Comparison of phenolic composition in date (Phoenix dactylifera L.) flesh and seeds extracted by an ultrasonic-assisted and conventional method

ABSTRACT Date palm (Phoenix dactylifera L.) is an important fruit of the Middle East and North Africa with a high content of phenolic compounds. This study applied ultrasonic-assisted and conventional methods to extract phenolic compounds from date flesh and seed. The efficacy of the extraction methods is compared in terms of quantity, content, and antioxidant capacity of phenolic compounds. The highest total phenolic content (TPC) was found in ultrasonic-assisted date seed samples extracted by ethanol (18.53 mg GAE/g). In contrast, the TPC content of conventionally extracted seed samples ranged from 1.30 to 14.46 mg GAE/g. The TPC and antioxidant capacity values of phenolic compounds extracted from date flesh samples were lower than those extracted from seed samples. Liquid chromatography electrospray ionization quadrupole time-of-flight mass spectrometry (LC-ESI-QTOF-MS/MS) was used to characterize phenolic compounds. A total of 47 different phenolic compounds were identified, including 21 phenolic acids, 21 flavonoids, and 5 other polyphenols. High-performance liquid chromatography equipped with a photodiode array detector (HPLC-PDA) was used to quantify the phenolic compounds in different date samples. The highest content was found in ultrasonic-assisted date seed samples extracted in methanol solvent. The extract also contained epicatechin (36.12 μg/g), coumaric acid (27.05 μg/g), p-hydroxybenzoic acid (21.03 μg/g), syringic acid (19.85 μg/g), and epicatechin gallate (11.67 μg/g). The results showed that ultrasonic-assisted extraction could significantly increase the quantities of functional components of the extracts obtained from fresh date fruit samples.


Introduction
Date palm (Phoenix dactylifera L.) is a kind of staple plant with a long history.For over 5000 years of cultivation, dates were mostly grown by people in arid regions.Nowadays, 90% of dates are produced in the Middle East and North Africa.Meanwhile, the land area of date palm in Australia is also expanding and becoming an important market part [1] .Date seeds are abundant in phenolic compounds and He and Sun [2] reported that date seed phenolics could act as free radical scavengers and reactive oxygen defenders to inhibit the oxidative process.However, date seeds as one of the main byproducts are often discarded.But low-quality date fruits and date seeds waste could be valorized by extraction of healthy functional components for health supplements.
A practical method is essential for the extraction of phenolic compounds from dates.Maceration extraction is a conventional method that soaks a pulverized sample in solvents including water, ethanol, and methanol, in a closed system, followed by continuous or sporadic shaking at specific temperatures. [3,4]However, maceration extraction is time-consuming, solvent-consuming, and laborintensive.Besides, Soxhlet extraction was introduced due to its timesaving and solvent-saving properties, but high temperature might trigger the decomposition of heat-labile phenolic compounds during extraction. [5]Previous studies also proposed other extraction methods, such as supercritical fluid extraction and microwave-assisted extraction, but they are too expensive to apply to large-scale commercialization. [6,7]ost recently, ultrasonic as an efficient and low-cost method, has been accepted as a more suitable phenolic compound extraction method.The ultrasonic-assisted extraction (UAE) method leverages strong physical effects originating from acoustic cavitation to facilitate mass transfer at a microscopic level. [8]In brief, micron-sized cavitation bubbles can grow, oscillate, and undergo violent implosion, from which microstreaming, micro-jetting, and strong shear forces can be generated near the biomaterial surfaces causing surface peeling, erosion, and particle breakdown. [9]As a result, the permeability of the plant cell membrane will increase and release the phenolic compounds from plants into solvents.Meanwhile, UAE is defined as a fast nonthermal technique that can maximally protect phenolic compounds from degradation. [10]ifferent types of solvents, including water, ethanol, methanol, and their aqueous mixtures are often used in phenolic compound extraction from plant materials. [11,12]However, extraction solvent selection highly depends on the properties of targeted phenolics. [13,14]In general, it is assumed that high-polarity organic solvents are more suitable for the extraction of phenolic compounds and methanol has a higher polarity than ethanol which may be more suitable for phenolic compound extraction.The investigation of the extraction solvent affinity for date samples is limited, but methanolic solvent (80%) was applied in most of the experiments. [15,16][19] By contrast, polyphenol recovery studies with the comprehensive concern of multiple factors at the laboratory scale have been only sparingly investigated.Therefore, it is imminent to determine how ultrasonic and solvents could affect the phenolic compound extraction in date samples.
Overall, this research aimed to find out the most suitable extraction condition and improve phenolics' extraction efficiency by exploring the combined effect of ultrasonic and solvent type on phenolic extraction from date samples.Specifically, two extraction methods (conventional and ultrasonic-assisted extraction) and three solvent types (ethanol, methanol, and water) were involved in the research.The extractants were subjected to detailed characterizations; the total phenolic content (TPC), total flavonoid content (TFC) and total tannin content (TTC) assay are used to quantify the content of phenolic compounds in dates.Antioxidant potential has been reported to be analyzed by 2,2′-diphenyl -1-picrylhydrazyl (DPPH) antioxidant assay, ferric reducing/antioxidant power (FRAP) assay, 2,2'-Azino-bis-3-ethylbenzothiazoline-6-sulfonic acid (ABTS) radical scavenging assay, reducing power assay (RPA), hydroxyl radical scavenging activity ( • OH-RSA), ferrous ion chelating activity (FICA) and total antioxidant capacity (TAC) assay.Identification and quantification of the phenolics were carried out with liquid chromatography coupled with electrospray-ionization quadrupole time-of-flight mass spectrometry (LC-ESI-QTOF-MS/MS) and high-performance liquid chromatography (HPLC). [20]

Sample preparation
The fully ripe Medjool variety of dates were used in this study.They were obtained from a local market in Melbourne (Victoria, Australia).A total of 5 kg of dates were used, and the flesh and seeds were separated.Fruit flesh was blended into a slurry and the seeds were dried in a fume hood for a week and then reduced to powder with an 800 g grinder (Laobenhang, model 400Y, Yongkang, Zhejiang, China).The flesh and seed samples were then stored at −20°C.

