Preparative Biochemistry and Biotechnology

ISSN: 1082-6068 (Print) 1532-2297 (Online) Journal homepage: https://www.tandfonline.com/loi/lpbb20

Optimization of the extraction of polyphenols and

antioxidant activity from Malva parviflora L. leaves
using Box–Behnken design

Ekram A. Abd El-Salam & Nashwa F. S. Morsy

To cite this article: Ekram A. Abd El-Salam & Nashwa F. S. Morsy (2019): Optimization
of the extraction of polyphenols and antioxidant activity from Malva�parviflora L.
leaves using Box–Behnken design, Preparative Biochemistry and Biotechnology, DOI:
10.1080/10826068.2019.1633667

To link to this article: https://doi.org/10.1080/10826068.2019.1633667

Published online: 27 Jun 2019.

Submit your article to this journal

Article views: 15

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=lpbb20
PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY
https://doi.org/10.1080/10826068.2019.1633667

Optimization of the extraction of polyphenols and antioxidant activity from

Malva parviflora L. leaves using Box–Behnken design
Ekram A. Abd El-Salam and Nashwa F. S. Morsy
Faculty of Agriculture, Food Science Department, Cairo University, Giza, Egypt

ABSTRACT KEYWORDS
Box–Behnken Design (BBD) was used to optimize the extraction conditions of polyphenols from Antioxidants; Box–Behnken
Malva (Malva parviflora L.) leaves. The effect of ethanol concentration (20–80%), solvent/leaf pow- design; Malva parviflora L.;
der ratio (10:1 to 30:1, v/w) and extraction time (5–45 min) on the polyphenols yield and the 2,2- polyphenols
diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity of the obtained extracts were investi-
gated. Quadratic models fit well. The optimal conditions (53.40% ethanol, solvent/leaf powder
ratio 20:1 (v/w), and 15 min) resulted in an extract with a maximum yield of polyphenols
(1098.4 mg GAE/100 g leaf powder) and high inhibition percentage of DPPH radical (33.31%) with
desirability 0.742. High Performance Liquid Chromatography (HPLC) analysis indicated that the
major identified polyphenol compounds extracted at the optimal conditions were naringenin,
q-coumaric acid, apigenin-7-glucoside, luteolin, and cinnamic acid. These findings indicate that M.
parviflora leaf extracts possess DPPH radical scavenging activity and could be used as a natural
source for bioactive products.

Introduction polyphenols using various extraction techniques from differ-

ent materials.[10,14–16] Box–Behnken Design (BBD) is a type
Numerous investigations focused on antioxidants from natural
of RSM that reduces the experiments required to find the
sources. Recently, attention has been paid to polyphenols as
optimal conditions and illustrates the interactions among
biologically active ingredients, due to its antioxidant, anti-
process variables. BBD is applied as an exploratory method
inflammatory and antitumor activities.[1] Polyphenols include
for second-order experiments to establish a quadratic
a lot of compounds that originated in the plant kingdom and
response surface.[17–19]
differ in their structures and functions.[2] Several studies were
Based on our knowledge, there is no available research
conducted to investigate the content and beneficial effects of
concerning the extraction optimization of Malva parviflora
polyphenols in different materials.[2–4] Extraction is the first
L. leaf for the efficient recovery of bioactive compounds,
step in the recovery of polyphenols. The optimum conditions
such as polyphenols, using aqueous ethanol. Preliminary sin-
for extracting polyphenols differ according to raw material
gle factor experiments were used to find suitable ranges of
and physicochemical factors.[5–11]
the extraction variables (extraction time, solvent/leaf powder
Malva parviflora L. (Malvaceae family) is a wild herb that
ratio, and ethanol concentration). BBD was used to establish
is prevalent in North Africa. It is known as little mallow
combinations of the selected ranges of the independent
and cheeseweed mallow.[12] Its common name in Egypt is
extraction factors to obtain polyphenols rich extract with
Khoubbiza. Malva leaves are traditionally cooked like spin-
high DPPH radical scavenging activity from Malva leaf.
ach and also used in the treatment of the flu, fever, dysen-
Furthermore, polyphenols extracted with the optimal condi-
tery, and wounds.[13]
tions were identified and quantified by High Performance
Conventional extraction methods of polyphenols with
Liquid Chromatography.
one factor at a time have disadvantages such as lengthy dur-
ation, low extraction efficiency, and decomposition of
thermolabile compounds. These methods neglect the inter- Materials and methods
action effect among the variables and lead to misinterpret-
Plant material
ation of results. The response surface methodology (RSM) is
an effective statistical technique for optimizing the responses Malva parviflora L. leaves were collected in October 2017,
that are affected by the independent variables. It can predict from Geziret El-Dahab, Giza, Egypt; Latitude 29
590
interaction effects of variables on process response. It is suc- 0.952800 (N), Longitude 31
130 19.613400 (E). The leaves were
cessfully used to optimize extraction conditions of manually separated and washed before drying at 40
C in an

