respectively, with no significance difference (p > 0.05) between them, whereas the same oils (citrus-peel and mixture oil) exhibited IC50 values of 0.71 and 0.68 mg/ml, respectively, when using SC-CO2 + ethanol at 300 bar. The IC50 values of citrus-seed oils were significantly (p < 0.05) high compared with those of citrus peels and mixture (0.78 and 1.31 mg/ml, respectively) for SC-CO2 + ethanol at 200 bar and 300 bar. On the other hand, the IC50 values showed a significant increase (p < 0.05) (decrease in antioxidant activity) for all the oils (compared with the case of SC-CO2 + ethanol) when neat SC-CO2 was used regardless of the pressure used. In this category, the lowest IC50 value was recorded for oil extracted from mixture and citrus-peel oil at 200 bar (0.89 and 0.9 mg/ml, respectively), whereas the same oils (mixture and citrus peel) exhibited IC50 values of 1.12 and 1.31 mg/ml, respectively, at 300 bar. Again, citrus-seed oils exhibited higher IC50 values. These results suggest that the IC50 value of extracted oils relied significantly on two parameters: extraction conditions and sample material. Regarding the extraction conditions, as discussed above, it was evident throughout this study, that an increase in pressure significantly (p < 0.05) increased the yield; however, at the same time, some quality characteristics (such as TPC, TFC, and tocopherols) decreased significantly. In fact, the phenolic and flavonoid compounds substantially contribute to the antioxidant activity owing to the redox properties of their hydroxyl groups [51], which intervene in scavenging free radicals or donating hydrogen atoms or electrons [52]. Likewise, tocopherols are important natural antioxidants that have been repeatedly reported for their antioxidant activity in foods and biological systems [53,54]. Given the fact that these compounds contribute a lot to antioxidant activity and relating this to our results, it is logical to assume that



Tocopherols

RT (min)

SC-CO2, 200 bar, 45 C Citrus seed

Citrus peel

Mixture

Citrus seed

Citrus peel

Mixture

Citrus seed

Citrus peel

Mixture

Citrus seed

Citrus peel

Mixture

a-Tocopherol b-Tocopherol g-Tocopherol d-Tocopherol

3.33 3.56 5.12 5.70

24.93 2.94 5.83 0.1 33.8 0.2cd

1.01 0.19 0.11 nd 1.31 0.1ab

13.5 1.05 2.08 0.13 16.76 0.6bc

19.4 1.77 3.31 0.1 24.58 0.3c

0.73 nd 0.2 nd 0.93 0.3a

7.16 0.82 1.53 0.15 9.66 0.2b

23.1 2.52 4.95 0.47 31.04 0.6cd

0.56 0.39 0.42 nd 1.37 0.2ab

10.67 1.25 2.03 0.25 14.2 0.7bc

16.7 1.29 4.02 0.49 22.5 1.1c

1.01 0.27 nd nd 1.28 0.1ab

5.88 0.65 1.96 nd 8.49 0.08b

Total

SC-CO2, 300 bar, 45 C



Values are presented as mean standard deviation (n = 3). Values with different letter in the same row are significantly different (P < 0.05). nd: not detected. RT: retention time.

Table 3 Phytosterols content (mg/100 g of oil) of extracted oils determined by HPLC. Phytosterols

RT (min)

SC-CO2, 200 bar, 45 C

SC-CO2, 300 bar, 45 C

Citrus seed

Citrus peel

Mixture

Citrus seed

Citrus peel

Mixture

Citrus seed

Citrus peel

Mixture

Citrus seed

Citrus peel

Mixture

Brassicasterol Campesterol Stigmasterol Sitosterol D-Avenasterol Total

18.27 22.49 23.13 26.38 27.64

nd 39.36 7.33 183.8 15.42 245.91 1e

na na na na na na

nd 15.74 2.85 66.52 5.85 90.96 0.02a

nd 40.79 11.48 219.54 16.17 287.98 0.8f

na na na na na na

nd 16.00 4.01 90.01 6.07 116.09 0.9b

0.21 44.15 8.79 255.04 19.5 327.69 0.3g

na na na na na na

nd 14.75 3.52 102.5 9.34 130.11 0.1c

0.68 49.16 23.78 276.08 18.06 367.76 1.7h

na na na na na na

nd 22.12 14.26 91.1 12.96 140.44 0.4d

Values are presented as mean standard deviation (n = 3). Values with different letter in the same row are significantly different (P < 0.05). nd: not detected. na: not analyzed. RT: retention time.





