Extraction from Castanea sativa leaves

Chemical Engineering Journal 203 (2012) 101–109
Contents lists available at SciVerse ScienceDirect
Chemical Engineering Journal
journal homepage: www.elsevier.com/locate/cej
Optimization of antioxidants – Extraction from Castanea sativa leaves
Beatriz Díaz Reinoso a,b, Diana Couto c, Andrés Moure a,b,⇑, Eduarda Fernandes c, Herminia Domínguez a,b,
Juan Carlos Parajó a,b
a
Departamento de Enxeñería Quimica, Universidade de Vigo (Campus Ourense), Edificio Politécnico, As Lagoas, 32004 Ourense, Spain
CITI-Universidade de Vigo, Parque Tecnolóxico de Galicia, Rúa Galicia no. 2, 32900 Ourense, Spain
c
REQUIMTE, Departamento de Ciências Químicas, Faculdade de Farmácia, Universidade do Porto, Rua de Jorge Viterbo Ferreira 228, 4050-313 Porto, Portugal
b
h i g h l i g h t s
" Response surface methodology was assessed to determine the optimum conditions of extraction.
" The highest phenolic release was performed with aqueous extraction.
" Antioxidant capacity of chestnut leaves extracts was screened using scavengers and chelating assays.
" Under optimum conditions the extracts show radical scavenging capacity similar to that of BHA.
a r t i c l e
i n f o
Article history:
Received 19 July 2011
Received in revised form 22 June 2012
Accepted 26 June 2012
Available online 4 July 2012
Keywords:
Castanea sativa leaves
Solvent extraction
Phenolic compounds
Antioxidant activity
a b s t r a c t
The extraction of Castanea sativa leaves (CsLs) with selected solvents (96% ethanol, methanol and acidified water) was assessed by means of a set of experiments following a Box–Wilson central composite circumscribed (CCC) design. CsL extraction was assessed on the basis of experimental results determined for
the yield in soluble compounds, the yield in phenolics and the antioxidant activity of extracts. The in vitro
antioxidant activity was measured in terms of the scavenging abilities for a,a-diphenyl-b-picrylhydrazyl
(DPPH) radical. Selected extracts were also assayed for scavenging activity against reactive oxygen species (ROSs) and reactive nitrogen species (RNSs). The highest extraction yields were obtained using methanol, whereas water allowed a selective extraction (with yields in phenolics up to 9 mg Gallic Acid
Equivalent/100 mg CsL) and provided the most concentrated (54 mg GAE/100 mg solid) and active
extracts. The major compounds identified in extracts were gallic acid, protocatechuic acid, 4-hydroxybenzoic acid, vanillic acid, rutin, quercetin and apigenin.
Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction
Castanea sativa Mill. is a species of the flowering plant family
Fagaceae, mainly cultivated in temperate regions (Asia, Southern
Europe and North of Africa). The seasonal fruit is a nut, commonly
named sweet chestnut or marron, which is collected in autumn. It
is a traditional basic foodstuff and can be processed into different
elaborated and progressively diversified food products. Recently,
the nuts became important in human health owing to their applications as a component of gluten-free diets and as a source of
essential fatty acids.
C. sativa leaves (CsL) are used in folk medicine to treat cough,
diarrhea and rheumatic conditions, lower back pain, and stiff joints
or muscles. As in other medicinal plants, some of the active sub⇑ Corresponding author at: Departamento de Enxeñería Quimica, Universidade de
Vigo (Campus Ourense), Edificio Politécnico, As Lagoas, 32004 Ourense, Spain. Tel.:
+34 988 368 892; fax: +34 988 387 001.
E-mail address: [email protected] (A. Moure).