CONTACT Ekram A. Abd El-Salam [email protected] Food Science Department, Faculty of Agriculture, Cairo University, Giza 12613, Egypt.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/lpbb.
ß 2019 Taylor & Francis Group, LLC
2 E. A. ABD EL-SALAM AND N. F. S. MORSY

oven (Heraeus electronics, Germany) for 20 hr. The dried All single factor experiments were performed in triplicate.
leaves were grounded in a bench top analytical mill (Cole- One-way ANOVA followed by Tukey’s test at p < 0.05 level
Parmer, USA), sieved through 0.500 mm stainless steel sieve was used to estimate the significant effect of different factor
and stored at 8
C until analysis. levels (Minitab 17, Minitab Inc., USA). Three factor levels
with the highest extracted polyphenols were used for further
RSM design.
Chemicals
Gallic acid, Folin-Ciocalteu reagent, and 2,2-diphenyl-1-pic-
rylhydrazyl (DPPH) and HPLC-grade acetonitrile were pur- Box–Behnken design
chased from Sigma Chemical Co., Ltd (St. Louis, MO, USA). Based on the results of the preliminary experiments (as
All other reagents and solvents were of analytical grade. shown in Figure 1) the levels of the independent variables
that were selected are depicted in Table 1: time (X1,
5–15 min), solvent/leaf powder ratio (X2, 20:1–30:1), and
Preliminary single factor extraction experiments
solvent concentration (X3, 40–60%).
The Malva leaf powder was extracted at ambient tempera- BBD with three factors (X1, X2, X3) at three levels (–1, 0,
ture using magnetic stirring (Medline Scientific, Model MS þ1) was used to optimize the extraction process of total pol-
300, UK). The leaf powder (3.33–10 g) was mixed with yphenols and its corresponding scavenging activity against
100 mL of ethanol solution (20–80%) corresponding to solv- DPPH radicals. The experimental design consists of 17 com-
ent/leaf powder ratio of 10:1–30:1 (v/w), for time ranged binations (Table 1) including five replicates of the central
from 5 to 45 min with 5 min intervals. The extract was fil- point to estimate the pure error. All the experiments were
tered through Whatman filter paper No. 1 before analysis. conducted in triplicate.

Figure 1. Preliminary experiments of each extraction parameter level’s effect on total polyphenols content [TPC (mg GAE/100 g dried leaf powder)] in Malva leaf
powder: (A) extraction time (min), (B) solvent/leaf powder ratio (mL/g), and (C) ethanol concentration (%).
 PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY 3

Table 1. The Box–Behnken design with experimental data for the extraction of bioactive compounds from Malva.
DPPH radical inhibition percentage of
Time Solvent ratio X Solvent conc. Yield of total polyphenols (mg Concentration of polyphenols 1 mL extract
Run (min) (mL): 1 solid (Ethanol %) GAE/ 100 g dried leaf powder) (mg GAE/ mL extract) the extract containing 1 mg GAE
1 5 (–1) 20 (–1) 50 (0) 786i ± 14 393.00gh ± 7 29.85b ± 0.12 76.0a ± 1.1
2 15 (þ1) 20 (–1) 50 (0) 1006ef ± 22.4 503.03a ± 11.21 35.28a ± 1.19 70.1bc ± 2.1
3 5 (–1) 30 (þ1) 50 (0) 1180a ± 12.12 393.33gh ± 4.04 22.63d ± 0.28 57.5d ± 1.2
4 15 (þ1) 30 (þ1) 50 (0) 1122ab ± 18.7 374.00hi ±6.24 17.66h ± 0.17 47.2g ± 0.8
5 5 (–1) 25 (0) 40 (–1) 1091bcd ± 31.3 436.50def ±12.5 20.59efg ± 0.18 47.2g ± 1.1
6 15 (þ1) 25 (0) 40 (–1) 1035de ± 25 414.00fg ±10 19.59g ± 0.45 47.4g ± 2.0
7 5 (–1) 25 (0) 60 (þ1) 892gh ± 15.61 357.00i ± 6.24 24.28c ± 0.04 68.0c ± 1.1
8 15 (þ1) 25 (0) 60 (þ1) 1046cde ± 21.3 418.67efg ± 8.5 20.18fg ± 0.48 48.2fg ± 0.8
9 10 (0) 20 (–1) 40 (–1) 949fg ± 25 474.50bc ± 12.5 20.62dfg ± 0.11 43.5h ± 1.0
10 10 (0) 30 (þ1) 40 (–1) 1074bcd ± 15 358.00i ± 5.00 17.75h ± 0.11 49.6efg ± 0.6
11 10 (0) 20 (–1) 60 (þ1) 975f ± 15 487.50ab ± 7.50 24.89c ± 0.25 51.1efg ± 0.9
12 10 (0) 30 (þ1) 60 (þ1) 888h ± 15.59 296.00j ± 5.20 21.34e ± 0.11 72.1b ± 1.3
13 10 (0) 25 (0) 50 (0) 1086bcd ± 14.22 413.00fg ± 8.00 21.53de ± 0.24 52.7e ± 0.9
14 10 (0) 25 (0) 50 (0) 1098bc ± 26.3 439.50def ± 10.5 21.20ef ± 0.07 48.3fg ± 1.1
15 10 (0) 25 (0) 50 (0) 1090bcd ± 12.5 448.00cd ± 17 21.73de ± 0.26 48.5fg ± 1.3
16 10 (0) 25 (0) 50 (0) 1089bcd ± 9.46 420.50defg ± 10.5 21.05ef ± 0.30 50.4efg ± 1.2
17 10 (0) 25 (0) 50 (0) 1104bc ± 16.65 442.50de ± 6.5 20.96ef ± 0.28 47.7fg ± 0.8
P (Brown-Forsythe) 0.993 0.960 0.609 0.999
P (ANOVA) < 0.001 < 0.001 < 0.001 <0.001
Means followed by different letters in the same column significantly differ based on Tucky’s test at p < 0.05.

