J. Korean Soc. Appl. Biol. Chem.

52(3), 295-300 (2009) Article

Optimization of Ultrasound Assisted Extraction of Phenolic Compounds

and Antioxidants from Grape Peel through Response Surface Methodology
Kashif Ghafoor and Yong Hee Choi*

Department of Food Science and Technology, Kyungpook National University, Daegu 702-701, Republic of Korea
Received March 26, 2009; Accepted April 8, 2009

Functional components from Campbell Early grape peel were extracted by ultrasound-assisted
extraction technology. The experiments were carried out according to a five level, three variable
central composite rotatable design. The best possible combinations of extraction variables were
obtained for the maximum phenolic compounds and antioxidant activities of grape peel extracts
using response surface methodology. The optimal conditions include 53.14% ethanol, 46.03 C o

temperature and 24.03 min time for the maximum total phenolic compounds (6.70 mg GAE/
100 mL) and 53.06% ethanol, 50.65 C temperature and 25.58 min time for the maximum
o

antioxidant activity (555.90 mg/L). Under these conditions, the experimental total phenolics were
6.67 mg GAE/100 mL and antioxidant activity was 554.84 mg/L of the grape peel extract, which is
well matched with the predicted values.
Key words: antioxidant activity, grape peel, response surface methodology, total phenols, ultrasound
assisted extraction

Grapes (Vitis vinifera) are among the most widely al., 1997].
consumed fruits and the demand for grapes and grape Extraction is a very important stage in the isolation,
products is increasing because of the associated health identification and use of phenolic compounds [Lapornik
benefits [Ghafoor et al., 2008]. Grapes are rich in et al., 2005]. The recovery of these components is
phenolic compounds [Sanchez-Alonso et al., 2008] with commonly performed through a solvent-extraction
approximately 75% of grape polyphenols existing in the procedure and the concentration of solvent, time and
skin and seeds. Grape skin phenols may be classified as temperature are important parameters to be optimized for
cell-wall phenols, which are bound to polysaccharides by maximum recovery of the targeted compounds [Spigno et
hydrophobic interactions and hydrogen bonds, non-cell- al., 2007]. Response surface methodology (RSM) has
wall phenols, encompassing phenols confined in the been applied in the industrial processes for optimization
vacuoles of plant cells, and phenols associated with the of response of interest [Wang et al., 2008]. Ultrasound-
cell nucleus [Pinelo et al., 2006]. The positive physiological assisted extraction is an inexpensive, simple and efficient
effects associated with the consumption of grape and alternative to conventional extraction techniques. The
grape derivatives are currently believed to be mainly due enhancement in extraction obtained by using ultrasound
to the antiradical and antioxidant properties of the is mainly attributed to the effect of acoustic cavitations
occurring phenolic species [Lurton, 2003]. Phenolic produced in the solvent by the passage of an ultrasound
compounds can be used in different therapeutic wave [Wang et al., 2008]. Ultrasound also offers a
procedures with the purpose of free radical neutralization mechanical effect allowing greater penetration of solvent
in biological systems [Yilmaz and Toledo, 2004] and into the sample matrix, increasing the contact surface area
oxidation of human low-density lipoproteins [Meyer et between the solid and liquid phase, as a result the solute
quickly diffuses from the solid phase to the solvent
*Corresponding author [Rostagno et al., 2003]. It has been reported that application
Phone: +82-53-950-5777; Fax: +82-53-950-6772 of ultrasound-assisted extraction in grape peel can
E-mail: yhechoi@knu.ac.kr enhance the recovery of functional compounds up to 30%
as compared to conventional solvent extraction [Cho et
Abbreviations: GAE, gallic acid equivalent; RSM, response sur- al., 2006]. In addition the use of ultrasound-assisted
face methodology; SD, standard deviation
extraction also prevents the possible chemical degradation
doi:10.3839/jksabc.2009.052 of targeted compounds [Wang and Weller, 2006].
296 Kashif Ghafoor and Yong Hee Choi

