Food Chemistry 217 (2017) 552559

Contents lists available at ScienceDirect

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

Ultrasound-assisted extraction of natural antioxidants from the flower

of Limonium sinuatum: Optimization and comparison with conventional
methods
Dong-Ping Xu a, Jie Zheng a, Yue Zhou a, Ya Li a, Sha Li b, Hua-Bin Li a,c,
a
Guangdong Provincial Key Laboratory of Food, Nutrition and Health, School of Public Health, Sun Yat-sen University, Guangzhou 510080, China
b
School of Chinese Medicine, The University of Hong Kong, Hong Kong, China
c
South China Sea Bioresource Exploitation and Utilization Collaborative Innovation Center, Sun Yat-sen University, Guangzhou 510006, China

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

Article history: Natural antioxidants are widely used as dietary supplements or food additives. An optimized method of
Received 23 March 2016 ultrasound-assisted extraction (UAE) was proposed for the effective extraction of antioxidants from the
Received in revised form 23 August 2016 flowers of Limonium sinuatum and evaluated by response surface methodology. In this study, ethanol
Accepted 3 September 2016
concentration, ratio of solvent to solid, ultrasonication time and temperature were investigated and opti-
Available online 6 September 2016
mized using a central composite rotatable design. The optimum extraction conditions were as follows:
ethanol concentration, 60%; ratio of solvent to solid, 56.9:1 mL/g; ultrasonication time, 9.8 min; and tem-
Keywords:
perature, 40 C. Under the optimal UAE conditions, the experimental values (483.01 15.39 lmol Trolox/
Limonium sinuatum
Flower
g DW) matched with those predicted (494.13 lmol Trolox/g DW) within a 95% confidence level. In addi-
Ultrasound-assisted extraction tion, the antioxidant activities of UAE were compared with those of conventional maceration and Soxhlet
Antioxidant extraction methods, and the ultrasound-assisted extraction could give higher yield of antioxidants and
markedly reduce the extraction time.
2016 Elsevier Ltd. All rights reserved.

1. Introduction reducing compounds (Frankel & Finley, 2008; Shahidi &

Excessive oxidative stress is implicated in the development of
several chronic diseases, such as diabetes, cardiovascular diseases,
cancer and ageing diseases (Kaur & Kapoor, 2001; Kumar & Singh,
2015; Young & Woodside, 2001). Many risk factors, such as cigar-
ette smoking, irradiation and environmental pollutants, may
induce excessive oxidative stress in human body, and the intake
of antioxidants is regarded as a preventative method to reduce
the negative effects induced by oxidative stress (Boots, Haenen, &
Bast, 2008; Li et al., 2015; Roleira et al., 2015). Antioxidants could
delay or inhibit the oxidation of biological macromolecules (such
as DNA, protein and lipid) in cells and tissues by inhibiting the ini-
tiation or propagation of oxidative chain reactions (Das, Das, Miyaji
& Deka, 2016; Young & Woodside, 2001). In addition to the capac-
ity of scavenging free radicals in the body, the natural antioxidants
could have multiple activities in food, including reducing
hydroperoxides into stable hydroxyl derivatives, inactivating metal
catalysts by chelation, and interacting synergistically with other
Corresponding author at: Guangdong Provincial Key Laboratory of Food,
Nutrition and Health, School of Public Health, Sun Yat-sen University, Guangzhou
510080, China.
E-mail address: lihuabin@mail.sysu.edu.cn (H.-B. Li).
Ambigaipalan, 2015). Fruits, vegetables, cereals, spices and herbs
are the main sources of natural antioxidants and many of them
have displayed strong free radical scavenging abilities and anti-
inflammatory activities. (Deng et al., 2012; Embuscado, 2015;
Hung, 2016; Kaur & Kapoor, 2001; Li et al., 2013; Shahidi &
Ambigaipalan, 2015). With the increased application of natural
antioxidants in medicine and food processing, the effective extrac-
tion of natural antioxidants from plant sources is required.
Limonium sinuatum belongs to the family Plumbaginaceae, and
has a worldwide distribution (Chen, Funnell, & Morgan, 2010;
Igawa, Hoshino, & Mii, 2002). Historically, the flower of Limonium
sinuatum was often used as a form of tea for the prevention of age-
ing. In previous studies, the flower of Limonium sinuatum has been
found to have strong antioxidant capacity in comparison with 51
edible flowers, and therefore could potentially be a rich resource
of natural antioxidants (Li et al., 2014). The antioxidants from the
flower of Limonium sinuatum could be used as dietary supplement
or food additive. Thus, maximum extraction of natural antioxidants
from the flower of Limonium sinuatum was investigated.
Extraction of antioxidant compounds from plants mainly utilise
conventional solvent and Soxhlet methods, which have long
extraction times, high usage of organic solvents and poor extrac-
tion efficiency. Recently, more efficient techniques have been

