Ultrasonics Sonochemistry xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Ultrasonics Sonochemistry
journal homepage: www.elsevier.com/locate/ultson

Effect of power ultrasound application on aqueous extraction of phenolic

compounds and antioxidant capacity from grape pomace
(Vitis vinifera L.): Experimental kinetics and modeling
M.R. González-Centeno ⇑, F. Comas-Serra, A. Femenia, C. Rosselló, S. Simal
Department of Chemistry, Universitat de les Illes Balears, Ctra Valldemossa, Km 7.5, 07122 Palma de Mallorca, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The kinetics of both conventional (mechanical stirring, 200 rpm) and acoustic (55 ± 5 kHz, 435 ± 5 W/L)
Received 6 March 2014 aqueous extraction of total phenolic content and antioxidant capacity from grape pomace by-products
Received in revised form 14 May 2014 (Vitis vinifera L.) have been experimentally evaluated and modeled at different extraction temperatures
Accepted 30 May 2014
(20, 35 and 50 °C). A gradual and significant increase of total phenolic content and antioxidant capacity
Available online xxxx
of the extracts was observed as the temperature increased, the highest values being obtained in the case
of the extraction assisted acoustically. According to the results, the acoustic assistance of the extraction
Keywords:
process led to aqueous extracts with phenolic and antioxidant characteristics similar to those obtained
Grape pomace
Winery by-products
with mechanical stirring, working under lower temperature conditions and during less operating time.
Ultrasound-assisted extraction Specifically, the conventional extraction of total phenolics at 35 and 50 °C did not differ significantly from
Weibull model extractions assisted with power ultrasound at 20 and 35 °C, respectively; and the acoustic process
Total phenolic content required approximately 3, 4 and 8 times less time, at 20, 35 and 50 °C, than the conventional extraction
Antioxidant capacity to obtain extracts with similar characteristics. The extraction curves obtained for total phenolic content
and antioxidant capacity, measured by the ABTS and FRAP methods, were properly represented by a
modified Weibull model for both conventional and acoustic extractions within the temperature range
20–50 °C, presenting an average percentage of explained variance P 97.9%, and an average mean relative
error 6 7.0%. A high correlation (r2 P 0.992) was observed between the experimental and simulated
values for all the quality attributes in study.
Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction makes conventional extraction unsuitable for industrial applica-

Nowadays, interest in the use of plant materials as a potential
source of bioactive compounds is mainly focused on (i) providing
for their increasing demand from the food, cosmetics and pharma-
ceutical industries; and (ii) upgrading the large quantity of agricul-
tural by-products generated annually [1]. Specifically, grape
pomace has gained on attention over the last few years, as a waste
by-product leftover from winemaking process which still contains
significant amounts of phenolics and antioxidants when discarded.
Within this context, the solid–liquid extraction, as a mass transfer
process of the target compounds from the plant material to a par-
ticular solvent, plays a key role in the recovery of those phenolics
and antioxidants retained in the vegetable matrix [2]. However,
the natural resistance of the plant structure to solvent penetration
⇑ Corresponding author. Tel.: +34 971259599; fax: +34 971173426.
E-mail address: reyes.gonzalez.centeno@gmail.com (M.R. González-Centeno).
tion due to long operating times, high energy demand and elevated
solvent consumption [3,4].
Over the past few decades, intensive research in this field has
enabled the development of more efficient and environmentally
friendly extraction techniques. Among them, ultrasound-assisted
extraction may be one of the most widely explored on both a lab-
oratory and an industrial scale, due probably to the relatively inex-
pensive equipment, simple operation and high efficiency derived
from the mechanical effects induced in the material to which they
are being applied [5]. Ultrasonic enhancement of the solid–liquid
extraction is mainly attributed to the phenomenon of acoustic cav-
itation, consisting of the formation, growth and implosive collapse
of microbubbles inside the liquid phase submitted to power ultra-
sound application [6]. According to Mason et al. [7], there are three
interconnected mechanisms through which power ultrasound
increase the efficiency of the extraction process by providing better
mass transfer: (i) breakdown of plant cells via the formation of

