Food Chemistry 318 (2020) 126450

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Simultaneous extraction and separation of bioactive compounds from apple T

pomace using pressurized liquids coupled on-line with solid-phase
extraction
Laise C. da Silvaa, Mariana C. Souzaa, Beatriz R. Sumerea, Luiz G.S. Silvaa, Diogo T. da Cunhaa,
⁎
Gerardo F. Barberob, Rosangela M.N. Bezerraa, Mauricio A. Rostagnoa,
a
Multidisciplinary Laboratory of Food and Health (LabMAS), School of Applied Sciences (FCA), University of Campinas (UNICAMP), Rua Pedro Zaccaria, n. 1300,
13484-350 Limeira, SP, Brazil
b
Department of Analytical Chemistry, Faculty of Sciences, University of Cadiz, Agrifood Campus of International Excellence (ceiA3), IVAGRO, P.O. Box 40, 11510 Puerto
Real, Cadiz, Spain

A R T I C LE I N FO A B S T R A C T

Chemical compounds studied in this article: The objective of this work was the development of an on-line extraction/fractionation method based on the
Gallic acid (PubChem CID: 370) coupling of pressurized liquid extraction and solid-phase extraction for the separation of phenolic compounds
Chlorogenic acid (PubChem CID: 1794427) from apple pomace. Several variables of the process were evaluated, including the amount of water of the first
Epicatechin (PubChem CID: 72276) stage (0–120 mL), temperature (60–80 °C), solid-phase extraction adsorbent (Sepra, Isolute, Strata X and Oasis)
Rutin (PubChem CID: 5280805)
and activation/elution solvent (methanol and ethanol). The best results were observed with the adsorbent Sepra.
Hyperoside (PubChem CID: 5281643)
The temperature had a small effect on recovery, but significant differences were observed for phlorizin and a
Quercetin rhamnoside (PubChem CID:
5353915) quercetin derivative. Results indicate that ethanol can be used to replace methanol as an activation, extraction/
Phloretin xylosil glucoside elution solvent. While using mostly green solvents (water, ethanol, and a small amount of methanol that could be
Phlorizin (PubChem CID: 6072) reused), the developed method produced higher or similar yields of acids (2.85 ± 0.19 mg/g) and flavonoids
(0.97 ± 0.11 mg/g) than conventional methods.
Keywords:
Apple pomace
Phenolics
Flavonoids
Pressurized liquid extraction
Solid-phase extraction

1. Introduction Martínez, 2009; Maleki, Crespo, & Cabanillas, 2019).

Among fruits, apple is known to contain large quantities of phenolic
Mounting evidences are suggesting that several naturally occurring compounds, including phenolic acids (such as gallic), flavanols (such as
chemical compounds in fruit residues have biological properties with epicatechin), flavonols (such as quercetin and its derivatives) and di-
potential to be used for the prevention and treatment of the main causes hydrochalcones (such as phloretin and phlorizin) (Suárez et al., 2010).
of death worldwide, such as cardiovascular diseases and cancer Evidence suggests that phlorizin is uniquely found in apples and apple
(Babbar, Oberoi, & Sandhu, 2015; Baiano, 2014; Rostagno and Prado, products (Lee, Chan, & Mitchell, 2017).
2013). Phenolics are one of the most studied bioactive compounds due The greater part of the apple phenolics is found in the cell vacuoles,
to their high antioxidant capacity, which is associated with health and their concentration is higher in the peel when compared to other
benefits and the prevention of oxidative stress. However, their effects tissues of the fruit. It is estimated that between 0.6 and 18% of the
on the organism are far more complex and may involve interference in phenolic compounds are extracted with the juice, and therefore most
cell cycle and signaling, anti-inflammatory properties, and physiolo- phenolics remain in the pomace (Grigoras, Destandau, Fougère, &
gical effects (García-Lafuente, Guillamón, Villares, Rostagno, & Elfakir, 2013). More importantly, the pomace can make up to 35% of

⁎
Corresponding author.
E-mail addresses: [email protected] (D.T. da Cunha), [email protected] (G.F. Barbero), [email protected] (R.M.N. Bezerra),
[email protected] (M.A. Rostagno).

https://doi.org/10.1016/j.foodchem.2020.126450
Received 19 August 2019; Received in revised form 18 February 2020; Accepted 18 February 2020
Available online 19 February 2020
0308-8146/ © 2020 Elsevier Ltd. All rights reserved.
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