Table 2 Tocopherol content (mg/100 g of oil) of extracted oils determined by HPLC.

7


8


Fig. 3. Antioxidant activity of the extracted oils expressed as IC50 values (mg/ml) determined by (a) DPPH and (b) ABTS assay. Values are presented as mean standard deviation (n = 3). Different letters on the histograms imply the significant difference (P < 0.05).

the decrease in antioxidant activity (higher IC50 value) might be linked to a reduction in TPC, TFC, and tocopherol content. Furthermore, the IC50 value was shown to decrease (increase of antioxidant activity) significantly (p < 0.05) when ethanol was added as a modifier. For instance, it dropped from 0.90 to 0.52 mg/ ml for citrus-peel oil at 200 bar when ethanol was added. This might be because ethanol helped extract more polar compounds (particularly phenolics, and flavonoids) which might contribute a lot to the antioxidant activity. Here it should be noted that although some less polar and lipophilic phenolic compounds can exert antioxidant activity, the more polar and hydrophilic phenolics can exhibit higher antioxidant activity [55]. Even though the individual phenolic and flavonoid compounds were not determined in this study, it can be suggested that addition of ethanol might extract even more polar and complex polyphenols, which might increase the antioxidant activity. Concerning the antioxidant activity of oils, the IC50 values for the oils from citrus peels and mixture were significantly lower (p < 0.05) than those for oils from citrus seeds. As observed in Table 1, it is quite remarkable that the TPC for the oils from citrus peels and mixture is higher than that for oil from citrus seeds; however, the same table also shows high TFC for citrus-seed oil compared with those for citrus-peel and mixture oils. Similarly, Table 2 markedly indicates that the tocopherol content for citrusseed oil is greater than that for citrus-peel and mixture oils. Nevertheless, the IC50 value was lower (high antioxidant activity) for citrus-peel and mixture oils than citrus-seed oil despite the high TFC and tocopherol content in citrus-seed oil. This demonstrates that the TPC might exert more activity than TFC and tocopherols. Interestingly, there was no significant difference (p < 0.05) between the citrus-peel and mixture oils in terms of antioxidant activity, even though the citrus-peel oils exhibited relatively higher TPC. This might presumably be ascribed to the synergistic effect, which might greatly contribute to mixture oils. In order to elucidate this effect, the oils composition can be taken into account. If their composition is compared, it is pretty obvious that mixture oils have additional compounds (such as phytosterols and tocopherols) compared with citrus-peel oils because they resulted from a combination of citrus peels and seeds. Therefore, these extra compounds (particularly tocopherols) might not only directly contribute to the antioxidant activity but might also act synergistically with other compounds, which may subsequently increase the overall antioxidant activity. This coincides with the results in previous studies [55], which demonstrated the synergistic effects of polyphenols and tocopherols as a