http://dx.doi.org/10.1016/j.foodchem.2016.09.013
0308-8146/ 2016 Elsevier Ltd. All rights reserved.
 D.-P. Xu et al. / Food Chemistry 217 (2017) 552559
developed to extract antioxidant compounds from plant materials,
including microwave-assisted extraction (Dahmoune, Nayak,
Moussi, Remini, & Madani, 2015), ultrasound-assisted extraction
(UAE) (Liu et al., 2015; Zou et al., 2014), pressurized liquid extrac-
tion (Povilaitis, Sulniute, Venskutonis, & Kraujaliene, 2015), and
supercritical fluid extraction (Diaz-Reinoso, Moure, Dominguez, &
Parajo, 2006). Among them, several UAE methods have been devel-
oped for the extraction of natural antioxidants from plant materi-
als (Bimakr et al., 2012; Liu et al., 2008; Peng et al., 2013; Shirsath,
Sonawane, & Gogate, 2012; Xu et al., 2016). UAE has the capacity to
improve the extraction efficiency by promoting the mass transfer
and possible rupture of cell wall due to its effect of acoustic cavita-
tion (Shirsath et al., 2012). The efficiency of an extraction method
is affected by many process parameters, e.g. solvent composition,
the ratio of solvent to solid, ultrasonication time and temperature
(Esclapez, Garcia-Perez, Mulet, & Carcel, 2011). Therefore, the opti-
mization of the extraction parameters is required to obtain the
maximum amount of antioxidants from plant materials. As a math-
ematical and statistical technique, response surface methodology
(RSM) is an effective tool for optimizing the process parameters
(Bimakr et al., 2012). The application of RSM may enable reduction
of the number of experimental trials and quantify the interactions
between multiple parameters.
In this paper, effects of extraction process parameters (ethanol
concentration, the ratio of solvent to solid, ultrasonication time
and temperature) were evaluated in order to optimise extraction
of antioxidants from Limonium sinuatum flower and analyzed using
RSM with a five-level, three-variable central composite rotatable
design (CCRD). Furthermore, the antioxidant activities of the
flower extracts using UAE were compared with conventional mac-
eration and Soxhlet extraction methods. Besides, the polyphenolic
compounds of extracts obtained under optimized conditions were
quantified using high-performance liquid chromatography-
photodiode array detector (HPLC-PAD).
2. Materials and methods
2.1. Plant materials
The flowers of Limonium sinuatum were bought from different
markets in Guangzhou, China. The flowers were dried to 1.8%
residual moisture utilizing freeze-drying technology and reduced
to a fine particle (<150 lm) in an electric grinder (RHP-100;
Ronghao Insustry & Trade Co., Ltd, Yongkang, China) for the subse-
quent experiments.
2.2. Reagents
The standards of ABTS (2,20 -azinobis(3-ethylbenothiazoline-
6-sulphonic acid) diammonium salt), Trolox (6-hydroxy-2,5,7,8-
tetramethylchroman-2-carboxylic acid) and polyphenolic standards
(epigallocatechin, cyanidin, chlorogenic acid, coffeic acid, p-coumaric
acid, rutin, resveratrol, 2-hydrocinnamic acid, quercetin, gallic acid,
ferulic acid) were purchased from SigmaAldrich (St. Louis, MO,
USA). Potassium persulfate was purchased from Tianjin Chemical
Factory (Tianjin, China). Deionized water was used. Ethanol of
analytical grade were used in the part of extraction, and solvents
(formic acid and methanol) of HPLC grade were used in the part of
HPLC analysis.
2.3. Extraction of antioxidant compounds
2.3.1. Ultrasound-assisted extraction
Ultrasound-assisted extraction was carried out in a domestic
ultrasonic bath device (Kj1012B; Guangzhou Kejin Ultrasonic
Instrument Factory, Guangzhou, China) with a constant power of
553
400 W, at a frequency of 40 kHz. The apparatus was equipped with
a digital control system for sonication time and temperature.
A ground particle sample (0.1 g) was placed in a capped plastic
tube (15 mL, round bottom) and mixed with an appropriate vol-
ume of ethanol solution. The tube with suspension was sonicated
at different time periods and temperatures in the ultrasonic device
containing 7 L water. All the samples in each extraction were
placed in the centre of ultrasonic device with the depth of six cen-
timeters and subsequent each batch of samples to be sonicated
was kept the same position. Besides, all the samples of each exper-
imental design were sonicated simultaneously in the ultrasonic
device. Each sample was done in triplicate. According to the differ-
ent levels of each experimental design, namely each extraction
parameter in single factor experiments and response surface
experiment, the number of tubes used in each extraction varied
from 18 to 60. After UAE, the mixtures were centrifuged and the
supernatants were obtained for the subsequent antioxidant activ-
ity determination.
2.3.2. Maceration extraction
A ground particle sample (0.1 g) was mixed with 60% ethanol
solution (5.69 mL). The extraction was performed in a shaking
water bath at 25 C for 24 h according to the method displayed
by Xu et al. (2016). Then, the extract was centrifuged and the
supernatants were obtained for the subsequent antioxidant
activity determination.
2.3.3. Soxhlet extraction
A ground particle sample (2.0 g) was weighed, and filtered using
Whatman filter paper. The extraction was carried out with 400 mL
of 60% ethanol at 95 C for 4 h, refluxing in a Soxhlet apparatus
according to the method displayed by Xu et al. (2016). Then, the
extract was collected for the subsequent assays.
2.4. Determination of antioxidant capacity
The assay of ABTS radical-scavenging capacity was performed
according to the method displayed by Jing, Dong, and Tong
(2015) previously. 7 mM ABTS solution was prepared and added
to the same volume of 2.45 mM potassium persulfate solution
(v/v), then, incubated for 16 h at room temperature in the dark.
The stock solution was diluted to make the absorbance of ABTS+
working solution be 0.710 0.05 at 734 nm before usage. For the
spectrophotometric assay, 3.8 mL of the ABTS+ working solution
and 100 lL of the diluted sample were mixed and the absorbance
was measured at 734 nm after 6 min of incubation. The results of
ABTS radical-scavenging assay were expressed as lmol Trolox/
g dry weight (DW).
2.5. HPLC analysis
Polyphenolics of extracts obtained under optimized conditions
were analyzed on the basis of the method described previously
with minor modification (Cai, Luo, Sun, & Corke, 2004). HPLC-
PAD equipped with a Waters (Milford, MA, USA) 1525 binary HPLC
pump and a Waters 2996 photodiode array detector was applied.
Separation was carried out using an Agilent Zorbax Extend-C18
column (250  4.6 mm, 5 lm) at 40 C with a gradient elution at
a flow rate of 0.8 mL/min. The mobile phase composed of solution
A (0.1% formic acidwater solution) and solution B (methanol),
which were delivered as follows: 0 min, 95% (A); 15 min,
80% (A); 20 min, 70% (A); 25 min, 63% (A); 40 min, 60% (A); and
60 min, 50% (A). The UV spectra were monitored between 200
and 600 nm. In this part, 11 polyphenolic standards (epigallocate-
chin, cyanidin, chlorogenic acid, coffeic acid, p-coumaric acid,
rutin, resveratrol, 2-hydrocinnamic acid, quercetin, gallic acid,
554 D.-P. Xu et al. / Food Chemistry 217 (2017) 552559