Extraction of phenolics
The conventional extracts of seeds and flesh were prepared by mixing 3 g of sample slurry and powder with 30 mL solvent (70% ethanol, 70% methanol, and Milli-Q® water).Formic acid (0.1%) was added to break the cell wall and increase the permeability [21] and then subjected to a shaking incubator (ZWYR-240, Labwit, Ashwood, VIC, Australia) at 120 rpm for 16 h (10°C).Alternatively, the ultrasonic-assisted extracts of seeds and flesh were made by mixing 4 g of sample with 40 mL same series of solution and then applying 5 min ultrasonic in an ice water bath with a cell disruptor (Branson, model Digital Sonifier 450) at an amplitude of 40%.After extraction, the date extracts were centrifugated with centrifuge (Hettich, ROTINA380R, Tuttlingen, Baden-Württemberg, Germany) at 8000 rpm for 15 min (4°C).The supernatant was collected after centrifugation and stored at −20°C.The design of this experiment is shown in Figure 1.The abbreviations of all procedures are also included.

Determination of Total Phenolic Content (TPC)
The TPC value of date seeds and flesh extracts was determined using the Folin -Ciocalteu method mentioned by Kılıç, Can, Yılmaz, Yıldız, and Turna [22] with some modifications.25 μL of extract was added to a 96-well microplate and then mixed with 200 μL of Milli-Q® water, 25 μL 25% (v/v) Folin -Ciocalteu reagent and incubated at 25°C for 5 min.Subsequently, 25 μL of 10% (w/w) sodium carbonate solution was added with continued incubation in a dark environment for 1 h at 25°C.Absorbance at 756 nm was then measured in triplicate.The calibration curve built in this test was based on ethanolic gallic acid (0-200 μg/mL).The results of the samples are expressed as milligram gallic acid equivalents (GAE) per fresh weight (mg GAE/g f.w. ) ± Standard deviation (SD).

Determination of Total Flavonoids Content (TFC)
The TFC value of date seeds and flesh extracts was determined using a modified AlCl 3 colorimetricbased assay described by Stavrou, Christou, and Kapnissi-Christodoulou. [23] Extract (80 μL) was added to a 96-well microplate and then 80 μL of 2% aluminum chloride solution and 120 μL of 50 g/L sodium acetate solution were added for 2.5 h incubation at room temperature.After incubation, the absorbance at 440 nm was determined in triplicate.The calibration curve built in this test is based on ethanolic quercetin (0-50 μg/mL).Extraction samples are expressed as milligram quercetin equivalents (QE) per fresh weight (mg QE/g f.w. ) ±SD.

Determination of Total Content of Tannins (TCT)
The TCT value of date seeds and flesh extracts was determined using a modified assay mentioned by Price, Van Scoyoc, and Butler. [24]Extract (25 μL) was added to a 96-well microplate, and then 150 μL 4% vanillin solution and 25 μL 32% ethanol diluted sulfuric acid were added.The solutions were incubated for 15 min at 25°C and then the absorbance at 500 nm in was measured in triplicate.The calibration curve built was based on methanolic catechin (0-1000 μg/mL), and extraction yields were expressed as milligram catechin equivalents (CE) per fresh weight (mg CE/g f.w. ) ± SD.

Free radical scavenging activity assay (DPPH)
The free radical scavenging activity of date seeds and flesh extracts was measured via DPPH assay. [25]5 μL of extract and 275 μL 0.1 mM DPPH radical methanol solution were added to a 96-well microplate and allowed to stand at 25°C for 30 min and then the absorbance at 517 nm was measured in triplicate.The calibration curve built was based on aqueous Trolox (0-200 μg/mL), and extraction yields were expressed as milligram Trolox equivalents (TE) per fresh weight (mg TE/g f.w. ) ± SD.

Ferric Reducing/Antioxidant Power assay (FRAP)
The FRAP value of date seeds and flesh extracts was measured by an adopted assay described by Benzie and Szeto. [26]FRAP reagent was prepared fresh daily by mixing 300 mM sodium acetate buffer, 10 mM TPTZ and 20 mM ferric chloride in a ratio of 10:1:1 (v/v/v).20 μL of extract and 280 μL of FRAP reagent were added to a 96-well microplate, allowed to stand at 37°C for 10 min, and then measured the absorbance at 593 nm in triplicate.The calibration curve was based on aqueous Trolox (0-200 μg/ mL), and extraction yields were expressed as mg TE/g f.w.± SD.

Chelating ability of Ferrous Ion Assay (FICA)
The ferrous ion chelating ability of date seeds and flesh extracts was determined using an adopted assay mentioned by Lee, Hwang, Chung, Cho, and Park. [27]15 μL of extract, 50 μL 1:15 (v/v) diluted 2 mM Fe[II] aqueous solution and 50 μL 1:6 (v/v) diluted 5 mM ferrozine aqueous solution was added to a 96-well microplate and allowed to stand at 25°C for 10 min and then the absorbance at 562 nm was measured in triplicate.The calibration curve was based on aqueous ethylenediaminetetraacetic acid (EDTA) (0-50 μg/mL), and extraction yields were expressed as milligram EDTA equivalents (EE) per fresh weight (mg EE/g f.w. ) ± SD.