In this study, ultrasound-assisted extraction parameters data of response variables such as total phenols ( ) and Y1

such as the solvent concentration, extraction temperature antioxidant activity ( ) obtained by triplicate observations
Y2

and extraction time were optimized using RSM, by as effected by the extraction conditions and was fitted into
employing a five level, three variable central composite an empiric second order polynomial model as shown in
rotatable design, in order to obtain the optimal conditions the following equation:
for the extraction of functional components from the peel 2 2 2

of Campbell Early grapes. Yn=b0+b1X1+b2X2+b3X3+b11X1 +b22X2 +b33X3 +b12X1X2

+b13X1X3+b23 X2X3 (1)

Materials and Methods where is the response variable, , and correspond
Yn X1 X2 X3

to the independent variables namely ethanol concentration,

Materials. Grapes were purchased from a local farm in extraction temperature and extraction time, respectively.
Kyungbuk, Province of Korea and the grape cultivar was The values represent corresponding regression coefficients.
bn

identified as ‘Campbell Early’. All the chemicals used Analysis for total phenolic compounds. The total
were of analytical grade and they were purchased either phenolic compounds were analyzed using the Folin
from Sigma Chemical Co. (St. Louis, MO) or Duksan Ciocalteu method with some modification [Singleton and
Pure Chemical Co. (Ansan, Korea). Rossi, 1965]. A 200 µL properly diluted sample or a
Ultrasonic-assisted extraction from grape peel. standard solution of varying concentrations were mixed
Grapes were excised from the stems and washed. Grapes with 400 µL Folin Ciocalteu reagent. The deionized
were manually cut into halves and the grape peels were water was used for dilution and control. The solution was
separated with a knife. Grape peel was oven dried at 50 Co
diluted to a total volume of 4.6 mL using deionized water
until the moisture level was constant. Dried grape peel then thoroughly mixed. After incubation for 10 min at
was ground to a powdered form using an electrical room temperature, 1 mL of 20% Na CO solution was
2 3

grinder. A sample of 2 g powdered grape peel was kept in added then immediately mixed and incubated for 2 h. The
a glass flask and the volume was made up to100 mL with absorbance was read at 765 nm on a spectrophotometer
the extraction solvent. Contents were dissolved by using a (TU-1800; Human Corporation, Seoul, Korea).
magnetic stirrer (KMC 130SH; Vision Scientific Co., Measurements were recorded in triplicates. Gallic acid of
Ltd., Daegu, Korea) for 5 min. Ultrasonic assisted 1 mg/mL was used as the standard and the total phenolic
extraction was performed in a sonication water bath (JAC compounds of the samples were expressed in milligram
Ultrasonic 2010P; Jinwoo Engineering Co., Ltd., gallic acid equivalent (GAE) per 100 mL (mg GAE/100
Hwasung, Korea) with a useful volume of 10 L. The mL).
working frequency was fixed at 40 KHz and the Determination of antioxidant activity. The antioxidant
temperature and time of extraction was controlled from activity of the grape peel extracts was evaluated by the
the panel. After extraction the flask was cooled to room phosphomolybdenum complex method [Prieto ., et al

temperature using cool water. The extract was filtered 1999]. In brief, 0.4 mL of sample solution (100 mL of
through filter paper No 5A under vacuum and the grape seed extract dissolved in 1 mL of methanol) was
solution was collected in a volumetric flask. It was then combined with 4 mL of reagent solution containing 0.6
used for the determination of total phenolics compounds, M sulphuric aicd, 2 mM sodium phosphate and 4 mM
antioxidants and anthocyanin contents. All the measurements ammonium molybdate. The blank solution contained 4
were carried out in triplicates and the data reported were mL of reagent solution and the 1 mL of methanol. Test
means±SD. tubes were caped and placed in hot water for 90 min at
Experimental design. A five level, three variable 95 C. After samples were cooled to room temperature
o