consequence of transfer of electrons from the polyphenols to the tocopherolxyl radical to regenerate tocopherol. In addition, the phytosterols (present in mixture oil but not in citrus-peel oil) might contribute to the scavenging capacity for free radicals. Moreover, it has been shown in earlier studies that phytosterols with an ethylidene group in the side chain (particularly avenasterol) can act effectively as antioxidants and as antipolymerization agents in oils [7,36] and suggested that a synergistic effect of those sterols with other oils compounds may occur. The use of SC-CO2 and vegetable oil (as a co-solvent) yields elevated extraction of carotenoids, which are lipophilic antioxidants, owing to the presence of triglyceride species in vegetable oil [24]. Although the carotenoids were not determined in this study, it might be expected that when citrus seeds and citrus peels are combined, the triglycerides in citrus seeds might help in extracting more carotenoids from peels, which might consequently contribute to the antioxidant activity of mixture oil. As far as the methods used are concerned, the DPPH assay showed significantly (p < 0.05) low IC50 values compared with the ABTS assay (Fig. 3(a) and (b)) for all extracted oils, and the correlation among those assays was not high (r2 = 0.685). This difference could be explained by analyzing the behaviors of those 2 assays. The ABTS assay is usually used to measure the antioxidant capacity of hydrophilic compounds [56]. However, the DDPH assay has been regularly applied in both aqueous-organic extracts of plant foods [57,58] and vegetable oils [59]. Based on the nature of our extracted oils, it seems that DPPH presented an advantage over ABTS because it might react with both hydrophilic and lipophilic antioxidant compounds present in a sample, which could have contributed to the augmentation of overall antioxidant activity for DPPH. Contrarily, ABTS is hydrophilic in nature; its ability to react might sometimes be limited only to the hydrophilic compounds in the extract, which in turn might contribute to a decrease in overall antioxidant activity compared with that for DPPH. Antimicrobial activity The results of the oils for the disk diffusion assay are depicted in Table 4. Generally, the oils showed significantly (p < 0.05) lower inhibition for gram-negative bacteria than that for gram-positive bacteria. Also, the increase in pressure caused a significant reduction (p < 0.05) in the inhibition capacity of the extracted oils. The highest antimicrobial activity was obtained for the oil from a mixture extracted using SC-CO2 + ethanol at 200 bar against S. aureus and B. cereus, with an inhibition zone diameter of 19 mm;




9

Table 4 The diameter of the zones of inhibition (in mm) of extracted oils tested against gram positive and gram negative bacteria; disk diameter (6.0 mm). Bacteria

SC-CO2, 200 bar, 45 C Citrus seed

Bacterial strains G (+) Staphylococcus 8 1.00a aureus Bacillus cereus 8 0.00a Bacterial strains G () Escherichia coli ni Salmonella ni typhimurium

SC-CO2, 300 bar, 45 C



Citrus peel Mixture

Citrus seed

Citrus peel Mixture

Citrus seed

14 1.00bc 15 1.00bc ni

9 1.00a

8 1.00a

16 0.00c

16 1.00c

Citrus peel Mixture

Citrus seed

9 1.00a

19 0.00cd ni

Citrus peel Mixture

9 1.00a

10 1.00ab

17 1.00c

ni

10 1.00ab 11 1.00ab 8 1.00a

18 1.00cd 19 0.00cd ni

14 2.00bc 14 2.00bc

11 1.00ab 13 1.00b 12 1.00b 13 1.00b

ni ni

8 0.00a 8 1.00a

15 1bc 14 1bc

8 1.00a 12 2.00b 11 0.00ab 10 0.00ab

8 0.00a ni

ni ni

15 1.00bc 16 1.00c

ni ni

ni: no inhibition. Values are presented as mean standard deviation (n = 3). Values with different letters in the same row are significantly different (P < 0.05). Values with different letters in the same column are significantly different (P < 0.05).