and ferulic acid) were used to identify polyphenolic compounds of
the extracts. These compounds were quantified by the peak area
under the maximum absorption wavelength. The polyphenolic
contents were expressed as mg/100 g dry weight of Limonium sin-
uatum flower.
2.6. Experimental design
2.6.1. Single-factor experiments
In this part, the effects of the following parameters were inves-
tigated: ethanol concentration (1090%), the ratio of solvent to
solid (10:170:1 mL/g), temperature (3080 C) and ultrasonica-
tion time (030 min). According to the single-factor experimental
results, major influence factors and their levels were obtained
and applied in the RSM design.
2.6.2. Response surface methodology experiments
Using a central composite rotatable design (CCRD), response
surface methodology (RSM) was conducted to maximize the
antioxidant activities of the extracts of Limonium sinuatum flower.
Based on above single-factor results, three major independent
variables, ethanol concentration (X1,%), the ratio of solvent to solid
(X2, mL/g), and ultrasonic time(X3, min), were selected. The inde-
pendent variables and their related codes and levels are displayed
in Table 1. The design matrix for a central composite rotatable
design for 3 variables (each one evaluated at 5 levels) involves
20 design experiments (Table 2). Six replicates were used to eval-
uate the pure error. After performing 20 different experiments, a
quadratic model was fitted to the response value using Design-
expert software (version 7.1.3). The variation of total antioxidant
activity (Y) versus the three retained variables X1, X2 and X3, were
displayed using a second-order polynomial model given by the
following equation:
X X 2
X effect on antioxidant activities of the extract, while other extrac-
Y b0 bi Xi bii Xi bij Xi Xj
where Y is the predicted value; bo is a constant; bi, bii and bij are the
linear, quadratic and interactive regression coefficients of the
model, respectively; Xi and Xj are the independent variables. Based
on the analysis of variance (ANOVA), the regression coefficients of
individual linear, interaction and quadratic terms were determined.
Numerical magnitude of the standardized model coefficients
reflected their importance in the model obtained. Among standard-
ized coefficients, the larger values are more effective. In addition to
the quadratic model, the interactive effects of the factors could also
be displayed in 3-D surface plots, which enable visualization of the
relationships between the variables in the plot and the response.
2.7. Statistical analysis
All experiments were performed in triplicate and the results
were expressed as mean value standard deviation. The statistical
analysis of model was carried out using Design Expert (version
7.1.3) and Excel 2007.
3. Results and discussion
3.1. Single factor experiment analysis
3.1.1. Effect of ethanol concentration
Most natural antioxidants from plants, like phenolic com-
pounds, flavonoids and anthocyanins, are easily dissolved by
low-polar organic solvents such as methanol, ethanol, and acetone,
with different levels of water (Carciochi, Manrique, & Dimitrov,
2015; Chavan & Singhal, 2013; Yang et al., 2010; Zou, Wang,
Gan, & Ling, 2011). Due to its low toxicity and economic accessibil-
ity, aqueous ethanol has been commonly employed to extract
antioxidants from plants. In this study, ethanol concentrations
ranging from 10% to 90% (v/v) were selected to investigate the
1
tion conditions remained constant with the ratio of solvent to solid

Table 1
Five levels of the independent variables of the extraction process.

Variable Units Symbol Code levels

1.68 1 0 1 1.68
Ethanol concentration % (v/v) X1 6.4 20 40 60 73.6
Solvent/material ratio mL/g X2 6.4 20 40 60 73.6
Ultrasound irradiation time min X3 1.6 5 10 15 18.4

Table 2
Response surface design and experimental results.