processed fruits, thus representing an enormous potential to be ex- 2. Material and methods
plored for the production of highly concentrated extracts to be used in
medicine, foods, and even cosmetics (Hyson, 2012). 2.1. Chemicals and solvents
However, in order to explore this potential, it is necessary to remove
the phenolic compounds from the pomace in a fast, selective way with HPLC grade acetonitrile and 96% ethanol were obtained from
the lowest cost and environmental impact as possible. There are several Panreac (Barcelona, Spain), while phosphoric acid was obtained from
extraction techniques and methods available, but conventional techni- Merck (Darmstadt, Germany). Ultra-pure water was supplied by a
ques are not efficient and selective, producing extracts with low con- Purelab Flex 3 system (Elga Veolia, High Wycombe, UK. The standards
centration while requiring additional processes such as filtration, frac- of the compounds analyzed by liquid chromatography were supplied
tionation, and concentration (Rabetafika, Bchir, Blecker, & Richel, from Sigma Aldrich (St. Louis, MO, USA). Standard solutions of each
2014; Reis, Rai, & Abu-Ghannam, 2012; Rostagno and Prado, 2013). substance were prepared in the recommended solvent according to the
A promising alternative is Pressurized-Liquid Extraction (PLE), supplier and stored at −20 °C.
which is a modern extraction technique based on the use of high tem-
perature and pressure to enhance the process and improve yields. The 2.2. Raw material
high temperature affects extraction kinetics by increasing mass transfer
rates and diffusion as well as the solubility of compounds while redu- The industrial residue from the production of apple juice was sup-
cing the viscosity of the solvent, among other effects. Furthermore, plied by the company (Macçã Desidratados e Congelados, Fraiburgo,
since the raw material is retained in the extraction cell, the process SC, Brazil). The residue was received dry and was stored at −20 °C
eliminates the need for filtering the extract after the process is com- until used as raw material for the extraction process.
pleted. Additionally, the higher efficiency of the technique allows re-
placing organic solvents, such as methanol or acetone, by an en- 2.3. Pressurized liquid extraction coupled on-line with solid phase
vironmentally friendly solvent such as water or ethanol. There are extraction
successful reports of the use of PLE for the extraction of different
polyphenols from apples, and it is parts (Alonso-Salces et al., 2001; The extractions were performed in the “Extract-US” system (FAPESP
Plaza, Abrahamsson, & Turner, 2013; Wijngaard & Brunton, 2009). 2013/04304-4, patent-pending) (Santos et al., 2019; Sumere et al.,
Advantages over conventional techniques include lower solvent con- 2018). The system consists of a liquid pump (PU-2080, Jasco, Tokyo,
sumption, higher efficiency, and elimination of additional post-extrac- Japan), needle valves (Autoclave Engineers, Erie, Pennsylvania, USA),
tion steps, among others (Rostagno, D’Arrigo, Martínez, & Martínez, back pressure regulator (Tescom, Ferguson, Missouri, USA) and five
2010; Rostagno, Villares, Guillamón, García-Lafuente, & Martínez, automatic two-position, 10-port valves (Waters, Milford, MA, USA)
2009). (Fig. 1 – supplementary material).
Despite being an efficient technique, PLE is not capable of separ- Extractions were carried out as follows: after a steel filter was placed
ating compounds from similar phenolic classes, and the produced ex- on the lower lid of the extraction cell, a layer of glass beads (2 mm) was
tracts contain a broad mix of components. Solid adsorbents (Solid Phase used to reduce the dead volume in the cell. Subsequently, the sample
Extraction – SPE) can be used to separate specific phenolic classes and (1.0 g) was weighed and transferred to the extraction cell (10 mL). The
compounds. The mechanism of SPE is similar to a chromatographic cell was then connected to the system, and the solvent inlet tubing
separation, where extracted compounds are passed through a column connected to the extraction cell. In parallel, the SPE column was filled
packed with the adsorbent and separated due to different interactions with the selected adsorbent (Sepra™ C18-E, particle size: 50 µm, pore
between the compounds, adsorbent, and solvent (Rostagno and Prado, size: 85 Å, Phenomenex, USA; Isolute C18-EC, particle size: 50 µm, pore
2013; Rostagno et al., 2010, 2009). size: 60 Å, Biotage, Sweden; Strata X C18, particle size: 33 µm, pore size:
Our hypothesis is that separation of the phenolic compounds pre- 85 Å, Phenomenex, USA; and Oasis HLB C18, particle size: 30 µm, pore
sent in the sample can be achieved using a two-stage process using size: 80 Å, Waters, USA) and connected to the system. Once the ex-
different solvents (water and methanol or ethanol) exploring the dif- traction cell and the column were connected to the system, the ex-
ferential migration of compounds through the adsorbent. The separa- traction procedure started. The extraction procedure was performed in
tion strategy consists of the extraction of highly polar compounds, such several steps:
as gallic acid and chlorogenic acid, in the first stage using water as the
solvent (Reis et al., 2012; Souza et al., 2020). As the solvent passes I. Activation and condition: Valves V1–V4 were set to position 1 for
through the sample and the adsorbent, polar compounds would be the activation and conditioning of the adsorbent with 30 mL of the
extracted, and due to the low interaction with the adsorbent, they selected solvent (methanol or ethanol for activation and water for
would quickly exit the column packed with the adsorbent. The less conditioning) at a flow rate of 3 mL.min−1.
polar compounds would interact more strongly with the adsorbent and II. 1st extraction stage: The system was reconfigured by setting the
take longer to exit the adsorbent column. Moderately polar compounds, automatic valves V1 and V4 to position 1 and the automatic valves
such as the flavonoids extracted in this stage, would be retained in the V2 and V3 in position 2 to initiate the first extraction stage, where
adsorbent. Afterward, a small volume of an organic solvent will be water was used as the solvent. The system was heated to the ex-
passed through the sample, and the adsorbent to extract the remaining perimental temperature (60–80 °C), and the extraction cell was
compounds from the sample matrix and simultaneously elute the re- filled with the solvent until the process pressure reached 100 bar.
tained compounds from the adsorbent while increasing the concentra- Once the experimental conditions were reached, the sample was
tion of the extract in a small volume (Qiu, Guo, & Chang, 2007; Soto, extracted in dynamic mode, maintaining a flow rate of 2 mL.min−1
Moure, Domínguez, & Parajó, 2011; Souza et al., 2020). and constant pressure (100 bar). Seven aqueous fractions were
In this context, the objective of this work was to evaluate the fea- collected during the dynamic extraction: Fraction 1: 0–15 mL;
sibility of coupling PLE and SPE on-line for the simultaneous extraction Fraction 2: 15–30 mL; Fraction 3: 30–60 mL; Fraction 4: 60–75 mL;
and fractionation of specific flavonoids from an industrial residue ob- Fraction 5: 75–90 mL; Fraction 6: 90–105 mL; Fraction 7:
tained in the production of apple juice focusing on the use of green 105–120 mL.
solvents. III. 2nd extraction stage: The second extraction stage was performed
with a solvent of lower polarity than that used in the first extraction
stage, namely methanol or ethanol. After the first stage, the second
extraction stage was initiated by setting the change of the solvent