central composite rotatable design [Cochran and Cox, absorbance was measured at 695 nm against a blank.
1992] was applied to determine the best combination of Antioxidant activity was expressed relative to that of
extraction variables for the extraction of total phenolic ascorbic acid.
compounds and antioxidants from grape peel. Three Statistical analysis. All the analysis was carried out in
independent variables selected for this study were the triplicates and the experimental results obtained were
concentration of solvent, the extraction temperature and expressed as means±SD. The responses obtained from
the extraction time. The factorial design consisted of eight the experimental design set were subjected to multiple
factorial points, six axial points (two axial points on the nonlinear regression analysis to obtain the coefficients of
axis of each design variable at a distance of 1.68 from the the second polynomial model. The quality of the fit of
design center) and four center points leading to 18 sets of polynomial model was expressed by the coefficient of
experiments. Regression analysis was performed on the determination , and its statistical significance was
R
2
 Ultrasonic Assisted Extraction from Grape Peel 297

checked using an F-test. The optimal extraction Table 2. Regression coefficients and analysis of the
conditions were estimated through three dimensional model for three response variables
response surface analyses of the three independent Coefficients estimated
variables and each dependent variable. Statistical analysis Coefficient
Total phenols Antioxidants
was performed by using the Statistical Analysis System
(SAS, version 9.1). Data was analyzed by analysis of b0 -3.143 -309.130
variance and the mean values were considered significantly b1 0.144b 14.447b
different at p<0.05. b2 0.113b 7.774c
b3 0.278b 22.225b
b11 -0.001b -0.120b
Results and Discussion
b22 -0.001b -0.098b
Modeling of the extraction process from grape peel. b33 -0.006a -0.538a
In order to optimize the extraction process with reference b12 -0.000 -0.022
to the extraction of phenolic and antioxidant components b13 -0.001 -0.023
b23 0.001 0.129
from grape peel under sonication a central composite Probability of F
design was developed as represented in Table 1. Table 1 value <0.001 <0.001
also presents the experimental values of total phenols and Probability of lack
antioxidant activities of grape peel extracts at various of fit 0.06 0.09
experimental conditions. The results of analysis of ‘a’ means p<0.001, ‘b’ means p<0.01, ‘c’ means p<0.05
variance, goodness of fit and the adequacy of the models
are summarized in Table 2. The data showed a good fit
with the Eq. (1) which were statistically acceptable at Effect of process variables on the total phenolics
p <0.05 and adequate with satisfactory R2 values. The full compounds. Solid-liquid extraction is a mass transport
model filled Eq. (1) was used for three dimensional plots phenomenon in which solids contained in a matrix
to predict the relationships between independent variables migrate into solvent brought into contact with the matrix.
and the dependent variables. This mass transport phenomenon can be enhanced with

Table 1. Experimental design of five-level, three-variable central composite design and total phenols and antioxidant
activities of ultrasonic assisted grape peel extracts
Extraction conditions and Analytical resultsa
Test set X1, Ethanol X2, Extraction X3, Extraction time Total phenols Antioxidant activity
concentration (%) temperature (C) (Min) (mg GAE/100 mL) (mg/L)
1 40 (-1) 30 (-1) 15 (-1) 5.74 455.23
2 40 (-1) 30 (-1) 25 (+1) 6.12 493.80
3 40 (-1) 50 (+1) 15 (-1) 5.98 476.49
4 40 (-1) 50 (+1) 25 (+1) 6.47 526.99
5 60 (+1) 30 (-1) 15 (-1) 6.10 484.47
6 60 (+1) 30 (-1) 25 (+1) 6.29 504.56
7 60 (+1) 50 (+1) 15 (-1) 6.09 483.03
8 60 (+1) 50 (+1) 25 (+1) 6.55 542.86
9 33 (-1.68) 40 (0) 20 (0) 5.95 475.23
10 67 (+1.68) 40 (0) 20 (0) 6.49 527.71
11 50 (0) 23 (-1.68) 20 (0) 5.88 476.49
12 50 (0) 57 (-1.68) 20 (0) 6.46 539.37
13 50 (0) 40 (0) 11 (-1.68) 5.51 444.28
14 50 (0) 40 (0) 29 (-1.68) 6.56 540.26
15 50 (0) 40 (0) 20 (0) 6.51 531.83
16 50 (0) 40 (0) 20 (0) 6.53 527.71
17 50 (0) 40 (0) 20 (0) 6.58 533.98
18 50 (0) 40 (0) 20 (0) 6.52 539.55
a
Analytical results are means±SD (n=3).
298 Kashif Ghafoor and Yong Hee Choi