Brown-Forsythe test and one-way ANOVA followed by Radical scavenging activity

Tukey’s test at p < 0.05 level were used to evaluate the
The scavenging activity of properly diluted Malva leaf pow-
homogeneity of variance for all responses and the significant
der extracts towards DPPH radicals was assessed spectro-
effect of different combinations, respectively (Minitab 17,
photometrically (UV-2000 Spectrophotometer, Unico, USA)
Minitab Inc., USA). To implement BBD, Design-Expert ver-
at 515 nm against a blank.[23] The inhibition percentage of
sion 7.0.0 (Stat-Ease, Inc., Minneapolis, MN, USA) was
DPPH radical was calculated using Eq. (2).
used.[20] The experimental data were fitted using the quad-
ratic model [Eq. (1)], that has the following equation: Inhibition percentage of DPPH radical
 
A517nm of control A517 nm of sampl (2)
X
3 X
3 XX
3 ¼  100
A517 nm of control
Y ¼ b0 þ bj Xj þ bjj Xj2 þ bij Xi Xj þ ei (1)
j¼1 j¼1 i <j ¼ 2 The correlation between inhibition percentage of DPPH
radical and its corresponding polyphenols at extract concen-
where, Y is the dependent variable. Xi and Xj are coded tration 1 mg GAE/mL extract was calculated using Eq. (3)
independent variables. b0, bj, bjj, and bij are variable regres- Inhibition percentage of DPPH radical of 1mg GAE=mL extract
sion coefficients for the model intercept, linear, quadratic, Inhibition percentage of DPPH radical
and interaction effects, respectively. The error is represented ¼  
Total polyphenols of the extract mg GAE=mL extract
by ei.[21] The adequacy of the generated mathematical
(3)
models was evaluated regarding their determination
coefficients R2; R2adj ; predicted R2, (CV%), and adequate
precision values.
HPLC analysis of polyphenols
Polyphenols of Malva leaf extract obtained at the optimum
Total polyphenols content
conditions were quantified.[24] HPLC-DAD (Agilent
The total polyphenols content (TPC) of the Malva leaf Technologies 1100 series liquid chromatograph, USA)
extract was determined by the Folin-Ciocalteu method at equipped with an auto sampler was used. The Eclipse XDB-
room temperature using Gallic acid as a standard.[22] C18 column (150  4.6 mm; 5 mm) was used with a C18 guard
Properly diluted sample (10 lL) was mixed using vortex column (Phenomenex, Torrance, CA). The mobile phase
with 50 lL of Folin-Ciocalteu reagent and 790 lL of distilled consisted of acetonitrile (solvent A) and 2% acetic acid in
water. After 1 min, 150 lL of 20% sodium carbonate solution water (v/v) (solvent B). Elution was performed at a flow rate
were added, vortexed and left to stand in a dark place for of 0.8 mL/min. The gradient program consisted of 0–100%
120 min. Absorption was read at 750 nm (UV-2000 A in B over 0–60 min. The injection volume was 50 mL, and
Spectrophotometer, Unico, USA). Calibration curve (r2 ¼ the peaks were monitored simultaneously at 280 and 320 nm
0.9990) of standard Gallic acid (50–800 mg/L) was used to for the benzoic acid and cinnamic acid derivatives, respect-
calculate the TPC. Results were expressed as mg Gallic acid ively as well as 360 nm for flavonoids. All samples were fil-
equivalents (GAE)/100 g dry leaf powder and mg Gallic acid tered through a 0.45 mm Acrodisc syringe filter (Gelman
equivalents (GAE)/mL extract. Laboratory, MI) before injection. The peaks were identified
4 E. A. ABD EL-SALAM AND N. F. S. MORSY