2
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

from water to 100% methanol or 100% ethanol, maintaining the Torrance, CA, USA). The peaks were recorded and integrated at
same flow (2 mL.min−1) and temperature (60–80 °C). Two fractions 260 nm. The column temperature was maintained at 40 °C. The mobile
of 20 mL of the solvent were collected. phase was composed of water (0.1% phosphoric acid) (solvent A) and
acetonitrile (0.1% phosphoric acid) (solvent B). The gradient profile
All extractions were carried out in duplicate. Fractions were stored with a flow-rate of 1.1 mL.min−1 was as follows: 0 min, 5% B; 0.5 min,
on amber glass vials and filtered through a syringe filter (nylon, 25 mm, 10% B; 2.0 min, 12.5% B; 3 min, 15% B; 4 min, 80% B; 5 min; 100% B;
0.22 µm, Analitica, São Paulo, Brazil) before being analyzed. Overall 6 min, 100% B; 7 min, 5% B. The equilibration time between runs was
extraction time was approximately 70 min (60 min for the 1st stage and 5 min and the injection volume was 5 μL. Identification of each com-
10 min for the 2nd stage). For the development of the method, a step- pound was based on the profile identified by the UHPLC-MS/MS ana-
by-step strategy was used. Initially, different adsorbents were evaluated lyses and by comparing the retention times and UV spectra of the se-
using fixed conditions, and once the most efficient adsorbent was se- parated compounds as well as by co-elution with authentic standards.
lected, different temperatures were evaluated. Finally, the replacement The standard curve (7 points – 0.1–100.0 mg L−1) of each compound
of the methanol used in the processes by ethanol was evaluated. The was prepared by plotting the concentration versus peak area.
results produced with the best conditions were compared to those ob- Quantification of these compounds was performed based on the cali-
tained with other techniques. bration curves of related compounds and expressed as equivalents (M.
A. Rostagno et al., 2011; Sumere et al., 2018). Phloretin xylosil glu-
2.4. Comparison with other extraction methods coside was quantified as phlorizin equivalents, while quercetin deri-
vatives 1–4 and quercetin rhamnoside were quantified as hyperoside
The extractions with ultrasound (US) were performed on an ultra- equivalents. The compound tentatively identified as protocatechuic
sonic bath (P60H, Elmasonic, Singen, Germany) operating at 37 kHz acid glucoside (PcAG) was quantified as gallic acid equivalents. All
and 100% of US power. Magnetic stirring extractions (Mag-Stir) were analyses were carried out in duplicate.
carried out on an IKA C-MAG stirrer (Werke GmbH & Co.KG/Germany)
set to speed 3. Shaking was performed on a MaxQ 4000 (Thermo Fisher 2.7. Statistical analysis
Scientific, Germering, Germany). Extractions with these techniques
were performed using 1.0 g of sample and 25 mL of solvent (water, All variables have undergone a compliance test to check the ap-
50–100% ethanol, and 50%–100% methanol). Extractions were carried proach with theoretical curves. The Shapiro-Wilk’s test and Levene test
out for 30 min at 60 °C. Pressurized Liquid Extraction (PLE) was also were used to verify the homoscedasticity. The one-way ANOVA with
carried out in the Extract-US system without the adsorbent column. the Bonferroni post hoc test was used to evaluate the results. In all tests,
Extractions were performed in static mode at a constant pressure it was considered the significance level of p < 0.05. The analyses were
(100 bar) using the same conditions of the conventional techniques performed on the software Statistical Package for Social Sciences (SPSS)
(60 °C, 25 mL of solvent, 30 min). All samples were centrifuged, filtered version 15.0.1.
through a syringe filter (nylon, 25 mm, 0.22 µm, Analitica, São Paulo,
Brazil), and stored at −20 °C before analysis. All extractions were 3. Results and discussion
carried out in duplicate.
3.1. Phenolic composition of the apple pomace
2.5. Ultra-high pressure liquid chromatography coupled to mass
spectrometry (UHPLC-MS/MS) The main compounds identified in the apple pomace were: Gallic
acid ([M−H]− 169.05, λmax: 271.0 nm; rt: 2.47 min; Peak #1), which
A UHPLC-MS XEVO G2-S QToF system (Waters, USA) was used to coeluted with a compound tentatively identified as protocatechuic acid
identify the compounds present in the sample. The system was glucoside (PcAG: [M−H]− 315.07, λmax: 287 nm; rt: 2.47 min; Peak
equipped with a diode detector coupled to a quadrupole mass spec- #1); chlorogenic acid ([M−H]− 353.1, λmax: 254.0, 326.1 nm; rt:
trometer with a negative and positive mode ESI ionization source. The 7.14 min; Peak #2); epicatechin ([M−H]− 279,1; λmax: 278.1 nm; rt:
separation was carried out on an Acquity C18 UPLC BECH column 10.54 min; Peak #3); rutin ([M−H]− 609.2, λmax: 255.1, 354.1 nm; rt:
(2.1 mm × 100 mm, 1.7 μm; Waters). Two solvents were used: solvent 13.88 min; Peak #4); hyperoside ([M−H]− 463.1; λmax: 354.1,
A (water with 2% formic acid) and solvent B (acetonitrile with 2% 354.1 nm; rt: 14.06 min; Peak #5); quercetin rhamnoside ([M−H]−
formic acid). The following gradient was used: 0 min, (100% A); 1 min, 447.1, λmax: 255.1, 349.1 nm; rt: 15.66 min; Peak #10); Phloretin
(100% UMA); 2 min, (95% A); 3 min, (90% A); 4 min, (85% A); 5 min, Xylosil Glucoside ([M−H]− 567.2, λmax: 285.1 nm; rt: 15.88 min; Peak
(85% A); 6 min, (80% A), 7 min, (80% A); 8 min (75% A); 9 min, (75% #11); and Phlorizin ([M−H]− 435.1, λmax: 284.1 nm; rt: 16.68 min;
A); 10 min, (70% A); 11 min, (0% A). The column was heated to 47 °C, Peak #12). Additionally, four quercetin derivatives ([M−H]− 433.1,
the solvent flow was 0.6 mL.min−1, and the injection volume was one λmax: 254.1, 353.1 nm; rt: 13.8–14.9 min; Peaks #6–9) were also de-
μL. UV absorbance was monitored between 210 and 400 nm. Peaks tected in the sample but their exact identification was not possible due
were integrated at 260 nm. The ESI method was adjusted with 3 kV to the structural similarities between them. The compounds with the
capillary voltage; source temperature at 120 °C; desolvation line tem- most evident peaks in the extracts were gallic acid/PcAG, phlorizin and
perature at 400 °C; cone gas flow at 10 L.h−1; desolvation line flow at hyperoside (Fig. 1). Phenolic acids were the compounds found in higher
850 L.h−1; between 100 and 1200 [M−H]−; 0.2 s probing time; and concentration followed by flavonoids.
25-volt cone voltage. The identification of the compounds was based on The identified compounds and their concentration in the apple po-
the masses [M−H]−, retention time and coelution with the standards mace used were in agreement with previous reports about the com-
available. pounds present in apple, apple pomace, and apple juice (Baskaran,
Pullencheri, & Somasundaram, 2016; Mari et al., 2010). The differences
2.6. Analysis by high-performance liquid chromatography in terms of composition and concentration of the sample used when
compared to previous works can be explained by the variation related
The analysis of the compounds present in the samples was per- to cultivar, growing conditions, fruit maturity, crop management,
formed by high-performance liquid chromatography (HPLC) in the drying conditions of the material, degradation, or by the method for
Extract-US system set to “analysis” mode (V1, V2, and V4 in position 1 identification and quantification of the compounds (Everette et al.,
and V3 in position 2). The separation of the compounds was performed 2010; Kalinowska, Bielawska, Lewandowska-Siwkiewicz, Priebe, &
on a Kinetex C18 column (2.6 μm, 100 A, 100 × 4.6 mm, Phenomenex, Lewandowski, 2014; Treutter, 2001).

3
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

(caption on next page)

4
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

Fig. 1. Chromatograms obtained with developed method. A: Reference extract where all compounds are presented (brute extract); B: Fraction 1 (0–15 mL), aqueous
fraction rich in gallic acid; C: Methanolic fraction (0–20 mL methanol). Identification of compounds: Peak #1: Gallic acid/Protocatechuic acid glucoside (PcAG); Peak
#2: Chlorogenic acid; Peak #3: Epicatechin; Peak #4: Rutin; Peak# 5: Hyperoside; Peak #6: Quercetin derivative 1; Peak #7: Quercetin derivative 2; Peak #8:
Quercetin derivative 3; Peak #9: Quercetin derivative 4; Peak #10: Quercetin rhamnoside; Peak #11: Phloretin xylosyl glucoside; 12: Phlorizin.

