J. of Supercritical Fluids 98 (2015) 172–182
Contents lists available at ScienceDirect
The Journal of Supercritical Fluids
journal homepage: www.elsevier.com/locate/supflu
Recovery of phenolics from apple peels using CO2 + ethanol
extraction: Kinetics and antioxidant activity of extracts
Audrey Massias a,b , Séverine Boisard c , Michel Baccaunaud a , Fernando Leal Calderon b ,
Pascale Subra-Paternault b,∗
a
Agrotec, site d’Agropole, BP102, 47931 Agen Cedex 9, France
CBMN UMR CNRS 5248, Université Bordeaux, IPB, Allée Geoffroy Saint Hilaire, Bat. 14B, 33600 Pessac, France
c
EA921 SONAS/SFR 4207 QUASAV, Université d’Angers, 16 Boulevard Daviers, 49045 Angers Cedex 01, France
b
a r t i c l e
i n f o
Article history:
Received 15 September 2014
Received in revised form
10 December 2014
Accepted 10 December 2014
Available online 18 December 2014
Keywords:
Apple peel
Phenolics
Supercritical extraction
Antioxidant
a b s t r a c t
Subcritical extraction (SFE) of dry and ground Golden delicious peels (30 g) was investigated at
25 MPa and 50 ◦ C using CO2 and ethanol (96%) in 75:25 mol ratio. As for conventional ethanol or
methanol/acetone/water extraction, nine phenolics were identified in SFE-extracts including the sugarbased phloridzin and quercetin derivatives. Extraction kinetics of the nine phenolics and of the global
yield were monitored via collection of fractions that were also characterized for their antioxidant activity
(ABTS antiradical activity). Kinetics showed a constant extraction rate up to 1.1 kg of fluid and a decreasing
rate afterwards, but the matrix was not exhausted after 3 h of extraction. Besides the classical continuous flow protocol, SFE was performed by introducing static periods between the dynamic collect of
fractions. Static periods did not yield significant improvement in the overall yield and in the individual
yield of most phenolics. Increasing the matrix loading did not improve the recovery either. Conversely,
extractions from 15 g provided the highest phenolics yield of 800 mg/100 gdry peels . For extracts tested for
antioxidant capacity (30 g loading), values up to 5–6 mg Equivalent Ascorbic Acid/gextract were obtained.
Activities were positively correlated with phenolics concentration in fractions only for static conditions.
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
A diet rich in fresh vegetables and fruits is generally recommended for a healthy lifestyle because they constitute important
sources of nutrients [1] like antioxidants [2], phytosterols [3] and
fibre [4]. Transformation of a raw material to foodstuff creates a
‘waste’ that comprises as well the edible food mass that is lost, discarded, or degraded in the different stages of the food supply chain.
Other remnants are peels, stems, cores, skins, seeds, husks, bran or
straw from cereals, fish skin, head and bones, mill wastewaters, etc.
Food ‘wastes’ have long been considered as undesirable materials
that were disposed of, in costly manners, via animal feed, landfill
or incineration, but nowadays, they are considered as promising
sources of valuable nutraceuticals [1,5].
Apples are one of the most consumed fruits worldwide and are
among the major sources of phytochemicals and antioxidants in
the human diet. Approximately 70 million tonnes of apples are produced worldwide (http://www.wapa-association.org). Apple fruits
∗ Corresponding author. Tel.: +33 05 40 00 68 32.
E-mail address: [email protected] (P. Subra-Paternault).
http://dx.doi.org/10.1016/j.supflu.2014.12.007
0896-8446/© 2014 Elsevier B.V. All rights reserved.
have a varied and well-balanced composition with a large diversity of vitamins, a high content of fibres compared to other fruits
and are moderately energetic in terms of caloric intake [6,7]. Prevention of various chronic diseases has been associated with apple
consumption [8], in particular cardiovascular disease, in relation
with the main bioactive compounds of apples, namely fibre and
polyphenols [9]. Apples contain over 60 different phenolic compounds [8]. The four major phenolic groups are hydroxycinnamic
acids (with chlorogenic acid as the most abundant representative),
dihydrochalcone derivatives (specially phloridzin), flavan-3-ols
(catechin as monomers or procyanidins as oligomers) and flavonols
(quercetin and quercetin glycosides) [10,11]. The distribution and
concentration of polyphenols vary greatly among apple cultivars
(range of 68–165 mg/100 g edible portion [6]) and within the apple
fruit. Apple peels have higher levels of total polyphenolic compounds than flesh or core and concentrate specially quercetin
glycosides, chlorogenic acid and phloridzin [11–15]. As example,
for eight apple cultivars, the total polyphenolics ranged from 1.0
to 2.3 mg/g of fresh weight in the peel, to be compared with the
0.33–0.93 mg/g of fresh weight in flesh [13]. Additionally, apple
peels polyphenols were shown to have beneficial actions on oxidative stress and inflammation [16].
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
In France, an average of 20 kg of apples per capita are consumed
[17] corresponding to a 20.4% market share, far more important
than that of the second most consumed fruits, bananas (14%) and
oranges (12%). Apple is the largest fruit produced in France with
1.5 million tonnes in 2013. Around 30% of the production is transformed in juices, compotes, concentrates, generating large volumes
of wastes including peels.
In this work, the use of compressed CO2 to recover phenolic
components from apple peel was investigated. Sub- and supercritical extraction of functional ingredients from natural sources
was reviewed by Herrero in 2006 [18], whereas Diaz-Reinoso [19]
focused on compounds with antioxidant activities and Marostica
[20] on phenolics from plant materials. Since 2006, more than 70
papers about ‘polyphenols’ and ‘supercritical extraction’ have been
published (scopus source). With special mention to food wastes,
grape residues produced by the wine or distilling industry are
among the most largely studied by-products. Residues comprised
grape skin or seeds [21,22], bagasse [23], marc [24,25] or pomace
[26,27]. As for other by-products, colouring anthocyanins were
extracted from eggplant peels [28], phenolic antioxidants from several berries pressing wastes [29] or oil containing phenolics from
cherry kernels [30] or grape seeds [31]. Fruits pulp or leaves were
investigated as well, in particular sweet cherries [32], jamun fruits
[33], pitanga leaves [34] and arrabidae chica leaves [35], for the
most recent studies. To the best of our knowledge, supercritical
extraction from apples was only reported once [36]. The by-product
was the apple pomace, i.e. skin and pulp residues remaining after
pressing the fruits for juice production. One gram of pomace
was extracted by CO2 + ethanol (14–20%) for 10–40 min varying pressure (20–60 MPa) and temperature (40–60 ◦ C). Extracts
were characterized for the total phenolic content and antioxidant
activity, but no identification and quantification of the extracted
phenolic species were reported.
The aim of this work was to evaluate the potentialities of
supercritical technology to extract phenolic compounds from apple
peels, focusing on three issues:
(1) Identification and quantification of the extracted polyphenols.
(2) Monitoring the extraction kinetics of polyphenols and of the
total extracted amount.
(3) Studying the impact of process conditions onto kinetics and
yields (implementation of static steps during extraction, variation of the amount of matrix loaded in the vessel).
In this work, dry apple peels were extracted by CO2 + 25% mol
cosolvent (ethanol at 96%) at 25 MPa and 50 ◦ C during 3 h. Being an
apolar fluid, CO2 has a limited capacity for dissolving polyphenolics so ethanol is required as cosolvent to overcome this limitation
[37–39]. The use of high proportion of ethanol was justified by the
fact that phloridzin and quercetin glycosides, abundant in apple
peels, contain a sugar moiety that is too polar to be soluble in
neat CO2 . The selected temperature and pressure are in the range
of those used for supercritical extraction of phenolic compounds
[20] and moderate temperature and high pressure are generally
associated with low thermal degradation and high solubility. Due
to the 25% of cosolvent, extractions were performed in subcritical conditions since 50 ◦ C is below the critical temperature of that
CO2 + cosolvent mixture [36]. During the extraction course, fractions were regularly collected in order to describe accurately the
extraction kinetics and possibly fractionate the phenolics pool.