this was followed by citrus-peel oil extracted under the same conditions, with an inhibition zone diameter of 16 and 18 mm, respectively, against S. aureus and B. cereus. As far as gram-negative bacteria are concerned, the inhibition zone diameter showed a significant decrease (p < 0.05) compared with that for grampositive bacteria with mixture oil (SC-CO2 + ethanol at 200 bar), showing the highest inhibition zones of 15 and 16 mm against E. coli and S. typhimurium, whereas the citrus-peel oil extracted under the same conditions showed inhibition zones of 15 and 14 mm against the same bacteria. Indeed, unlike gram-positive bacteria, the citrus-seed oils showed no inhibition against gram-negative bacteria regardless of the extraction conditions. The MIC results for the tested bacteria are presented in Table 5. Similarly to the results of the disk diffusion assay, gram-positive bacteria were more sensitive than gram-negative bacteria, and the MIC values ranged from 0.20 to 1.35 mg/ml depending on the extracted oil. In this category, the lowest MIC value (highest antimicrobial activity) was exhibited by mixture oil (extracted using SC-CO2 + ethanol at 200 bar) at 0.20 and 0.25 mg/ml against B. cereus and S. aureus, respectively, whereas citrus-peel oil extracted under the same conditions exhibited an MIC value of 0.25 mg/ml for both bacteria. On the other hand, the MIC values for gram-negative bacteria varied between 0.27 and 1.75 mg/ml, with the mixture and citrus-peel oils extracted using SC-CO2 + ethanol at 200 bar exhibiting the lowest MIC value. According to both results (disk diffusion assay and MIC), it is quite obvious that gramnegative bacteria were less susceptible to the oils than grampositive bacteria. This less susceptibility of gram-negative bacteria to the action of extracted oils was perhaps attributed to the presence of an external membrane enclosing the cell wall, which limits the diffusion of lipophilic compounds of the oil through its lipopolysaccharide covering, thereby reducing antimicrobial activity. This has also been shown by other studies that reported that gram-negative bacteria are less sensitive to plant extracts than

gram-positive bacteria [12]. Regarding the type of extracted oils, it was evident that the oils obtained at 200 bar (either using neat or modified SC-CO2) exhibited higher antimicrobial activity for both disk diffusion assay (high diameter inhibition zone) and MIC (low MIC values) than the corresponding oils obtained at 300 bar. This reduction in antimicrobial activity resulting in augmentation of pressure can be explained as follows. According to our data, it was clear that increasing pressure affected the yield positively, but simultaneously affected the quality. It not only affects the unsaponifiable matters (such as tocopherols) but also reduces the volatile fraction of the extracted oils [15]. The volatile fraction in citrus oils mainly comprises of monoterpene hydrocarbons, sesquiterpene hydrocarbons, and oxygenated monoterpenes. These compounds play an important role in bacterial inhibition owing to their hydrophobic property, which enables them to partition the lipids of the bacterial cell membrane and mitochondria, distracting the structures and rendering them more permeable, which leads to an outflow of proteins and other cell contents, and therefore the bacteria dies [60]. Moreover, it was revealed in our study that the composition of sesquiterpene hydrocarbons and oxygenated monoterpenes decreased significantly when the pressure increased. For example, the composition of g-terpinene (sesquiterpene hydrocarbon) exhibited a decrease from 13.35% and 9.85% to 7.65% and 5.47%, respectively, for citrus-peel and mixture oil (data not shown) when the pressure was increased from 200 bar to 300 bar (for neat SC-CO2). Similarly, the composition of linalool (oxygenated monoterpene) decreased from 3.74% and 3.66% to 1.81% and 1.06%, respectively, for citrus-peel and mixture oil extracted under the same conditions (data not shown). By correlating these results with those for antimicrobial activity, it can be stated that the decrease in antimicrobial activity as a result of increased pressure could be linked to the reduction of volatile compounds in extracted oils, which consequently might provoke a decrease in antimicrobial activity. Moreover, the addition of

Table 5 Minimal inhibitory concentration (MIC) (mg/ml) of extracted oils. Bacteria

SC-CO2, 200 bar, 45 C

SC-CO2, 300 bar, 45 C



Citrus seed Citrus peel Mixture Citrus seed Citrus peel Mixture Citrus seed Citrus peel Mixture Citrus seed Citrus peel Mixture Bacterial strains G (+) Staphylococcus aureus Bacillus cereus

>2.00 >2.00

Bacterial strains G () Escherichia coli nt Salmonella typhimurium nt

0.375 0.40

0.50 0.35

nt nt

1.35 1.25

1.35 1.00

>2.00 >2.00

0.25 0.25

0.25 0.20

nt nt

1.00 0.70

1.00 0.75

0.675 0.675

0.575 0.625

nt nt

>2.00 1.75

>2.00 nt

nt nt

0.275 0.325

0.325 0.30

nt nt

>2.00 1.50

1.35 >2.00

nt: not tested.