Run X1 (Concentration X2 (Solvent/material X3 (Ultrasound irradiation Y (TEAC value, lmol

of ethanol, %) ratio, mL/g) time, min) Trolox/g DW)
1 60 60 5 494.38
2 73.6 40 10 457.46
3 40 40 10 449.16
4 60 20 5 387.33
5 40 40 1.6 394.35
6 20 60 15 364.85
7 40 73.6 10 435.32
8 20 20 15 241.19
9 40 40 18.4 425.91
10 40 6.4 10 237.74
11 20 20 5 191.36
12 40 40 10 425.91
13 40 40 10 425.91
14 40 40 10 415.94
15 60 20 15 391.48
16 40 40 10 389.37
17 6.4 40 10 266.48
18 20 60 5 349.90
19 40 40 10 442.51
20 60 60 15 501.85
 D.-P. Xu et al. / Food Chemistry 217 (2017) 552559 555

500 500

400 400

TEAC Value (mol Trolox/g DW)

TEAC Value (mol Trolox/g DW)

300 300

200 200

100 100

0 0
10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70
Ethanol concentration (%) Ratio of s olvent to solid (mL/g)

(A) (B)
500 500

400 400
TEAC Value (mol Trolox/g DW)

TEAC Value (mol Trolox/g DW)

300 300

200 200

100 100

0 0
0 5 10 15 20 30 30 40 50 60 70 80
Ultrasonic time (min) Temperature ( oC)

(C) (D)
Fig. 1. Effects of extraction parameters on the antioxidant activities of flower extracts: (A) ethanol concentration; (B) ratio of solvent to solid; (C) ultrasonic time; and (D)
temperature.

of 100 mL/g, ultrasonication time of 30 min, and temperature at 10, 15, 20, 30 min) were employed, and three other extraction
30 C. As shown in Fig. 1A, the antioxidant activities were conditions were fixed as follows: ethanol concentration, 40%
increased from 267.12 13.18 lmol Trolox/g DW to 439.00 5.43 (v/v); the ratio of solvent to solid, 40 mL/g; and temperature,
lmol Trolox/g DW when the ethanol concentration changed from 30 C. As shown in Fig 1C, the increase of sonication time from 0
10% to 40%. However, a downward trend was observed when the to 10 min led to an enhancement of the extraction yield (from
ethanol concentration increased from 50% (431.94 12.10 lmol 325.16 7.49 lmol Trolox/g DW to 409.30 3.32 lmol Trolox/g DW),
Trolox/g DW) to 90% (361.77 16.48 lmol Trolox/g DW). Accord- then followed by a decrease after 15 min. This was an indication
ing to the principle of similarity and intermiscibility, when that extending the extraction time beyond 15 min was unneces-
polarities of solvent and solute are similar, the solute is easily sary and also could result in degradation of antioxidants in the
dissolved from plant cells (Yang et al., 2010; Zhang et al., 2007). flower extracts. Therefore, 10 min was selected for the following
When the ethanol concentration reached 40%, the polarity of experiments.
extraction solvent was more close to the polarity of antioxidants,
and the antioxidant activities achieved a maximum value. Thus, 3.1.4. Effect of temperature
40% ethanol was used in the subsequent experiments. With the aim of maximizing the extraction efficiencies, the dif-
ferent temperature (3080 C) were also evaluated with three
3.1.2. Effect of the ratio of solvent to solid other fixed factors: ethanol concentration, 40% (v/v); ultrasonic
The effects of the ratio of solvent to solid on the antioxidant time, 10 min; and the ratio of solvent to solid, 40 mL/g. As shown
activities of the extract of Limonium sinuatum flowers were in Fig 1D, the antioxidant activities of the extract of Limonium
investigated at seven levels ranging from 10:1 to 70:1 mL/g, with sinuatum flowers increased as the temperature increased from
other conditions kept as follow: ethanol concentration, 40% (v/v); 30 C to 40 C. This was because higher temperature may decrease
ultrasonic time, 30 min; and temperature, 30 C. As can be seen solvent viscosity, give rise to higher molecular motion and increase
in Fig. 1B, the antioxidant activities (from 228.44 4.39 to the solubility (Wang, Geng, Li, Hu, & Li, 2013). However, when
370.00 13.32 lmol Trolox/g DW) enhanced with the increase of the temperature was higher than 40 C, a decline was observed.
the ratio of solvent to solid (from 10:1 to 40:1 mL/g). When the The reason was that higher temperature could lead to degradation
ratio of solvent to solid was higher than 40:1 mL/g, the antioxidant of antioxidants (Rostagno, Palma, & Barroso, 2007). Therefore,
activities were almost unchanged. A possible explanation is that 40 C was selected as the optimal temperature.
with the increase of the ratio of solvent to solid, the contact area
between material and solvent is also increased enhancing the solu- 3.2. Analysis of response surface methodology
bility of antioxidant components within the plants cells (Yang et al.,
2010). When the ratio of solvent to solid reached to 40:1 (mL/g), the 3.2.1. Fitting the model
dissolution process reached its equilibrium. Thus, 40:1 mL/g was The response of ABTS radical-scavenging capacities for extracts
used for further optimization of extraction conditions. of Limonium sinuatum flower using UAE were optimized based on
the central composite rotatable design, and the data obtained from
3.1.3. Effect of ultrasonication time the CCRD were fitted to a second order polynomial equation. The
To study the effects of ultrasonic time on the extraction yield of significance of the coefficients of the models was determined by
antioxidants from Limonium sinuatum flowers, different time (0, 5, analysis of variance (ANOVA) and presented in Table 3. The results
556 D.-P. Xu et al. / Food Chemistry 217 (2017) 552559

Table 3
ANOVA for response surface quadratic model.