changes in diffusion coefficients induced by ultrasounds

and extraction temperature [Corrales ., 2009].
et al

Solvent concentration and extraction time also play a

significant role in extraction of phenolic compounds from
plant materials [Wang ., 2008]. The use of ultra-
et al

sonication process was due to the fact that ultrasonic

waves break the cells of the vegetal matrix and the cells’
contents are released into the extraction medium
[Vinatoru ., 1997]. In our experiments, the total
et al

phenols of grape peel extracts obtained by ultrasound-

assisted extraction based on the central composite design
are shown in Table 1. Multiple regression analysis was
performed on the experimental data and the coefficients
of model were evaluated for significance. The effect of
extraction time was highly significant ( <0.001) on the
p

extraction of phenolics and it was consistent with the

findings of Revilla . [1998] who obtained higher
et al

yields of phenolics from grape skins when extraction was

done for a longer time. The values of the coefficients as
presented in Table 2 were used for final predictive
equation neglecting the non-significant cross terms as
given below:
1=−3.143+0.144 1+0.113 2+0.278 3−0.001 1
2
Y X X X X

−0.001 2 -0.006 3
2
X
2
X (2)
To determine the optimal levels of variables for the
ultrasound-assisted extraction of total phenols from grape
peel, three-dimensional surface plots (Fig. 1) were
constructed according to the Eq. (2). Extraction process
variables significantly effected ( <0.05) the extraction of
p

total phenols from peel of Campbell Early grapes. Fig.

1A shows the effect of ethanol concentration and extraction
time on the content of the total phenolic compounds. The
total phenolic contents increased slowly with the increase
of ethanol concentration at a fixed extraction temperature Fig. 1. The response surface plots of the total phenolic
contents of grape peel extract as affected by ethanol
and nearly reached a peak at the highest ethanol concentration, temperature, and extraction time in
concentration tested. Similarly, the increase in extraction ultrasonic-assisted extraction. Where (A) is ethanol
temperature at a fixed ethanol concentration led to a concentration and temperature (time 20 min); (B) is
gradual increase in the total phenolic content, and reached ethanol concentration and time (temperature 40oC); (C) is
a maximum at the highest extraction temperature tested. temperature and time (ethanol concentration 50%).
The plot of total phenols as affected by ethanol
concentration and extraction time (Fig. 1B) demonstrates extraction of antioxidant components from grape peel at
a marked increase in phenolic contents with the increase various extraction conditions are presented in Table 1.
of ethanol concentration at a fixed extraction time while The highest contents of antioxidants (542.86 mg/L) were
an increase in extraction time at a fixed ethanol observed in experimental run 8 with 60% ethanol
concentration also led to marked increase in total phenol concentration, 50 C extraction temperature and 25 min
o

contents. A similar linear increase in total phenolic extraction time. The lowest yield of antioxidants (444.28
contents with the increase of extraction temperature at a mg/L) was observed in experimental run 13. Statistical
fixed extraction time, while an obvious quadratic effect of analysis revealed that most relevant variable with
extraction time were both observed (Fig. 1C). p<0.001 was extraction time. The extraction time has this
Effect of process variables on the antioxidant kind of effect on the antioxidant activity because
activity. The mean experimental data showing the increasing the contact time of the solvent with solids may
 Ultrasonic Assisted Extraction from Grape Peel 299