10


ethanol as a modifier caused a slight increase in the antimicrobial activity of the oils. This may be due to the augmentation of polyphenols, resulting from polarity increase, which can contribute to the inhibition [61]. The mixture oil showed slightly high antimicrobial activity compared with that of the citrus-peel oil, although in most cases there was no significant difference (p > 0.05). This might be due to the presence of additional compounds (such as tocopherol, sterols, and other lipophilic compounds) present in mixture oils, which could contribute to synergism with the volatile compounds and increase the overall antimicrobial activity [60]. Overall, the mechanism on how plant extracts can inhibit the microorganisms is complex because both major and minor compounds can contribute to the inhibition. The mechanism might involve synergism or antagonism effects between those compounds or specificity of some compounds to certain microorganisms [60,62]; therefore, it is difficult to predict the exact individual compound for inhibition of such bacteria. Conclusion The processing of citrus fruits generates by-products that are rich sources of bioactive substances. In this study, the bioactive compounds and the antioxidant and antimicrobial activities of oils extracted from a mixture of citrus peels and seeds were studied. The results demonstrated that the antioxidant and antimicrobial activities might be attributed but not merely limited to the phenolic, flavonoid, tocopherol and phytosterol content of the extracted oils since there might be other minor components playing a role through synergistic or mutual effects, which consequently might contribute to the overall activity. In addition, the extraction conditions (pressure) qualitatively and quantitatively affected the extracted oils, which might be associated with SC-CO2 density and solubility. Moreover, addition of ethanol as a modifier boosted the antioxidant and antimicrobial activities of the extracted oils. More importantly, it appeared that the mixture oil had the same or even higher potentiality than the citrus-peel oil, which might reflect how the mixture oil can be customized to many applications. Overall, this study showed that the extraction of oils from a mixture of citrus seeds and peels, which are considered as waste, using SC-CO2 and/or modified SCCO2 might result in oils with high bioactivity and be a promising approach in many areas since SC-CO2 extraction is an environment-friendly extraction technique. Acknowledgements The authors gratefully acknowledge the financial support for the research work provided by Business for Cooperative R&D between Industry, Academy, and Research Institute funded by Korea Small and Medium Business Administration in 2015. References [1] F. Mehl, G. Marti, J. Boccard, B. Debrus, P. Merle, E. Delort, L. Baroux, V. Raymo, M.I. Velazco, H. Sommer, Food Chem. 143 (2014) 325. [2] G. Laufenberg, B. Kunz, M. Nystroem, Bioresour. Technol. 87 (2) (2003) 167. [3] K. Fisher, C. Phillips, Trends Food Sci. Technol. 19 (3) (2008) 156. [4] B. Matthaus, M. Özcan, Grasas Aceites 63 (3) (2012) 313. [5] K. Hayat, X. Zhang, H. Chen, S. Xia, C. Jia, F. Zhong, Sep. Purif. Technol. 73 (3) (2010) 371. [6] J.A. Manthey, K. Grohmann, J. Agric. Food Chem. 49 (7) (2001) 3268. [7] Y.-H. Lee, A.L. Charles, H.-F. Kung, C.-T. Ho, T.-C. Huang, Ind. Crops Prod. 31 (1) (2010) 59. [8] L. Espina, M. Somolinos, S. Lorán, P. Conchello, D. García, R. Pagán, Food Control. 22 (6) (2011) 896. [9] M. Calvo, E. Angulo, P. Costa-Batllori, C. Shiva, C. Adelantado, A. Vicente, Biotechnology 5 (2) (2006) 137.