Source Sum of squares df Mean square F value p Value Significant

Model 1.385E+005 9 15384.93 38.15 <0.0001 Significant
X1 (ethanol concentration) 65934.34 1 65934.34 163.50 <0.0001
X2 (ratio of solvent to material) 50677.46 1 50677.46 125.67 <0.0001
X3 (ultrasound irradiation time) 1227.17 1 1227.17 3.04 0.1117
X1X2 524.33 1 524.33 1.30 0.2808
X1X3 353.01 1 353.01 0.88 0.3715
X2X3 124.44 1 124.44 0.31 0.5908
X21 7233.30 1 7233.30 17.94 0.0017
X22 14205.98 1 14205.98 35.23 0.0001
X23 416.60 1 416.60 1.03 0.3334
Residual 4032.67 10 403.27
Lack of fit 1789.58 5 357.92 0.80 0.5949 Not significant
Pure error 2243.09 5 448.62
Cor total 1.425E+005 19
R-squared 0.972
Adj R-squared 0.946

showed that the linear coefficients of ethanol concentration and
the ratio of solvent to solid were significant (p < 0.001) on the
antioxidant activity, whereas, no significant effect of ultrasonica-
tion time was found on the response. Furthermore, ethanol con-
centration and the ratio of solvent to solid have a significant
quadratic effects (p < 0.001) for the response (negative effect on
the total antioxidant activity) and all the interaction parameters
were insignificant (p > 0.05). Removing the non-significant
variables, the final predictive equation (Eq. (2)) is given as follows:
Y 166:37 9:43X1 10:53X2  0:06X21  0:08X22
The validity of the model was confirmed using ANOVA and
summarized in Table 3. The analysis showed that the model was
significant at F-value of 38.15 (p < 0.0001). The Lack of Fit
F-value of 0.80 implies that the lack-of-fit was not significant
relative to pure error (p = 0.60) verifying the validity of the model.
The determination coefficient (R2) was 0.972, which implied that
97.2% of the variations could be explained by the fitted model.
The adjusted determination coefficient (R2Adj) was 0.946, which
was comparable to R2, indicating a high degree of correlation
between the observed and predicted values. Furthermore, a low
value of the coefficient of variation (C.V. = 5.22%) indicated that the
variation in the mean value was low and the experimental values
with a high degree of precision were reliable. All these results
revealed that the model had adequately represented the real
relationship between the response and independent variables.
3.2.2. Effect of independent variables on antioxidant activities in the
RSM model
To illustrate the interactive effects of the independent variables
on the antioxidant activities of the extract of Limonium sinuatum
flowers, three dimensional response surface and contour plots
were displayed in Fig. 2AC. The plots were generated by plotting
the response (TEAC value) using the z-axis against two indepen-
dent variables while keeping the other independent variable at
the fixed level. Fig. 2A shows the response surface plot for the
interactive effect of ethanol concentration and the ratio of solvent
to solid on the response value at a fixed ultrasonication time. An
increase of ethanol concentration or the ratio of solvent to solid
both led to an rapid enhancement of antioxidant activities and
the antioxidant activities could reach a peak value with 60% etha-
nol concentration and 60 mL/g the ratio of solvent to solid. Fig. 2B
shows the response surface plot for the interactive effect of etha-
nol concentration and ultrasonic time on the respone value at a
fixed ratio of solvent to solid. An increase of ethanol concentration
from 20% to 60% improved the antioxidant activities markably,
while the antioxidant activities increased slightly with the
increase of ultrasonic time from 5 to 15 min. Fig. 2C shows the
response surface plot for the interactive effect of the ratio of sol-
vent to solid and ultrasonication time on the response value at a
fixed ethanol concentration. The antioxidant activities increased
markedly with the increase of the ratio of solvent to solid and
reached a peak value with 60% ethanol concentration, while the
extension of ultrasonic time almost did not change the antioxi-
dant activities. The above results of response surface analysis
were in accordance with the ANOVA (Table 3), which showed that
2 the quadratic and linear effects of the two variables were signifi-
cant (p < 0.01). These results indicated that the ethanol concentra-
tion and the ratio of solvent to solid were the major factors
affecting the responses at the level of p < 0.01.
3.2.3. Validation of the predicted value
The suitability of the model equation for predicting the
response values was tested at the optimal ultrasonic conditions:
ethanol concentration, 60%; the ratio of solvent to solid,
56.9:1 mL/g; temperature, 40 C; and ultrasonic time, 9.8 min.
Under the optimal extraction conditions, the experimental value
(483.01 15.39 lmol Trolox/g) was obtained, which was well
matched with the predicted value (494.13 lmol Trolox/g)
(p > 0.05). The good correlation between the experimental and
predicted values confirmed that the model obtained by CCRD could
accurately predict the antioxidants extraction yield of Limonium
sinuatum flowers using ultrasound-assisted extraction.
3.3. Comparison of ultrasound-assisted extraction with maceration
and Soxhlet extraction methods
To estimate and validate the efficiency of ultrasound on antiox-
idants extraction from Limonium sinuatum flowers, a comparison
was made between UAE method and two conventional extraction
techniques (maceration and Soxhlet extraction) for three samples
from different sources. The results in Table 4 showed that the
antioxidant activities obtained using UAE were significantly
higher than those using Soxhlet extraction. This enhancement
could be attributed to cavitation and thermal effects of the ultra-
sound technique, which cause disruption of the cell wall and
intensification of mass transfer (Esclapez et al., 2011; Vajic
et al., 2015). Besides, the antioxidants from Limonium sinuatum
flower may be more easily destroyed under the high temperature
of Soxhlet extraction. In addition, although the antioxidant
activity of UAE was similar to that of maceration extraction,
UAE (9.8 min) could greatly shorten the extraction time compared
 D.-P. Xu et al. / Food Chemistry 217 (2017) 552559 557