3.2. Comparison of different adsorbents suggesting an excellent potential of the extraction/elution strategy used
with the combination of PLE and an adsorbent.
We initially sought to compare several adsorbents with different Excellent results were also observed in the recovery of most of the
characteristics with previously tested conditions, namely the adsorbents flavonoids present, but it was dependent on the target compound and
Sepra, Isolute, Strata X, and Oasis, to evaluate the feasibility of our adsorbent. Significant differences were observed in terms of total fla-
strategy. Initial conditions were determined based on a series of tests vonoids analysis. Interestingly, conventional adsorbents were, in gen-
using different amounts of water to elute compounds other than fla- eral, also more efficient than polymeric for the recovery of flavonoids
vonoids from the sample. Based on these tests, sample collection con- (p < 0.001).
sisted in the removal of 8 fractions: the first seven fractions using water Epicatechin (Fig. 2.B1), for instance, was only detected in the ex-
as solvent (15 mL, 30 mL, 60 mL, 75 mL, 90 mL, 105 mL, and 120 mL), tracts of the organic fraction of the conventional adsorbents with a
and the last fraction of 20 mL of methanol. Initially, methanol was used significantly higher recovery produced by the adsorbent Sepra than by
as activation and elution solvent since previous reports use this solvent, the adsorbent Isolute (p < 0.01). Epicatechin was not detected in any
and it is the standard solvent used for this task in analytical purposes of the aqueous or organic fractions produced with Oasis adsorbents,
(Mauricio A. Rostagno, Palma, & Barroso, 2005; Simpson, 1991). The which may also be due to the gradual elution of these compounds in the
temperature was maintained at 60 °C. The recovery of compounds from aqueous fraction in low concentration (below detection levels) leading
apple pomace obtained with the different adsorbents is shown in Fig. 2. to its loss in a similar way to chlorogenic acid. However, in this case,
The recovery of total phenolic acid was not statistically significant, epicatechin was not adequately retained by the adsorbent instead of
and conventional adsorbents (Sepra and Isolute) revealed a similar re- being retained more strongly, as was the case of chlorogenic acid.
tention capacity to those of polymeric adsorbents (Strata X and Oasis) Higher recovery of hyperoside (Fig. 2.B3) was also observed with
(Fig. 2.A), which are more advanced and usually produce better results. the conventional adsorbents Sepra and Isolute than with polymeric
However, some interesting trends were observed when comparing the ones, while the adsorbent Strata X produced the lowest recovery
results in detail through the recovery of individual compounds. As can (p < 0.001). Conventional adsorbents were also more efficient for the
be seen in Fig. 2.A1 and 2.A2, depending on the acid, the tested ad- recovery of phloretin xylosyl glucoside (Fig. 2.B9) than polymeric
sorbents showed very different retention behavior. Strata X (p < 0.001). Interestingly, for this compound, the polymeric
For gallic acid/PcAG, a significantly higher recovery was obtained adsorbent Oasis was more efficient than Strata X (p = 0.009). Similar
with the polymeric adsorbents (Stata X and Oasis) in comparison with results were also obtained for phlorizin (Fig. 2.B10), with higher re-
the conventional adsorbents (p < 0.031). However, significant dif- covery produced by conventional adsorbents (p < 0.001), but there
ferences were also observed in the retention of chlorogenic acid. was not a significant difference between both polymeric adsorbents.
Chlorogenic acid was not retained by the adsorbent Oasis, and only Furthermore, higher recovery of quercetin derivatives #3 (Fig. 2.B6)
small amounts were recovered using the adsorbent Strata X, which was was observed when using the adsorbent Isolute while the lowest were
significantly lower than those produced with conventional adsorbents produced with the adsorbent Sepra and Strata X (p < 0.001).
(p < 0.001). The highest recovery was obtained with the adsorbent However, in some cases, there were no significant differences be-
Sepra, followed by the adsorbent Isolute. tween adsorbents, such as for quercetin derivatives #1 and #2 and
The lower recoveries of chlorogenic acid observed with the poly- quercetin rhamnoside (Fig. 2.B4, 2.B5, and 2.B8, respectively). In other
meric adsorbents and gallic acid with the conventional adsorbents were cases, the conventional adsorbents were not able to retain compounds,
probably associated with the gradual elution of these compounds in the such as rutin (Fig. 2.B2), which was only detected in the organic frac-
aqueous fractions collected during the first extraction stage. Since the tion of the polymeric adsorbents, although much higher recovery was
same extraction conditions were used, both acids were extracted from achieved with the adsorbent Strata X than with the adsorbent Oasis.
the raw material matrix and reached the adsorbent with the same Poor retention of these compounds and its breakthrough in the aqueous
concentration, and therefore differences are associated with the reten- fractions coupled with low concentration may also explain the lack of
tion capacity of the adsorbent. The higher recovery of chlorogenic acid recovery with conventional adsorbents. Finally, quercetin derivative #4
with the conventional adsorbent is probably due to a lower retention (Fig. 2.B7) showed a different behavior and was not detected in the
capacity by this type of adsorbent, prompting the fast elution and its organic fraction of the adsorbents Isolute and Oasis, while the highest
collection in higher concentration in the first aqueous fractions (al- recovery was observed with the adsorbent Sepra (p < 0.001)
lowing its detection and quantitation). Therefore, although all adsorbents share the same basic structure
In contrast, due to it is higher retention capacity, polymeric ad- (C18), due to differences in their chemical characteristics, they will have
sorbents interact more with chlorogenic acid and gradually elute it different retention capacities for similar compounds. These differences
distributing throughout the aqueous fractions. Due to the high volume in retention capacity are more acute when considering compounds with
of water used, the concentration of chlorogenic acid was too low to be very different chemical structures. Overall, the best results were pro-
detected and quantified in these fractions and therefore were not ac- duced with the adsorbent Sepra, which produced the highest recoveries
counted for, resulting in a lower recovery. The different behavior ob- for most flavonoids and phenolic acids. Thus, this adsorbent was se-
served depending on the acid is probably related to their structural lected to be used in the further development of the method.
characteristics. While gallic acid is a trihydroxy benzoic acid, chloro-
genic acid is derived from cinnamic acid, resulting in different inter-
actions with the adsorbent and the solvent, and therefore affecting their 3.3. Extraction temperature
recovery. The lack of statistical difference in terms of total phenolic
acid between all adsorbents recovery is due to the higher recovery of Temperature is one of the main factors influencing the extraction
gallic acid by polymeric adsorbents compensating the lower recovery of processes and is particularly relevant for PLE and the behavior of ad-
chlorogenic acid. Nevertheless, high amounts of gallic acid were re- sorbents. The importance of the temperature is due to the enhancement
covered from the raw material independently of the adsorbent, of the mass transfer of analytes from the raw material matrix to the
solvent improving extraction yield. It also can affect the retention of

5
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

Fig. 2. Recovery (mg/g) of phenolics acids (A), and flavanoids (B) with different adsorbents (Sepra, Isolute, Strata X and Oasis).

6
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

Table 1
Recovery (mg/g) of phenolics compounds from apple pomace using different extraction temperatures. Gallic/PcAG: Gallic acid and Protocatechuic acid glucoside;
Que-Drv: Quercetin-derivative; Que-Rhn: Quercetin rhamnoside; Pht-Xyl-Glc: Phloretin xylosyl glucoside.
Class Peak # Compound Recovery (mg/g)

60 °C 70 °C 80 °C

Acids 1 Gallic/PcAG 1.867a ± 0.096 1.925a ± 0.068 1.755a ± 0.087

2 Chlorogenic 0.718a ± 0.176 0.153b ± 0.103 0.000b ± 0.000
– Total Acids 2.727a ± 0.352 2.221b ± 0.179 1.755b ± 0.087
Flavonoids 3 Epicatechin 0.115a ± 0.004 0.114a ± 0.002 0.077a ± 0.051
4 Rutin 0.000a ± 0.000 0.000a ± 0.000 0.008a ± 0.017
5 Hyperoside 0.113a ± 0.005 0.085a ± 0.030 0.091a ± 0.010
6 Que-Drv 1 0.555a ± 0.001 0.052a ± 0.000 0.052a ± 0.005
7 Que-Drv 2 0.048a ± 0.000 0.047a ± 0.000 0.045a ± 0.002
8 Que-Drv 3 0.054a ± 0.002 0.056a ± 0.000 0.049b ± 0.002
9 Que-Drv 4 0.067a ± 0.002 0.065a ± 0.001 0.061a ± 0.005
10 Que-Rhn 0.038a ± 0.001 0.028a ± 0.019 0.036a ± 0.009
11 Pht-Xyl-Glc 0.168a ± 0.003 0.154a ± 0.018 0.168a ± 0.015
12 Phlorizin 0.228b ± 0.017 0.295a ± 0.050 0.285a ± 0.064
– Total Flavonoids 0.895a ± 0.033 0.920a ± 0.055 0.880a ± 0.081