2,2'-azino-bis-3-ethylbenzothiazoline-6-sulfonic acid radical scavenging assay (ABTS)
The ABTS radical scavenging ability of date seeds and flesh extracts was determined using an adopted assay described by del Castillo, Ames, and Gordon. [28]ABTS + was prepared at room temperature via the reaction between 1.25 mL 7 mM ABTS and 22 μL 140 mM potassium persulfate solution and allowed to stand in a dark environment for 16 h.0.5 mL ABTS + solution is diluted in 45 mL ethanol to obtain absorbance between 0.75 and 0.78 at 734 nm. 10 μL of extract and 290 μL of ABTS + solution were added to a 96-well microplate and allowed to stand at 25°C for 6 min and then the absorbance at 734 nm was measured in triplicate.The calibration curve built in this test was based on aqueous TROLOX (0-500 μg/mL).Extraction yields are expressed as mg TE/g f.w.± SD.

Hydroxyl Radical Scavenging Activity assay ( • OH-RSA)
The • OH-RSA value of date seeds and flesh extracts was determined using an adapted assay as described by Hernández-Ledesma, Amigo, Netto, and Miralles. [29]50 μL of extract and 50 μL 6 mM H 2 O 2 were added to a 96-well microplate, and allowed to stand at 25°C for 10 min.Subsequently, 50 μL 6 mM 3-Hydroxybenzonic acid solution was added, allowed to stand at the same temperature in a dark environment for 10 min, and then measured the absorbance at 510 nm in triplicate.The calibration curve built in this test is based on aqueous Trolox (0-300 μg/mL).Extraction yields are expressed as mg TE/g f.w.± SD.

Reducing Power Assay (RPA)
The RPA value of date seeds and flesh extracts was determined using an adopted assay described by Oyaizu. [30]Buffer was prepared by mixing 0.2 M Na 2 HPO 4 •7 H 2 0 and 0.2 M Na 2 HPO 4 •H 2 0 (3.74 mL/ 6.24 mL). 10 μL of extract, 10 μL buffer, and 25 μL of 1% (w/v) K 3 [Fe(CN) 6 ] aqueous solution were added to a 96-well microplate, and allowed to stand at 25°C for 20 min.Subsequently, 25 μL of trichloroacetic acid (10.0%, w/v) was added to terminate the reaction and then adding 85 μL of Milli-Q® water, 8.5 μL ferric chloride (0.1% w/v) and allowed to stand at the same temperature in dark environment for 15 min then the absorbance at 750 nm was measured in triplicate.The calibration curve built in this test was based on aqueous Trolox (0-500 μg/mL).Extraction yields are expressed as mg TE/g f.w.± SD.

Determination of Total Antioxidant Capacity (TAC)
The TAC value of date seed and flesh extracts was determined using an adopted assay. [31]40 μL of extract and 260 μL of phosphomolybdate reagent (0.6 M H 2 SO 4 , 0.028 M sodium phosphate, and 0.004 M ammonium molybdate) were added to a 96-well microplate, and allowed to stand at 95°C for 90 min.Then, cooled at 25°C for 10 min and measured the absorbance at 695 nm in triplicate.The calibration curve built in this test was based on ethanolic ascorbic acid (0-300 μg/mL).Extracts are expressed as milligram ascorbic acid equivalents (AAE) per fresh weight (mg AAE/g f.w. ) ± SD.

Characterization of phenolic compounds using LC-ESI-QTOF-MS/MS analysis
Agilent 1200 series HPLC (Agilent Technologies, Santa Clara, CA, USA) was used for phenolic compound characterization along with an Agilent 6520 Accurate-Mass Q-TOF LC/MS (Agilent Technologies, Santa Clara, CA, USA).The separation was conducted using a Synergi Hydro-RP 80A, LC column (250 mm × 4.6 internal diameter, 4 μm particle size) (Phenomenex, Lane Cove, NSW, Australia) at a column temperature of 25°C and sample temperature of 10°C.The mobile phase was compromised of mobile phases A and B, which are A: aqueous acetic acid (98:2, v/v) and B: acetonitrile/water/acetic acid (100:99:1, v/v/v). [32]The gradient profile was accomplished over 85 min in a wide range of conditions (Time (min): Mobile Phase (A)(%): Mobile Phase (B)(%)) -0 min: 90% A: 10% B; 20 min: 75% A: 25% B; 30 min: 65% A: 35% B; 40 min: 60% A: 40% B; 70 min: 45% A: 55% B; 75 min: 20% A: 80% B; 77-79 min: 0% A: 100% B; 82-84 min: 90% A: 10% B. The mobile phase flow rate was set at 0.8 mL/min with a sample injection volume of 6 μL.Peaks were identified in both positive and negative ion modes with the capillary and nozzle voltage set to 3.5 kV and 500 V, respectively.A complete mass scan ranging from m/z 50 to 1300 was used, and MS/MS analyses were carried out in automatic mode with collision energy (10, 15, and 30 eV) for fragmentation.Both positive and negative modes were applied for peak identification.At the same time, the compounds were identified in date seed and fruit flesh based on their m/z value and MS 2 spectral data using MassHunter workstation software (Qualitative Analysis, version B.03.01) (Agilent Technologies, Santa Clara, CA, USA) and Personal Compound Database and Library (PCDL) with the database of the Kansas State University, USA.Compounds with scores of higher than 80 (PCDL Score) and mass error < ±5 ppm were selected for m/z verification and MS/MS identification purposes.

Quantification of polyphenols through HPLC-PDA analysis
The quantification of phenolic compounds present in date samples was carried out by HPLC (Waters Alliance 2690, Chromatograph Separation Module) along with a photodiode array (PDA) detector (Model 2998, Waters), which was set at λ 280, 320 and 370 nm with 1.25 scan/s (peak width = 0.2 min). [33]The sample volume used in this test was 20 μL under the same column conditions used in the LC-ESI-QTOF-MS/MS analysis.Individual phenolic compounds were determined using calibration curves generated from standards produced from 29 commonly found phenolic compounds present in dates.The results were expressed as μg/g of sample.All aspects of instrument control, data acquisition, and chromatography processing were conducted with Empower Software (2010).Venn diagrams based on the phenolic compounds quantified through HPLC-PDA were built, accompanied by a Heat Map that could assist in visualizing the correlation between the content of phenolic compounds and the extraction methods.