Extracts were characterized in terms of global yield, phenolic composition and antioxidant activity. A new procedure of extraction
that alternated static and dynamic periods was also investigated
with the aim of giving longer time for mass transfer and diffusion
to occur.
173
2. Materials and methods
2.1. Apple peel preparation and chemicals
Golden delicious apples were purchased from a local conventional orchard and were stored at 1 ◦ C. Apples were peeled
mechanically (Kali, France) and the obtained peels were immediately packed into polyethylene bags and frozen at −18 ◦ C for 24 h.
Samples were then freeze-dried during 48 h (Heto Lab Equipment,
Heto FD 2.5, Denmark) and were further stored in the dark under
vacuum at room temperature. The moisture content of the dried
apple peels was measured by weight loss at 68 ◦ C in oven under
vacuum (Multilab20, Le Matériel Physico-Chimique, France) and
was in the range of 5–7%. For extractions, the dried apple peels
were ground using a kitchen-type grinder (Moulinex, France). The
obtained ground material was not sieved so the samples consisted
in grains of various sizes below 1 mm.
Carbon dioxide (CO2 , 99.5 wt.%, Air Liquide, France) and ethanol
(EtOH, 96%, Xylab, France) were used for supercritical extractions.
Solvents for liquid chromatography were of HPLC grade (acetonitrile, 99.98%; acetic acid, 99.5%) and were purchased from Fisher
Chemical and Acros Organics, respectively. Several standards of
polyphenols (HPLC-grade, purity higher than 97%) were purchased
from Sigma–Aldrich and Extrasynthese (France): (+)-catechin, (−)epicatechin, phloridzin, chlorogenic acid, quercetin-3-d-glucoside,
quercitrin (quercetin-3-O-arabinoside), hyperoside (quercetin-3O-galactoside).
2.2. Supercritical fluid extraction
Extractions were performed using a home-made system which
consists of an extractor vessel of 490 cm3 (length of 25 cm and
inner diameter of 5 cm) heated by heating mantle (Watlow) and
two Gilson pumps for fluids admission (model 305, heads of 25SC
and 10SC for CO2 and cosolvent, respectively). The CO2 and cosolvent flow rates were checked by a gas-meter and the cosolvent
volume consumption over time, respectively. Multiple stop valves
(Autoclave France) enable to bypass the extractor to stabilize the
extracting flux, and a metering valve is used to control the overall flow rate. The set-up comprises a pre-cooling unit (Julabo,
Germany) for CO2 , a pre-heater cartridge (TOP Industrie, France),
a relief valve and various pressure and temperature sensors. The
extractions were performed at 50 ◦ C and 25 MPa which is close to
the pressure limit of the equipment. The extractor was filled with
a weighted amount of dried apple peels, using alternated beds of
matrix and glass beads of 2 mm to avoid caking. The extracting
CO2 + cosolvent mixture circulated from bottom to top. Dissolved
species were recovered in a home-made cyclonic collector whose
bottom was plunged in ice and that operated at atmospheric pressure. Fractionation of extracts was performed by changing regularly
the collector bottom.
Extractions were carried out following two procedures. As common steps, the vessel was charged with the matrix, heated, and
pressurized with CO2 up to 25 MPa. A static period of 20–30 min
was applied before starting the extraction by activating the CO2
and cosolvent pumps. The first fraction was collected until a continuous flow of ethanol appeared in the collector, typically after
20–25 min, for an overall CO2 /EtOH flow of 10 g/min. This fraction,
labelled F0, corresponds to the ethanol breakthrough and contains
mostly the matter that was solubilized during the static period in
neat CO2 . In the so-called dynamic procedure, the extraction continued with regular collects of extracts every 20 min, i.e. for one
breakthrough time, over 200 min. The extractor was sometimes
subjected to 10 min of static step when recovering F0 and F1 in
order to wash the collector. After 200 min, the cosolvent pump was
stopped and neat CO2 was used to flush the matrix from ethanol
174
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
until no solvent traces were detected at the separator entrance, typically after 1 h. In the second procedure, called ‘static’, a static period
was introduced before collecting each fraction, i.e. at the end of a
20 min dynamic collection, the vessel was isolated and the flux was
stopped for 30 min. The static/dynamic alternation was carried out
for 4 × 20 min, before stopping the ethanol flow and flushing the
matrix with pure CO2 as in the dynamic procedure. Extractions by
the two procedures were repeated twice to check reproducibility.
The collected fractions were analyzed for total mass by gravimetry after evaporating ethanol under nitrogen flux. The total
extracted mass was determined from the sum of the fraction
masses collected over the extraction time. Fractions were analyzed for antioxidant activity via the Ascorbic Acid Equivalent
Antioxidant Capacity (AAEAC) method. Phenolics were identified
and quantified by liquid chromatography (HPLC). In Tables, the
amounts of phenolics or of global extracts are expressed in mg
or g per 100 gdry peels , in which the subscript ‘dry peels’ designates
the lyophilized apple peels containing 5–7 wt% moisture. In figures,
kinetics data are given as cumulated amount as a function of the
mass of extracting fluid. The latter variable was preferred to time
since the amount of fluid used to collect a fraction was the same
whatever the number and duration of static periods.
2.3. Conventional solvent extraction
Conventional extractions were carried out on dry peels following two different methods. In the first one, derived from
Colin-Henrion [11] and named hereafter MeOH/Acetone, peels
were first mixed with methanol (14 mL for 0.6 g). After filtration
under vacuum, peels were re-suspended in an acetone:water mixture 70:30 v:v (8 mL) and filtrated again. The two resulting extracts
were combined and dried using a rotary evaporator under vacuum
at 30 ◦ C. The final concentrate was redissolved in 10 mL methanol
for HPLC analysis. In the second method, peels were macerated in
ethanol (EtOH 96%) at 45 ◦ C for 3 h under magnetic stirring (100 mL
of solvent per 10 g of peels). Maceration was performed for the sake
of comparison with supercritical extractions and was thus carried
out with a volume of ethanol identical to that passed with 1.1 kg of
{CO2 + ethanol} mixture. The supernatant was recovered by filtration under vacuum and directly injected in HPLC.
2.4. Analytical methods
Identification of polyphenols was performed by coupling
chromatographic separation and photodiode array or mass spectrometry detectors [40]. Analytical HPLC was run on a 2695
Waters (Guyancourt, France) separation module equipped with
a diode array detector 2996 Waters. Separation was achieved
on a Phenomenex (Le Pecq, France) Luna column 100 C18
(250 mm × 4.6 mm i.d., 5 ␮m) protected with a Phenomenex cartridge (C18, 4 mm × 3 mm i.d.) at a temperature of 30 ◦ C. The mobile
phase consisted of 0.1% formic acid in water (eluent A) and acetonitrile (eluent B) and the separation was performed using the
following gradient: 4–10% B (0–5 min), 10–45% B (5–40 min), at a
flow rate of 1 mL/min. Injection volume was 10 ␮L and UV detection
was studied at two wavelengths: 280 and 350 nm. The MS analyses
were performed on a Bruker (Bremen, Germany) ESI/APCI ion-trap
Esquire 3000+ in negative mode, with the following conditions:
collision gas, He; collision energy amplitude, 1.3 V; nebulizer and
drying gas, N2 , 7 L/min; pressure of nebulizer gas, 30 psi; dry temperature, 340 ◦ C; flow rate, 1.0 mL/min; solvent split ratio 1:9; scan
range, m/z 100–1000. Identification of polyphenols was achieved
by comparison of UV profiles and mass spectra with literature data
and/or using pure standards.