Design-Expert?Software
TEAC Value
TEAC Value 108.00
250
Design points above predicted value 300
Design points below predicted value
501.854
350
191.364 92.00

X1 = A: Ethanol concentration

B: Solvent/material ratio
X2 = B: Solvent/material ratio 500

Actual Factor
C: Time = 10.00 76.00
432.5
T E A C V a lu e

400

365
60.00
450

297.5

44.00
230 6
60.00 350

50.00 300
60.00 28.00
50.00 40.00 20.00 28.00 36.00 44.00 52.00 60.00

30.00 A: Ethanol concentration

40.00
30.00

B: Solvent/material ratio 20.00 20.00 A: Ethanol concentration

(A)
Design-Expert?Software

TEAC Value
Design points above predicted value
TEAC Value
12.03
Design points below predicted value
501.854

191.364
510 8.03
X1 = C: Time
X2 = A: Ethanol concentration

Actual Factor 467.5

T E A C V a lu e

C : T im e
B: Solvent/material ratio = 60.00
4.03

425
350 300
400

382.5 0.03

450
340

-3.97

60.00 400
50.00 15.00
12.50
40.00 -7.97
10.00
A: Ethanol concentration 30.00 7.50
52.67 68.67 84.67 100.67 116.67 132.67

20.00 5.00 C: Time

(B) A: Ethanol concentration

Design-Expert?Software
TEAC Value
TEAC Value 16.07
Design points above predicted value
501.854

191.364

X1 = B: Solvent/material ratio 12.07

510
X2 = C: Time

Actual Factor 6
C : T im e

475
T E A C V a lu e

A: Ethanol concentration = 60.00

8.07
450
440
350 300

405 4.07

370

0.07
350

15.00
12.50 60.00 400
10.00 50.00 -3.93
40.00
C: Time 7.50
30.00
28.00 44.00 60.00 76.00 92.00 108.00

5.00 20.00 B: Solvent/material ratio

(C) B: Solvent/material ratio

Fig. 2. Response surface analysis for the antioxidant activities from Limonium sinuatum flowers with ultrasound-assisted extraction with respect to ethanol concentration and
solvent/material ratio (A); ethanol concentration and ultrasonic time (B); solvent/material ratio and ultrasonic time (C).

Table 4
The comparison of UAE with maceration and Soxhlet extraction.

Extracting method Ethanol concentration Temperature Time TEAC value (lmol Trolox/g DW)
Sample 1 Sample 2 Sample 3
UAE 60% 40 C 9.8 min 483.01 15.39 506.24 34.73 447.69 26.40
Maceration extraction 60% 25 C 24 h 449.16 28.27 476.23 17.15 401.54 24.38
Soxhlet extraction 60% 95 C 4h 228.32 16.39 231.67 19.03 205.88 14.61