Bold values indicate significant differences (one-way ANOVA) p < 0.05; Different letters in the same line indicate significant differences (Bonferroni’s test)
p < 0.05.

compounds by the adsorbent and solvent properties, including viscosity enhanced solubility of gallic acid/PcAG in the solvent could also con-
and dielectric constant. It is also essential to consider that several tribute to the observed increase in the recovery. The net result from
phenolic compounds can be subjected to degradation if an excessive these effects can explain the lack of differences in the results observed
temperature is used (Manchón et al., 2010; Rostagno and Prado ,2013; at different temperatures for gallic acid/PcAG.
Rostagno et al., 2010). In terms of total flavonoid recovery, there was no significant dif-
In this context, the temperature of the process was increased from ference between the tested temperatures. Although most flavonoids
60 °C to 70 °C and 80 °C using the adsorbent Sepra, as discussed in the were not affected by the temperature, the recovery of some compounds
previous section. The results shown in Table 1 suggest that compounds was greatly influenced by this variable. The recovery of epicatechin,
were differently affected by the temperature. While the temperature did rutin, hyperoside, quercetin derivatives #1, #2, and #4, quercetin
not influence the recovery of most compounds, in some cases, it re- rhamnoside, and phlorizin-xylosil glucoside was not significantly af-
duced the efficiency of the adsorbent to retain compounds. fected by the increase in the extraction temperature. Although the re-
Some phenolic acids were particularly affected by the increase in covery of rutin was not statistically different in all temperatures, this
temperature, and the total recovery was significantly reduced if the compound was only detected in the organic solvent fraction of the
temperature used in the process was above 70 °C (p = 0.04). When process carried out at 80 °C. Differences were not statistically sig-
analyzing individual data, the main factor determining total acid yield nificant due to the low concentration of this compound and to the
was the recovery of chlorogenic acid, which was drastically reduced at observed variability in the retention of the compound by the adsorbent
80 °C (p < 0.001). These results can be explained by the degradation at 80 °C.
of chlorogenic acid caused by the higher temperature, among other Among the most relevant results, it is essential to highlight that the
factors. In general, the stability of the acids and other phenolics is af- recovery of the main flavonoid found in apples, Phlorizin, significantly
fected by temperature, light and extraction time, because degradation increased with the increase of the temperature from 60 °C to 70 °C
reactions such as hydrolysis, reduction or polymerization occur to the (p = 0.007). However, a further increase in the extraction temperature
detriment of these factors, as well as the release of oxidative enzymes to 80 °C did not have a significant effect. Recovery is determined in
present in the samples, altering the final amounts of the compounds in great part by the retention of the compounds by the adsorbent, which is
the extraction fractions (Liazid, Palma, Brigui, & Barroso, 2007). influenced by the temperature (Garcia-Salas, Morales-Soto, Segura-
Degradation is a likely factor influencing recovery since it has been Carretero, & Fernández-Gutiérrez, 2010; Gómez Caravaca, Carrasco
reported that chlorogenic acid may be susceptible to losses during Pancorbo, Cañabate Díaz, Segura Carretero, & Fernández Gutiérrez,
processing at high temperatures (Milić, Stojanović, Vučurević, & Turčić, 2005; Palma, Piñeiro, & Barroso, 2002). Therefore, higher temperatures
1968). Light is not a factor to be considered since PLE is carried-out on (i.e., 80 °C) could reduce the retention capacity to the extent of a
a closed system not exposed to UV radiation. Since it is a dry material, it breakthrough, leading to the loss of the compound in the aqueous phase
is also not likely that enzymatic activity is taking place at such high instead of the organic phase. Due to the high volume of the aqueous
levels to explain the observed drastic reduction in the recovery of this phase and to the gradual elution from the adsorbent, the concentration
compound in such a short time. Therefore, it is possibly a combination in the extract was too low to allow detection, and therefore, these are
of factors taking place, including hydrolysis, reduction or polymeriza- considered losses explaining the lack of difference between 70 and
tion reactions, and enzymatic activity are involved in promoting the 80 °C.
degradation of chlorogenic acid during the extraction, and thus redu- Another explanation, besides thermal degradation, is that quanti-
cing it is recovery. tative recovery was already achieved at 70 °C, and there was not any
The increase in the temperature did not affect the recovery of gallic Phlorizin available to be extracted by the enhanced extraction condi-
acid/PcAG, possibly due to it is high solubility in water, and the high tions promoted by the higher temperature. The quantitative recovery
volume used, prompting an effective extraction not leaving room for would also explain the lack of effect by the increase in the temperature
improvement by the increase in the temperature. However, those two on the recovery of all other flavonoids since Phlorizin is the flavonoid
effects could be simultaneously taking place caused by the increase in found in the highest concentration in the sample. The lack of significant
the temperature. As temperature increases, degradation of gallic acid/ difference in the total flavonoids recovery between all temperatures is
PcAG could take place (Réblová, 2012) while the same time enhanced due to smaller accumulative non-statistically significant differences of
the mass transfer from the raw material matrix to the solvent. The other flavonoids, such as Hyperoside and Quercetin-derivative 1,