[10] O.-H. Lee, B.-Y. Lee, Bioresour. Technol. 101 (10) (2010) 3751. [11] R. Casquete, S.M. Castro, A. Martín, S. Ruiz-Moyano, J.A. Saraiva, M.G. Córdoba, P. Teixeira, Innov. Food Sci. Emerg. Technol. 31 (2015) 37. [12] F. Anwar, R. Naseer, M. Bhanger, S. Ashraf, F.N. Talpur, F.A. Aladedunye, J. Am. Oil Chem. Soc. 85 (4) (2008) 321. [13] E.I. Adeyeye, A.J. Adesina, Int. J. Curr. Microbiol. Appl. Sci. 4 (5) (2015) 537. [14] G. Raman, G. Jayaprakasha, M. Cho, J. Brodbelt, B.S. Patil, Sep. Purif. Technol. 45 (2) (2005) 147. [15] V. Illés, H. Daood, S. Perneczki, L. Szokonya, M. Then, J. Supercrit. Fluids 17 (2) (2000) 177. [16] Y. Yamini, F. Sefidkon, S. Pourmortazavi, Flavour Fragr. J. 17 (5) (2002) 345. [17] S.M. Pourmortazavi, S.S. Hajimirsadeghi, J. Chromatogr. A 1163 (1) (2007) 2. [18] U. Salgın, O. Döker, A. Çalımlı, J Supercrit. Fluids 38 (3) (2006) 326. [19] B. Díaz-Reinoso, A. Moure, H. Domínguez, J.C. Parajó, J. Agric. Food Chem. 54 (7) (2006) 2441. [20] X.-S. Shu, Z.-H. Gao, X.-L. Yang, Fitoterapia 75 (7) (2004) 656. [21] J. Yu, D.V. Dandekar, R.T. Toledo, R.K. Singh, B.S. Patil, Food Chem. 105 (3) (2007) 1026. [22] H. Ueno, M. Tanaka, S. Machmudah, M. Sasaki, M. Goto, Food Bioprocess Technol. 1 (4) (2008) 357. [23] R. Aparicio, L. Roda, M.A. Albi, F. Gutiérrez, J. Agric. Food Chem. 47 (10) (1999) 4150. [24] M. Sun, F. Temelli, J. Supercrit. Fluids 37 (3) (2006) 397. [25] J. Ndayishimiye, A.T. Getachew, B.S. Chun, Waste Biomass Valorization (2016) 1, doi:http://dx.doi.org/10.1007/s12649-016-9697-8. [26] A. Meda, C.E. Lamien, M. Romito, J. Millogo, O.G. Nacoulma, Food Chem. 91 (3) (2005) 571. [27] F. Brahmi, B. Mechri, G. Flamini, M. Dhibi, M. Hammami, Acta Physiol. Plant. 35 (4) (2013) 1061. [28] R. Re, N. Pellegrini, A. Proteggente, A. Pannala, M. Yang, C. Rice-Evans, Free Radic. Biol. Med. 26 (9) (1999) 1231. [29] C. Mann, J. Markham, J. Appl. Microbiol. 84 (4) (1998) 538. [30] M.D.L. De Castro, M. Valcarcel, M.T. Tena, Analytical Supercritical Fluid Extraction, Springer Science & Business Media, 2012. [31] G.-B. Lim, S.-Y. Lee, E.-K. Lee, S.-J. Haam, W.-S. Kim, Biochem. Eng. J. 11 (2) (2002) 181. [32] I. Ignat, I. Volf, V.I. Popa, Food Chem. 126 (4) (2011) 1821. [33] H.R. De Moraes Barros, T.A.P. De Castro Ferreira, M.I. Genovese, Food Chem. 134 (4) (2012) 1892. [34] S. Gorinstein, Z. Zachwieja, M. Folta, H. Barton, J. Piotrowicz, M. Zemser, M. Weisz, S. Trakhtenberg, O. Màrtín-Belloso, J. Agric. Food Chem. 49 (2) (2001) 952. [35] C.G. Pereira, M.a.A. Meireles, Food Bioprocess Technol. 