As presented in the three dimensional plots for antioxidant

contents (Fig. 2), the extraction process variables effected
the extraction of antioxidants in a similar way as in the
case of total phenolic compounds. This is due to the fact
that antioxidant activities of grapes are closely associated
with the phenolic compounds [Lu and Foo, 2001].
Antioxidant activity increased with the increase of
ethanol concentration at a fixed temperature and also
increased significantly ( <0.05) with the increase of
p

extraction temperature at a fixed ethanol concentration

(Fig. 2A). Similarly antioxidant value increased with the
increase of ethanol concentration at a fixed time and it
rapidly increased with the increase of extraction time at a
fixed ethanol concentration as represented in Fig. 2B. Fig.
2C shows a linear increase in antioxidant activity with the
increase of extraction temperature at fixed time and a
similar increase in antioxidant activity with the increase
of extraction time at a constant extraction temperature.
Optimization. The estimated levels of optimum
extraction conditions for maximum response of total
phenols and antioxidant activities of grape peel extracts
obtained by ultrasonic-assisted extracts are summarized
in Table 3. The predicted ultrasonic-assisted extraction
conditions were 53.14% ethanol concentration, 46.03 C o

extraction temperature and 24.03 min extraction time for

the maximum total phenols (6.70 mg GAE/100 mL) and
53.06% ethanol, 50.65 C temperature and 25.58 min time
o

for maximum antioxidant activity (555.90 mg/L) The 2, R

adjusted 2 values of the model are 0.959, 0.936 and

R

0.948, 0.920 for total phenols and antioxidants respectively

which represents that the model had adequately represented
the real relation between the parameters chosen. To
compare the predicted results with experimental values,
experimental rechecking was performed for each
Fig. 2. The response surface plots of the antioxidant response using the optimum extraction conditions. Mean
activity of grape peel extract as affected by ethanol values of 6.67 mg GAE/100 mL total phenols and 554.84
concentration, temperature and extraction time in
ultrasonic assisted extraction. Where (A) is ethanol mg/L antioxidants obtained from real experiment validated
concentration and temperature (time 20 min); (B) is the RSM model. The good correlation between these
ethanol concentration and time (temperature 40oC); (C) is results confirmed that the response model was adequate
temperature and time (ethanol concentration 50%). in reflecting the expected optimization (Table 3).
This study indicated that the Campbell Early grape peel
improve the diffusion of the compounds (Corrales et al., is a good source of phenolic compounds and the use of
2009). The results of multiple regression analysis showed ultrasonication for the extraction of total phenols and
that the antioxidant contents of grape peel were antioxidants was an effective method because it could
significantly ( <0.05) affected by the linear and quadratic
p greatly decrease the extraction time compared with other
terms of ethanol concentration, extraction temperature extraction methods. Lapornik et al. [2005] found that
and extraction time (Table 2). The final predictive increasing the time from 12 to 24 h during conventional
equation for antioxidant activity of grape peel extract by solvent extraction was found to have statistically significant
using significant terms is as follows: effects on the yields of total phenols and antioxidant
activities from red grape pressed marc. The efficiency of
2= −309.130+14.447 1+7.774 2+22.225 3−0.120 1
2
Y

−0.098 2 −0.538 3
2 2
X X X X

(3) ultrasonication could be explained by the fact that

X X
sonication simultaneously enhanced the hydration and
300 Kashif Ghafoor and Yong Hee Choi

Table 3. Estimated optimum conditions, predicted and experimental values of responses under these conditions
Optimum extraction conditions Maximum value
Response variables R2 R2-adjust F-value p-value
Ethanol (%) Temp ( C) Time (min) Estimated Experimentala
o

Total phenols
(mg GAE/100 mL) 0.959 0.936 20.85 0.0001 53.14 46.03 24.03 6.70 006.67±0.03
Antioxidant activities 0.948 0.920 20.94 0.0001 53.06 50.65 25.58 555.90 554.84±0.56
(mg/L)
a
Means±standard deviation (n=3).