7
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

Table 2 compensating the significant difference in the recovery of Phlorizin in

Recovery (mg/g) of phenolics compounds from apple pomace using different the overall total. It is also possible that a combined effect of the thermal
activation and extraction/elution solvents. Gallic/PcAG: Gallic acid and degradation of flavonoids and the lower retention and breakthrough in
Protocatechuic acid glucoside; Que-Drv: Quercetin-derivative; Que-Rhn: the aqueous phase is compensating for the enhanced extraction of these
Quercetin rhamnoside; Pht-Xyl-Glc: Phloretin xylosyl glucoside. M = M:
compounds from the sample matrix, explaining the lack of significant
Adsorbent activated with methanol and extraction/elution with methanol;
effect of the temperature in recovery.
M = E: Adsorbent activated with methanol and extraction/elution with
ethanol; E-E: Adsorbent activated with ethanol and extraction/elution with
ethanol;
3.4. Activation and extraction/elution solvent
Compounds Recovery (m/g)
Due to the toxic nature, higher cost, and environmental impact of
Activation and Extraction/Elution Solvent
methanol, ethanol was evaluated as an alternative to be used activation
M-M M-E E-E and extraction/elution solvent. Thus, ethanol was used to replace me-
thanol, first as the activation solvent, while maintaining methanol as
a a
Gallic/PcAG 1.867 ± 0.968 2.024 ± 0.134 1.915a ± 0.076 extraction/elution solvent (M−E). Ethanol was also used as both acti-
Chlorogenic acid 0.718a ± 0.176 0.824a ± 0.053 0.862a ± 0.064
Total Acids 2.727a ± 0.352 2.851a ± 0.187 2.786a ± 0.223
vation and extraction/elution solvent (E-E), and results were compared
Epicatechin 0.115a ± 0.004 0.116a ± 0.011 0.104a ± 0.006 with those produced using methanol in both stages (M−M). The ad-
Rutin 0.000b ± 0.000 0.000b ± 0.000 0.035a ± 0.002 sorbent Sepra was used, and the extraction temperature was 60 °C. The
Hyperoside 0.113a ± 0.005 0.135a ± 0.024 0.135a ± 0.021 recovery of phenolic acids and flavonoids obtained with the different
Que-Drv #1 0.055a ± 0.001 0.062a ± 0.006 0.065a ± 0.008
solvents is shown in Table 2.
Que-Drv #2 0.048a ± 0.000 0.053a ± 0.004 0.051a ± 0.001
Que-Drv #3 0.054b ± 0.002 0.104a ± 0.011 0.097a ± 0.006 The results reveal that the activation and extraction/elution solvent
Que-Drv #4 0.067a ± 0.002 0.000b ± 0.000 0.000b ± 0.000 can significantly influence the recovery of target compounds. As ex-
Que-Rhn 0.038a ± 0.001 0.118a ± 0.161 0.035a ± 0.000 pected, the recovery of phenolic acids was not affected by replacing
Pht-Xyl-Glc 0.168a ± 0.004 0.177a ± 0.022 0.169a ± 0.012 methanol by ethanol in neither of both stages since these compounds
Phlorizin 0.228a ± 0.018 0.278a ± 0.098 0.309a ± 0.050
Total Flavonoids 0.895a ± 0.033 0.970a ± 0.111 1.012a ± 0.101
are recovered in the first fractions with water and are poorly retained
by the adsorbent and, therefore the activation and extraction/elution
Bold values indicate significant differences (one-way ANOVA) p < 0.005; solvent should not affect their recovery.
Different letters in the same line indicate significant differences (Bonferroni’s On the other hand, surprisingly good results were obtained when
test) p < 0.05. replacing methanol as extraction/elution solvent by ethanol since dif-
ferences in the recovery of phenolic acids and most flavonoids were not
statistically significant, suggesting that it is possible to reduce the use of
toxic solvents in the process. However, replacing methanol by ethanol

Fig. 3. Comparison of the recovery (mg/g) of phenolics acids and flavanoids-produced by the developed method (PLE-SPE) and those obtained with different
conventional techniques (PLE: Pressurized Liquid Extraction; US: Ultrasound-assisted Extraction; Mag-Stir: Magnetic Stirring) and solvents (water, ethanol, methanol
and mixtures with water). Different letters in the same solvent indicate significant differences (Bonferroni’s test) p < 0.05 when compared with the developed
method (PLE-SPE).

8
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

as activation solvent negatively affected the recovery of some flavo- waves. Due to these variations, the differences between the US and PLE-
noids, such quercetin derivative #4 (p < 0.001), while at the same SPE were not statistically significant. In general, US provided the best
time improved the recovery rutin (p < 0.001). Since the activation of results among the conventional techniques, but the performance of all
adsorbent is fundamental to the ability of the adsorbent to retain techniques was dependent on the solvent used.
compounds, activation solvent plays an essential role in the recovery of In this context, the data support our hypothesis and indicate an
moderately polar compounds such as flavonoids, which depend on their excellent performance of the developed method when compared to
retention to be recovered. If the adsorbent does not correctly retain other techniques. Besides of high recoveries with green solvents, it was
them, they may break through and be lost in the aqueous fractions of possible to separate compounds in different fractions and obtain very
the first extraction step. These results indicate that ethanol is a suitable clean chromatograms while eliminating the need of post-extraction
solvent for the extraction/elution of phenolic compounds from apple processes, such as centrifugation and concentration, and protecting the
pomace and can be used to replace methanol. Furthermore, the data sample from light and oxygen, preventing degradation (Azmir et al.,
indicate that ethanol can be used for the activation of the adsorbent but 2013; Rostagno et al., 2010; Wang & Weller, 2006). Furthermore, it is
it is important to point out that it may negatively affect the retention of relevant to point out that it is possible to combine ultrasound and an
some compounds. expansion gas with PLE, which could speed up the process and further
Overall, the results obtained with the developed method are ex- improve recovery and maximize performance as suggested by previous
cellent, considering that it uses green solvents (and a small amount of studies (Santos et al., 2019; Sumere et al., 2018).
methanol, which can be recycled) while achieving high extraction
yields and good separation between compounds in different fractions. 4. Conclusions
As can be seen in Fig. 1.B and 1.C, there is only marginal contamination
between well-defined fractions and clean chromatograms are produced, The data provided in all experiments support the benefits of the on-
revealing an enormous potential to be explored in similar applications. line PLE-SPE method for the simultaneous extraction and fractionation
of phenolic compounds from natural products. However, depending on
3.5. Comparison with conventional techniques the target compounds, it will be necessary to carefully select extraction/
elution conditions in order to produce high recoveries. Furthermore,
To evaluate the performance of the developed method, the results the use of a gradient from water to the organic solvent may also allow
were compared with those produced with different solvents (water, separation between different flavonoids, and it is a promising area of
50–100% methanol and 50–100% ethanol) and techniques (Ignat, Volf, research. Conventional adsorbents revealed surprisingly good results
& Popa, 2011; Rostagno, Palma, & Barroso, 2004; Rostagno, Palma, & when compared to polymeric adsorbents with consistent higher per-
Barroso, 2003). Water, ethanol, and methanol 50%, 80%, and 100% formance in terms of the recovery of target compounds. Most com-
were selected to provide an intermediary solvent strength in order to pounds were adequately retained by conventional adsorbents allowing
allow the simultaneous extraction of all compounds despite their dif- their separation and recovery in separate fractions, but the higher
ferences in polarity and solubility in the solvents. performance was observed with the adsorbent Sepra. The temperature
As can be seen in Fig. 3, PLE-SPE provided higher or similar re- was not a significant parameter affecting the process, possibly due to
covery of both phenolic acids and flavonoids than most techniques, the dynamic extraction process using different solvents with varying
independently of the solvent used. Recovery of phenolic acids by the polarity. The results also suggest that ethanol can be used as an alter-
developed method (which used water and ethanol 100%) was 3,69 to native to methanol for the extraction/elution. The developed method
1,45 times higher (p < 0.001) than those produced by PLE, US, allowed the separation of phenolics compounds in different fractions
Magnetic Stirring (Mag-Stir) and Shaker. Without a doubt, the larger while at the same time providing higher recovery than most conven-
volume of water used in the first stage of PLE-SPE was a key factor tional methods. The excellent recoveries suggest that the proposal of
influencing the results since it can extract several phenolics from apple combining PLE with Solid-Phase Extraction using green solvents in a
pomace (Fernandes et al., 2019; Reis et al., 2012). For the extraction of two-step extraction/fractionation process can be explored for analytical
phenolic acids, the worst results were observed with Shaker and Mag- purposes, for the initial preparative production of analytical standards
Stir using 100% ethanol as solvent. Water was relatively efficient for or even for the production of extracts to be used by the food and
the extraction of phenolic acids independently of the technique used pharmaceutical industry.
and provided similar results to those obtained with 50–80% EtOH and
50% MeOH. However, the highest recovery of phenolic acids for all CRediT authorship contribution statement
conventional techniques was achieved using 80% MeOH.
Methanol and water mixtures are regarded as excellent solvents for Laise C. da Silva: Conceptualization, Data curation, Formal ana-
the extraction of phenolics, which combined with an efficient extraction lysis, Software, Supervision, Validation, Visualization, Writing - ori-
technique such as ultrasound, provide high recoveries and illustrates ginal draft, Writing - review & editing. Mariana C. Souza: Formal
the difficulty of replacing it by another cheaper and environmentally analysis, Methodology. Beatriz R. Sumere: Formal analysis,
compatible solvent. The combination of the technique and solvent is Methodology. Luiz G.S. Silva: Formal analysis, Methodology. Diogo T.
particularly relevant in the case of flavonoids. As can be seen in Fig. 3, da Cunha: Formal analysis, Methodology, Writing - original draft,
water was not capable of extracting any flavonoids from the sample Writing - review & editing. Gerardo F. Barbero: Conceptualization,
using conventional techniques. There are some reports in the literature Data curation, Formal analysis, Investigation, Methodology, Writing -
suggesting that water is not efficient for the extraction of flavonoids, original draft, Writing - review & editing. Rosangela M.N. Bezerra:
especially quercetin glucosides, which is in agreement with our results Conceptualization, Methodology, Supervision, Validation, Writing -
(Reis et al., 2012). PLE-SPE provided higher recovery of flavonoids than review & editing. Mauricio A. Rostagno: Conceptualization, Data
most techniques using 80%, 100% EtOH, and 100% MeOH curation, Formal analysis, Funding acquisition, Investigation,
(p < 0.001), but differences between all techniques were not statisti- Methodology, Project administration, Resources, Software, Supervision,
cally significant when using 50% EtOH. It is essential to highlight that Validation, Visualization, Writing - original draft, Writing - review &
combining US and 80% MeOH allowed achieving a higher recovery of editing.
flavonoids than the developed method. It is also important to indicate
that it also provided the highest variation of the results among all Declaration of Competing Interest
techniques, which could be related to the presence of lower energy
points in the ultrasonic bath and uneven propagation of ultrasonic The authors declare that they have no known competing financial