Routine analysis of produced extracts was performed using
an Accela HPLC system (Thermo Scientific, USA) consisting of an
Accela 1250 pump with a degasser, an autosampler, and a photodiode array detector. A C18 pyramid column (250 mm × 4.6 mm,
5 ␮m particle size; Macherey-Nagel, France) was used for chromatographic separations at 30 ◦ C. A mobile phase constituted by
acetic acid (distilled water:acetic acid 97.5:2.5 (v/v), eluent A) and
acetonitrile (eluent B) was used with a discontinuous gradient
of 3–9% B (0–5 min), 9–16% B (5–15 min), 16–33% B (15–30 min),
33–50% B (30–37 min), 50–90% B (37–40 min), at a flow rate of
1 ml/min. Injection volume was 10 ␮L. Quantification was performed via calibration curves that plot the peak area versus known
concentrations of standards in the range of 0.05–1 mg/mL. Calibration curves derived for 6 data minimum were satisfactorily fitted
with linear regression equation (correlation coefficients larger
than 0.997). Standards of catechin, epicatechin, phloridzin, chlorogenic acid, and several quercetin derivatives (quercetin-glucoside,
-arabinoside, -galactoside) were used for quantification. Quercetin
xyloside and quercetin rhamnoside were quantified using the
quercetin-arabinoside calibration curve, so data are expressed
as arabinoside equivalent (mg Eq.Arabi). Quantification was performed at = 360 nm for quercetin derivatives and = 280 nm for
other phenolics.
The determination of the extracts antioxidant capacity (AAEAC)
is based on the ability of antioxidants to scavenge the radical cation
of ABTS (2,2 -azino-bis-3-ethylbenzothiazoline-6-sulfonate). The
methodology here employed followed that described by Re et al.
[41] with some modifications. Briefly, ABTS radical cation (ABTS+• )
was produced by reacting 7 mmol/L ABTS stock solution with
140 mmol/L manganese dioxide (MnO2 ) in dark at room temperature for 12 h. The ABTS+• solution was diluted with distilled
water to an absorbance of 0.70 ± 0.02 at = 734 nm. After the addition of 40 ␮L of the sample to analyze to 1960 ␮L of the diluted
ABTS+• solution, absorbance readings were taken every 1 min using
Helios ␥ spectrophotometer (Thermo Electron Corporation, USA). lAscorbic acid (AA, Sigma–Aldrich, USA) was used as reference. The
percentage inhibition of absorbance at 734 nm was calculated as a
function of the concentration of extracts and ascorbic acid. Results
were expressed as mg of ascorbic acid per g of extract.
3. Results and discussion
3.1. Polyphenol profile of apple peels extracts
Chromatograms of apple peel extracts obtained by
CO2 + 25%EtOH extraction are given in Fig. 1.
HPLC analysis of extracts obtained by EtOH maceration and
MeOH/Acetone extraction are provided as well for comparison.
Phenolics were identified by UV spectra/mass spectrometry data
and/or comparison with retention time of standards (Table 1).
Whatever the extract, the identified phenolics were catechin, epicatechin, chlorogenic acid, phloridzin and quercetin glycosides
(quercetin-3-O-glucoside, -galactoside, -arabinoside, -xyloside, rhamnoside) which are the major phenolics of apples [11,16,42].
It is worthwhile noting that HPLC profiles of supercritical extracts
exhibited all peaks of phenolics indicating that the conditions were
suitable to extract even the sugar-based phloridzin and quercetin
glycosides.
3.2. Supercritical extraction: static versus dynamic procedure
Extractions with CO2 + EtOH mixture were carried out by two
methods, a conventional dynamic and a stepwise procedure that
involved static periods between collecting fractions. The evolution
of the extracted amount in each fraction is illustrated Fig. 2 for the
static procedure. The F0 fraction rarely contained any phenolics
since it collected what was solubilized during the static period in
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
175
Fig. 2. Amount of each phenolic in fractions collected during SFE at 25 MPa, 50 ◦ C,
and CO2 + 25% cosolvent (EtOH 96%), static procedure, from 30 g of ground apple
peels, Ftotal = 9.7 ± 0.8 g/min. E13 label is the sample reference in our database.
Fig. 1. Phenolic profiles at = 280 nm of various samples: (A) SFE static on
30 g-loading, fraction F3; (B) SFE dynamic 30 g-loading, fraction F3; (C) EtOH
maceration; (D) MeOH/acetone extraction. Peak attribution: (1) catechin; (2)
chlorogenic acid; (3) epicatechin; (4) quercetin-3-O-glucoside; (5) quercetin3-O-galactoside; (6) quercetin-3-O-arabinoside; (7) quercetin-3-O-xyloside; (8)
quercetin-3-O-rhamnoside; (9) phloridzin.
neat CO2 . All phenolics were usually present in the F1 fraction indicating that all components were readily accessible in the ground
apple peels matrix. Their concentration increased up to F2 or F3
before decreasing in F4 and eventually being constant when extraction was continued for longer duration as in the dynamic case.
Cumulated data are given in Fig. 3 with deviation bars provided
by duplication of extraction experiments.
The reproducibility was critical in dynamic conditions for catechin, chlorogenic acid, quercetin-glucoside and -rhamnoside.
Catechin and chlorogenic acid are eluted in chromatography as
small peaks among other peaks, which could complicate detection
and therefore the quantification in complex extracts. For chlorogenic acid [11,43,44] and catechin [45], their susceptibility to the
polyphenol oxidase enzyme could introduce an additional source
of deviation. None of the hypothesis holds for the two quercetinderivatives. Besides, the extracting fluid is a ternary mixture of
CO2 /ethanol/water, whose components can absorb separately and
differently onto the solid matrix and induce therefore a different
partitioning of extractable species between the solid and the fluid
phases along the extractor bed. Looking closer at the amount collected in each individual fraction, the main deviations occurred
in fractions F1 and F2, i.e. collecting what was eluted by a volume of fluid corresponding to twice the ethanol breakthrough. The
permanent regime of flow and of the complex fluid composition
might be not yet attained. When time was given for conditions
to settle via static periods, reproducibility was better. Regarding
now the extraction profiles, the curves shape was similar whatever the compound and proceeded mostly at a constant extraction
rate up to 1.1 kg fluid. Note that the last point of curves corresponds to the fraction obtained during the final flush with pure
CO2 , so the stronger apparent falling rate discernable at the end
of curves is more likely due to poorer solubilization rather than
to matrix exhaustion. Therefore, the extraction of phenolics was
not finished after 2 kg of fluid, i.e. after a fluid/matrix ratio of 67.
This is consistent with Hasbay Adil results on apple pomace that
reported that a value of 80 g of fluid/g of matrix was necessary to
reach a plateau on the total phenolics extraction curve [36]. Taking
reproducibility deviations into account, the introduction of static
periods between the dynamic collections did not yield appreciable
improvement of the extracted rate, excepted for phloridzin. Static
periods were investigated as a way to give more time for species
to diffuse within the peel structure, the material grains or in the
extracting fluid. An insignificant impact indicated that mass transfer was not a significant limiting mechanism for the extraction of
most phenolics.
Cumulated amounts of phenolics extracted at 1.1 kg of
fluid from the 30 g-loading are summarized in Table 2. For
Table 1
Phenolics identified from apple peels by diode array and mass spectroscopy detectors coupled to HPLC separation. tR : retention time, max : wavelength of maximal absorption,
[M−H]− : ion and fragment in mass spectrometry, Mw : molecular weight.