with maceration method (24 h). Thus, in comparison to the extraction method in the extraction of antioxidants from
conventional extraction methods (Soxhlet and maceration extrac- Limonium sinuatum flower with shorter time, higher yield and
tion), ultrasound-assisted extraction method was a more efficient reduced temperature.
558 D.-P. Xu et al. / Food Chemistry 217 (2017) 552559
3.4. Qualitative and quantitative determination of polyphenolic
compounds
Qualitative and quantitative determination of polyphenolic
components of the extracts obtained under optimal extraction con-
ditions were conducted using HPLC-PAD. Three phenolic com-
pounds and two flavonoids were identified in the extract. The
highest content of polyphenolic components detected were gallic
acid (835.63 11.5 mg/100 g dw, Rt = 8.080 min, kmax = 270 nm),
then followed by ferulic acid (73.53 5.31 mg/100 g dw,
Rt = 29.146 min, kmax = 320 nm), quercetin (20.35 0.74 mg/100 g dw,
Rt = 49.336, kmax = 250 nm), rutin (10.44 0.83 mg/100 g dw,
Rt = 32.467 min, kmax = 250 nm) and epigallocatechin (9.22
0.78 mg/100 g dw, Rt = 17.426, kmax = 250 nm). The strong antioxi-
dant activity of the extract may be related to the synergistic effect
of these polyphenolic ingredients and others present in the extract,
which act as free radical scavengers. In other studies, gallic acid,
ferulic acid and quercetin have been confirmed effective in
treatment of several free radical mediated diseases, such as cancer,
diabetes, cardiovascular, and Alzheimers diseases (Kawabata,
Mukai, & Ishisaka, 2015; Sgarbossa, Giacomazza, & di Carlo,
2015; Subramanian et al., 2015).
4. Conclusions
In this paper, an UAE method was successfully optimized for the
extraction of natural antioxidants from the flower of Limonium
sinuatum using a RSM coupled with central composite rotatable
design. In this work, four extraction factors (ethanol concentration,
the ratio of solvent to solid, ultrasonic time and temperature) have
been evaluated, and the results indicated that ethanol concentra-
tion and the ratio of solvent to solid had significant effects
on the yield of antioxidant extraction. The optimal antioxidant
activities (483.01 15.39 lmol Trolox/g DW) was obtained when
the extraction process was performed at ethanol concentration of
60%, the ratio of solvent to solid of 56.9:1 mL/g, temperature
of 40 C and ultrasonic time of 9.8 min. The R2 values (0.972)
and the no-significant difference between predicted value
(483.01 15.39 lmol Trolox/g DW) and experimental value
(494.13 lmol Trolox/g DW) confirmed the validity of the second-
order polynomial model proposed. Additionally, the comparative
studies revealed that the antioxidant activities of the flower
extracts using UAE were significantly higher than that of Soxhlet
extraction, and its extraction time was markably shorter than that
of maceration extraction. The results demonstrated that UAE was
an efficient and feasible method in the extraction of desired natural
antioxidants from Limonium sinuatum flower and also indicate the
utilization of Limonium sinuatum flower.
Acknowledgments
This work was supported by the National Natural Science Foun-
dation of China (No. 81372976), Key Project of Guangdong Provin-
cial Science and Technology Program (No. 2014B020205002), and
the Hundred-Talents Scheme of Sun Yat-Sen University.
References
Bimakr, M., Rahman, R. A., Taip, F. S., Adzahan, N. M., Sarker, M. Z. I., & Ganjloo, A.
(2012). Optimization of ultrasound-assisted extraction of crude oil from winter
melon (Benincasa hispida) seed using response surface methodology and
evaluation of its antioxidant activity, total phenolic content and fatty acid
composition. Molecules, 17(10), 1174811762.
Boots, A. W., Haenen, G., & Bast, A. (2008). Health effects of quercetin: From
antioxidant to nutraceutical. European Journal of Pharmacology, 585(23),
325337.
Cai, Y. Z., Luo, Q., Sun, M., & Corke, H. (2004). Antioxidant activity and phenolic
compounds of 112 traditional Chinese medicinal plants associated with
anticancer. Life Sciences, 74(17), 21572184.
Carciochi, R. A., Manrique, G. D., & Dimitrov, K. (2015). Optimization of antioxidant
phenolic compounds extraction from quinoa (Chenopodium quinoa) seeds.
Journal of Food Science and Technology-Mysore, 52(7), 43964404.
Chavan, Y., & Singhal, R. S. (2013). Ultrasound-assisted extraction (UAE) of
bioactives from arecanut (Areca catechu L.) and optimization study using
response surface methodology. Innovative Food Science & Emerging Technologies,
17, 106113.
Chen, J. Y., Funnell, K. A., & Morgan, E. R. (2010). A model for scheduling flowering of
a Limonium sinuatum x Limonium perezii Hybrid. HortScience, 45(10), 14411446.
Dahmoune, F., Nayak, B., Moussi, K., Remini, H., & Madani, K. (2015). Optimization
of microwave-assisted extraction of polyphenols from Myrtus communis L.
leaves. Food Chemistry, 166, 585595.
Das, A. J., Das, G., Miyaji, T., & Deka, S. C. (2016). In vitro antioxidant activities of
polyphenols purified from four plant species used in rice beer preparation in
Assam India. International Journal of Food Properties, 19(3), 636651.
Deng, G. F., Shen, C., Xu, X. R., Kuang, R. D., Guo, Y. J., Zeng, L. S., et al. (2012).
Potential of fruit wastes as natural resources of bioactive compounds.
International Journal of Molecular Sciences, 13(7), 83088323.
Diaz-Reinoso, B., Moure, A., Dominguez, H., & Parajo, J. C. (2006). Supercritical CO2
extraction and purification of compounds with antioxidant activity. Journal of
Agricultural and Food Chemistry, 54(7), 24412469.
Embuscado, M. E. (2015). Spices and herbs: Natural sources of antioxidants A mini
review. Journal of Functional Foods, 18, 811819.
Esclapez, M. D., Garcia-Perez, J. V., Mulet, A., & Carcel, J. A. (2011). Ultrasound-
assisted extraction of natural products. Food Engineering Reviews, 3(2), 108120.