fragmentation process while facilitating mass transfer of Prieto P, Pineda M, and Aguilar M (1999) Spectrophotomet-
solutes to the extraction solvent (Toma ., 2001).
at al ric quantitation of antioxidant capacity through the for-
mation of a phosphomolybdenum complex: specific
Acknowledgments. This study was supported through application to the determination of vitamin E. Anal Bio-
Grape Research Projects’ Group by Agriculture R&D chem 269, 337-341.
Promotion Center, Ministry of Agriculture and Forestry, Revilla E, Ryan JM, and Martin OG (1998) Comparison of
South Korea. several procedures used for the extraction of anthocya-
nins from red grapes. J Agric Food Chem 46, 4592-
4597.
References Rostagno MA, Palma M, and Barroso CG (2003) Ultra-
sound-assisted extraction of soy isoflavones. J Chro-
Cho YJ, Hong JY, Chun, HS, Lee SK, and Min HY (2006) matogr A 1012, 119-128.
Ultrasonication-assisted extraction of resveratrol from Sánchez-Alonso I, Jiménez-Escrig A, Saura-Calixto F, and
grapes. J Food Eng 77, 725-30. Borderías AJ (2008) Antioxidant protection of white
Cochran WG and Cox GM (1992) Some methods for the grape pomace on restructured fish products during fro-
study of response surfaces. In Experimental Designs, pp. zen storage. Food Sci Technol-Leb 41, 42-50.
335-375. Wiley, New York, NY, U.S.A. Singleton VL and Rossi JJ (1965) Colorimetry of total phe-
Corrales M, Garcia AF, Butz P, and Tauscher B (2009) nolics with phosphomolybdic phosphotungstic acid
Extraction of anthocyanins from grape skins assisted by reagents. Am J Enol Vitic 16, 144-158.
high hydrostatic pressure. J Food Eng 90, 415-421. Spigno G, Tramelli L, and De-Faveri DM (2007) Effects of
Ghafoor K, Jung JE, and Choi YH (2008) Effects of gel- extraction time, temperature and solvent on concentra-
lan, xanthan, and ë-carrageenan on ellagic acid sedimen- tion and antioxidant activity of grape marc phenolics. J
tation, viscosity, and turbidity of ‘Campbell Early’ grape Food Eng 81, 200-208.
juice. Food Sci Biotechnol 17, 80-84. Toma M, Vinatoru M, Paniwnyk L, and Masom TJ (2001)
Lapornik B, Prosek M, and Wondra AG (2005) Compari- Investigation of the effects of ultrasound on vegetal tis-
son of extracts prepared from plant by-products using sues during solvent extraction. Ultrason Sonochem 8,
different solvents and extraction time. J Food Eng 71, 137-142.
214-222. Vinatoru M, Toma M, Radu O, Filip PI, Lazurca D, and
Lu Y and Foo YL (2001) Antioxidant activities of polyphe- Mason TJ (1997) The use of ultrasound for the extrac-
nols from sage (Solvia officinalis). Food Chem 75, 197- tion of bioactive principles from plant materials. Ultra-
202. son Sonochem 4, 135-139.
Lurton L (2003) Grape polyphenols: New powerful health Wang J, Sun B, Cao Y, Tian Y, and Li X (2008) Optimisa-
ingredients. Innovations Food Technol 18, 28-30. tion of ultrasound-assisted extraction of phenolic com-
Meyer AS, Yi OS, Pearson, DA, Waterhouse AL, and pounds from wheat bran. Food Chem 106, 804-810.
Frankel EN (1997) Inhibition of low-density lipoprotein Wang L and Weller CL (2006) Recent advances in extrac-
oxidation in relation to composition of phenolic antioxi- tion of nutraceuticals from plants. Trends Food Sci Tech-
dants in grapes (Vitis vinifera). J Agric Food Chem 45, nol 17, 300-312.
1638-1643. Yilmaz Y and Toledo RT (2004) Health aspects of func-
Pinelo M, Arnous A, and Meyer AS (2006) Upgrading of tional grape seed constituents. Trends Food Sci Technol
grape skins: Significance of plant cell-wall structural 15, 422-433.
components and extraction techniques for phenol release.
Trends Food Sci Technol 17, 579-590.