by retention time after matching with the standards. The Response surface methodology analysis
linear calibration curve of the standards was used for
Average values of polyphenols and inhibition percentage of
quantification.
DPPH radical are listed in Table 1. Brown-Forsythe’s test
indicated that all responses were homoscedastic (p  0.6),
Results and discussion while significant differences (p ˂ 0.001) were found for the
dependent variables using one-factor ANOVA.
Preliminary single factor experiments The yield of polyphenols varied significantly (p < 0.001)
Extraction time from 786 ± 14.00 to 1180 ± 12.12 (mg GAE/100 g dry leaf pow-
The effect of extraction time (5–45 min) on the yield of poly- der). This result is higher than that obtained by Afolayan
phenols from the Malva leaf powder was investigated using et al.[32] who found that TPC was 290 mg GAE/100 g dried M.
ethanol concentration 80% (v/v); and solvent/leaf powder ratio parviflora whole plant. This difference may be explained by
20:1 (v/w). Results are illustrated in Figure 1A. Extending part of the plant being extracted, type of solvent and the
extraction time from 5 to 10 min improved significantly (p ˂ extraction technique used. They used shaking technique dur-
0.05) the extraction efficiency of polyphenols, after which no ing extraction of whole plant with methanol for 24 hr.
Results in Table 1 indicate that the inhibition percentage of
further increase was obtained. This could be due to the small
DPPH radical of the obtained extracts varied significantly
particle size leaf powder that enhanced mass transfer process
(p < 0.001) from 17.75 ± 0.11 to 35.28 ± 1.19% according to the
and decreased the time for maximum extraction.[25] The max-
extraction conditions used. Results showed that M. parviflora
imum yield of polyphenols was found to be 581.33 ± 10.26 mg
leaf extract (1 mg GAE/mL) that was obtained at ethanol
GAE/100 g leaf powder. Based on the obtained results, 5, 10,
(50%)/leaf powder ratio of 20:1 (v/w) or at a higher ratio using
and 15 min were selected for RSM experiments.
an ethanol concentration of 60% exerted a high scavenging
activity against DPPH radicals that reached 76%. This finding
Solvent/leaf powder ratio (v/w) exceeded that obtained by Afolayan et al.[32] They found that
Polyphenol yields with different solvent/leaf powder ratios 0.5 mg GAE/mL extract of M. parviflora. inhibited only 9.3% of
(10:1 to 30:1, v/w) were evaluated at extraction time 10 min the DPPH radicals. Bouriche et al.[33] indicated that 89.03 lg/
and ethanol concentration 80% (v/v). As shown in Figure mL of lyophilized M. parviflora leaf powder methanol extract
1B, increasing solvent/material ratio from 10:1 to 20:1 (v/w) possessed the ability to scavenge 50% of the DPPH radicals.
increased significantly (p ˂ 0.05) the yield of polyphenols The correlation between the total polyphenols content of
from 423.3 ± 21.5 to 556.67 ± 9.02 mg GAE/100 g leaf pow- the extracts and their scavenging activity against DPPH radi-
der. Mass transfer of the extractable material is highly cals is a significant (p < 0.001) negative correlation (r ¼
dependent on its concentration in the solvent until it –0.457). This could be due to structure, concentration, and
the interaction between the antioxidants in the extract.[34]
reached equilibrium.[26,27] Results revealed that the continual
To achieve regression models, linear, interactive (2 FI), quad-
increase of solvent/material ratio to 30:1 (v/w) did not
ratic and cubic models were fitted to the experimental data and
enhance significantly (p ˂ 0.05) the extraction of polyphe-
the results are listed in Table 2. The data outlined in Table 2
nols. When the mass transfer of the extractable material is
reveal that, the most adequate model to represent experimental
maximized, any further increase of solvent/material ratio
data was quadratic model. As, the values of determination coef-
(v/w) does not enhance the extraction yield.[28,29] Thus, the
ficient (R2), adjusted determination coefficient (Adj R2) and pre-
solvent/leaf powder ratio in the RSM experiments was deter-
dicated determination coefficient (Pred. R2) of quadratic model
mined at 20:1, 25:1, and 30:1 (v/w). were the highest and highly significant (p < 0.001) among other
models except cubic model, which was aliased or confounding.
Effect of solvent concentration Hence, the following second order polynomial equations with
The effect of different concentrations of aqueous ethanol interaction term were used to represent the relationship between
(from 20 to 80%) was evaluated using 10 min extraction independent factors and responses in terms of coded factors.
 
time and solvent/leaf powder ratio of 20:1 (v/w). As shown YTPC mg GAE=100 g DW
in Figure 1C, increasing the ethanol concentration to 40% ¼ 1093:75 þ 32:50X1 þ 68:49X2  43:39X3  69:52X1 X2
increased significantly (p ˂ 0.05) the yield of polyphenols to
the maximum level (990.66 ± 13.29 mg GAE/100 g leaf pow- þ 52:60X1 X3  53:00X2 X3  12:69X12  57:54X22  64:71X32
der). Extracts obtained with solvent concentrations of 40% (4)
and 50% had significantly (p ˂ 0.05) the same level of poly-
phenols. Increasing the ethanol concentration to levels Yinhibition percentage of DPPH radical
higher than 50% exhibited a significant (p ˂ 0.05) decrease ¼ 21:29  0:58X1  3:91X2 þ 1:52X3  2:60X1 X2
in the extraction efficiency of polyphenols. The extraction of
 0:78X1 X3  0:17X2 X3 þ 2:54X12 þ 2:53X22  2:67X32
polyphenols is improved using binary solvent systems (etha-
nol/water) instead of mono-solvent systems due to their (5)
relative polarity.[30,31] Therefore, ethanol concentration of The ability of the obtained equations to describe the
40, 50, and 60% were used for RSM design. response’s variability was determined using multiple regression
 PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY 5

Table 2. Adequacy of the model tested.

Source Std. Dev. R2 Adjusted R2 Predicted R2 PRESS Prob < F Remarks
Yield of total polyphenols (mg GAE/ 100 g dried leaf powder)
Linear 83.55 0.3582 0.3172 0.2277 3.948 Eþ005 0.0001
2FI 67.96 0.6025 0.5483 0.4585 2.768 Eþ005 <0.0001
Quadratic 49.00 0.8074 0.7652 0.6619 1.728 Eþ005 <0.0001 Suggested
Cubic 18.57 0.9744 0.9663 0.9498 25,654.86 <0.0001 Aliased
DPPH radical inhibition percentage of the extract
Linear 3.20 0.4718 0.4380 0.3482 593.90 <0.0001
2FI 2.99 0.5689 0.5101 0.3597 583.45 0.0288
Quadratic 1.91 0.8357 0.7996 0.7092 265.02 <0.0001 Suggested
Cubic 0.40 0.9933 0.9912 0.9873 11.59 <0.0001 Aliased
Quadratic models are underlined and the model is labeled as “Suggested” by the Design-Expert software. Suggestion is based on a subjective
scoring system that uses a combination of selected metrics to propose that quadratic model looks best compared with other models

Table 3. ANOVA analysis and statistical parameters of the model.