1385-8947/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cej.2012.06.122
stances have been identified as phenolic compounds. The most
abundant of them in CsL are ellagic acid and gallic acid derivatives
[1]. These compounds present biological activities, including
ability for protecting against oxidative stress-related diseases and
antimicrobial activity. The capacity of aqueous, methanol, and
ethyl acetate extracts from CsL to scavenge a,a-diphenyl-b-picrylhydrazyl (DPPH), superoxide radical ðO
2 Þ, and hydroxyl radical
(HO) was already reported [2–5]. Water:acetone CsL extracts were
further washed with ethyl acetate to separate phenolic compounds
from tannins. This fraction showed antioxidant activity comparable to the ones of reference pure antioxidants, and higher than
the ones determined for solvent purified fractions [3]. Ethanol:water extracts of CsL showed scavenging activities against
hydrogen peroxide (H2O2) and singlet oxygen (1O2) comparable
to that of ascorbic acid, and were more active against O
2 than Trolox. In comparative terms, extracts were more active for scavenging
reactive nitrogen species (RNSs) than for reactive oxygen species
(ROSs) [6]. Water extracts of CsL also showed reducing power
and ability to cause the inhibition of b-carotene bleaching, lipid
102
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
peroxidation and hemolysis [4,7]. The same properties were confirmed for extracts obtained with 50% ethanol [5], which also
showed ability to protect against DNA damage in rat pancreatic
b-cells [5].
CsL extracts have been reported to prevent changes related to
the fluidity and integrity of the erythrocyte membranes [1]. The
ethyl acetate soluble fraction of the CsL aqueous extracts displayed
potent antibacterial action against Gram-positive and Gram-negative bacteria, which was ascribed to the presence of flavonoids [8];
whereas ethanol extracts showed strong antibacterial activity
against Gram-positive bacteria [9].
The utilization of solvent extracts from CsL in cosmetic compositions has been claimed [10]. A CsL extract obtained with ethanol–
water (containing rutin, ellagic acid, hyperoside, isoquercitrin and
chlorogenic acid) was considered safe and suitable for preventing
and treating oxidative stress-related diseases and photoageing
[11]. Flavonoid glycosides obtained from CsL have been considered
useful for the treatment and prevention of insulin-resistance diseases [12].
The extraction and purification of biologically active compounds
from natural sources is becoming a growing research field, finding
applications in the formulation of ingredients and additives for
food, pharmaceuticals and cosmetics. The yields in bioactive compounds obtained from different plant materials are strongly influenced by the extraction and purification procedures. The type of
solvent is an influential variable: when CsL phenolics are the target
products, water [7], methanol [10], acetone, ethylacetate [3], ethanol-aqueous solutions [1,6,9,11] are suitable agents for extraction.
Recent literature has been reported on the extraction of antioxidant phenolic compounds from CsL, with emphasis on the type of
solvent, characterization of the extracts, and comparative evaluation of different parts of the plant. Despite the increased extraction
yields and antioxidant potential of phenolics from skins [1,7],
leaves can be easily collected and possess a traditional background.
To our knowledge, the combined effects of the major operational
variables (temperature, contact time and liquid–solid ratio) on
the extraction of phenolics from CsL has not been reported before.
This study provides an experimental assessment of the effects
of selected operational variables on the extraction of phenolic compounds with antioxidant activity from CsL. The composition of CsL
was measured, and the extraction of target fractions was assessed
using a set of experiments with a Box–Wilson central composite
circumscribed (CCC) design. In experiments, the yields in soluble
solids and phenolics, as well as the antioxidant activities of extracts, were measured. The in vitro antioxidant activity was characterized using the DPPH radical scavenging test. Selected extracts
were also assayed for scavenging activity against ROS (O
2 , H2O2,
peroxyl radical ROO, hypochlorous acid HOCl, and 1O2) and RNS
(nitric oxide NO and peroxynitrite ONOO).
ide, sodium hypochlorite solution containing 4% available chlorine,
diethylenetriaminepentaacetic acid (DTPA), 3-(aminopropyl)-1hydroxy-3-isopropyl-2-oxo-1-triazene (NOC-5), b-nicotinamide
adenine dinucleotide (NADH), phenazine methosulfate (PMS),
nitroblue tetrazolium chloride (NBT) and lucigenin were obtained
from Sigma–Aldrich (St. Louis, USA). a,a0 -Azodiisobutyramidine
dihydrochloride (AAPH), histidine and Trolox were obtained from
Fluka Chemie GmbH (Steinheim, Germany). Fluorescein sodium
salt was obtained from Aldrich (Milwaukee, USA).
2.2. Solid–liquid extraction
Ground CsL samples were in contact for the desired time with
the selected solvent (aqueous ethanol (96%) denoted as ES, methanol denoted as MS, or acidified water denoted as AWS) at the selected liquid to solid ratio and temperature in capped Erlenmeyer
flasks kept in an orbital shaker at 120 rpm. Phase separation was
accomplished by filtration. The liquid phase was vacuum evaporated and freeze-dried to obtain a crude extract. The extraction
yield was determined gravimetrically.