Statistics analysis
One-way analysis of variance (ANOVA) and Tukey's honestly significant differences (HSD) multiple rank test at p ≤ 0.05 were used to analyze the mean differences between samples.ANOVA was carried out by Minitab for Windows version 19.0 (Minitab, LLC, State College, PA, USA).The results are shown as mean ± standard deviation (SD).Pearson's correlation coefficient at p ≤ 0.05 and a principal component analysis (PCA) graph were applied to analyze the correlations between the content of phenolic compounds and antioxidant activities.

Polyphenol estimation (TPC, TFC, and TCT)
The phenolic compound content of each date sample was determined by the TPC, TFC, and TCT; the results are displayed in Table 1.Overall, the TPC, TFC, and TCT values were higher in date seeds than in date flesh.Regarding extraction method, UAE showed significantly higher TPC values than conventional extraction in both seed and flesh samples.However, differences were observed over the best solvent.In seed samples, ethanol-extracted samples obtained the highest phenolic content.In contrast, an aqueous solvent had the highest efficiency in date flesh samples.It was proved by the TPC value of CDFW and the TFC value of UDFW, which were significantly higher than other methods.
In the TPC assay, UDSE had the highest value (18.53 mg GAE/g), which was statistically higher than CDSE (14.46 mg GAE/g), which indicated UAE method could significantly increase TPC value in ethanolic extraction.Simultaneously, the TPC value of ethanolic extracted date flesh samples ranged from 0.81 to 1.19 mg GAE/g, while it presented from 0.85 to 0.96 mg GAE/g in the methanol group.But for aqueous extracted date flesh samples, the UA method decreased the TPC value from 1.40 to 1.18 mg GAE/g.UDSM (12.51 mg GAE/g) has a lower content than CDSM (14.41 mg GAE/g) in methanolic extraction.But the UAE method decreased the TPC value from 1.40 to 1.18 mg GAE/g for aqueous extracted date flesh samples.Previous studies had reported that the TPC value of date seed was significantly higher than that of date flesh, which matches the results of this assay. [34,35]or TCT, the value of UDSE (40.26 mg CE/g) was the highest, followed by UDSM, CDSE, and CDSM (32.74 mg CE/g), 22.70 mg CE/g; 21.58 mg CE/g, respectively.Consistent with previous studies, all methods had a poor ability to extract condensed tannins in date flesh, [19,36] because of its low content or due to the complexity of the matrix of raw material (Medjool dates) (Ghnimi, Umer, Karim, and Kamal-Eldin. [37]n relation to TFC, ultrasonic had no benefit for phenolic compound extraction in date seed.CDSE (0.53 mg QE/g) showed the highest content of flavonoids among all the seed samples, followed by CDSM (0.52 mg QE/g) and UDSE (0.52 mg QE/g).Focusing on date flesh, a meaningful difference demonstrated that UDFM and UDFW extract more flavonoids than others (0.02 mg QE/g; 0.02 mg QE/g, respectively).Compared with previous studies, [38] the values were lower which may be due to two reasons.(i) 70% acetone may be a better solvent for flavonoid extraction than the three solvents applied in our assay, and (ii) the difference of conditions in date growing, processing, and storage, which may affect the total flavonoid content in date fruits.

Antioxidant estimation (DPPH, FRAP, ABTS, • OH-RSA, RAP, FIAC, and TAC)
The antioxidant efficacy of date samples was carried out using DPPH, FRAP, ABTS, RPA, • OH-RSA, FICA, and TAC assays.The results (Table 1) indicated the potential to scavenge free radicals by different date samples.
Adjimani and Asare [39] have mentioned that reactive oxygen species (ROS) cause damage to biomolecules, and the antioxidants in dates can scavenge the ROS during metabolic activities.In other words, they can act as free radical scavengers, metal chelating agents, and hydrogen atom donators to inhibit oxidative stress and prevent damage.In this project, parts of antioxidant experiments were used to evaluate phenolic compounds' free radical scavenging activity, including DPPH, • OH-RSA, and ABTS assays.FICA was used to determine the metal chelation properties.FRAP determination focused on the capacity of donating electrons (reduce Fe 3+ -TPTZ complex to Fe 2+ -TPTZ complex).
For seed samples, DPPH values of CDSE, CDSM, and UDSE were the highest (24.63 mg TE/g, 24.15 mg TE/g, 24.63), while CDSW and UDSW had significantly lower values of DPPH (4.03 mg TE/ g and 5.12 mg TE/g).The value of FRAP and ABTS also showed a similar trend.It might attribute to the higher solubility of phenolic compounds in organic polar solvents. [11]he RPA (41.72 mg TE/g) and FICA (0.12 mg TE/g) values of CDSM were higher than compared to those observed in other methods.Concerning • OH-RSA, the highest value was found in the UDSM sample (116.79 mg TE/g).Also, a higher scavenging ability was shown when compared with other methods applied samples.Additionally, CDSE was recorded as the highest value (18.69 mg AAE/g) in the TAC assay.However, the valve of TAC values in the previous report was higher, which might be because of the pre-treatment by freeze-drying, which increased the extraction efficiency and antioxidant capacity. [40]

Distribution of phenolic compounds -venn diagram
To explore the distribution of phenolic compounds in date palm, Venn diagrams were made among factors including solvent selection, extraction method, and fruit parts (Figure 2), as well as the phenolic compound distribution among different factors in date seed and flesh (Figure 3).