Compounds
tR (min)
UV max (nm)
[M−H]− (m/z)
Mw (g/mol)
Catechin (1)
Chlorogenic acid (2)
Epicatechin (3)
Quercetin-3-O-glucoside (4)
Quercetin-3-O-galactoside (5)
Quercetin-3-O-arabinoside (6)
Quercetin-3-O-xyloside (7)
Quercetin-3-O-rhamnoside (8)
Phloridzin (9)
14.40
15.37
17.29
22.50
22.76
23.89
24.85
25.29
27.00
280
298, 325
280
256, 355
256, 355
256, 355
256, 354
256, 350
284
nd
353
289
463, frag 301
463, frag 301
433, frag 301
433
447, frag 301
435, frag 273
290
354
290
464
464
434
434
448
436
nd: not detected.
176
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
Fig. 3. Cumulated amounts (in mg) of phenolics obtained by SFE using the dynamic (dark symbols) or the static (open symbols) procedure from 30 g of ground apple peels.
Amounts of quercetin 3-xyloside and 3-rhamnoside are expressed as equivalent quercetin 3-arabinoside.
comparison with other extraction methods or literature, data
are given for 100 g of matrix. By the supercritical technique, extracted amounts of phenolics were in the range
600 mg/100 gdry peel in which quercetin glycosides group represented 77–78%. As individual compounds, catechin, chlorogenic
acid, phloridzin, quercetin-galactoside were extracted in the
range of 6–36 mg/100 gdry peel whereas epicatechin and the other
quercetin derivatives were around 55–140 mg/100 gdry peel . Compared to ethanol maceration, the ethanol + CO2 combination
enhanced the extraction of all phenolics (except chlorogenic
acid), a trend that was evidenced as well by Farias-Compomanes
et al. in case of grape bagasse extraction [23]. Total phenolics
yield in the macerated sample was of 177 mg/100 gdry peel and
the glycosides-derivatives accounted for only 50% of the total.
When extracted with methanol/acetone/water, polyphenols totalized 792 mg/100 gdry peel in which the pool of quercetin glycosides
accounted for 50% and chlorogenic acid for 40%. Abundance of
chlorogenic acid and quercetin-derivatives especially of galactoside moiety was reported in literature for Golden fresh peels
extracted by methanol [13], but epicatechin and phloridzin were
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
177
Table 2
Conditions of extractions and yield of phenolic compounds extracted. EtOH: ethanol, MeOH: methanol, Ace: acetone; SFE: subcritical fluid extraction.
Method
SFE static-30 g
SFE dynamic-30 g
SFE-15 g c
Maceration
Solvent
Solvent (mol)
Solvent/peels ratio (wt basis)a
T/P
Global yield a (g/100 g dry
peel)d
CO2 :EtOH:H2 O75:22:3
37
50 ◦ C/25 MPa
15.8 ± 0.2
CO2 :EtOH:H2 O75:22:3
37
50 ◦ C/25 MPa
22.0 ± 0.7
CO2 :EtOH:H2 O75:22:3
73
50 ◦ C/25 MPa
30.1 ± 7
EtOH:H2 O88:12
8
45 ◦ C/0.1 MPa
65 ± 2
MeOH + Ace/H2 O
30
20 ◦ C/0.1 MPa
3.2
603 ± 182
6.0 ± 6.0
36.2 ± 23.0
78.7 ± 11.4
13.2 ± 3.0
113.7 ± 60.4
21.2 ± 5.7
54.6 ± 5.7
139.8 ± 14.2
134.2 ± 46.4
27.4
800 ± 25
27.0 ± 3.1
12.8 ± 3.1
118.3 ± 4.8
21.7 ± 0.3
147.1 ± 4.5
43.7 ± 2.0
110.1 ± 1.2
183.5 ± 4.6
135.6 ± 1.4
26.7
177 ± 32
0
58.3 ± 11.8
28.5 ± 4.6
2.3 ± 0.28
26.1 ± 4.9
3.7 ± 0.05
35.0 ± 6.0
9.6 ± 2.8
13.6 ± 2.0
2.5
792 ± 26
25.7 ± 0.2
319.7 ± 18.0
43.5 ± 1.9
15.3 ± 0.5
86.6 ± 2.3
97.1 ± 1.7
49.1 ± 1.9
69.8 ± 2.6
85.6 ± 13.6
247
Phenolics yielda (mg/100 g dry peeld )
550 ± 115
Total phenolics
11.4 ± 2.3
Catechin
Chlorogenic acid
16.4 ± 4.5
71.6 ± 3.4
Epicatechin
Phloridzin
21.5 ± 3.0
81.2 ± 23.0
Quercetin-3-glucoside
26.5 ± 10.4
Quercetin-3-galactoside
67.6 ± 8.3
Quercetin-3-arabinoside
b
Quercetin-3-xyloside
144.5 ± 27.5
109.3 ± 32.8
Quercetin-3-rhamnosideb
34.7
Total phenolics concentration
(g phenolics/kg extract)
a
b
c
d
1.1 kg of extracting fluid for supercritical extractions.
Expressed as equivalent quercetin 3-arabinoside.
Data correspond to mean values between static and dynamic results since they were similar.
Dry peel = lyophilized peels that still contain 5–7% humidity.
less abundant in our sample than usually reported [46–49]. In
other hand, composition and concentration of phenolics in fruits
vary with fruits type, growing season, geographic location, genetic
variation [48,50,51] and conditions of extraction (sample to solvent ratio, time of exposure, temperature, etc.) influence the
quantification as well [52]. The particular abundance of chlorogenic acid in our extract prepared by the methanol/acetone/water
method could also come from the use of acetone that by inhibiting the polyphenol oxidase enzyme preserves the original content
of the fruit [6]. Regarding quercetin-derivatives presence, the
sum of total phenolics quantified by HPLC from a Gala apple
pomace extracted by acetone/water, was 7.24 g/kg dry matter,
of which more than 50% was quercetin glycosides [42]. The
quercetin-derivatives pool also represented the main phenolic
constituents of methanol/water extracts from 7 apple varieties
peels, with a content in the range of 0.1–0.9 g/kg fresh weight,
which taking in account the peel humidity, corresponded to
40–360 mg/100 g of dry matter [12]. The level of quercetin glycosides in our various extracts and their high contribution to the total
extracted phenolic pool are therefore consistent with literature
data.
Regarding consistency with the only study of supercritical
extraction from apple [36], let first remind that individual phenolic concentrations were not measured by HPLC and that the total
phenolic content was estimated by the Folin–Ciocalteu method.
Expressed as gallic acid equivalent (GAE), contents of 171 mg
GAE/100 g sample and 47 mg GAE/100 g sample was reported
for ethanol and (CO2 + 20 wt% EtOH) extractions, respectively,
which is an opposite trend to our own findings. However, the
Folin–Ciocalteu reagent does not only measure phenols and react
with any reducing substance. It therefore measures the total reducing capacity of a sample, not just phenolic compounds. Besides
phenols, nitrogen-containing compounds, trioses glyceraldehyde,
vitamin derivatives or proteins are reactive towards this reagent
[53]. For apple purees and juice, the contribution of the interfering substances can be between 31% and 48% [36]. Moreover,
Hasbay Hadil investigations [36] were carried out on apple pomace
(flesh + skin residues after pressing) from Starking and Amasya
apple varieties, whereas our own study is performed on fresh skin
solely and Golden Delicious variety, two differences that induce fluctuations in quantity of antioxidants [48].
Fig. 4. Global amounts of extracts obtained by SFE using the dynamic (dark symbols)
or the static (open symbols) procedure on 30 g of ground apple peels. Deviation bars
for the static case are too small to be viewed.