2.3. Experimental plan
The experimental plan followed a three-factor, central composite circumscribed (CCC) involving 20 experiments with six replicates in the central point of the experimental domain. The
independent variables, their nomenclatures in dimensional and
dimensionless terms and their variation ranges were as follows:
contact time (t, X1, 30–90 min); temperature (T, X2, 25–50 °C)
and liquid to solid ratio (L/S, X3, 15–25 v/w). The particle size distribution (in the range 0.25–1 mm) was constant in all experiments. Three solvents (96% ethanol, methanol, acidified water)
were employed in individual sets of experiments. The range selected for the extraction time covered the conditions of practical
interest, whereas the temperature range was fixed keeping in mind
a compromise between extraction yield, extraction rate and limitation of thermal denaturation. The values assayed for the liquid to
solid ratio was in a wide interval where mass transfer is not limited. Table 2 lists the operational conditions assayed for individual
experiments.
The measured effects were the Soluble Extraction Yield (SEY, mg
soluble solids/100 mg dry CsL), the phenolic extraction yield (PEY,
mg GAE/100 mg dry CsL), and the DPPH radical scavenging capacity
(DPPH RS), which was selected to assess the antioxidant activity.
The experimental data determined for each of the three solvents
were adjusted according to the following second order equation:
Y ¼ b0 þ
k
X
i¼1
2. Materials and methods
2.1. Materials
Chestnut CsL from the 2005 harvest were locally collected. Airdried to 9.7% average moisture, ground, sieved to pass 1 mm, and
stored in sealed plastic bags in a dry and dark. All the chemicals
and reagents were of analytical grade. Ethanol, hexane and methanol were purchased from Panreac (Drogallega, S.L, Spain). Butylhydroxy anisol (BHA) and butyl hydroxytoluene (BHT) were
purchased from Analema (Drogallega, S.L, Spain). Gallic acid, protocatechuic acid, 4-hydroxybenzoic acid, vanillic acid, quercetin and
apigenin were obtained from Sigma–Aldrich (St. Louis, USA) and
rutin from Extrasynthèse (Lyon Nord, France). Dihydrorhodamine
123 (DHR), 4,5-diaminofluorescein (DAF-2), 30% hydrogen perox-
bi X i þ
k
k1 X
k
X
X
bii X 2i þ
bij X i X j
i¼1
ð1Þ
i¼1 j¼2
i<j
where X1, X2, . . ., Xk are the independent, dimensionless variables
affecting the responses (Y0 s); b0, bi (i = 1, 2, . . ., k), bii (i = 1, 2, . . ., k),
and bij (i = 1, 2, . . ., k; j = 1, 2, . . ., k) are the regression coefficients
for intercept, linear, quadratic, and interaction terms, respectively;
k is the number of variables. The Statsoft vs 5.0 software was used
in calculations.
2.4. Analytical methods
Ash content was determined by calcination at 550 °C. Extractives (in hexane, 80% ethanol, and toluene:ethanol) were determined gravimetrically after Soxhlet extraction. Ground CsL
samples were subjected to moisture determination and to quantitative acid hydrolysis with 72% sulfuric acid, following standard
methods [15]. The solid residue after hydrolysis was considered
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
as Klason lignin. Monosaccharides and acetic acid were determined
by HPLC using a Hewlett–Packard chromatograph fitted with a
refractive index detector (temp., 40 °C). Other analysis conditions
were: column, ION-300 (Transgenomic Inc., USA); mobile phase,
3 mM H2SO4; flow, 0.4 mL/min. The yield in soluble compounds
was determined gravimetrically. The phenolic content was determined by the Folin–Ciocalteau method using gallic acid as a standard. Uronic acids were determined spectrophotometrically [16] as
galacturonic acid equivalents (GAEs).
Other analyses were carried out using an Agilent HPLC 1100
instrument equipped with a Waters Spherisorb ODS-2 column
and DAD detector, operating at a flow rate of 1 mL/min. Elution followed the following nonlinear gradient: solvent A (acetonitrile/
5%(v/v) formic acid in water, 10:90); solvent B (acetonitrile/5%(v/
v) formic acid in water, 90:10); elution periods: 0 min, 100% A,
0% B; 40 min, 85% A, 15% B; 45 min, 0% A, 100% B; 55 min, 0% A,
100% B; 60 min, 100% A, 0% B; 65 min, 100% A, 0% B.