Frankel, E. N., & Finley, J. W. (2008). How to standardize the multiplicity of methods
to evaluate natural antioxidants. Journal of Agricultural and Food Chemistry, 56
(13), 49014908.
Hung, P. V. (2016). Phenolic compounds of cereals and their antioxidant capacity.
Critical Reviews in Food Science and Nutrition, 56(1), 2535.
Igawa, T., Hoshino, Y., & Mii, M. (2002). Efficient plant regeneration from cell
cultures of ornamental statice, Limonium sinuatum Mill. Vitro Cellular &
Developmental Biology-Plant, 38(2), 157162.
Jing, C. L., Dong, X. F., & Tong, J. M. (2015). Optimization of ultrasonic-assisted
extraction of flavonoid compounds and antioxidants from alfalfa using response
surface method. Molecules, 20(9), 1555015571.
Kaur, C., & Kapoor, H. C. (2001). Antioxidants in fruits and vegetables the
millenniums health. International Journal of Food Science and Technology, 36(7),
703725.
Kawabata, K., Mukai, R., & Ishisaka, A. (2015). Quercetin and related polyphenols:
new insights and implications for their bioactivity and bioavailability. Food &
Function, 6(5), 13991417.
Kumar, A., & Singh, A. (2015). A review on mitochondrial restorative mechanism of
antioxidants in Alzheimers disease and other neurological conditions. Frontiers
in Pharmacology, 6.
Li, S., Li, S. K., Gan, R. Y., Song, F. L., Kuang, L., & Li, H. B. (2013). Antioxidant
capacities and total phenolic contents of infusions from 223 medicinal plants.
Industrial Crops and Products, 51, 289298.
Li, A. N., Li, S., Li, H. B., Xu, D. P., Xu, X. R., & Chen, F. (2014). Total phenolic contents
and antioxidant capacities of 51 edible and wild flowers. Journal of Functional
Foods, 6, 319330.
Li, S., Tan, H. Y., Wang, N., Zhang, Z. J., Lao, L., Wong, C. W., & Feng, Y. (2015). The role
of oxidative stress and antioxidants in liver diseases. International Journal of
Molecular Sciences, 16(11), 2608726124.
Liu, S., Liang, Y. Z., Luo, J., Cao, D. S., Li, X. Z., Lei, P., & Yuan, D. L. (2008). Optimization
of liensinine, isoliensinine and neferine extraction from the embryo of the seed
of Nelumbo nucifera GAERTN. Separation Science and Technology, 43(14),
36373651.
Liu, L., Shen, B. J., Xie, D. H., Cai, B. C., Qin, K. M., & Cai, H. (2015). Optimization of
ultrasound-assisted extraction of phenolic compounds from Cimicifugae
rhizoma with response surface methodology. Pharmacognosy Magazine, 11(44),
682689.
Peng, L. X., Zou, L., Zhao, J. L., Xiang, D. B., Zhu, P., & Zhao, G. (2013). Response surface
modeling and optimization of ultrasound-assisted extraction of three
flavonoids from tartary buckwheat (Fagopyrum tataricum). Pharmacognosy
Magazine, 9(35), 210215.
Povilaitis, D., Sulniute, V., Venskutonis, P. R., & Kraujaliene, V. (2015). Antioxidant
properties of wheat and rye bran extracts obtained by pressurized liquid
extraction with different solvents. Journal of Cereal Science, 62, 117123.
Roleira, F., Tavares-da-Silva, E. J., Varela, C. L., Costa, S. C., Silva, T., Garrido, J., &
Borges, F. (2015). Plant derived and dietary phenolic antioxidants: Anticancer
properties. Food Chemistry, 183, 235258.
Rostagno, M. A., Palma, M., & Barroso, C. G. (2007). Ultrasound-assisted extraction of
isoflavones from soy beverages blended with fruit juices. Analytica Chimica Acta,
597(2), 265272.
Sgarbossa, A., Giacomazza, D., & di Carlo, M. (2015). Ferulic acid: A hope for
Alzheimers disease therapy from plants. Nutrients, 7(7), 57645782.
Shahidi, F., & Ambigaipalan, P. (2015). Phenolics and polyphenolics in foods,
beverages and spices: Antioxidant activity and health effects A review. Journal
of Functional Foods, 18, 820897.
Shirsath, S. R., Sonawane, S. H., & Gogate, P. R. (2012). Intensification of extraction of
natural products using ultrasonic irradiationsA review of current status.
Chemical Engineering and Processing, 53, 1023.
 D.-P. Xu et al. / Food Chemistry 217 (2017) 552559
Subramanian, A. P., John, A. A., Vellayappan, M. V., Balaji, A., Jaganathan, S. K.,
Supriyanto, E., & Yusof, M. (2015). Gallic acid: prospects and molecular
mechanisms of its anticancer activity. RSC Advances, 5(45), 3560835621.
Vajic, U. J., Grujic-Milanovic, J., Zivkovic, J., Savikin, K., Godevac, D., Miloradovic, Z.,
et al. (2015). Optimization of extraction of stinging nettle leaf phenolic
compounds using response surface methodology. Industrial Crops and
Products, 74, 912917.
Wang, J. M., Geng, Y., Li, P., Hu, F., & Li, L. Y. (2013). Optimization of ultrasound-
assisted extraction procedure to determine total isoflavones in Chinese
soybean cheese by Box-Behnken Design. Food Analytical Methods, 6(1),
221226.
Xu, D. P., Zhou, Y., Zheng, J., Li, S., Li, A. N., & Li, H. B. (2016). Optimization of
ultrasound-assisted extraction of natural antioxidants from the flower of
Jatropha integerrima by response surface methodology. Molecules, 21(1). Article
No. 18.
559
Yang, L., Cao, Y. L., Jiang, J. G., Lin, Q. S., Chen, J., & Zhu, L. (2010). Response surface
optimization of ultrasound-assisted flavonoids extraction from the flower of
Citrus aurantium L. var. amara Engl. Journal of Separation Science, 33(9),
13491355.
Young, I. S., & Woodside, J. V. (2001). Antioxidants in health and disease. Journal of
Clinical Pathology, 54(3), 176186.
Zhang, Z. S., Li, D., Wang, L. J., Ozkan, N., Chen, X. D., Mao, Z. H., & Yang, H. Z. (2007).
Optimization of ethanol-water extraction of lignans from flaxseed. Separation
and Purification Technology, 57(1), 1724.
Zou, T. B., Wang, M., Gan, R. Y., & Ling, W. H. (2011). Optimization of ultrasound-
assisted extraction of anthocyanins from mulberry, using response surface
methodology. International Journal of Molecular Sciences, 12(5), 30063017.
Zou, T. B., Xia, E. Q., He, T. P., Huang, M. Y., Jia, Q., & Li, H. W. (2014). Ultrasound-
assisted extraction of mangiferin from mango (Mangifera indica L.) leaves using
response surface methodology. Molecules, 19(2), 14111421.