Phenolic compound distribution in different factors.
Fruit parts were divided into two groups (date seeds and date flesh), extraction methods were divided into two groups (conventional and UAE) and solvents were divided into three groups (water, ethanol, and methanol).
A total of 162 phenolic compounds were found in all date extract samples through LC-ESI-QTOF-MS as shown in Figure 2. In this study, methanol extracts contained the largest diversity of phenolic compounds among the three solvents followed by ethanol.According to Figure 2(a,b), 24.1% of the compounds were identified in both date seeds and flesh and 35.8% were identified in both extraction methods.It was verified that data seeds contained more phenolic compounds than flesh, which was constant with Allaith [34] and H. Liu et al. [41] 97 (59.99%) unique phenolic compounds in date seeds were identified, while only 26 (16%) phenolic compounds were found in fruit flesh.Interestingly, UAE methods extracted 80 (49.4%) phenolics, three-folder higher than conventional-extracted.This result proved that UAE could more efficiently extract the phenolic compounds from plants, with greater quantity and quality. [42]Besides, among extraction solvents expressed in Figure 2c, 13% of them were identified in all three solvents including aqueous, ethanol, and methanol.Thereby, methanol is solvent with the most phenolic compounds identified (40.7%), which was much more than aqueous (6.2%) and ethanol (9.9%) which was proven by a previous study that higher polarity of extraction solvent could trigger higher solubility of plant phenolic compounds. [43]enolic compound distribution in date seed and flesh.Since the complexity of the similar and unsimilar phenolic compounds existed in samples, two Venn diagram in Figure 3 was developed to visualize the distribution of phenolic compounds in date fruit and date seed among all the five factors, including conventional and ultrasonic-assisted extraction, solvents including ethanol, methanol, and water.
In Figure 3a, the number of the phenolic compounds present in all factors was 14; meantime 13 phenolics were identified in all factors except the aqueous method.Moreover, significant numbers of phenolic compounds appeared only in ultrasonic and methanol.It could support that methanol was an effective solvent for phenolic compound extraction, although the substances of phenolic compounds in date flesh were limited.A total of 57 phenolic compounds were found in data seed extracts extracted in methanol solution and simultaneously applied by ultrasonic in Figure 3b.This result was much higher than other factors showing the advantages of methanolic extraction and ultrasonic assistance.A similar situation also appeared in Figure 3b; 22 types in total of phenolic compounds were identified in all factors followed by 20 compounds identified in all factors except the aqueous method.