Yield of total polyphenols (mg GAE/ 100 g dried leaf powder) DPPH radical inhibition percentage of the extract
Source RC SS p Value RC SS p Value
Model 1093.75 4.13 Eþ05 < 0.0001 21.29 761.46 < 0.0001
X1 32.5 25,346.75 0.0023 –0.58 8.07 0.1448
X2 68.49 1.13 Eþ05 < 0.0001 –3.91 366.58 < 0.0001
X3 –43.39 45,175.07 < 0.0001 1.52 55.21 0.0004
X1 X2 –69.52 57,990.8 < 0.0001 –2.6 80.91 < 0.0001
X1 X3 52.6 33,206.38 0.0006 –0.78 7.23 0.1668
X2 X3 –53 33,708 0.0006 –0.17 0.35 0.759
X12 –12.69 2034.01 0.3628 2.54 81.18 < 0.0001
X22 –57.54 41,826.73 0.0002 2.53 80.59 < 0.0001
X32 –64.71 52,887.14 < 0.0001 –2.67 89.86 < 0.0001
Residual 98,447.47 149.74
Lack of fit 85,338.68 < 0.0001 143.67 < 0.0001
Pure error 13,108.79 6.07
Cor total 5.11 Eþ05 911.19
Std. dev. 49 1.91
Mean 1030.25 22.42
C.V.% 4.76 8.52
PRESS 1.73 Eþ05 265.02
R2 0.8074 0.8357
Adj R2 0.7652 0.7996
Pred R2 0.6619 0.7092
Adeq precision 13.059 19.359

analysis and analysis of variance (ANOVA) as outlined in variables (extraction time, solvent/leaf powder ratio (v/w)
Table 3. The obtained models are highly significant, as its prob- and ethanol %) on the response variables. Generally, as
ability values are very low (p < 0.0001). Except regression coeffi- shown in Figure 2A–C, increasing extraction time and solv-
cient of X12 for TPC and regression coefficients of X1 ; X1 X3 ; ent/leaf powder ratio (v/w) increased the yield of polyphe-
and X2 X3 for DPPH radical inhibition percentage, all regression nols, while increasing the ethanol % above 50% decreased
coefficient are significant (p  0.0023). The values of R2 and the yield of polyphenols. The linear terms of extraction time
adjusted R2 are greater than 0.80 and 0.75, respectively, which and solvent/leaf powder ratio (v/w) were significant
indicate the adequacy of the obtained models to predict the (p ¼ 0.0023 and p < 0.0001, respectively) and positive (Table
experimental data.[27] The lack of fit (F-value) of 82.46 for TPC 3), whereas its interactions and the quadratic term for solv-
and 299.92 for DPPH radical scavenging activity implies the ent/leaf powder ratio (v/w) only were significant
lack of fit is significant. There is only a 0.01% chance that a (p  0.0002) but negative. Hence, as appeared in Figure
“Lack of Fit F-value” this large could occur due to noise. 2A–C, there is a curvilinear increase in the polyphenols yield
Adequate precision measures the signal to noise ratio. A ratio for all solvent/leaf powder ratios (v/w) used. On the other
larger than four is desirable. Similarly, adequate precision for hand, all terms of ethanol % were significant (p  0.0006)
the responses (Table 3) are greater than 13, which also indicates and negative except the interaction term of ethanol % and
the adequacy of the signals.[35] As shown in Table 3, the CV of solvent/leaf powder ratio (v/w), which was positive. Thus,
the model is < 10, indicating that the model explains the there was a curvilinear decrease in the yield of polyphenols
response adequately and indicates that the experimental values for all ethanol concentrations used (Figure 2B,C).
are associated with a high degree of precision.[21] Figure 2D–F illustrates the 3D surface plots of the DPPH
radical scavenging activity model. The interaction term of
extraction time, solvent/leaf powder ratio (v/w), and quad-
Effect of process variables on the yield of polyphenols
ratic term of extraction time were statistically significant
and DPPH radical scavenging activity
(p ˂ 0.0001) (Table 3), which result in a curvilinear change
Equations (4) and (5) were used to generate (3D) plots in the inhibition percentage of DPPH radical for all extrac-
(Figure 2) to illustrate the interactive effect of independent tion time investigated. Figure 2D shows that, the inhibition
6 E. A. ABD EL-SALAM AND N. F. S. MORSY

Figure 2. Response surface plots showing interaction effects of process variables (Extraction time (min), solvent/leaf powder ratio (mL/g) and ethanol %) on the
responses [TPC (mg GAE/100 g dried leaf powder) (A–C), inhibition percentage of DPPH radical (D-F)].

percentage of DPPH radical increased as extraction time The highest inhibition percentage of DPPH radical could be
increased at low solvent/leaf powder ratio (v/w). However, observed at solvent/leaf powder ratio of 20:1 (v/w) (Figure 1D).
at high solvent/leaf powder ratio (v/w), the inhibition per- The linear and quadratic terms of ethanol % were significant
centage of DPPH radical decreased as extraction time (p  0.0004), whereas its interactive terms were insignificant
increased. Except the interaction term of the solvent/leaf (p  0.1668). The highest inhibition percentage of DPPH radical
powder ratio (v/w) and ethanol %, all terms of solvent/leaf could be obtained at an ethanol concentration of 50–55%
powder ratio (v/w) were statistically significant (p ˂ 0.0001). (Figure 1E,F). The optimum conditions to obtain Malva leaf
 PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY 7