Regarding global yields (Fig. 4, Table 2), the overall extracted
amounts were in the range of gram, and the extraction profiles
exhibited, as phenolics, a linear part up to 1.1 kg of fluid.
When the process was run above 1.1 kg fluid as for dynamic
case, the extraction continued at a lower rate. It is interesting to
note that this falling extraction rate occurs at the same fluid consumption than the one reported by Farias-Campomanes [23] for
similar matrix amount and vessel capacity (20 g of grape bagasse,
415 cm3 vessel). The introduction of static periods yields a slight
decrease of the global extracted amount, in the range of 1.5 g at
1.1 kg of fluid (reproducibility bars for static conditions are too
small to appear in the graph). In other words, giving time for diffusion penalized the overall extraction whereas it had almost no
impact for phenolics, maybe because they contributed for only few
percentiles to the extract (see below). Less extracted amount in
case of static period means a smaller transported concentration
through the bed. Though not very pronounced, the behaviour suggests competitive reactions to extraction, such as degradation of
extractable species in less soluble compounds or a re-absorption of
the extractable species. Because of the complexity of the extracting
mixture (CO2 : 77 mol%, ethanol: 22 mol%, water: 3 mol%), the fluid
components can absorb separately and in different extend onto the
solid matrix, inducing a different partitioning of extractable species
between the solid and the fluid phases over the extraction duration.
When experiments were carried out in a dynamic mode, there was
178
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
Fig. 5. Effect of the matrix loadings on the global amount obtained by SFE using
dynamic or static procedure (15 g, 30 g and 55 g of ground apple peels).
not enough time for the competitive redistribution or degradation
to occur, so extractables were extracted.
As mentioned before, the range of extracted amounts of phenolics and global extracts is very different. More precisely, with
values of 27–35 g/kg of extract at 1.1 kg of fluid (Table 2), phenolics
represent only 3% of the global extract. Although the quantification took only in account the identified compounds, the difference
is too large to come from non-identified polyphenols solely. The
presence of other compounds in extracts is inherent of the use
of ethanol that is far from selective. Similar range varying from
gram to ten of grams per kg of extract was reported in literature
indicating a substantial extraction of multiple components besides
the phenolics whatever the matrix studied [23,29,32]. Generally
speaking, polyphenols are secondary metabolites of plants so they
are present at a much lower level than constitutive molecules of
cells like lipids, proteins, carbohydrates, and, as minor compounds,
minerals and vitamins. Among possible extracted components,
lipids and carbohydrates are potential candidates although one
cannot discard a possible extraction by leaching of any molecules.
Lipids, including phospholipids are well extracted by CO2 + 25%
ethanol [54]. The sugar content of apples is very high (10 g over
100 g of fresh edible portion that contains 83% of water) with dfructose and in less extent d-glucose as predominant sugars [6].
Though scCO2 is non-polar and hence not suitable for carbohydrates extraction, studies have shown that using a polar cosolvent
as ethanol can increase notably their solubility: a solubility of
0.1–0.8 mg/g in CO2 + 21%ethanol (at 96%) was reported for four
prebiotic fructose- and galactose-based carbohydrates [55]. Comparing scCO2 + cosolvent and solvent extractions and commenting
respective global yields [56], Pinelo et al. hypothesized that scCO2
conditions could be able to promote a breakdown of the sugar
structure of the plant matrix and the consequent solubility of
the resulting monomeric and oligomeric polysaccharides previously ravelled in the cell wall framework. Moreover, apple peels
CO2 -treated in this work were analyzed for glucose and fructose
contents, and after the subcritical treatment, the levels were half
the initial values, which indicated that these compounds were
indeed extracted.
3.3. Effect of apple peels loadings
SFE was performed by charging various amounts of apple peels
in the extractor, 15, 30 and 55 g for the static procedure, and 15
and 30 g for the dynamic. Increasing the loading did not yield an
enhancement of the global extracted amount in the same proportion than the loading increase (Fig. 5).
For the first fractions (up to 0.6 kg of fluid), the global extracted
amount was in the same range whatever the mass loaded or the
procedure. Afterwards, the increase from 15 to 30 g did increase the
extracted amount but not in a two-fold factor. Increasing further the
loading from 30 to 55 g (in case of static procedure) did not yield
more extracted amount since curves were almost superimposed.
The poor sensitivity of the extracted amount with the matrix weight
might come from a poor packing of the solid in the extractor and/or
the saturation of the fluid. Although the bed was made by alternated
layers of matrix and glass beads, the thickness of the apple layers
were increased to accommodate the increase of matrix loading.
When channelling occurs, only parts of the matrix are in contact with the extracting fluid; higher loadings could increase only
slightly the matrix in contact letting a larger portion not extracted.
Saturation of the fluid, especially at the early times of the extraction and above 15 g of matrix could contribute as well to the poor
effect of loadings. Similar extracted amount means similar transported concentration through the bed, e.g. a fluid already enriched
with extractables coming for instance from the first 15 g of matrix
at the bed bottom cannot completely solubilize the extractables
available from the other 15 g of the bed exit. After a certain time,
the negative effect of saturation progressively vanished allowing
for other molecules to be extracted and for regaining effects of
operational parameters. A saturation that postponed extraction of
other molecules was encountered by Uquiche et al. [57] dealing
with extraction of oil and minor lipids from seeds cakes. Specially,
sterols and carotenoids that were present at lower concentration
in cake and were less soluble than oil were extracted in the later
stages of extraction when there was not enough oil to saturate the
CO2 phase. Therefore, a time fractionation effect between the major
oil and the minor lipids was observed.
For phenolics (Fig. 6), it is worthwhile noting that a 15 g of loading allows for obtaining similar kinetics in static and in dynamic
modes and for attaining a plateau, visible in dynamic conditions
since extraction was performed over longer time. The extraction
rate fell down notably after 1.1 kg fluid so that at 1.8 kg of fluid, i.e.
for a solvent/matrix ratio of 120, the extraction of all phenolics was
almost finished.
The overall amount of phenolics extracted in those conditions
was in the range of 140 mg for 15 g of feed. At 1.1 kg fluid, the
amount extracted was 120 mg, which gives the highest value
of 0.8 g of phenolics/100 gdry peel (Table 2). Whatever the protocol, the extracts were mainly composed by epicatechin and
quercetin derivatives (Table 2), with the pool of quercetin derivatives accounting for 76–79%. As observed for the global extracted
amount, an increase of loading did not yield a correspondingly
increase of extracted phenolics, excepted for phloridzin. All kinetics were in the same range whatever the procedure and the loading
up to 0.5–0.7 kg fluid, which is the same behaviour than observed
for overall extract. Beyond that value, effects of operational parameters were regained, but phenolics did not respond in the same
extend to the variation of loading. For static conditions that were
more reproducible than dynamic ones, quercetine-glucoside and
-galactoside were only slightly affected, whereas the extracted
amounts of other quercetin derivatives, epicatechin and phloridzin
were more regularly augmented. Chlorogenic acid still behave differently, with an extracted amount that was lower at 60 g than from
30 g. In dynamic conditions and for compounds of higher reproducible data (quercetine-xyloside, -arabinoside, -epicatechin), the
extracted amount was increased as well providing however that
more than 1.1 kg of fluid was used.