LC-ESI-QTOF-MS/MS based characterization of phenolic compounds
In this study, a qualitative analysis of the phenolic compounds from date seed and date flesh extracts has been conducted using LC-ESIQTOF-MS/MS in negative and positive ionization modes.47 different phenolic compounds were identified in all date samples including 21 phenolic acids, 21 flavonoids, and 5 other polyphenols (shown in Table 2).
Hydroxybenzoic acids.A total of 6 hydroxybenzoic acids were identified in this test, and 5 out of 6 were identified in both conventional and UA methods, 4 out of 6 were identified in methanolic extracted seed samples.Ellagic acid glucoside (Compound 1) was tentatively identified in CDSW, which was based on the observed [M -H] − m/z at 463.0501 with the product ions at m/z 301, which as Escobar-Avello et al. [44] mentioned in their work was the indication of the loss of hexosyl moiety (162 Da) from precursor ions.However, according to the previous study by Chaira et al., [15] this compound was not found in dates.Two phenolic acids conjugated with glucoside were also found, and tentatively identified as protocatechuic acid 4-O-glucoside (Compound 2) in both CDSW and UDSM and 4-hydroxybenzoic acid 4-O-glucoside (Compound 3) only in UDFW based on the product ions at m/z 153 and m/z 137, these two compounds were confirmed by Ma, Dunshea, and Suleria. [45]ompounds 4 and 5 were tentatively identified as 2,3-Dihydroxybenzoic acid in CDSE, CDSM, UDSE, and UDSM and 2-Hydroxybenzoic acid in UDSM, CDSE, and CDSM showing product ions at m/z 109 and m/z 93 but these two compounds were not found in previous studies.One gallic acid derivative was also identified only in UDSM, which is gallic acid 4-O-glucoside (Compound 6), but previous studies did not mention this compound existed in dates.
Hydroxycinnamic acids.Hydroxycinnamic acids were the largest group among all the subclass of phenolic acids detected in this study.A total of 14 hydroxycinnamic acids were identified and 5 out of 14 hydroxycinnamic acids (Compound 11, 12, 16, 17, 20) were detected in both conventional and UA   [46] mentioned in their study represented the loss of SO 3 and CH 3 .Several studies confirmed this compound as a primary phenolic acid found in dates. [45,47]Compounds 11, 12, 16, and 20 were characterized in both modes, tentatively identified as cinnamic acid, caffeoyl glucose, m-Coumaric acid, and 2-Hydroxy-2-phenylacetic acid.In the MS 2 spectra, the loss of the CO 2 [M -H − 103] was observed in cinnamic acid. [48]In terms of caffeoyl glucose, the product ion at m/z 179 and m/z 161 corresponds to the loss of glucoside (162 Da), glucuronide (176 Da), tartaric acid fission (150 Da) [49] X. [50] J. Wang et al. [50] reported that m-Coumaric acid produced product ion at m/z 119 was related to the loss of CO 2 (44 Da) from the parent ion.And for 2-Hydroxy-2-phenylacetic acid, produced product ions at m/z 136 and m/z 92 corresponding to the loss of H 2 O (18 Da) and CO 2 (44 Da). [51]ompounds 7, 13, 14, 15, and 18 were only detected in UA extracts and were all characterized in negative mode.Compound 7 was tentatively identified as 3-p-coumaroylquinic acid according to the [M -H] − m/z at 337.0917, which was also found by Ma et al. [45] In the MS 2 experiment of 3-p-coumaroylquinic acid, the product ions at m/z 265, m/z 173 and m/z 162 were due to the respective loss of 4 H 2 O, C 9 H 7 O 3 and C 7 H 11 O 5 . [52]Compound 13 was tentatively identified as 3-Caffeoylquinic acid, confirmed by the product ions of m/z 253, m/z 190 and m/z 144. [52]The presence of sinapic acid (Compound 14) was confirmed by the product ions at m/z 233 and m/z 179 representing the consecutive loss of COO from the precursor. [53]Caffeic acid (Compound 15) was confirmed in the MS 2 spectrum, which showed the product ions of m/z 143 and m/z 133. [52]ompound 18 was tentatively identified as rosmarinic acid, confirmed by the product ion at m/z 179. [54]According to the study done by Mansouri, Embarek, Kokkalou, and Kefalas, [55] sinapic acid, and caffeic were determined in ripe Algerian date fruits, while Sheikh et al. [56] mentioned rosmarinic acid could be determined in Ajwah cultivar of date palm.
For compounds detected in conventional extracts, there were only four (Compounds 8, 9, 10, 19), and they were all characterized in negative mode.Compound 8 was tentatively identified as p-Coumaroyl tartaric acid confirmed by the [M -H] − m/z at 295.0452.Compound 9 was tentatively identified as p-Coumaric acid 4-O-glucoside, confirmed by the product ion at m/z 163 [M -H − 162, loss of glucoside]. [57]Due to the fragment of m/z 193 [M -H − glucoside, loss of 162 Da], Compound 10 with the [M -H] − m/z at 337.0917 was confirmed as ferulic acid 4-O-glucoside (X. [50]Compound 19 was tentatively identified as 3,4-Dihydroxyphenylacetic acid ([M -H] − m/z at 167.0345), confirmed by the product ion at m/z 149 and m/z 123.Both ferulic acid 4-O-glucoside and 3,4-Dihydroxyphenylacetic acid were previously reported by Ma et al. [45] that were determined in date palms.
Hydroxyphenylpentanoic acids.In this study, only one compound was found in this subclass of phenolic acid, compound 21 was detected in both the conventional and UA extracts and is characterized in negative mode.Compound 21 was only found in date flesh and only extracted by aqueous solvent, tentatively identified as dihydroferulic acid 4-sulfate because of the [M -H] − m/z at 275.0231.In the MS 2 spectra, the product ion at m/z 195, m/z 151, and m/z 177 corresponded to the loss of glucuronide (176 Da) moiety, C 8 H 8 O (120 Da), C 6 H 6 O (659 Da), respectively, was observed in dihydroferulic acid 4-sulfate. [58,59]This compound was not found in previous studies.
Dihydroflavonols.Compound 23 was tentatively identified as Dihydromyricetin 3-O-rhamnoside which was detected in the negative mode of both extraction methods with [M -H] − m/z at 449.1099 confirmed by the product ion at m/z 301.Dihydromyricetin 3-O-rhamnoside (Compound 24) was detected only in UDSM in negative mode with [M -H] − m/z at 465.1036.This compound as Getasetegn [60] reported in their reasearch was previously found in Ethiopia's khat plant.In addition, these two dihydroflavonols have been reported in the literature by Ma et al.. [45] Flavanols.Flavanols are one of the largest groups among all the subclass of flavonoids detected in this study.In the present work, a total of 5 flavanols were identified.Three out of 5 flavanols (Compound 25, 26, 27) were detected in both conventional and UA-extracted samples and detected in both modes.Interestingly, all five compounds were identified in UDSM.Compound 25 was allocated for (+)-Catechin confirmed by the product ion at m/z 245, m/z 205 and m/z 179.Compound 26 and Compound 27 were tentatively identified as procyanidin trimer C1 and procyanidin dimer B1 based on the [M -H] − m/z at 865.1995 and [M -H] − m/z at 577.134.But due to our knowledge, procyanidin trimer C1 and procyanidin dimer B1 have never been reported in previous literature.
Compounds 28 and 29 were only detected in UDSM and were both characterized in negative mode.Compound 28 was tentatively identified as theaflavin 3,3'-O-digallate according to the [M -H] − m/z at 867.1401.Compound 29 was tentatively identified as (+)-Catechin 3-O-gallate, confirmed by the product ions of m/z 289, m/z 169 and m/z 125 from the parent ion.However, these two compounds were also not recorded in previous studies about the phenolic compounds in date palms.
Flavones.In this study, four flavones were detected, and two were detected in both methods of extracted samples (Compound 31, Compound 33) and in both modes.Compound 31 was tentatively identified as chrysoeriol 7-O-glucoside due to the [M + H] + m/z at 463.1244.In the MS 2 spectra, based on the study done by Liao et al., [61]  This compound was reported by Khallouki et al. [16] and Ma et al. [45] that existed in date palms.Apigenin 6-C-glucoside (Compound 33) was detected with [M -H] − m/z at 431.0974.The identification of apigenin 6-C-glucoside which was also reported by Ma et al. [45] and was confirmed by the product ion detected at m/z 413 [M -H − 18], m/z 341 [M -H − 90] and m/z 311 [M -H − 120], resulting from the loss of H 2 O, 90 and 120 Da loss of cross-ring cleavages of the glycoside moiety, respectively. [62]ompounds 32 and 34 were only detected in UDSM and were both characterized in negative mode.Compound 32 was tentatively identified as rhoifolin according to the [M -H] − m/z at 577.1561.In the MS 2 fragmentation, peaks at m/z 413 [M -H -rhamnose moiety -H 2 O, loss of 164 Da] and m/z 269 [M -H -rhamnose moiety -glucose moiety, loss of 308 Da] were observed in rhoifolin. [59]Compound 34 was tentatively identified as 6-Hydroxyluteolin 7-O-rhamnoside ([M -H] − m/z at 447.0918), confirmed by the product ions of m/z 301 [M -H -rhamnoside] from the parent ion. [63]These two compounds were not identified in previous literature and only existed in UA method samples.
Flavonols.In the present work, a total of 3 flavonols were identified.All flavonols (Compounds 35, 36, and 37) were detected in UA-extracted samples.Myricetin 3-O-rhamnoside (Compound 35) presenting in UDSM and UDSE in both modes was tentatively identified based on the MS 2 peaks at 317, representing the loss of rhamnoside (146 Da) from parent ion. [64]This compound was also identified in the study done by Ma et al. [45] Isorhamnetin (Compound 36) presenting also in UDSM and UDSE in both modes was tentatively identified by the product ions at m/z 300 and m/z 271.Isorhamnetin was identified in several studies about the phenolic compound in dates. [18,45,47]ompound 37 was only detected in UDSM and was characterized in negative mode as kaempferide.
Compound 41 was only detected in conventional extract and was characterized in positive mode.This compound was tentatively identified as 2-Dehydro-O-desmethylangolensin.Ma et al. [45] mentioned that this compound was previously found in palm fruits.
Other flavonoids.Other flavonoids include one chalcone (Compound 22) and one flavanone (Compound 30).Compound 22 was tentatively identified as xanthohumol which was detected in the UDSM sample and was characterized in negative mode.Compound 30 was tentatively identified as hesperetin 3'-O-glucuronide was detected in both conventional and UA samples in negative mode.This compound was achieved by the product ions at m/z 301, m/z 175, m/z 113 and m/z 85.However, Ma et al. [45] reported that this compound was detected in jelly palm fruit but not recorded in date palm.
Other polyphenols.In date extracts, several polyphenols other than phenolic acids and flavonoids were detected in both methods, and a total of 5 polyphenols were identified, including one furanocoumarin, one hydroxybenzaldehyde, one hydroxyphenylpropene, one phenolic terpene, and one other polyphenol.
A total of 3 out of 5 compounds of these polyphenols were detected in UA extracts.Only furanocoumarin (Compound 43) was detected in the UDFE sample and was characterized in positive mode.According to the [M + H]+ m/z at 247.0611, this compound was identified as isopimpinellin.In the MS 2 fragmentation, peaks at m/z 232, m/z 217, m/z 205 and m/z 203, representing the loss of CH 3 (15 Da), 2CH 3 (30 Da), CH 2 (42 Da), and CO 2 (44 Da), were observed in isopimpinellin. [67]Compound 44 was detected only in conventional extracts in negative mode and was tentatively identified as a hydroxybenzaldehyde (4-Hydroxybenzaldehyde) due to [M -H] − at m/z 121.0292. [68]One hydroxyphenylpropene (Compound 45) was detected in positive mode and observed in both conventional and UA extracts as 2-Methoxy-5-prop-1-enylphenol.Carnosic acid (Compound 46) was the only phenolic terpene detected in UDSM in negative mode, identified based on [M -H] − m/z at 331.1905, and was found in banana, lime, and avocado peels. [51]Salvianolic acid B (Compound 47) with precursors at [M -H] − m/z at 717.1456 is one other polyphenol detected in CDFW in positive mode.All five compounds, were not detected in previous studies of date palms.
According to our results, the solvent and extraction method applied to the date seeds and flesh would influence the profile of phenolics in samples.After sonication, ethanolic and methanolic extraction samples indicated increasing effective compounds.Meanwhile, the characterization of polyphenolic compounds confirmed the potential antioxidant present in date samples.Many detected compounds have been reported to have substantial health benefits, such as antioxidant, anti-diabetic, anti-hypertensive, anti-inflammatory, and anti-inflammatory anti-cancer activities. [69]These antioxidant compounds present in dates illustrate that these products are a prospective source of bioactive ingredients with antioxidant potential.
LC-ESI-QTOF-MS/MS analysis was an effective tool to characterize and identify the phenolic compounds in 12 date samples.47 phenolics were found under the application of LC-MS/MS with their fragment ions.Based on our research, few researchers have reported all these compounds in previous studies, and some compounds were recognized in the dates for the first time.These initially identified date phenolic components with significant antioxidant potential could be tested in vivo in follow-up studies for their health benefits and further applications in the food, feed, and pharmaceutical industries.