9
L.C. da Silva, et al. Food Chemistry 318 (2020) 126450

interests or personal relationships that could have appeared to influ- Maleki, S. J., Crespo, J. F., & Cabanillas, B. (2019). Anti-inflammatory effects of flavo-
ence the work reported in this paper. noids. Food Chemistry. https://doi.org/10.1016/j.foodchem.2019.125124.
Manchón, N., D’Arrigo, M., García-Lafuente, A., Guillamón, E., Villares, A., Ramos, A., ...
Rostagno, M. A. (2010). Fast analysis of isoflavones by high-performance liquid
Acknowledgments chromatography using a column packed with fused-core particles. Talanta. https://
doi.org/10.1016/j.talanta.2010.08.050.
Mari, A., Tedesco, I., Nappo, A., Russo, G. L., Malorni, A., & Carbone, V. (2010). Phenolic
Support from São Paulo Research Foundation (FAPESP) is ac- compound characterisation and antiproliferative activity of “Annurca” apple, a
knowledged (projects 2013/043044 and 2018/14582-5). BRS is southern Italian cultivar. Food Chemistry. https://doi.org/10.1016/j.foodchem.2010.
grateful for the fellowship from the National Council for Scientific and 04.023.
Milić, B., Stojanović, S., Vučurević, N., & Turčić, M. (1968). Chlorogenic and quinic acids
Technological Development (CNPq). LCS, MCS and MPS are grateful to in sunflower meal. Journal of the Science of Food and Agriculture. https://doi.org/10.
São Paulo Research Foundation (FAPESP 2019/24537-0, 2016/19930-6 1002/jsfa.2740190211.
and 2018/17089-8) for their fellowships. MAR is thankful to CNPq for Palma, M., Piñeiro, Z., & Barroso, C. G. (2002). In-line pressurized-fluid extraction-solid-
phase extraction for determining phenolic compounds in grapes. Journal of
the productivity grant funding (303568/2016-0). This study was fi-
Chromatography A. https://doi.org/10.1016/S0021-9673(02)00823-3.
nanced in part by the Coordenação de Aperfeiçoamento de Pessoal de Plaza, M., Abrahamsson, V., & Turner, C. (2013). Extraction and neoformation of anti-
Nível Superior – Brasil (CAPES) Finance Code 001 and Process oxidant compounds by pressurized hot water extraction from apple byproducts.
88887.310558/2018-00. The authors also would like to thank MACÇÃ Journal of Agricultural and Food Chemistry. https://doi.org/10.1021/jf400584f.
Qiu, N., Guo, S., & Chang, Y. (2007). Study upon kinetic process of apple juice adsorption
and Sioji for donating the industrial apple pomace used in this study. de-coloration by using adsorbent resin. Journal of Food Engineering. https://doi.org/
10.1016/j.jfoodeng.2006.10.030.
Appendix A. Supplementary data Rabetafika, H. N., Bchir, B., Blecker, C., & Richel, A. (2014). Fractionation of apple by-
products as source of new ingredients: Current situation and perspectives. Trends in
Food Science and Technology. https://doi.org/10.1016/j.tifs.2014.08.004.
Supplementary data to this article can be found online at https:// Réblová, Z. (2012). Effect of temperature on the antioxidant activity of phenolic acids.
doi.org/10.1016/j.foodchem.2020.126450. Czech Journal of Food Sciences.
Reis, S. F., Rai, D. K., & Abu-Ghannam, N. (2012). Water at room temperature as a solvent
for the extraction of apple pomace phenolic compounds. Food Chemistry. https://doi.
References org/10.1016/j.foodchem.2012.06.068.
Rostagno M. A., Prado, J. M. (2013). Natural Product Extraction: Principles and
Applications.
Alonso-Salces, R. M., Korta, E., Barranco, A., Berrueta, L. A., Gallo, B., & Vicente, F.
Rostagno, M. A., D’Arrigo, M., Martínez, J. A., & Martínez, J. A. (2010). Combinatory and
(2001). Pressurized liquid extraction for the determination of polyphenols in apple.
hyphenated sample preparation for the determination of bioactive compounds in
Journal of Chromatography A. https://doi.org/10.1016/S0021-9673(01)01212-2.
foods. TrAC – Trends in Analytical Chemistry. https://doi.org/10.1016/j.trac.2010.02.
Azmir, J., Zaidul, I. S. M., Rahman, M. M., Sharif, K. M., Mohamed, A., Sahena, F., ...
015.
Omar, A. K. M. (2013). Techniques for extraction of bioactive compounds from plant
Rostagno, M. A., Manchón, N., D’Arrigo, M., Guillamón, E., Villares, A., García-Lafuente,
materials: A review. Journal of Food Engineering. https://doi.org/10.1016/j.jfoodeng.
A., ... Martínez, J. A. (2011). Fast and simultaneous determination of phenolic
2013.01.014.
compounds and caffeine in teas, mate, instant coffee, soft drink and energetic drink
Babbar, N., Oberoi, H. S., & Sandhu, S. K. (2015). Therapeutic and nutraceutical potential
by high-performance liquid chromatography using a fused-core column. Analytica
of bioactive compounds extracted from fruit residues. Critical Reviews in Food Science
Chimica Acta. https://doi.org/10.1016/j.aca.2010.11.031.
and Nutrition. https://doi.org/10.1080/10408398.2011.653734.
Rostagno, M. A., Palma, M., & Barroso, C. G. (2004). Pressurized liquid extraction of
Baiano, A. (2014). Recovery of biomolecules from food wastes – A review. Molecules.
isoflavones from soybeans. Analytica Chimica Acta. https://doi.org/10.1016/j.aca.
https://doi.org/10.3390/molecules190914821.
2004.05.078.
Baskaran, R., Pullencheri, D., & Somasundaram, R. (2016). Characterization of free, es-
Rostagno, M. A., Villares, A., Guillamón, E., García-Lafuente, A., & Martínez, J. A. (2009).
terified and bound phenolics in custard apple (Annona squamosa L) fruit pulp by
Sample preparation for the analysis of isoflavones from soybeans and soy foods.
UPLC-ESI-MS/MS. Food Research International. https://doi.org/10.1016/j.foodres.