3.4. Antioxidant activity of extracts
The antioxidant capacity of the various fractions collected during the static procedure (30 g loading) is illustrated Fig. 7. Deviation
bars accounted for the analysis of the duplicated SFE experiments. The F2 and F3 fractions exhibited the highest antioxidant
capacity (5–6 mg AAEq./gextract ) with a trend that paralleled the
evolution of the phenolic content in fractions (Fig. 7). Looking for a
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
179
Fig. 6. Effect of the matrix loadings on the phenolic amount obtained by SFE using dynamic or static procedure (15 g, 30 g and 55 g of ground apple peels). Amounts of
quercetin 3-xyloside and 3-rhamnoside are expressed as equivalent quercetin 3-arabinoside.
correlation, the antioxidant activity was plot as function of phenolics concentration in fractions (Fig. 8), considering dynamic results
obtained at 30 g loading as well. Static results exhibited a good
trend (R2 = 0.916), whereas data from dynamic procedure were
more scattered (R2 < 0.15). The two data out of the trend contour
for dynamic results corresponded to fractions F1 and F2 that were
those at the early stage of the extraction. The scattering of data
and the two different correlations depending on procedure seems
to indicate that either non-phenolic compounds contribute to the
antioxidant activity, or only some of phenolics are responsible for
it rather than the whole pool.
In literature, antioxidant activities did not correspond necessarily to high content of phenols. A positive correlation was found
for 56 vegetables by Deng et al. [58] whereas no relationship was
evidenced by Ismail et al. for 9 types of vegetables [59]. The different
methods used to produce extracts and to determine the phenolic
180
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
4-oxo function in the C ring, the 3- and 5-hydroxyl groups with
the 4-oxo function in A and C rings. The free 3-hydroxyl group
is important for antioxidant activity. Even if the 3-glycosylation
may reduce the activity when compared with the corresponding aglycone [60], quercetin glycosylated derivatives still have the
other attributes. When tested as inhibitors of fish oil oxidation,
quercetin-3-O-glucoside, -galactoside and -rhamnoside exhibited
similar or slightly higher effectiveness than the aglycone quercetin
[61], whereas for phloridzin, its 2-glycosylation yielded a lower
inhibition than the aglycone phloretin [62]. In case of apples, an
activity in the range of 820 ␮mol Trolox eq./100 g of fruit were measured for Golden variety by Serra et al. [63]. Eight other varieties
of apple were studied as well, and when submitting the overall
results to statistical analysis, catechin, epicatechin, procyanidin B1
and quercetin-3-glucoside were identified as major contributors
to the antioxidant activities (flesh + peels extracts; ORAC, HORAC
and LDL assays). For an industrial residue coming as a pomace of
juice-squeezed Fuji, Granny Smith and Qinguan apples mixture, the
antioxidant activity, expressed as ABTS radical inhibition rate, was
positively correlated with the phenolics content of various fractions
and specially to the presence of procyanindins B, chlorogenic acid
and quercetin [64]. Looking for a correlation with the chlorogenic
concentration or quercetin-derivative pools in our extracts will be
investigated in future.
4. Conclusions
Fig. 7. Ascorbic acid equivalent antioxidant capacity (AAEAC, mg/g of dry extract)
and total phenolic concentration (mg/g of dry extract) of fractions produced during
SFE (static procedure, 30 g of ground apple peels, duplicate extractions).
content and the antioxidant capacity could have generated different selectivity and responses: ethanol extraction, Folin–Ciocalteu
reagent, coupled reaction of linoleic acid and ␤-carotene in Ismail
[59], two steps extraction with methanol–acetic acid–water mixture for obtaining a hydrophilic fraction, Folin–Ciocalteu reagent,
Trolox equivalent antioxidant capacity in Deng [58]. In another
hand, all phenolics did not exhibit the same antioxidant capacity.
Phenolic compounds possess chemical structural requirements for
antioxidant activity because they have phenolic hydroxyl groups
able to donate a hydrogen atom or an electron to a free radical and
extended conjugated aromatic system to delocalize an unpaired
electron. As discussed by Dai and Mumper [60], chlorogenic acid (an
hydroxycinnamic acid derivative), is considered as a potent antioxidant agent as it possesses a catechol moiety (ortho-dihydroxy
structure) linked to a CH CH COO which allows the delocalization of electron by resonance. Among flavonoids, quercetin exhibits
high antioxidant activity because it combines all the requirements: a catechol moiety in the B ring, the 2,3-double bond with a
Fig. 8. Test of correlation between antioxidant capacity and concentration of phenolics (mg/g of dry extract) in fractions obtained during SFE static and dynamic
procedure (30 g of ground apple peels, duplicate extractions).
The aim of this work was to evaluate the potentialities of supercritical technology to recover the valuable polyphenolics from a
fruit waste, namely the apple peels. Compared to the only study
reported in literature, the present work identified and quantified
the major extracted phenolics, monitored the extraction kinetics of the polyphenols and of the global extract amount, and
investigated several process parameters in the view of scale-up
(amount of matrix charged in the vessel, introduction of static
steps during extraction). All principal phenolics known to be
present in apple peels were extracted by the conditions used, i.e.
CO2 + 25 mol% cosolvent (EtOH 96%), 25 MPa, 50 ◦ C, including the
polar sugar-based quercetin derivatives and phloridzin. Extractions were carried out by collecting fractions over a time related
to the ethanol breakthrough, in order to improve the accuracy of
kinetics and the understanding of the permanent regime establishment. Extraction of phenolics started rapidly with no delay between
components, indicating that they were all readily accessible. Over
the duration of the extractions, the exhaustion of the matrix was
usually not attained, except for experiment performed on a 15 g
loading treated with 1.8 kg of fluid, i.e. for a solvent/feed ratio of
120. Increasing the matrix loading did not increase the extracted
amounts in the same ratio, a poor efficiency that could be related to
the fluid saturation, bed packing, permanent regime of extraction
not established yet, or a combination of those factors. Introduction
of static periods was envisaged as way to give times for species
to diffuse within the material. For phenolics, the difference was
not significant, whereas for the overall extract, static conditions
led to a slightly smaller extracted amount. Although each phenolic was responding diversely to the effect, the differences were not
significant enough to envisage separation between groups.
So far, SCF technology was efficient to recover the phenolics at a
level of 120 mg from 15 g of peels with only 1.1 kg of extracting fluid.
Compared to methanol/acetone/water or ethanol extracts, contribution of quercetin derivatives to the phenolic pool in SC extracts
was particularly high with value in the range of 75–80 wt% instead
of 50%. However, phenolics constituted less than 5% wt of the whole
extract, a low content inherent to the use of ethanol that is far from a
selective solvent. Supercritical extracts were exhibiting antioxidant
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
activities but, despite a positive correlation between antioxidant
capacity and total phenolics content in static conditions, the larger
scattering when all data were considered indicated that not every
phenolics contributed to the antioxidant activity.
Acknowledgements
Authors acknowledge Aquitaine Region and National Association of Research and Technology – France, for financial support. Dr
Ingrid Freuze from “Plateforme d’Ingénierie et d’Analyses Moléculaires” (PIAM) and Dr David Guilet from “Plateau Technique
PHYTO”, University of Angers, are thanked for HPLC/UV/MS analyses.
References
[1] N. O’Shea, E. Arendt, E. Gallagher, Dietary fibre and phytochemical characteristics of fruit and vegetable by-products and their recent applications as novel
ingredients in food products, Innovative Food Science and Emerging Technology 16 (2012) 1–10.
[2] A. Morales-Soto, P. García-Salas, C. Rodríguez-Pérez, C. Jimenez-Sanchez, M.
Cadiz-Gurrea, A. Segura-Carretero, A. Fernández-Gutiérrez, Antioxidant capacity of 44 cultivars of fruits and vegetables grown in Andalusia, Spain, Food
Research International 58 (2014) 35–46.
[3] C. Martins, F. Fonseca, C. Ballus, A. Figueiredo-Neto, A. Meinhart, H.T. de Godoy,
M.C. Izar, C. Martins, Common sources and composition of phytosterols and
their estimated intake by the population in the city of São Paulo, Brazil, Nutrition 29 (2013) 865–871.