HPLC analysis
As shown in Table 3, HPLC chromatograms of date extracts presented the mass of various phenolic compounds in samples.Caftaric acid, chlorogenic acid, and epicatechin concentrations were determined to be about 39 μg/g, 36 μg/g and 34 μg/g, respectively.These three compounds are the main compounds in this sample.UDSM had the highest phenolic compound content while UDSW had the lowest.As a result, UDSM extracts had the most various types of phenolic acids and the highest content.Also, we found more types of phenolic acids extracted from date seeds by applying UAE.This result corroborates the idea of Chemat, Lagha, AitAmar, Bartels, and Chemat, [70] who suggested that UAE is more effective in phenolic compound extraction than conventional extraction.Three phenolic acids, namely, p-hydroxybenzoic acid, syringic acid, and coumaric acid, were qualified as the highest types in UDSM.The same number of phenolic acids also qualified highest in CDSM (p-hydroxybenzoic acid, chlorogenic acid, and syringic acid).In addition, caffeic acid can only be qualified in CDSM.
For the flavonoids in seeds, we found quercetin-3-glucuronide can be extracted with the assistance of ultrasonic.Quercetin concentration was determined 17.97 μg/g in UDSE and 17.98 μg/g in UDSM.As Boots, Haenen, and Bast [71] mentioned in their work, quercetin is a member of the flavonoids family which relates to antioxidant, anticancer, and antiaging activities.Moreover, epicatechin, epicatechin gallate, polydatin, and quercetin-3-glucuronide can be extracted by UDSM, but CDSM cannot, indicating a significant increase of the released flavonoids from dates after sonication.
For flesh extracts, CDFW can extract more types of phenolic acids from date flesh and gallic acid extracted by this method is the highest (18.89 μg/g) while no phenolic acids can be extracted by UDFE.Besides, epicatechin gallate is the only flavonoid that could be extracted by UDFE and UDFM from date flesh, which means ultrasonic can raise the extraction of flavonoids and the solvent difference did not affect on the content of epicatechin gallate (11.77 μg/g and 11.73 μg/g).