2.5. Antioxidant activity assays
In vitro antioxidant assays were performed using a microplate
reader (Synergy HT, BIO-TEK), for fluorescence, absorbance in
UV/vis and luminescence measurements, equipped with a thermostat. Control assays with commercial synthetic antioxidants were
run in parallel for comparative purposes.
2.5.1. DPPH (a,a-Diphenyl-b-picrylhydrazyl) radical scavenging assay
Two milliliters of a 3.6 105 M methanolic solution of DPPH
were added to 50 lL of a methanolic solution of the considered extract. The decrease in absorbance at 515 nm was recorded after
16 min. The Half Maximal Effective Concentration (EC50) was calculated as the amount of methanolic extract causing a 50% inhibition of the DPPH radical.
103
cent DHR [18]. 1O2 was generated by the thermal decomposition of
endoperoxide disodium 3,30 -(1,4-naphthalene) bispropionate
(NDPO2) [19]. Each study corresponds to four experiments, performed in triplicate.
2.5.7. Peroxyl radical (ROO) scavenging assay
The ROO scavenging activity measured the effects of the tested
extracts on the fluorescence decay resulting from ROO-induced
oxidation of fluorescein, and the results were expressed in terms
of the ‘‘Oxygen Radical Absorbance Capacity’’ (ORAC) [18]. Each
study corresponds to four experiments, performed in triplicate.
2.5.8. Nitric oxide (NO) scavenging assay
The NO scavenging activity measured the inhibition percentage
of NO-induced oxidation of non-fluorescent 4,5-diaminofluorescein (DAF-2) to the fluorescent triazolofluorescein (DAF-2T) [18].
Each study corresponds to four experiments, performed in
triplicate.
2.5.9. Peroxynitrite (ONOO) scavenging assay
The ONOO scavenging activity measured the effects of the
tested extracts on the inhibition percentage of the ONOO-induced
oxidation of non-fluorescent DHR to fluorescent rhodamine 123
[18]. ONOO was synthesized as described before [20]. Parallel assays were performed in the presence of 25 mM NaHCO3 in order to
simulate the physiological CO2 concentrations, where the reaction
between ONOO and bicarbonate is predominant, with a very fast
rate constant (k2 = 3–5.8 104 M1 s1) [21]. Each study corresponds to four experiments, performed in triplicate.
3. Results and discussion
3.1. Composition
2.5.2. Trolox equivalent antioxidant capacity (TEAC)
ABTS radical cation ðABTSþ Þ was produced according as described previously [17]. Assays were run with solvent blanks and
extracts, and the percentage inhibition of absorbance was referred
to the concentration of extracts and Trolox.
2.5.3. Superoxide radical ðO
2 Þ scavenging assay
The O
2 scavenging activity was determined by monitoring the
effects of the tested extracts on the O
2 -induced reduction of
nitroblue tetrazolium (NBT) at 560 nm [18]. O
2 was generated
by the b-nicotinamide adenine dinucleotide/phenazine methosulfate/oxygen (NADH/PMS/O2) system. The results were expressed
in terms of inhibition percentage of NBT reduction to diformazan.
Each result corresponds to four experiments, performed in
triplicate.
2.5.4. Hydrogen peroxide (H2O2) scavenging assay
The H2O2 scavenging activity measured the inhibition (in percentage) of the H2O2-induced oxidation of lucigenin [18]. Each
study corresponds to four experiments, performed in triplicate.
2.5.5. Hypochlorous acid (HOCl) scavenging assay
The HOCl scavenging activity measured the inhibition percentage of HOCl-induced oxidation of dihydrorhodamine 123 (DHR) to
rhodamine 123 [18]. HOCl was prepared by adjusting the pH of a
1% (m/v) solution of NaOCl to 6.2 with dropwise addition of
10%H2SO4. Each study corresponds to four experiments, performed
in triplicate.