http://dx.doi.org/10.1016/j.ultsonch.2014.05.027
1350-4177/Ó 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
2 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
Nomenclature
a parameter of Eq. (6) (1/K)
AG agitation or mechanical stirring
C extraction yield (mg/100 g dm)
C0 initial extraction yield (mg/100 g dm)
Ceq equilibrium extraction yield (mg/100 g dm)
C eq0 parameter of Eq. (6) (mg/100 g dm)
Cp specific heat capacity of the solvent (J/kg °C)
CI confidence interval
dm dry matter basis
Ea activation energy (J/mol)
GAE gallic acid equivalent
m solvent mass (kg)
MRE mean relative error (%)
n Number of experimental data Eq. (4)
P ultrasonic power (W)
R universal gas constant (8.31 J/mol K)
Scalc standard deviation of the calculated values Eq. (5)
Sexp standard deviation of the experimental values Eq. (5)
SE standard error
t Time (s)
T absolute temperature (K)
microjets due to asymmetrical bubble collapse near a solid surface,
which increases the permeability of the plant tissue and releases
the intracellular material; (ii) improved analytes solubility and
solvent penetration, caused by the localized increase in tempera-
ture and pressure at the zone of the bubble implosion; and (iii)
enhanced diffusion as a consequence of the microstreaming
generated by ultrasound.
In any case, extraction rates of the target compounds will
depend on the effect of the different operating experimental
conditions, there being a renewed interest in optimizing all these
process variables to the best utilization of energy, time, raw mate-
rial and/or solvent [8]. From a food engineering point of view,
mathematical modeling has been widely used for this purpose in
the last decades, since it provides a quick and inexpensive determi-
nation of the effects of the experimental conditions on the outcome
of the extraction process. Analyses of the kinetic parameters and
their dependence on the operating conditions have led to a
better understanding of the mass transfer mechanism, facilitating
the design, simulation, optimization and control of industrial
processes [9].
Kinetic models of solid–liquid extraction may be based on
either phenomenological or empirical equations. The latter, formu-
lated from experimental data, attempt to determine the relevant
variables of the process and the relationships among them.
Although they do not physically explain the extraction process,
the empirical models may indeed be very useful for certain indus-
trial applications or for simplifying the study of complex systems,
hard to formulate and/or solve by phenomenological models [10].
Such is the case of the extraction of bioactive compounds from
plant materials, due to the heterogeneity of the samples and the
multicomponent complexity of the mass transfer mechanism.
Thus, the main aim of the present research was to investigate
the effect of both temperature and presence (or not) of power
ultrasound on the extraction rate of phenolics and antioxidant
capacity from grape pomace by-products. Furthermore, this study
aimed at proposing an empirical model to simulate the extraction
kinetics of both acoustic and agitation experiments, that properly
relates changes in phenolic and antioxidant capacity of the aque-
ous extracts to the ultrasonic field, operating temperature and
extraction time variables.
Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
TE trolox equivalent
TP total phenolic content (mg GAE/100 g dm)
US power Ultrasound
VAR percentage of explained variance (%)
Xcalc calculated value Eq. (4)
Xexp experimental value Eq. (4)
Greek symbols
a kinetic reaction constant (Weibull model, Eqs. (2) and
(3) (s)
a0 pre-exponential factor Eq. (3) (s)
b shape parameter of the Weibull model Eqs. (2) and (7)
b0 empirical parameter of Eqs. (7) and (8)
b0a empirical parameter of Eq. (8)
b0b empirical parameter of Eq. (8) (K1)
b1 empirical parameter of Eqs. (7) and (9) (s1)
b1a empirical parameter of Eq. (9) (s1)
b1b empirical parameter of Eq. (9) (K1)
2. Materials and methods
2.1. Samples
The raw material consisted of grape pomace from Syrah red
grape variety (Vitis vinifera), provided by Ca’n Majoral winery
(Mallorca, Balearic Islands, Spain) during the 2011 harvest.
Samples, consisting of skins and seeds together, were obtained
after pressing the grape and macerating for 2 weeks at 25 °C. Grape
pomace samples presented a ratio seeds/skins of 0.2 and a mois-
ture content, determined according to the AOAC official method
[11], of 1.18 g H2O/g dm. Samples were stored vacuum-packed at
80 °C until further processing and analysis.
2.2. Conventional and acoustic extraction procedures
Before extraction, grape pomace samples were sieved to a par-
ticle size of 5 ± 1 mm. Water was used as extraction solvent at a
ratio of 1:20 (w/v, g/mL) with a total extraction volume of
200 mL. Extractions were performed in a vessel with a double glass
layer to control temperature throughout the process. For this pur-
pose, a peristaltic pump (VitaTech 600, Vitakraft, Germany) recir-
culated a 50% (v/v) ethylene glycol solution from the cooling
reservoir, equipped with a chiller unit (Frigedor, J.P. Selecta, Spain),
through the jacketed extraction vessel. The experimental setup
used to carry out the extraction experiments is shown in Fig. S1.
Both conventional and ultrasound-assisted extractions were
performed and compared at three different temperatures
(20 ± 2 °C, 35 ± 3 °C and 50 ± 4 °C). Conventional extractions (AG
experiments) were conducted with mechanical agitation. Specifi-
cally, a stirrer (RZR 2021, Heidolph, Germany) equipped with a
4-blade propeller (50 mm diameter, 200 rpm), and placed at the
central point of the total extraction volume (X = 2.4 cm from both
the liquid interface and the bottom of the jacketed vessel), was
used. For the extractions with acoustic assistance (US experi-
ments), the mechanical stirring was replaced by an ultrasonic
device consisting of a probe system (UP400S, Hielscher Ultrasound
Technology, Germany) working at 55 ± 5 kHz of acoustic fre-
quency, 435 ± 5 W/L of acoustic power density, 22.9 ± 0.1 W/cm2
 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
of ultrasonic intensity and cycles of 0.5 s. The sonotrode or
ultrasonic emitter, with 3.8 cm2 of emitter surface, was immersed
Y = 1.3 cm into the extraction solution (Fig. S1). Four replicates
were carried out for each temperature and type of extraction.
Extract aliquots (1 mL) were taken at 2.5, 5, 7.5, 10, 12.5, 15, 20,
25, 30, 40, 50 and 60 min, filtered (RC-membrane, 0.45 lm,
Sartorius Stedim Biotech GmbH, Germany) and stored in
eppendorfs at 4 °C until the total phenolic and antioxidant capacity
analyses were performed the same day as the extraction
experiment.
2.3. Characterization of the acoustic field
Prior to the ultrasonic extraction experiments, a calorimetric
approach was used out to measure the effective ultrasonic power
transferred into the medium, in accordance with the procedure
described by González-Centeno et al. [12] in the absence of sample
and without thermostating the system. Briefly, the temperature of
the solvent was logged every second for the first 5 min of ultra-
sound application by using two K-type thermocouple probes con-
nected to a data acquisition equipment HP 34970A Data Logger
(Hewlett-Packard, Spain). The experimental temperature rise,
caused by the dissipation of the acoustic waves, was used to calcu-
late the effective ultrasonic power applied (P) as expressed in Eq.
(1). The ultrasonic power was measured in triplicate for the acous-
tic conditions applied:
dT
PðWÞ ¼ m  C p 
dt
Acoustic power density (W/L) was calculated as the ratio
between the ultrasonic power applied (P) and the total extraction
volume, whereas acoustic intensity (W/cm2) was mathematically
determined as the ratio between the ultrasonic power applied
and the emitter surface of the probe system.
2.4. Determination of total phenolic content and antioxidant capacity
Total phenolic content and antioxidant capacity were the qual-
ity attributes evaluated on the grape pomace extracts. No dilution
was conducted for the measurements. The total phenolic content
(TP) was spectrophotometrically evaluated by means of the Folin
Ciocalteu method in accordance with González-Centeno et al.
[13]. Results, expressed as mg of gallic acid equivalent (GAE) per
100 g of grape pomace sample (on a dry matter basis, dm), were
a mean of six determinations.
As no single method is able to totally assess the antioxidant
capacity of a sample, two different spectrophotometric assays were
used: ABTS and FRAP. The antioxidant capacity of the grape
pomace extracts was determined according to the procedures
previously described by González-Centeno et al. [13]. Results were
expressed as mg of trolox equivalent (TE) per 100 g of grape
pomace sample (dm).
2.5. Mathematical modeling
A mathematical model was proposed with the aim of establish-
ing a methodology to analyze the mass transfer process during the
conventional and ultrasonic-assisted extraction of bioactive com-
pounds from grape pomace. The extraction kinetics of phenolic
compounds and the antioxidant capacity of the extracts, according
to both ABTS and FRAP methods, were mathematically described
by using the Weibull model Eq. (2):
C  C eq
t b
¼ e½ðaÞ 
C 0  C eq
Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
3
The terms a, Ceq and b were the fitting parameters of the model.
In general, the parameter a is interpreted as a kinetic reaction con-
stant, dependent on the temperature according to an Arrhenius
type equation Eq. (3):
Ea
a ¼ a0  eð RT Þ ð3Þ
For each quality attribute of the grape pomace extracts (TP,
ABTS and FRAP), all the experimental data at different extraction
times and temperatures were simultaneously used to carry out
the identification of the extraction kinetic parameters of the
Weibull model. The initial values were easily established by using
linearization methods.
2.6. Statistical analysis
All the experimental results represent the mean value and stan-
dard deviation from at least four replicates. The identification of
the Weibull parameters was carried out by using the ‘nlinfit’ func-
tion of the optimization toolbox of Matlab 7.5 (The MathWorks
Inc., USA) which estimates the coefficients of a nonlinear regres-
sion function and the residuals using least squares. To determine
the 95% confidence intervals and the standard error of the esti-
mated parameters, the ‘nlparci’ function and the covariance matrix
were used, respectively.
The mean relative error (MRE) Eq. (4) and the percentage of
explained variance (VAR) Eq. (5), estimated by the comparison of
experimental and simulated data, were calculated to statistically
ð1Þ evaluate the accuracy of the proposed mathematical models to
simulate the extraction kinetics.
n 
X 
jX exp X calc j
X exp
i¼1
MRE ¼  100 ð4Þ
n
 