Table 4. Identified phenolic and flavonoid compounds in the Malva leaf Conclusions
extract obtained at optimum conditions.
Compound lg/g leaf powder BBD was successfully used to optimize the extraction of pol-
Phenolic acids yphenols with high DPPH radical scavenging activity from
Gentisic acid 101.55 Malva parviflora L. leaves. Effect of independent variables
Rosmarinic acid 57.85
Vanillic acid 11.33 [ethanol concentration, and solvent/leaf powder ratio (v/w)]
Ferulic acid 3.99 on the responses was significant (p  0.0004). The absolute
Sinapic acid 86.44 errors between predicted and experimental values under
p-coumaric acid 371.86
Cinnamic acid 126.07 optimal conditions were low, which indicate the suitability
Flavonoids of the generated models to predict responses. The optimal
Apigenin-7-glucoside 346.28 extraction conditions were found to be extraction time
Catechin 35.78
Luteolin 271.66 15 min, solvent/leaf powder ratio 20:1 (v/w), and aqueous
Naringenin 557.33 ethanol concentration 53.40%. HPLC analysis indicated that
Apigenin 10.80 the extract obtained at optimal conditions was rich in narin-
Kaempferol 27.54
Chrysin 6.04 genin and p-coumaric acid. These findings indicate that M.
parviflora leaf extracts possess DPPH radical scavenging
activity and could be considered as a good source of natural
antioxidants.
extract with high DPPH radical scavenging activity were extrac-
tion time of 15 min, leaf powder/solvent ratio of 1: 20 (w/v),
and ethanol % of 53.40% with desirability of 0.742. DPPH rad- Disclosure statement
ical scavenging activity of the extract may be attributed to its
No potential conflict of interest was reported by the authors.
hydrogen donation potency. The antioxidant activity of each
phenolic compound depends on the number of hydroxyl
groups and their position in the structure. References
[1] Khan, M.K.; Ahmad, K.; Hassan, S.; Imran, M.; Ahmad, N.; Xu,
Validation of the model C. Effect of Novel Technologies on Polyphenols During Food
Processing. Innov. Food Sci. Emerg. Technol. 2018, 45, 361–381.
For validation of the model, the experiments were carried out DOI: 10.1016/j.ifset.2017.12.006.
under the optimized conditions obtained by Box–Behnken [2] Khalifa, I.; Zhu, W.; Li, K.; Li, C. Polyphenols of Mulberry
Fruits as Multifaceted Compounds: Compositions, Metabolism,
experiments.
Health Benefits, and Stability—A Structural Review. J. Funct.
The predicted results that obtained by RSM for the yield Foods. 2018, 40, 28–43. DOI: 10.1016/j.jff.2017.10.041.
of TPC of Malva leaves and DPPH radical scavenging activ- [3] Solari-Godi~ no, A.; Perez-Jimenez, J.; Saura-Calixto, F.;
ity of the extract were 1071 mg GAE/100 g leaf powder and Borderıas, A.J.; Moreno, H.M. Anchovy Mince (Engraulis ring-
32.28%, respectively, while the validated values were ens) Enriched with Polyphenol-Rich Grape Pomace Dietary
1098.4 mg GAE/100 g leaf powder and 33.31%, respectively. Fibre: In Vitro Polyphenols Bioaccessibility, Antioxidant and
Physico-Chemical Properties. Food Res. Int. 2017, 102, 639–646.
DOI: 10.1016/j.foodres.2017.09.044.
[4] Rojas, R.; Gonzalez-Hernandez, M.D.; Castro-L opez, C.;
HPLC analysis 
Ventura-Sobrevilla, J.M.; Ascacio-Valdes, J.A.; Martınez-Avila,
Malva parviflora L. leaf powder extract that obtained at opti- G.C.G.; Aguilar, C.N. Impact of Extraction Techniques on
Antioxidant Capacities and Phytochemical Composition of
mum conditions was analyzed to identify and quantify phen- Polyphenol-Rich Extracts. Food Chem. 2017, 237, 1139–1148.
olic compounds and the results are presented in Table 4. The DOI: 10.1016/j.foodchem.2017.06.032.
main identified phenolic compounds were p-coumaric acid [5] 
Zivkovic, J.; Savikin, K.; Jankovic, T.; Cuji
 c, N.; Menkovic, N.
(371.86 lg/g), cinnamic acid (126.07 lg/g), and gentisic acid Optimization of Ultrasound-Assisted Extraction of Polyphenolic
(101.55 lg/g), whereas naringenin (557.33 lg/g), apigenin-7- Compounds from Pomegranate Peel Using Response Surface
Methodology. Sep. Purif. Technol. 2018, 194, 40–47. DOI: 10.
glucoside (346.28 lg/g), and luteolin (271.66 lg/g) were the
1016/j.seppur.2017.11.032.
dominated flavonoids. Terninko et al.[36] studied the phenolic [6] Rodrigues, S.; Fernandes, F.A.N.; de Brito, E.S.; Sousa, A.D.;
profile of Malva sylvertris L. leaf ethanolic extract. They identi- Narain, N. Ultrasound Extraction of Phenolics and
fied luteolin, chlorogenic acid, apigenin, rutin, and quercetin. Anthocyanins from Jabuticaba Peel. Ind. Crops Prod. 2015, 69,
Ben Saad et al.[37] also showed that epicatechic and gallic acids 400–407. DOI: 10.1016/j.indcrop.2015.02.059.
[7] Prakash Maran, J.; Manikandan, S.; Vigna Nivetha, C.; Dinesh,
were the main phenolic acids in M. sylvestris methanol extract
R. Ultrasound Assisted Extraction of Bioactive Compounds
whereas the major flavonoids were rutin and kaempferol. from Nephelium lappaceum L. Fruit Peel Using Central
Liquid chromatography analysis showed that Kaempferol 3-O- Composite Face Centered Response Surface Design. Arab. J.
glucoside, quercetin-3-O-rutinoside, isorhamnetin, quercetin- Chem. 2017, 10, S1145–S1157. DOI: 10.1016/j.arabjc.2013.02.
3-O-glucoside, and isorhamnetin-7-O-glucoside were the 007.
[8] Pompeu, D.R.; Silva, E.M.; Rogez, H. Optimisation of the
major flavonoids of ethanolic extract of Malva aegyptiaca
Solvent Extraction of Phenolic Antioxidants from Fruits of
leaves.[13] This difference in the phenolic composition of the Euterpe oleracea Using Response Surface Methodology.
extracts could be attributed to plant species, part of the plant Bioresour. Technol. 2009, 100, 6076–6082. DOI: 10.1016/j.bio-
being extracted and solvent type. rtech.2009.03.083..
8 E. A. ABD EL-SALAM AND N. F. S. MORSY