[4] J.W. Anderson, P. Baird, R.H. Davis, S. Ferreri, M. Knudtson, A. Koraym, V. Waters,
C.L. William, Health benefits of dietary fiber, Nutrition Reviews 67 (2009)
188–205.
[5] C. Galanakis, Recovery of high added-value components from food wastes: conventional, emerging technologies and commercialized applications, Trends in
Food Science and Technology 26 (2012) 68–87.
[6] R. Feliciano, C. Antunes, A. Ramos, A. Serra, M. Figueira, C. Duarte, A. de Carvalho, M. Bronze, Characterization of traditional and exotic apple varieties from
Portugal: Part 1. Nutritional, phytochemical and sensory evaluation, Journal of
Functional Food 2 (2010) 35–45.
[7] Aprifel, Agence fruits et legumes frais, http://www.aprifel.com/fiches
[8] V. Rupasinghe, S. Thilakarathna, S. Nair, Polyphenols of apples and their potential health benefits, in: J. Sun, K.N. Prasad, A. Ismail, B. Yang, X. You, L. Li (Eds.),
Polyphenols: Chemistry, Dietary Sources and Health Benefits, Nova Science
Publishers Inc., Hauppauge, NY, USA, 2012, pp. 333–368.
[9] A.-T. Serra, J. Rocha, B. Sepodes, A. Matias, R. Feliciano, A. de Carvalho, M. Bronze,
C. Duarte, M. Figueira, Evaluation of cardiovascular protective effect of different
apple varieties – correlation of response with composition, Food Chemistry 135
(2012) 2378–2386.
[10] K. Kahle, M. Kraus, E. Richling, Polyphenol profiles of apple juices, Molecular
Nutrition and Food Research 49 (2005) 797–806.
[11] M. Colin-Henrion, De la pomme à la pomme transformée: impact du procédé
sur deux composes d’intérêt nutritionnel (Ph.D. thesis), Université d’Angers,
France, 2008.
[12] S. Karaman, E. Tütem, K.S. Baskan, R. Apak, Comparison of antioxidant capacity
and phenolic composition of peel and flesh of some apple varieties, Journal of
Science of Food and Agriculture 93 (2013) 867–875.
[13] R. Tsao, R. Yang, C. Young, H. Zhu, Polyphenolic profiles in eight apple cultivars
using high-performance liquid chromatography (HPLC), Journal of Agricultural
and Food Chemistry 51 (2003) 6347–6353.
[14] A. Leccese, S. Bartolini, R. Viti, Antioxidant properties of peel and flesh
in ‘GoldRush’ and ‘Fiorina’ scab-resistant apple (Malus domestica) cultivars, New Zealand, Journal of Crop and Horticultural Science 37 (2009)
71–78.
[15] V. Rupasinghe, C. Kean, Polyphenol concentrations in apple processing
by-products determined using electrospray ionization mass spectrometry,
Canadian Journal of Plant Science 88 (2008) 759–762.
[16] M.-C. Denis, A. Furtos, S. Dudonne, A. Montoudis, C. Garofalo, Y. Desjardins, E. Delvin, E. Levy, Apple peel polyphenols and their beneficial
actions on oxidative stress and inflammation, PLOS ONE 8 (1) (2013) e53725,
http://dx.doi.org/10.1371/journal.pone.0053725.
[17] lapomme, Association Nationale Pommes Poires, http://www.lapomme.org/
[18] M. Herrero, A. Cifuentes, E. Ibañez, Sub- and supercritical fluid extraction
of functional ingredients from different natural sources: plants, food-byproducts, algae and microalgae – a review, Food Chemistry 98 (2006) 136–148.
[19] B. Diaz-Reinoso, A. Moure, H. Dominguez, J. Parajo, Supercritical CO2 extraction
and purification of compounds with antioxidant activity, Journal of Agricultural
and Food Chemistry 54 (2006) 2441–2469.
[20] M. Marostica Jr., A. Leite, N. Dragano, Supercritical fluid extraction and stabilization of phenolic compounds from natural sources – review, Chemical
Engineering Journal 4 (2010) 51–60.
181
[21] L. Fiori, D. de faveri, A. Casazza, P. Perego, Grape by-products: extraction of
polyphenolic compounds using supercritical CO2 and liquid organic solvent: a
preliminary investigation, CyTA – Journal of Food 7 (3) (2009) 163–171.
[22] J.L. Marqués, G. Della Porta, E. Reverchon, J.A.R. Renuncio, A.M. Mainar,
Supercritical antisolvent extraction of antioxidants from grape seeds after vinification, Journal of Supercritical fluids 82 (2013) 238–243.
[23] A. Farias-Campomanes, M. Rostagno, A. Meireles, Production of polyphenols
from grape bagasse using supercritical fluids: yield, extract composition and
economic evaluation, Journal of Supercritical Fluids 77 (2013) 70–78.
[24] C. Da Porto, D. Decorti, A. Natolino, Water and ethanol as co-solvent in supercritical fluid extraction of proanthocyanidins from grape marc: a comparison
and a proposal, Journal of Supercritical Fluids 87 (2014) 1–8.
[25] C. Da Porto, A. Natolino, D. Decorti, Extraction of proanthocyanidins from grape
marc by supercritical fluid using CO2 as solvent and ethanol–water mixture as
co-solvent, Journal of Supercritical Fluids 87 (2014) 59–64.
[26] D. Oliveira, A. Salvador, A. Smânia, M. Maraschin, S. Ferreira, Antimicrobial activity and composition profile of grape (Vitis vinifera) pomace
extracts obtained by supercritical fluids, Journal of Biotechnology 164 (2013)
423–432.
[27] M. Pinelo, A. Ruiz-Rodriguez, J. Sineiro, F.J. Senorans, G. Reglero, M.J. Nunez,
Supercritical fluid and solid–liquid extraction of phenolic antioxidants from
grape pomace: a comparative study, European Food Research and Technology
226 (2007) 199–205.
[28] D. Chatterjee, N. Jadhav, P. Bhattacharjee, Solvent and supercritical carbon dioxide extraction of color eggplants: characterization and food applications, LWT
– Food Science and Technology 51 (2013) 319–324.
[29] L. Laroze, B. Diaz-Reinoso, A. Moure, M. Zuniga, H. Dominguez, Extraction
of antioxidants from several berries pressing wastes using conventional and
supercritical solvents, European Food Research and Technology 231 (2010)
669–677.
[30] C. Yilmaz, V. Gokmen, Compositional characteristics of sour cherry kernel and
its oil as influenced by different extraction and roasting conditions, Industrial
Crops and Products 49 (2013) 130–135.
[31] N. Rombaut, R. Savoire, B. Thomasset, T. Belliard, J. Castello, E. Van-Hecke, J.L. Lanoisellé, Rape seed oil extraction: interest of supercritical fluid extraction
and gas assisted mechanical extraction for enhancing polyphenol co-extraction
in oil, Comptes Rendus Chimie 17 (2014) 284–292.
[32] A.-T. Serra, I. Seabra, M. Braga, M. Bronze, H. de Sousa, C. Duarte, Processing
cherries (Prunus avium) using supercritical fluid technology: Part 1. recovery
of extract fractions rich in bioactive compounds, Journal of Supercritical Fluids
55 (2010) 184–191.
[33] J. Maran, B. Priya, S. Manikandan, Modeling and optimization of supercritical
fluid extraction of anthocyanin and phenolic compounds from Syzygium cumini
fruit pulp, Journal of Food Science and Technology 51 (2014) 1938–1946.