Sexp
VAR ¼ 1  100 ð5Þ
Scalc
The normality of the residuals frequency distribution of each
quality attribute of the grape pomace extracts was tested by using
the Shapiro test at the 5% significance level.
3. Results and discussion
3.1. Conventional and acoustic extraction kinetics
Experimental results for the extraction kinetics of the total phe-
nolic content from grape pomace at different temperatures, with
mechanical stirring (20 °C AG, 35 °C AG and 50 °C AG) and with
acoustic assistance (20 °C US, 35 °C US and 50 °C US), are depicted
in Fig. 1A. The corresponding experimental extraction curves for
the antioxidant capacity of the extracts, according to the ABTS
and FRAP assays, are also shown (Fig. 1B and C). Data from both
antioxidant capacity methods revealed a high, significant and posi-
tive correlation (r P 0.97, p < 0.05).
A gradual and significant increase of the phenolic content and
antioxidant capacity of the extracts as the temperature increased,
was observed throughout the extraction process, the highest val-
ues being obtained when the extraction was acoustically assisted.
Regardless of the kind of extraction, this temperature effect on
the extraction efficiency is due to the enhanced solubility of ana-
lytes within the extraction solvent and their improved diffusion
rate from the solid matrix as temperature increases [14,15]. Thus,
the yields of phenolic content and antioxidant capacity of the
extracts obtained from the AG extraction process performed at
ð2Þ 20 °C, were the lowest among all the experimental conditions
tested (p < 0.05). Specifically, after 1 h of extraction at this
4 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx

A 800

20 ºC AG

TP (mg GAE/100 g dm)

600
35 ºC AG
50 ºC AG
400
20 ºC US
35 ºC US
200
50 ºC US

0
0 1000 2000 3000 4000
t (s)

B 800

20 ºC AG
ABTS (mg TE/100 g dm)

600
35 ºC AG

50 ºC AG
400
20 ºC US

35 ºC US
200
50 ºC US

0
0 1000 2000 3000 4000
t (s)

C 800
FRAP (mg TE/100 g dm)

20 ºC AG
600
35 ºC AG
50 ºC AG
400
20 ºC US
35 ºC US
200 50 ºC US

0
0 1000 2000 3000 4000
t (s)

Fig. 1. Experimental extraction curves of grape pomace at different temperatures under agitation (20 °C AG, 35 °C AG, 50 °C AG) and acoustic (20 °C US, 35 °C US, 50 °C US)
conditions, for total phenolic content (A) and antioxidant capacity according to ABTS (B) and FRAP (C) assays.

temperature, total phenolic content was 230.1 ± 24.6 mg GAE/
100 g dm and antioxidant capacity, 278.7 ± 39.6 mg TE/100 g dm
and 180.3 ± 29.4 mg TE/100 g dm, assessed by ABTS and FRAP
methods, respectively.
Meanwhile, the greatest quality attributes were obtained after
1 h of acoustic extraction at 50 °C. Specifically, the total phenolic
content of the extracts was around 3.3-fold higher (770.9 ±
77.5 mg GAE/100 g dm) than in conventional extraction at 20 °C. In
regard to the antioxidant capacity, it was nearly 2.5 (705.9 ±
41.7 mg TE/100 g dm) and 4.0-fold higher (722.4 ± 41.0 mg TE/
100 g dm), for ABTS and FRAP measurements, respectively.
When calculating the ratio between the experimental results
obtained in the acoustic extraction and those resulting from the
conventional process at the same temperature, it was observed
that, regardless of the extraction time considered, ultrasound assis-
tance allowed to obtain extracts with higher total phenolic content
and antioxidant capacity than the AG experiments (ratio > 1.0).
This phenomenon was especially accentuated in the first few min-
utes of the extraction process (Fig. 1). Specifically, for the temper-
ature range from 20 to 50 °C and 2.5 min of extraction time, the
acoustic extraction rate was from 2.3 to 3.0-fold higher than the
conventional one for the total phenolic content, from 2.1 to

Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
500
Confidence intervals (95%)
Prediction limits (95%)
Mathematical adjustment
375
TPcalc (mg GAE/100 g dm)
20 ºC AG
35 ºC AG
50 ºC AG
250
125
0 0
0 125 250
TPexp (mg GAE/100 g dm)
800
Confidence intervals (95%)
Prediction limits (95%)
Mathematical adjustment
TPcalc (mg GAE/100 g dm)
20 ºC US
550
35 ºC US
50 ºC US
300
50
50 300
TPexp (mg GAE/100 g dm)
Fig. 2. Simulated vs. experimental total phenolic content (TP) of the extracts, for
experiments carried out with agitation (A) and with the application of ultrasound
(B) at different temperatures.
2.4-fold higher for the antioxidant capacity evaluated by the ABTS
assay, and from 2.2 to 3.4-fold higher, when analyzed by the FRAP
method.
Although being lower than that observed in the present
research, important improvements of the extraction process by
the application of US have been previously described in the litera-
ture. For example, Corrales et al. [16] reported the enhancing US
effect (35 kHz, with an ultrasonic bath) on the extraction of total
phenolics and on the antioxidant capacity of the extracts from
red grape peels in comparison with the conventional extraction
(with mechanical stirring, no specified conditions), achieving 1.8
and 1.6-fold higher concentrations, respectively. Khan et al. [17]
observed an increase of 30–40% of the total phenolic content and
antioxidant capacity of extracts from orange peels when working
with US (25 kHz, 90 W, with an ultrasonic transducer plate). Virot
et al. [18] also compared ultrasound-assisted extraction (25 kHz,
150 W, with an ultrasonic bath) and conventional extraction (with
mechanical stirring, no specified conditions), reporting 1.2, 1.4 and
Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
5
700
Confidence intervals (95%)
Prediction limits (95%)
Mathematical adjustment
ABTScalc (mg TE/100 g dm)
525
20 ºC AG
35 ºC AG
50 ºC AG
350
175
A A
375 500 0 175 350 525 700
ABTSexp (mg TE/100 g dm)
800
Confidence intervals (95%)
Prediction limits (95%)
Mathematical adjustment
ABTScalc (mg TE/100 g dm)
20 ºC US
550
35 ºC US
50 ºC US
300
B B
50
550 800 50 300 550 800
ABTSexp (mg TE/100 g dm)
Fig. 3. Simulated vs. experimental antioxidant capacity of the extracts (ABTS assay),
for experiments carried out with agitation (A) and with the application of
ultrasound (B) at different temperatures.
1.3-fold greater concentrations for the acoustic experiments, with
regard to the phenolic, proanthocyanidins and flavan-3-ols content
of apple pomace extracts. And, among others, Ahmad-Qasem et al.
[5] described a significant effect of US (260 ± 6 W/L, with an
ultrasonic probe of 3.8 cm2 of emitter surface) on the extraction
rate of total phenolics from olive leaves, with a 1.9-fold higher
value than conventional extraction (with mechanical stirring,
170 rpm).
According to the experimental extraction curves, conventional
extraction of total phenolics at 35 and 50 °C did not differ signifi-
cantly (p > 0.05) from ultrasound-assisted extractions performed
at 20 and 35 °C, respectively (Fig. 1A). These phenomena reveal
that acoustic assistance of the extraction allows working at lower
temperatures than the conventional process without reducing
the extraction rate or even improving it, as previously reported
by other authors [4,7]. According to the literature, Cho et al. [19]
observed that only 1 min of acoustic extraction (47 kHz, 14 W/L,
with an ultrasonic bath) at room temperature (not specified) led
6 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
600
Confidence intervals (95%)
Prediction limits (95%)
Mathematical adjustment
FRAPcalc (mg TE/100 g dm)
450 20 ºC AG
35 ºC AG
50 ºC AG
300
150
0 aqueous conventional extraction of phenolic compounds from
0 150 300
FRAPexp (mg TE/100 g dm)
800
Confidence intervals (95%)
Prediction limits (95%)
Mathematical adjustment
FRAPcalc (mg TE/100 g dm)
600
20 ºC US
35 ºC US
50 ºC US
400
200
0
0 200 400
FRAPexp (mg TE/100 g dm)
Fig. 4. Simulated vs. experimental antioxidant capacity of the extracts (FRAP assay),
for experiments carried out with agitation (A) and with the application of
ultrasound (B) at different temperatures.
to the recovery of the same amounts of resveratrol from grape
stems as 30 min of conventional extraction (with mechanical
stirring, no specified conditions) at 60 °C, by using EtOH:H2O
(80:20, v/v) as extraction solvent.
Furthermore, as observed in Fig. 1, US application also led to a
significant reduction of the extraction time needed to obtain aque-
ous extracts with similar phenolic and antioxidant characteristics.
In particular, the acoustic process required about 3, 4 and 8 times
less time than the conventional process (60 min) to obtain the
same total phenolic recovery at 20 °C (230 mg GAE/100 g dm),
35 °C (360 mg GAE/100 g dm) and 50 °C (475 mg GAE/
100 g dm), respectively. To a lesser extent, a similar behavior was
observed for the antioxidant capacity of the extracts.
In line with the present research, Ye et al. [20] reported that the
ethanolic recovery of corn carotenoids after 20 min of ultrasonic-
assisted extraction (20 kHz, 2500 W/L, no specified equipment)
was higher than that obtained with conventional extraction (with
mechanical stirring, no specified conditions) during 240 min. Khan
Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
et al. [17] observed that the acoustic extraction (25 kHz, 90 W,
with an ultrasonic transducer plate) of phenolics from orange peels
during 10 min, achieved extracts with the same concentration as
the Soxhlet process in 60 min. Meanwhile, Virot et al. [18] obtained
similar phenolic recoveries from apple pomace with acoustic
(25 kHz, 150 W, with an ultrasonic bath) and conventional extrac-
tion (with mechanical stirring, no specified conditions), by using
2.3 times less time at the ultrasonic experiments. Chittapalo
et al. [21] concluded that the application of US (20 kHz,
400–1000 W/L, with an ultrasonic probe) reduced from 120 (con-
ventional extraction) to 10 min, the time required for protein
extraction from rice bran, using distilled water at pH of 11 as
solvent.
3.2. Mathematical modeling
A The Weibull model was used to mathematically describe the
450 600 grape pomace and the antioxidant capacity of the extracts at differ-
ent operating temperatures. This model was also applied to the
acoustic experiments.
By using the experimental results obtained at different extrac-
tion times and considering the initial extraction yield equal to zero
(C0 = 0) in all cases, the parameters a, b and Ceq of the kinetic mod-
els were identified for each tested extraction condition and quality
attribute of the grape pomace extracts.
Once the three parameters were identified for each experiment,
it was confirmed that a followed an Arrhenius-type relationship
with the extraction temperature, as previously proposed in the lit-
erature [22]. It was also observed that, in both conventional and
acoustic experiments, extraction yield in the equilibrium (Ceq) var-
ied linearly with the operating temperature (K). To illustrate this
effect, the proposed equation was as follows Eq. (6), considering
C eq0 and a as constants:
C eq ¼ C eq0 þ aT ð6Þ
On the other hand, the b parameter showed two different
behaviors depending on the type of extraction. When modeling
the extraction curves obtained under the conventional conditions,
B b remained almost constant and independent of the operating tem-
600 800 perature. Thus, in this case, for the extraction kinetics of each qual-
ity attribute of the grape pomace extracts, five parameters were
simultaneously identified (C eq0 , a, a0, Ea, b) by using all the exper-
imental results obtained at different extraction times and temper-
atures (20, 35 and 50 °C).
However, b exhibited a clear dependence on both the extraction
time and the temperature of the process in the acoustic experi-
ments. This behavior was represented by Eqs. (7)–(9), considering
b0a , b0b , b1a and b1b as constants:
b ¼ b0 eðb1 tÞ ð7Þ
b0 ¼ b0a þ b0b T ð8Þ
b1 ¼ b1a þ b1b T ð9Þ
Thus, for the extraction kinetics assisted by ultrasound of each
quality attribute of the grape pomace extracts, a total of eight dif-
ferent parameters of the proposed modified Weibull model were
simultaneously identified (C eq0 , a, a0, Ea, b0a , b0b , b1a , b1b ) also using
the experimental results obtained at different extraction times and
temperatures (20, 35 and 50 °C).
Results obtained in these two identifications are shown in
Table 1, together with the corresponding confidence intervals
and the standard error associated with each estimated parameter.
As observed, some confidence intervals and standard errors were
 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx 7