[9] Chen, S.; Zeng, Z.; Hu, N.; Bai, B.; Wang, H.; Suo, Y. and Anthrones. Pharmacology. 1993, 47, 77–85. DOI: 10.1159/
Simultaneous Optimization of the Ultrasound-Assisted 000139846.
Extraction for Phenolic Compounds Content and Antioxidant [24] Kim, K.H.; Tsao, R.; Yang, R.; Cui, S.W. Phenolic Acid Profiles
Activity of Lycium ruthenicum Murr. Fruit Using Response and Antioxidant Activities of Wheat Bran Extracts and the
Surface Methodology. Food Chem. 2018, 242, 1–8. DOI: 10. Effect of Hydrolysis Conditions. Food Chem. 2006, 95, 466–473.
1016/j.foodchem.2017.08.105. DOI: 10.1016/j.foodchem.2005.01.032.
[10] Prabhu, A.A.; Jayadeep, A. Optimization of Enzyme-Assisted [25] Rai, A.; Mohanty, B.; Bhargava, R. Supercritical Extraction of
Improvement of Polyphenols and Free Radical Scavenging Sunflower Oil: A Central Composite Design for Extraction
Activity in Red Rice Bran: A Statistical and Neural Network- Variables. Food Chem. 2016, 192, 647–659. DOI: 10.1016/j.food-
Based Approach. Prep. Biochem. Biotechnol. 2017, 47, 397–405. chem.2015.07.070.
DOI: 10.1080/10826068.2016.1252926. [26] Xu, D.P.; Zheng, J.; Zhou, Y.; Li, Y.; Li, S.; Li, H.B. Ultrasound-
[11] Alara, O.R.; Abdurahman, N.H.; Olalere, O.A. Optimization of Assisted Extraction of Natural Antioxidants from the Flower of
Microwave-Assisted Extraction of Flavonoids and Antioxidants Limonium sinuatum: Optimization and Comparison with
from Vernonia amygdalina Leaf Using Response Surface Conventional Methods. Food Chem. 2017, 217, 552–559. DOI:
Methodology. Food Bioprod. Process. 2018, 107, 36–48. DOI: 10.1016/j.foodchem.2016.09.013.
10.1016/j.fbp.2017.10.007. [27] Yang, L.; Cao, Y.L.; Jiang, J.G.; Lin, Q.S.; Chen, J.; Zh, L.
[12] Michael, P.J.; Steadman, K.J.; Plummer, J.A. The Biology of Response Surface Optimization of Ultrasound-Assisted
Australian Weeds 52. Malva parviflora L. Plant Prot. Q. 2009, Flavonoids Extraction from the Flower of Citrus aurantium L.
24, 1–9. var. amara Engl. J. Sep. Sci. 2010, 33, 1349–1355. DOI: 10.1002/
[13] Fakhfakh, N.; Abdelhedi, O.; Jdir, H.; Nasri, M.; Zouari, N. jssc.200900776.
Isolation of Polysaccharides from Malva aegyptiaca and [28] Ly, M.; Margaritis, A.; Jajuee, B. Effect of Solvent
Evaluation of Their Antioxidant and Antibacterial Properties. Concentration on the Extraction Kinetics and Diffusivity of
Int. J. Biol. Macromol. 2017, 105, 1519–1525. DOI: 10.1016/j. Cyclosporin a in the Fungus Tolypocladium inflatum.
ijbiomac.2017.07.105. Biotechnol. Bioeng. 2007, 96, 67–79. DOI: 10.1002/bit.21198.
[14] € urek, M.; G€
Bilgin, M.; Elhussein, E.A.A.; Ozy€ uçl€
u, K.; Şahin, S. [29] Xu, Y.; Pan, S. Effects of Various Factors of Ultrasonic
Optimizing the Extraction of Polyphenols from Sideritis mon- Treatment on the Extraction Yield of All-Trans-Lycopene from
tana L. using Response Surface Methodology. J. Pharm. Red Grapefruit (Citrus paradise Macf.). Ultrason. Sonochem.
Biomed. Anal. 2018, 158, 137–143. DOI: 10.1016/j.jpba.2018.05. 2013, 20, 1026–1032. DOI: 10.1016/j.ultsonch.2013.01.006.
039. [30] Mustafa, A.; Turner, C. Pressurized Liquid Extraction as a
[15] Escobedo-Flores, Y.; Chavez-Flores, D.; Salmeron, I.; Molina-
Green Approach in Food and Herbal Plants Extraction: A
Guerrero, C.; Perez-Vega, S. Optimization of Supercritical Fluid
Review. Anal. Chim. Acta. 2011, 703, 8–18. DOI: 10.1016/j.aca.
Extraction of Polyphenols from Oats (Avena sativa L.) and
2011.07.018.
Their Antioxidant Activities. J. Cereal Sci. 2018, 80, 198–204. Savikin, K.P.; Zdunic, G.M.; D - ord-evic, V.B.; Bugarski, B.M.;
[31]
DOI DOI: : 10.1016/j.jcs.2018.03.002.