Heat map and hierarchical clustering phenolic compound analysis
This heat map was constructed to analyze the hierarchical clustering of the phenolic compounds in the 12 extract samples of dates (Figure 4).The correlation between samples and compounds was represented by each sample's measurable distance and average concentration, the rows and columns clustered.The tightest clusters will be clustered in tree order first.
Based on the hierarchical clustering, two clusters in rows and columns were generated and highlighted in Figure 4; the significant differences among phenolic profiles were represented as samples in different clusters.The results show that UDFW, CDFM, CDFW, UDFM, and CDFE were clustered together in the first group showing similar phenolic contents.All the extracts in this cluster had red color areas for gallic acid, meaning they all have a higher content of this compound.The rest of the extracts as the second group had more red color areas, including CDSM, CDSE, UDSM, UDSE, CDSW, UDSW, and UDFE.In this group, UDSM had the highest content of quercetin-3-glucuronide while UDSE and CDSW had the highest content of polydatin.Compared with the first group, more phenolic compounds with higher content could be found in date seed extracts.The higher concentration could be found in UAE of date seed, which matched the expectation that ultrasonic can increase the type and content of phenolic compounds extracted from plants.

Correlation between phenolic compounds and antioxidant assays
The correlation between phenolic content (TPC, TFC, and TCT) and antioxidant activities (DPPH, FRAP, ABTS, ABTS, • OH-RSA, RAP, FICA, and TAC) was performed with Pearson's correlation test (Table 4).A principal components analysis (PCA, Figure 5) was also performed to determine the similarities and differences among all the studied fractions and the relationship among the antioxidant potential assays.A total of 93.92% of initial data variability can be explained by the first two factors (F1 and F2) in Figure 5. Regarding antioxidant assays, DPPH, ABTS, FRAP, RPA, and TAC were strongly correlated with each other (p ≤ 0.01).This positive correlation was proved by Floegel, Kim, Chung, Koo, and Chun [72] as ABTS and DPPH assays both can be used as a tool to evaluate the free radical scavenging ability.The high correlation among DPPH, ABTS, FRAP, and TAC indicated that phenolic compounds present in date extracts showed a strong antioxidant capacity, and these assays are strongly related.However, the figure also showed that FICA and • OH-RSA had very little correlation with other assays and TFC did not show a strong relationship with other antioxidant assays like FRAP and FICA (r = 0.632; r = 0.342).
TPC had a strong correlation with most antioxidant assays except • OH-RSA and FICA (r = 0.494; r = 0.059) which suggested that phenolic compounds are strongly related to the antioxidant of the date samples which was also proven by previous studies. [51]Strong correlations were also observed between TCT and antioxidant assays except for FICA (r = −0.067),which indicates that tannin also significantly contributes to the antioxidant activities in dates.However, the correlation between TFC and FRAP, • OH-RSA, and FICA (r = 0.632; r = −0.057;r = 0.342, respectively) was not significant, indicating flavonoids contribute less to the antioxidant potential of date samples.From Table 4, we also found that FICA had non-significant correlations with the phenolic, flavonoid, or tannin content.different extracts, and a total of 162 phenolic compounds were found, most of which were identified from UDSM.According to the phenolic compounds quantified by HPLC-PDA, date samples, especially date seeds, are a good source of phenolic compounds.With this context, ultrasonic-assisted method and methanolic solvent has been proven to improve the value of date extracts by maximizing the extraction of phenolic compounds, which has the potential to be a food processing agent and nutritional supplement.

Figure 1 .
Figure 1.Flowchart of the experimental design with the abbreviation of each procedure.

Figure 2 .
Figure 2. Venn diagram of total phenolic compounds presented in different date samples.(a) Comparison of total phenolic compounds presents in date seeds and flesh.(b) Comparison of total phenolic compounds presents in different extraction methods.(c) Comparison of total phenolic compounds presents in different solvents.

Figure 3 .
Figure 3. Venn diagram of total phenolic compounds presented in different date samples grown.(a) shows comparison of total phenolic compounds present in date flesh.(b) shows comparison of total phenolic compounds present in date seed.

Figure 4 .
Figure 4. Heatmap showing phenolic compounds distribution and concentration among twelve date samples.Red boxes' mean concentrations are higher than the mean value among samples.Blue boxes mean lower concentrations.

Table 1 .
The antioxidant assays of date seeds and date flesh.
aThe data are shown as mean ± standard deviation (n = 3); a, b, c, d, e indicate the means in a row with significant difference (p < .05)using a one-way analysis of variance (ANOVA) and Tukey's test.The standards and samples were mentioned in abbreviations.GAE, gallic acid equivalents; QE, quercetin equivalents; CE, catechin equivalents; AAE, ascorbic acid equivalents; TE, Trolox equivalents; EE, EDTA equivalents.

Table 2 .
Characterization of phenolic compounds extracted by different methods in date seeds and flesh by LC-ESI-QTOF-MS/MS.

Table 2 .
(Continued).Compound was detected in more than one seed/flesh samples, data presented in this table are from asterisk sample.**Compounds were detected in both negative [M-H] − and positive [M+H] + mode of ionization while only single mode data was presented.Date seed and flesh samples were mentioned in abbreviations.extracted samples while 5 out of 14 of them (Compound 7, 13, 14, 15, 18) were only detected in UA samples and 4 out of 14 (Compound 8, 9, 10, 19) were only detected in conventional samples.And methanolic extraction was most suitable for hydroxycinnamic acids because 10 out of 14 samples can be identified in methanolic extracted samples.Only Compound 17 was characterized in negative mode, which is tentatively identified as ferulic acid with observed [M -H] − m/z at 193.0503 and further confirmed by the product ions at m/z 178 (M -H-95 Da) and m/z 193 (M -H − 80 Da), which as Piazzon et al.

Table 3 .
Quantitative analysis in phenolic compounds of date seeds and flesh extracts.