Table 1
Extraction of total phenolic content from grape pomace and antioxidant capacity of the extracts at different temperatures under agitation or acoustic conditions. Identified
parameters of the proposed modified Weibull model, and the corresponding confidence interval (CI) and standard error (SE) associated with each parameter.

TP ABTS FRAP
Value CI SE Value CI SE Value CI SE
AGa a0 (s) 12.430 [32.732, 16.552 6.653 104 [6.248 103, 2.534 103 1.579 104 [5.874 103, 2.221 103
57.592] 7.578 103] 6.189 103]
Ea (kJ/mol) 12.536 [2.816, 22.256] 3.563 37.634 [9.438, 65.829] 10.334 48.286 [53.372, 149.943] 37.437
C eq0 (mg/ 2228 [2630, 1825] 148 2331 [3615, 1047] 471 2221 [17305, 12863] 5555
100 g dm)
a (K1) 8.658 [6.893, 10.422] 0.647 9.378 [3.515, 15.241] 2.149 10.939 [58.177, 80.055] 25.453
b 0.645 [0.577, 0.712] 0.025 0.623 [0.507, 0.738] 0.042 0.598 [0.482, 0.715] 0.043
USb a0 (s) 584.926 [123.701, 169.855 196.960 [737.290, 340.465 31.616 [113.465, 176.696] 52.871
1046.151] 1.131.210]
Ea (kJ/mol) 4.601 [2.615, 6.588] 0.731 8.536 [3.871, 20.943] 4.521 11.489 [0.2224, 23.202] 4.268
C eq0 (mg/ 3371 [3659, 3083] 106 804 [3227, 4834] 1469 4278 [5.565 ,3.301] 529
100 g dm)
a (K1) 13.971 [12.734, 15.207] 0.455 1.707 [16.970, 20.383] 6.806 17.591 [10.828, 24.354] 2.465
b0 a 3.081 [2.631, 3.532] 0.165 4.173 [3.108, 5.238] 0.388 3.529 [2.317, 4.742] 0.442
b0b (103) (K1) 8.679 [10.640, 0.720 12.230 [17.137, 7.322] 1.788 10.230 [15.657, 4.802] 1.978
6.718]
b1a (103) (s1) 2.082 [0.544, 3.620] 0.565 1.208 [3.204, 0.788] 0.727 1.525 [0.744, 3.793] 0.827
b1b (106) (K1) 5.758 [10.815, 1.856 5.104 [1.222, 11.430] 2.305 3.695 [10.961, 3.571] 2.648
0.701]
a
Experiments with mechanical stirring.
b
Experiments with ultrasound assistance.

high as a result of the simultaneous identification of a high number
of parameters. In any case, simulations obtained by disregarding
some of them were not satisfactory; therefore, it was decided to
keep all of them in the models.
According to the results shown in Table 1, a values ranged, for
temperatures between 20 and 50 °C, from 35.6 to 22.1 min for con-
ventional extraction of total phenolics; and from 56.7 to 13.5 min,
and from 1066.7 to 169.4 min for antioxidant capacity of the
extracts obtained in AG experiments, according to ABTS and FRAP
measurements, respectively. On the other hand, in US-assisted
extractions, a values varied between 64.5 and 54.1 min, 109.1
and 78.8 min, and 58.9 and 38.0 min, for total phenolic content
and antioxidant capacity according to ABTS and FRAP analysis,
respectively. As models used to simulate conventional and acoustic
extractions were not equivalent, because of the temperature and
time dependence of the b parameter in the second ones, a values
could not be considered comparable between both types of
extraction.
In any case, a decreased with increased operating temperature
regardless of the type of extraction and the quality attribute con-
sidered, as previously observed in extraction modeling of saponins
from quinoa seeds [23] and b-carotene from rose hip [24], by using
the Weibull model within the temperature ranges of 20–60 °C and
25–55 °C, respectively.
This behavior of the a parameter may be justified by taking into
account the following two premises: the a parameter is related to
the inverse of the extraction rate constant [24], and the extraction
rate is directly proportional to the temperature [25]. Thus, from a
higher extraction rate (higher temperature) the lower a parameter
is expected. According to the experimental results, both
conventional and acoustic extractions occurred faster at 50 °C. This
increase in the extraction rate with the temperature is in agree-
ment with the findings of Bucic-Kojic et al. [26] for conventional
phenolic extraction from grape seeds within a temperature range
from 20 to 80 °C; Sant’Anna et al. [27,28] for conventional
extraction of phenolics and anthocyanins from red grape pomace
between 25 and 50 °C; and Sturzoiu et al. [24] for conventional
b-carotene extraction from rose hip at temperatures from 25 to
45 °C. Nevertheless, it is important to point out that temperature
only has a positive effect when it is not too high, as most of
Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
bioactive compounds are susceptible to degradation at higher
temperatures, reducing the observed extraction rate [29].
In any case, the experimental a values presented the same order
of magnitude as those observed by Sant’Anna et al. [27,28] for
phenolic and anthocyanin extraction from red grape pomace, also
simulated at different temperatures (25–60 °C) by the Weibull
model.
The a parameter was considered as a kinetic constant depen-
dent on the temperature according to the Arrhenius equation Eq.
(3). This approach allows estimation of the activation energy (Ea)
for the extraction processes, which may be described as the energy
barrier that bioactive compounds need to cross in order to be
removed from the grape pomace matrix. The estimated Ea, respec-
tively, for the conventional and US-assisted extractions were of
12.5 and 4.6 kJ/mol for total phenolic content, 37.6 and 8.5 kJ/
mol for antioxidant capacity according to the ABTS method and,
48.3 and 11.5 kJ/mol for antioxidant capacity according to the FRAP
method. As expected, acoustic assistance significantly reduced the
Ea of the total phenolics and antioxidant capacity extraction kinet-
ics. Similar behavior was observed by Chemat et al. [30] and Khan
et al. [17], for extraction of carvona terpenoid from caraway seeds
and flavanones from orange peels, respectively.
According to Wang et al. [31], if the activation energy is greater
than 40 kJ/mol, the extraction is solubilization reaction controlled,
whereas if the activation energy is lower than 20 kJ/mol, the
extraction is managed by diffusion. For activation energies
between 20 and 40 kJ/mol the extraction is governed by a mixed
regime of both reaction and diffusion. According to this classifica-
tion, the estimated Ea values suggest a controlling diffusion regime
for the conventional extraction of total phenolics, as well as for the
acoustic extraction of all the three quality attributes in the study,
within the temperature range 20–50 °C.
The Ea values may depend on different factors in the extraction
process, such as the food matrix, the target biocompounds, the pre-
treatment of the sample and the solvent used, among others [27].
In any case, the experimental values obtained in the present study
are in line with those described in the literature for the extraction
kinetics of different bioactive compounds from products and/or by-
products of plant origin. Bucic-Kojic et al. [26] proposed activation
energies from 1.1 to 7.7 kJ/mol for the solid–liquid extraction of
8 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx

phenolic compounds from white grape seeds, at different particle
sizes and temperatures between 25 and 80 °C. Sant’Anna et al.
[28] and Bucic-Kojic et al. [1] reported activation energies of 23.0
and 24.2 kJ/mol, respectively, for total phenolics extraction from
red grape seeds, within the temperature range of 25–60/80 °C.
Khan et al. [17] and Ye et al. [20] noted activation energies of
34.2 and 44.6 kJ/mol, respectively, for the solid–liquid extraction
of flavanones from orange peels (10–40 °C) and for the acoustic
extraction of corn carotenoids (28–38 °C).
In the Weibull model Eq. (2), b is a shape parameter which rep-
resents a behavior index of the raw material during the extraction
process [24]. When this parameter is equal to 1, the model corre-
sponds to a first order kinetic, with a constant extraction rate
[22]. Nevertheless, when b takes values below and above 1, this
parameter denotes the convexity (decreasing extraction rate over
time) or concavity (increasing extraction rate over time) of the
extraction curve, respectively [32]. The shape of the extraction
kinetics is much more pronounced as b values move away from
the unit.
In the case of the AG experiments, b remained constant, since
no significant differences (p > 0.05) were detected over the temper-
ature range 20–50 °C for any of the attributes studied. The esti-
mated b values were 0.645, 0.623 and 0.598 for total phenolics
and antioxidant capacity according to ABTS and FRAP assays,
respectively. The fact that all these three b values identified for
AG experiments were lower than 1 means that the extraction rate
decreased over time for the temperature range considered. These
experimental results were similar to those reported for saponin
extraction from quinoa seeds (b = 0.206–0.305) [23] and for b-car-
otene recovery from rose hip (b = 0.424–0.644) [24]. Furthermore,
the observed behavior of the extraction rate has been previously
published in the literature for different extraction systems
[23,24,26–28].
The only experimental difference between conventional and
acoustic extractions was the replacement of mechanical stirring
by a power ultrasound system. Thus, the different behavior of b
ought to be directly attributed to the ultrasonic effect. According
to the Eqs. (7)–(9), it can be seen that b depended on the extraction
time and the operating temperature in the acoustic experiments.
Since the b parameter is a characteristic of the raw material itself
and power ultrasound are able to modify its structure and proper-
ties, it is understandable that b might vary during the extraction
process.
The estimated b values for acoustic experiments ranged,
depending on time and extraction temperature, according to the
intervals plotted in Fig. S2. As observed, b changed significantly
over time and this variation was greater as the operating temper-
ature decreased. Thus, this phenomenon suggests that US effect
was more pronounced at lower temperatures.
Furthermore, during the extraction process, b gradually chan-
ged from lower to greater values than the unit for each operating
temperature. When comparing the b evolution (Fig. S2) with the
experimental extraction curves (Fig. 1), it is observed that the con-
vexity of the extraction curves began to acquire a slightly concave
In line with the present research, Odriozola-Serrano et al. [32]
also reported an exponential dependence for the b parameter of
the Weibull model, in particular, on the pulsed electric field
applied to tomato juice to avoid phenolic and antioxidant degrada-
tion during storage.
3.3. Simulation of the extraction curves
Experimental extraction curves of total phenolic content and
antioxidant capacity from grape pomace were simulated by using
the identified parameters from the modified Weibull model
(Table 1), for both acoustic and conventional extraction experiments
at the three temperatures in study.
Simulated versus experimental values are depicted in Figs. 2–4
for the total phenolic content and the antioxidant capacity of the
extracts according to ABTS and FRAP assays, respectively.
Mathematical adjustment of the modified Weibull model, and
the corresponding confidence intervals (p < 0.05) and prediction
limits (p < 0.05) were also included in these figures. In all cases, the
y-intercepts of the mathematical adjustments did not differ signif-
icantly from zero (p > 0.05) and slopes presented values close to
the unit (0.989 6 x 6 0.999).
Regardless of the experimental variability, mainly attributed to
the heterogeneity of the raw material and the error associated with
the analytical determinations, a high correlation coefficient
between the proposed model and the experimental data was
observed in all cases (r2 P 0.992).
Furthermore, to statistically evaluate the accuracy of the
proposed mathematical models and, therefore, their capacity to
simulate the experimental results and to predict variation within
the system, mean relative error (MRE, Eq. (4)) and percentage of
explained variance (VAR, Eq. (5)) were calculated for all the extrac-
tion conditions tested by comparing the experimental and calcu-
lated values. According to the literature, the lower the MRE and
the closer the VAR to 100%, the better the simulation fit provided
by the model [33].
As observed in Table 2, MRE 6 7.0% and VAR P 97.9% in all
cases. From the average values of both statistical parameters, it
can be stated that the proposed models successfully simulated
the extraction of phenolics and antioxidant capacity from grape
pomace for all temperatures in study, under both ultrasonic
conditions and mechanical stirring.
The quality of the model fits and regression analyses were fur-
ther assessed by randomness and normality of the residuals
inspection. Fig. S3 shows the residuals vs. simulated values in a
normalized form, for both total phenolic content and antioxidant
capacity results at different temperatures and with/without
Table 2
Statistical parameters (mean relative error, MRE, and percentage of explained
variance, VAR) estimated by comparison of experimental and simulated extraction
kinetics of total phenolic content and antioxidant capacity from grape pomace at
different temperatures, without/with US assistance.