[34] T. Garmus, L. Paviani, C. Queiroga, P. Magalhães, F. Cabral, Extraction of phenolic
compounds from pitanga (Eugenia uniflora L.) leaves by sequential extraction
in fixed bed extractor using supercritical CO2 , ethanol and water as solvents,
Journal of Supercritical Fluids 86 (2014) 4–14.
[35] J.T. Paula, L.C. Paviani, M.A. Foglio, I. Sousa, G. Duarte, M. Jorge, M. Eberlin,
F. Cabral, Extraction of anthocyanins and luteolin from Arrabidaea chica by
sequential extraction in fixed bed using supercritical CO2 , ethanol and water
as solvents, Journal of Supercritical Fluids 86 (2014) 100–107.
[36] I. Hasbay Adil, H. Cetin, M. Yener, A. Bayindirli, Subcritical (carbon dioxide + ethanol) extraction of polyphenols from apple and peach pomaces, and
determination of the antioxidant activities of the extracts, Journal of Supercritical Fluids 43 (2007) 55–63.
[37] A. Berna, A. Chafer, J.B. Monton, S. Subirats, High-pressure solubility data of
system ethanol (1) + catechin (2) + CO2 (3), Journal of Supercritical Fluids 20
(2001) 57–162.
[38] A. Chafer, A. Berna, J. Monton, R. Munoz, High-pressure solubility data of system
ethanol (1) + epicatechin (2) + CO2 (3), Journal of Supercritical Fluids 24 (2002)
103–109.
[39] A. Chafer, T. Fornari, A. Berna, R.R. Stateva, Solubility of quercetin in supercritical CO2 + ethanol as a modifier: measurements and thermodynamic modeling,
Journal of Supercritical Fluids 32 (2004) 89–96.
[40] S. Boisard, A.M. Le Ray, J. Gatto, M.-C. Aumond, P. Blanchard, S. Flurin, P.
Richomme, Chemical composition, antioxidant and anti-AGEs activities of a
French poplar type Propolis, Journal of Agricultural and Food Chemistry 62
(2014) 1344–1351.
[41] R. Re, N. Pellegrini, A. Proteggente, A. Pannala, M. Yang, C. Rice-Evans, Antioxidant activity applying an improved ABTS radical cation decolorization assay,
Free Radical Biology and Medicine 26 (1999) 1231–1237.
[42] Y. Lu, L.Y. Foo, Identification and quantification of major polyphenols in apple
pomace, Food Chemistry 59 (1997) 187–194.
[43] J.J. Nicolas, F.C. Richard-Forget, P.M. Goupy, M.-J. Amiot, S.Y. Aubert, Enzymatic browning reactions in apple and apple products, Critical Reviews in Food
Science and Nutrition 34 (1994) 109–157.
[44] K. Amaki, E. Saito, K. Taniguchi, K. Joshita, K.M. Murata, Role of chlorogenic
acid quinone and interaction of chlorogenic acid quinone and catechins in the
enzymatic browning of apple, Bioscience, Biotechnology, and Biochemistry 75
(2011) 829–832.
[45] J.-H. Yi, M.-L. Li, Z.-B. Zhu, Influence of apple polyphenols on the properties of polyphenol oxidase, Modern Food Science and Technology 29 (2013)
2601–2606.
[46] M. Ceymann, E. Arrigoni, H. Schärer, A. Bozzi Nising, R. Hurrell, Identification of
apples rich in health-promoting flavan-3-ols and phenolic acids by measuring
182
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
A. Massias et al. / J. of Supercritical Fluids 98 (2015) 172–182
the polyphenol profile, Journal of Food Composition and Analysis 26 (2012)
128–135.
C.G. Grigoras, E. Destandau, L. Fougère, C. Elfafir, Evaluation of apple pomace
extracts as a source of bioactive compounds, Industrial Crops and Products 49
(2013) 794–804.
B. Lata, K. Tomala, Apple peel as a contributor to whole fruit quantity of potentially healthful bioactive compounds: cultivar and year implication, Journal of
Agricultural and Food Chemistry 55 (2007) 10795–10802.
G.M. Huber, H.P.V. Rupasinghe, Phenolic profiles and antioxidant properties of
apple skin extracts, Journal of Food Science 74 (9) (2009) C693–C700.
B. Lata, A. Trampczynska, J. Paczesna, Cultivar variation in apple peel
and whole fruit phenolic composition, Scientia Horticulturae 121 (2009)
176–181.
C. Manach, A. Scalbert, C. Morand, C. Rémésy, L. Jiménez, Polyphenols: food
sources and bioavailability, American Journal of Clinical Nutrition 79 (2004)
727–747.
V. Rupasinghe, P. Kathirvel, G. Huber, Ultrasonication-assisted solvent extraction of quercetin glycosides from ‘idared’ apple peels, Molecules 16 (2011)
9783–9791.
J. Everette, Q. Bryant, A. Green, Y. Abbey, G. Wangila, R. Walker, Thorough study
of reactivity of various compound classes toward the Folin–Ciocalteu reagent,
Journal of Agricultural and Food Chemistry 58 (2010) 8139–8144.
P. Subra-Paternault, H. ThongDeng, A. Grélard, M. Cansell, Extraction of phospholipids from scallop by-product using supercritical CO2 /alcohol mixtures,
LWT – Food Science and Technology 60 (2015) 990–998.
F. Montanés, T. Fornari, R.P. Stateva, A. Olano, E. Ibànez, Solubility of carbohydrates in supercritical carbon dioxide with (ethanol + water) cosolvent, Journal
of Supercritical Fluids 49 (2009) 16–22.
[56] M. Pinelo, A. Ruiz-Rodriguez, J. Sineiro, F. Senorans, G. Reglero, M. Nunez, Supercritical fluid and solid–liquid extraction of phenolic antioxidants from grape
pomace: a comparative study, European Food Research Technology 226 (2007)
199–205.
[57] E. Uquiche, X. Fica, K. Salazar, J. del Valle, Time fractionation of minor lipids
from cold-pressed rapeseed cake using supercritical CO2 , Journal of American
Oil Chemist’s Society 89 (2012) 1135–1144.
[58] G.-F. Deng, X.-R. Lin, L.-L. Gao, J.-F. Xie, H.-B. Li, Antioxidant capacities and
total phenolic contents of 56 vegetables, Journal of Functional Foods 5 (2013)
260–266.
[59] A. Ismail, Z. Marjan, C. Foong, Total antioxidant activity and phenolic content
in selected vegetables, Food Chemistry 87 (2004) 581–586.
[60] J. Dai, R. Mumper, Plant phenolics: extraction, analysis and their antioxidant
and anticancer properties, Molecules 15 (2010) 7313–7352.
[61] G.M. Huber, H.P.V. Rupasinghe, F. Shahidi, Inhibition of oxidation of omega-3
polyunsaturated fatty acids and fish oil by quercetin glycosides, Food Chemistry
117 (2009) 290–295.
[62] V. Rupasinghe, A. Yasmin, Inhibition of oxidation of aqueous emulsions of
omega-3 fatty acids and fish oil by phloretin and phloridzin, Molecules 15
(2010) 251–257.
[63] A.-T. Serra, A. Matias, R. Frade, R. Duarte, R. Feliciano, M. Bronze, M. Figueira, A.
de Carvalho, C. Duarte, Characterization of traditional and exotic apple varieties
from Portugal: Part 2. Antioxidant and antiproliferative activities, Journal of
Functional Foods 2 (2010) 46–53.
[64] X. Bai, H. Zhang, S. Ren, Antioxidant activity and HPLC analysis of polyphenolenriched extracts from industrial apple pomace: analysis and antioxidant
activity of industrial apple pomace extract, Journal of the Science of Food and
Agriculture 93 (2013) 2502–2506.