God-evac, D.M.; Jovanovic, A.A.; Pljevljakusic, D.S.
[16] Zhou, Z.; Shao, H.; Han, X.; Wang, K.; Gong, C.; Yang, X. The
Optimization of the Extraction Process of Polyphenols from
Extraction Efficiency Enhancement of Polyphenols from Ulmus
Thymus serpyllum L. herb Using Maceration, Heat- and
pumila L. Barks by Trienzyme-Assisted Extraction. Ind. Crops
Ultrasound-Assisted Techniques. Sep. Purif. Technol. 2017, 179,
Prod. 2017, 97, 401–408. DOI: 10.1016/j.indcrop.2016.12.060.
_ Modeling and Optimization I: Usability of 369–380. DOI: 10.1016/j.seppur.2017.01.055.
[17] Baş, D.; Boyacı, IH.
[32] Afolayan, A.J.; Aboyade, O.M.; Sofidiya, M.O. Total Phenolic
Response Surface Methodology. J. Food Eng. 2007, 78, 836–845.
DOI: 10.1016/j.jfoodeng.2005.11.024. Content and Free Radical Scavenging Activity of Malva parvi-
[18] Rambo, D.F.; Biegelmeyer, R.; Toson, N.S.B.; Dresch, R.R.; flora L. (Malvaceae). J. Biol. Sci. 2008, 8, 945–949. DOI: 10.
Moreno, P.R.H.; Henriques, A.T. Box–Behnken Experimental 3923/jbs.2008.945.949.
Design for Extraction Optimization of Alkaloids from Erythrina [33] Bouriche, H.; Meziti, H.; Senator, A.; Arnhold, J. Anti-
verna Vell. trunk Barks and LC Method Validation. Ind. Crops Inflammatory, Free Radical-Scavenging, and Metal-Chelating
Prod. 2019, 133, 250–258. DOI DOI: .org/10.1016/j.indcrop. Activities of Malva parviflora. Pharm. Biol. 2011, 49, 942–946.
2019.03.030. DOI: 10.3109/13880209.2011.558102.
[19] dos Santos, W.N.L.; da Silva, E.G.P.; David, J.M.; Ferreira, [34] Piluzza, G.; Bullitta, S. Correlations between Phenolic Content
S.L.C.; Brand~ao, G.C.; dos Reis, P.S.; Matos, G.D.; Portugal, and Antioxidant Properties in Twenty-Four Plant Species of
L.A.; Souza, A.S.; Bruns, R.E.; et al. Box-Behnken Design: An Traditional Ethnoveterinary Use in the Mediterranean Area.
Alternative for the Optimization of Analytical Methods. Anal. Pharm Biol. 2011, 49, 240–247. DOI: 10.3109/13880209.
Chim. Acta. 2007, 597, 179–186. DOI: 10.1016/j.aca.2007.07. [35] Jeganathan, P.M.; Venkatachalam, S.; Karichappan, T.;
011.. Ramasamy, S. Model Development and Process Optimization
[20] Granato, D.; de Ara ^
ujo Calado, V.O.M.; Jarvis, B. Observations for Solvent Extraction of Polyphenols from Red Grapes Using
on the Use of Statistical Methods in Food Science and Box-Behnken Design. Prep. Biochem. Biotechnol. 2014, 44,
Technology. Food Res. Int. 2014, 55, 137–149. DOI: 10.1016/j. 56–67. DOI: 10.1080/10826068.2013.791629.
foodres.2013.10.024. [36] Terninko, I.I.; Nemyatykh, O.D.; Sakipova, Z.B.; Kuldyrkaeva,
[21] Myers, R.H.; Montgomery, D.C.; Anderson-Cook, C.M. E.V.; Onishschenko, U.E. Phytochemical and Pharmacological
Response Surface Methodology: Process and Product Vectors from Malva sylvestris L. for Application in
Optimization Using Designed Experiment; Wiley: Hoboken, New Dermatological Practice. Pharm. Pharm. Chem. J. 2017, 50,
Jersey, 2009. 805–809. DOI: -10.1007/s11094-017-15360. DOI: 10.1007/
[22] Arnous, A.; Makris, D.P.; Kefalas, P. Correlation of Pigment s11094-017-1536-0.
and Flavanol Content with Antioxidant Properties in Selected [37] Ben Saad, A.; Rjeibi, I.; Brahmi, D.; Smida, A.; Ncib, S.; Zouari,
Aged Regional Wines from Greece. J. Food Compos. Anal. N.; Zourgui, L. Malva sylvestris Extract Protects upon Lithium
2002, 15, 655–665. DOI: 10.1006/jfca.2002.1070. Carbonate-Induced Kidney Damages in Male Rat. Biomed.
[23] Malterud, K.E.; Farbrot, T.L.; Huse, A.E.; Sund, R.B. Pharmacother. 2016, 84, 1099–1107. DOI: 10.1016/j.biopha.
Antioxidant and Radical Scavenging Effects of Anthraquinones 2016.10.026.