shape at the moment that b changed from values <1 to values >1. In T (°C) TP ABTS FRAP
accordance with this, the extraction rate, which had been decreas- MRE (%) VAR (%) MRE (%) VAR (%) MRE (%) VAR (%)
ing over time until that moment, started to increase over time. It is
AGa 20 2.4 99.6 7.2 97.5 8.1 97.2
important to point out that as b increased at the first stage, the 35 1.4 99.9 5.0 98.3 8.7 97.6
extraction rate gradually declined in a lesser extent up to the 50 2.5 99.5 2.6 99.2 4.4 99.0
moment that b was equal to the unit; at this moment, the extrac- Average 2.1 ± 0.6 99.7 ± 0.2 4.9 ± 2.3 98.3 ± 0.9 7.0 ± 2.3 97.9 ± 1.0
tion rate started to increase with the further rise of b. This phe- USb 20 5.1 98.6 4.6 98.5 6.9 98.0
nomenon may be explained by two countered effects: the 35 3.8 98.9 2.7 99.0 4.2 98.3
decreasing availability of bioactive compounds and the increasing 50 1.7 99.4 1.6 99.4 2.2 99.2
Average 3.5 ± 1.7 99.0 ± 0.4 3.0 ± 1.5 99.0 ± 0.5 4.4 ± 2.4 98.5 ± 0.6
mass transfer provided by power ultrasound as a consequence of
the mechanical impact on the material to which it was being a
Experiments with mechanical stirring.
b
applied. Experiments with ultrasound assistance.

Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
 M.R. González-Centeno et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
acoustic assistance. Data normalization was performed by dividing
residuals by the predicted final concentration (t = 1 h) at each tem-
perature. The residuals appeared randomly scattered close around
zero (0.10 < x < 0.10) and Shapiro test confirmed the normal dis-
tribution of the residuals (p > 0.05) for the three temperatures and
both types of extraction in study.
In accordance with the high agreement between the experi-
mental data and the values simulated by the proposed models
(Table 2 and Figs. 2–4), and the randomness and normality of the
residuals values (Fig. S3), it may be concluded that the proposed
modified Weibull model is suitable to accurately predict the
extraction kinetics of phenolics and antioxidant capacity from
grape pomace within the range of temperatures 20–50 °C, for both
conventional and acoustic experiments.
4. Conclusions
On the whole, the experimental data discussed above clearly
suggest that the application of power ultrasound improves the effi-
ciency and/or the yield of the extraction process when compared to
the conventional solid–liquid procedure. As expected, the increas-
ing temperature had a positive effect on the extraction of the total
phenolic content and antioxidant capacity from grape pomace by-
products, the highest values reported in the case of the extraction
assisted acoustically. As observed, the acoustic process required
less extraction time and lower operating temperatures to obtain
extracts with similar phenolic and antioxidant characteristics than
those resulting from conventional extraction.
The proposed modified Weibull model satisfactorily described
both the conventional and acoustic aqueous extraction kinetics of
total phenolic content and antioxidant capacity (average
MRE 6 7.0% and average VAR P 97.9%) within the temperature
range from 20 to 50 °C. It is important to point out that the b
parameter of the modified Weibull model showed that in the case
of the ultrasonic-assisted extraction, the process is not only faster
but also accelerates over time.
Acknowledgements
The authors acknowledge the financial support of the Spanish
Government (MICINN) and European Regional Development Fund
(FEDER) for the research projects DPI2012-37466-C03-02 and
AGL 2012-34627, and the Balearic Government (57/2011). The
authors also thank Ca’n Majoral S.L. winery for supplying the grape
pomace samples.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ultsonch.2014.05.
027.
References
[1] A. Bucic-Kojic, H. Sovova, M. Planinic, S. Tomas, Temperature-dependent
kinetics of grape seed phenolic compounds extraction: experiment and model,
Food Chem. 136 (2013) 1136–1140.
[2] M. Pinelo, J. Sineiro, M.J. Núñez, Mass transfer during continuous solid-liquid
extraction of antioxidants from grape byproducts, J. Food Eng. 77 (2006) 57–
63.
[3] M. Cissé, P. Bohuon, F. Sambe, C. Kane, M. Sakho, M. Dornier, Aqueous
extraction of anthocyanins from Hibiscus sabdariffa: Experimental kinetics and
modeling, J. Food Eng. 109 (2012) 16–21.
[4] S.R. Shirsath, S.H. Sonawane, P.R. Gogate, Intensification of extraction of
natural products using ultrasonic irradiations. A review of current status,
Chem. Eng. Process 53 (2012) 10–23.
Please cite this article in press as: M.R. González-Centeno et al., Effect of power ultrasound application on aqueous extraction of phenolic compounds and
antioxidant capacity from grape pomace (Vitis vinifera L.): Experimental kinetics and modeling, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.05.027
9
[5] M.H. Ahmad-Qasem, J. Cánovas, E. Barrajón-Catalán, V. Micol, J.A. Cárcel, J.V.
García-Pérez, Kinetic and compositional study of phenolic extraction from
olive leaves (var. Serrana) by using power ultrasound, Innovative. Food Sci.
Emerg. Technol. 17 (2013) 120–129.
[6] M.D. Esclapez, J.V. García-Pérez, A. Mulet, J.A. Cárcel, Ultrasound-assisted
extraction of natural products, Food Eng. Rev. 3 (2011) 108–120.
[7] T.J. Mason, F. Chemat, M. Vinatoru, The extraction of natural products using
ultrasound or microwaves, Curr. Org. Chem. 15 (2011) 237–247.
[8] K. Radha Krishnan, M. Sivarajan, S. Babuskin, G. Archana, P. Azhagu Saravana
Babu, M. Sukumar, Kinetic modeling of spice extraction from S. aromaticum
and C. Cassia, J. Food Eng. 117 (2013) 326–332.
[9] S. Simal, M.C. Garau, A. Femenia, C. Rosselló, A diffusional model with a
moisture-dependent diffusion coefficient, Drying Technol. 24 (2006) 1365–
1372.
[10] M. Blasco, J.V. García-Pérez, J. Bon, J.E. Carreres, A. Mulet, Effect of blanching
and air flow rate on turmeric drying, Food Sci. Technol. Int. 12 (2006) 315–323.
[11] AOAC Official methods of analysis of AOAC International, Gaithersburg
(EE.UU.): AOAC International, 1997.
[12] M.R. González-Centeno, K. Knoerzer, H. Sabarez, S. Simal, A. Femenia, C.
Rosselló, Effect of acoustic frequency and power density on the aqueous
ultrasonic-assisted extraction of grape pomace (Vitis vinifera L.) – A response
surface approach, Ultrason. Sonochem. (2014), http://dx.doi.org/10.1016/
j.ultsonch.2014.01.021.
[13] M.R. González-Centeno, M. Jourdes, A. Femenia, S. Simal, C. Rosselló, P.-L.
Teissedre, Proanthocyanidin composition and antioxidant potential of the
stem winemaking byproducts from 10 different grape varieties (Vitis vinifera
L.), J. Agric. Food Chem. 60 (2012) 11850–11858.
[14] K. Ghafoor, Y.H. Choi, J.Y. Jeon, I.H. Jo, Optimization of ultrasound-assisted
extraction of phenolic compounds, antioxidants, and anthocyanins from grape
(Vitis vinifera) seeds, J. Agric. Food Chem. 57 (2009) 4988–4994.
[15] R. González-Montelongo, M.G. Lobo, M. González, The effect of extraction
temperature, time and number of steps on the antioxidant capacity of
methanolic banana peel extracts, Sep. Purif. Technol. 71 (2010) 347–355.
[16] M. Corrales, S. Toepfl, P. Butz, D. Knorr, B. Tauscher, Extraction of anthocyanins
from grape by-products assisted by ultrasonics, high hydrostatic pressure or
pulsed electric fields: a comparison, Innovative Food Sci. Emerg. Technol. 9
(2008) 85–91.
[17] M.K. Khan, M. Abert-Vian, A.-S. Fabiano-Tixier, O. Dangles, F. Chemat,
Ultrasound-assisted extraction of polyphenols (flavanone glycosides) from
orange (Citrus sinensis L.) peel, Food Chem. 119 (2010) 851–858.
[18] M. Virot, V. Tomao, C. Le Bourvellec, C.M.C.G. Renard, F. Chemat, Towards the
industrial production of antioxidants from food processing by-products with
ultrasound-assisted extraction, Ultrason. Sonochem. 17 (2010) 1066–1074.
[19] Y.-J. Cho, J.-Y. Hong, H.S. Chun, S.K. Lee, H.-Y. Min, Ultrasonication-assisted
extraction of resveratrol from grapes, J. Food Eng. 77 (2006) 725–730.
[20] J. Ye, L. Feng, J. Xiong, Y. Xiong, Ultrasound-assisted extraction of corn
carotenoids in ethanol, Int. J. Food Sci. Technol. 46 (2011) 2131–2136.
[21] T. Chittapalo, A. Noomhorm, Ultrasonic assisted alkali extraction of protein
from defatted rice bran and properties of the protein concentrates, Int. J. Food
Sci. Technol. 44 (2009) 1843–1849.
[22] V. Eim, D. Urrea, C. Rosselló, J.V. García-Pérez, A. Femenia, S. Simal,
Optimization of the drying process of carrot (Daucus carota v. Nantes) on the
basis of quality criteria, Drying Technol. 31 (2013) 951–962.
[23] I. Quispe-Fuentes, A. Vega-Gálvez, M. Miranda, R. Lemus-Moncada, M. Lozano,
K. Ah-Hen, A kinetic approach to saponin extraction during washing of quinoa
(Chenopodium quinoa willd.) seeds, J. Food Process Eng. 36 (2013) 202–210.
[24] A. Sturzoiu, M. Stroescu, A.S. Guzun, T. Dobre, Empirical models applied for
kinetics extraction of beta-carotene from Rosa canina, Rev. Chim. 62 (2011)
344–348.
[25] J.E. Cacace, G. Mazza, Mass transfer process during extraction of phenolic
compounds from milled berries, J. Food Eng. 59 (2003) 379–389.
[26] A. Bucic-Kojic, M. Planinic, S. Tomas, M. Bilic, D. Velic, Study of solid-liquid
extraction kinetics of total polyphenols from grape seeds, J. Food Eng. 81
(2007) 236–242.
[27] V. Sant’Anna, A. Brandelli, L.D. Ferreira Marczak, I.C. Tessaro, Kinetic modeling
of total polyphenol extraction from grape marc and characterization of the
extracts, Sep. Purif. Technol. 100 (2012) 82–87.
[28] V. Sant’Anna, L. Damasceno, F. Marczak, I.C. Tessaro, Kinetic modeling of
anthocyanin extraction from grape marc, Food Bioprocess Technol. 6 (2013)
3473–3480.
[29] M. Hojnik, M. Skerget, Z. Knez, Extraction of lutein from Marigold flower petals
– experimental kinetics and modelling, LWT – Food Sci. Technol. 41 (2008)
2008–2016.
[30] S. Chemat, A. Lagha, H. AitAmar, P.V. Bartels, F. Chemat, Comparison of
conventional and ultrasound-assisted extraction of carvone and limonene
from caraway seeds, Flavour Fragrance J. 19 (2004) 188–195.
[31] Y.G. Wang, S.T. Yue, D.Q. Li, M.J. Jin, C.Z. Li, Kinetics and mechanism of Y(III)
extraction with CA-100 using a constant interfacial cell with laminar flow,
Solvent Extr. Ion Exch. 20 (2002) 345–358.
[32] I. Odriozola-Serrano, R. Soliva-Fortuny, V. Gimeno-Ano, O. Martín-Belloso,
Modeling changes in health-related compounds of tomato juice treated by
high-intensity pulsed electric fields, J. Food Eng. 89 (2008) 210–216.
[33] D. Bas, I.H. Boyaci, Modeling and optimization I: usability of response surface
methodology, J. Food Eng. 78 (2007) 836–845.