Journal of Chromatography A. https://doi.org/10.1016/j.chroma.2008.11.035.
2016.02.001.
Rostagno, M., Palma, M., & Barroso, C. G. (2003). Ultrasound-assisted extraction of soy
Everette, J. D., Bryant, Q. M., Green, A. M., Abbey, Y. A., Wangila, G. W., & Walker, R. B.
isoflavones. Journal of Chromatography. A.. https://doi.org/10.1016/s0021-9673(03)
(2010). Thorough study of reactivity of various compound classes toward the folin-
01184-1.
Ciocalteu reagent. Journal of Agricultural and Food Chemistry. https://doi.org/10.
Rostagno, Mauricio, A., Palma, M., & Barroso, C. G. (2005). Solid-phase extraction of soy
1021/jf1005935.
isoflavones. Journal of Chromatography A. https://doi.org/10.1016/j.chroma.2005.
Fernandes, P. A. R., Le Bourvellec, C., Renard, C. M. G. C., Nunes, F. M., Bastos, R.,
04.045.
Coelho, E., ... Cardoso, S. M. (2019). Revisiting the chemistry of apple pomace
Santos, M. P., Souza, M. C., Sumere, B. R., da Silva, L. C., Cunha, D. T., Bezerra, R. M. N.,
polyphenols. Food Chemistry. https://doi.org/10.1016/j.foodchem.2019.05.006.
& Rostagno, M. A. (2019). Extraction of bioactive compounds from pomegranate peel
García-Lafuente, A., Guillamón, E., Villares, A., Rostagno, M. A., & Martínez, J. A. (2009).
(Punica granatum L.) with pressurized liquids assisted by ultrasound combined with
Flavonoids as anti-inflammatory agents: Implications in cancer and cardiovascular
an expansion gas. Ultrasonics Sonochemistry. https://doi.org/10.1016/j.ultsonch.
disease. Inflammation Research. https://doi.org/10.1007/s00011-009-0037-3.
2019.02.021.
Garcia-Salas, P., Morales-Soto, A., Segura-Carretero, A., & Fernández-Gutiérrez, A.
Simpson, N. (1991). Solid-phase extraction – Principles, Techniques and Applications. In
(2010). Phenolic-compound-extraction systems for fruit and vegetable samples.
Acta chemica Scandinavica (Copenhagen, Denmark: 1989). https://doi.org/10.3891/
Molecules. https://doi.org/10.3390/molecules15128813.
acta.chem.scand.45-1047.
Gómez Caravaca, A. M., Carrasco Pancorbo, A., Cañabate Díaz, B., Segura Carretero, A., &
Soto, M. L., Moure, A., Domínguez, H., & Parajó, J. C. (2011). Recovery, concentration
Fernández Gutiérrez, A. (2005). Electrophoretic identification and quantitation of
and purification of phenolic compounds by adsorption: A review. Journal of Food
compounds in the polyphenolic fraction of extra-virgin olive oil. Electrophoresis.
Engineering. https://doi.org/10.1016/j.jfoodeng.2011.02.010.
https://doi.org/10.1002/elps.200500202.
Souza, M. C., Santos, M. P., Sumere, B. R., Silva, L. C., Cunha, D. T., Martínez, J., ...
Grigoras, C. G., Destandau, E., Fougère, L., & Elfakir, C. (2013). Evaluation of apple
Rostagno, M. A. (2020). Isolation of gallic acid, caffeine and flavonols from black tea
pomace extracts as a source of bioactive compounds. Industrial Crops and Products.
by on-line coupling of pressurized liquid extraction with an adsorbent for the pro-
https://doi.org/10.1016/j.indcrop.2013.06.026.
duction of functional bakery products. LWT. https://doi.org/10.1016/j.lwt.2019.
Hyson, D. A. (2012). A Comprehensive review of apples and apple components and their
108661.
relationship to human health. Advances in Nutrition. https://doi.org/10.3945/an.111.
Suárez, B., Álvarez, Á. L., García, Y. D., del Barrio, G., Lobo, A. P., & Parra, F. (2010).
000513.
Phenolic profiles, antioxidant activity and in vitro antiviral properties of apple po-
Ignat, I., Volf, I., & Popa, V. I. (2011). A critical review of methods for characterisation of
mace. Food Chemistry. https://doi.org/10.1016/j.foodchem.2009.09.073.
polyphenolic compounds in fruits and vegetables. Food Chemistry. https://doi.org/10.
Sumere, B. R., de Souza, M. C., dos Santos, M. P., Bezerra, R. M. N., da Cunha, D. T.,
1016/j.foodchem.2010.12.026.
Martinez, J., & Rostagno, M. A. (2018). Combining pressurized liquids with ultra-
Kalinowska, M., Bielawska, A., Lewandowska-Siwkiewicz, H., Priebe, W., &
sound to improve the extraction of phenolic compounds from pomegranate peel
Lewandowski, W. (2014). Apples: Content of phenolic compounds vs. variety, part of
(Punica granatum L.). Ultrasonics Sonochemistry. https://doi.org/10.1016/j.ultsonch.
apple and cultivation model, extraction of phenolic compounds, biological proper-
2018.05.028.
ties. Plant Physiology and Biochemistry.. https://doi.org/10.1016/j.plaphy.2014.09.
Treutter, D. (2001). Biosynthesis of phenolic compounds and its regulation in apple. Plant
006.
Growth Regulation. https://doi.org/10.1023/A:1013378702940.
Lee, J., Chan, B. L. S., & Mitchell, A. E. (2017). Identification/quantification of free and
Wang, L., & Weller, C. L. (2006). Recent advances in extraction of nutraceuticals from
bound phenolic acids in peel and pulp of apples (Malus domestica) using high re-
plants. Trends in Food Science and Technology. https://doi.org/10.1016/j.tifs.2005.12.
solution mass spectrometry (HRMS). Food Chemistry. https://doi.org/10.1016/j.
004.
foodchem.2016.07.166.
Wijngaard, H., & Brunton, N. (2009). The optimization of extraction of antioxidants from
Liazid, A., Palma, M., Brigui, J., & Barroso, C. G. (2007). Investigation on phenolic
apple pomace by pressurized liquids. Journal of Agricultural and Food Chemistry.
compounds stability during microwave-assisted extraction. Journal of
https://doi.org/10.1021/jf902498y.
Chromatography A. https://doi.org/10.1016/j.chroma.2006.11.040.

10