3 (3) (2010) 340. _ Şanal, E. Bayraktar, Ü. Mehmetog lu, A. Çalımlı, J. Supercrit. Fluids 34 (3) [36] I. (2005) 331. [37] E.S. Choi, M.J. Noh, K.-P. Yoo, J. Chem. Eng. Data 43 (1) (1998) 6. _ [38] I.H. Adil, M.E. Yener, A. Bayındırlı, Sep. Sci. Technol. 43 (5) (2008) 1091. [39] I.H. Adil, H. Cetin, M. Yener, A. Bayındırlı, J. Supercrit. Fluids 43 (1) (2007) 55. [40] M.J. Cocero, L. Calvo, J. Am. Oil Chem. Soc. 73 (11) (1996) 1573. [41] E. Reverchon, I. De Marco, J. Supercrit. Fluids. 38 (2) (2006) 146. [42] S.-N. Lou, Y.-C. Lai, Y.-S. Hsu, C.-T. Ho, Food Chem. 197 (2016) 1. [43] Y. Zhang, Y. Sun, W. Xi, Y. Shen, L. Qiao, L. Zhong, X. Ye, Z. Zhou, Food Chem. 145 (2014) 674. [44] E. Karimi, E. Oskoueian, R. Hendra, A. Oskoueian, H.Z. Jaafar, Molecules 17 (2) (2012) 1203. [45] X. Chen, K. Yuan, H. Liu, J. Food Agric. Environ. 8 (2) (2010) 150. [46] K. Ghasemi, Y. Ghasemi, M.A. Ebrahimzadeh, Pak. J. Pharm Sci. 22 (3) (2009) 277. [47] V. Goulas, G. Manganaris, Food Chem. 131 (1) (2012) 39. [48] K.L. Nyam, C.P. Tan, O.M. Lai, K. Long, Y.B.C. Man, Food Bioprocess Technol. 4 (8) (2011) 1432. [49] T.H. Beveridge, B. Girard, T. Kopp, J.C. Drover, J. Agric. Food Chem. 53 (5) (2005) 1799. [50] E. Lazos, D. Servos, Grasas Aceites 39 (4–5) (1988) 232. [51] M. Materska, I. Perucka, J. Agric. Food Chem. 53 (5) (2005) 1750. [52] R. Amarowicz, R. Pegg, P. Rahimi-Moghaddam, B. Barl, J. Weil, Food Chem. 84 (4) (2004) 551. [53] A. Kamal-Eldin, L.-Å. Appelqvist, Lipids 31 (7) (1996) 671. ska, J. Korczak, A. Gramza, E. Wa˛sowicz, P.C. Dutta, J. AOAC Int. 87 (2) [54] M. Rudzin (2004) 499. [55] G. Pande, C.C. Akoh, Food Chem. 120 (4) (2010) 1067. [56] J. Pérez-Jiménez, S. Arranz, M. Tabernero, M.E. Díaz-Rubio, J. Serrano, I. Goñi, F. Saura-Calixto, Food Res. Int. 41 (3) (2008) 274. [57] R. Llorach, F.A. Tomás-Barberán, F. Ferreres, J. Agric. Food Chem. 52 (16) (2004) 5109. [58] H.-Y. Chen, Y.-C. Lin, C.-L. Hsieh, Food Chem. 104 (4) (2007) 1418. [59] C.I. Tuberoso, A. Kowalczyk, E. Sarritzu, P. Cabras, Food Chem. 103 (4) (2007) 1494. [60] S. Burt, Int. J. Food Microbiol. 94 (3) (2004) 223. [61] D.A. Oliveira, A.A. Salvador, A. Smânia, E.F. Smânia, M. Maraschin, S.R. Ferreira, J. Biotechnol. 164 (3) (2013) 423. [62] F. Bakkali, S. Averbeck, D. Averbeck, M. Idaomar, Food Chem. Toxicol. 46 (2) (2008) 446.