2.5.6. Singlet oxygen (1O2) scavenging assay
The 1O2 scavenging activity was measured by monitoring the
inhibition (in percentage) of 1O2-induced oxidation of non-fluores-
Table 1 lists compositional data of CsL. Lignin was the major
structural fraction, followed by hemicelluloses, which were mainly
constituted of xylose and uronic acids. In comparison with the data
reported for green leaves from South West England [22], the results determined for CsL showed similar cellulose content and a
significantly higher lignin content. Extractives in 80% ethanol accounted for 13.5% of the dry material, below the ranges reported
for 80% methanol (17–20%) [10] and hot water (20.9%) [7].
Extractables in the ethanol solution (80% v/v) yield an extract
with a phenolic concentration intermediate between those reported for hot water [3,7], methanol [3] and aqueous ethanol
[9,11] and 80% ethanol was a solvent more selective for phenolic
compounds than toluene:ethanol and hexane. The yield in soluble
solids, phenolic content and scavenging capacity for pro-oxidant
species obtained from less polar solvents than acidified water solvent (AWS) was lower, corroborating previously reported findings
[3]. Reduced yields and activities have been reported for extracts
obtained after sequential extractions with hexane, chloroform,
ethyl acetate, methanol, and water [8], although ethyl acetate
extracts eluted by medium-pressure liquid chromatography with
methanol were highly active [3]. The general trend of antioxidant
activity is closely dependent on the solvent polarity, the antioxidant potential and the sequence found in this work was
BHA > 80% ethanol > toluene:ethanol (2:1) ffi BHT hexane.
3.2. Optimization of the extraction conditions
In order to achieve maximum yields of active extracts, the joint
optimization of the most influential variables was addressed. The
liquid phases for extraction were selected on the basis of previous
data and in practical aspects (including availability, cost, toxicity,
104
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
Table 1
Composition of Castanea sativa leaves (CsLs). Data expressed on dry basis.
Structural components
Content (g/100 g CsL)
Cellulose
Hemicelluloses
Xylan
Arabinan
Acetyl groups
Uronic acids
Klason lignin
19.4
24.3
9.41
3.28
1.66
9.39
37.2
Non-structural components
Content (g/100 g CsL)
Ash
Extractives
4.34
80% Ethanol
14.0 g GAE/100 g dry extract
EC50, DPPH = 1.49 mg extract/mL
13.5
Hexane
0.42 g GAE/100 g dry extract
EC50, DPPH = 16.42 mg extract/mL
9.48
Toluene:ethanol (1:2)
3.04 g GAE/100 g dry extract
EC50, DPPH = 2.73 mg extract/mL
12.8
Table 2
Operational conditions assayed, expressed in terms of dimensional and dimensionless
independent variables.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Exp.
SEY (mg soluble solids/100 mg dry CsL)
ES
GAEs: Gallic Acid Equivalents.
Exp.
Table 3
Experimental and calculated values of the solubles extraction yield (SEY).
Variables
t min (X1)
T °C (X2)
L/S v/w (X3)
30 (1)
30 (1)
30 (1)
30 (1)
90 (1)
90 (1)
90 (1)
90 (1)
9.6 (1.68)
110.4 (1.68)
60 (0)
60 (0)
60 (0)
60 (0)
60 (0)
60 (0)
60 (0)
60 (0)
60 (0)
60 (0)
25 (1)
50 (1)
25 (1)
50 (1)
25 (1)
50 (1)
25 (1)
50 (1)
37.5 (0)
37.5 (0)
16.5 (1.68)
58.5 (1.68)
37.5 (0)
37.5 (0)
37.5 (0)
37.5 (0)
37.5 (0)
37.5 (0)
37.5 (0)
37.5 (0)
25 (1)
25 (1)
15 (1)
15 (1)
25 (1)
25 (1)
15 (1)
15 (1)
20 (0)
20 (0)
20 (0)
20 (0)
11.6 (1.68)
28.4 (1.68)
20 (0)
20 (0)
20 (0)
20 (0)
20 (0)
20 (0)
and extracting capacity). Ethanol was chosen owing to its frequent
food applications [6,9]. Methanol is a reference solvent with high
extracting ability [10] and leads to fractions of high radical scavenging capacity [3]. Acidified water is widely used to extract polar
compounds [23] and it is interesting because it could serve for the
development of alternative, solvent-free processes of limited
environmental impact, and increases the extraction of tannins
[3]. CsL were extracted with each solvent under the experimental
conditions shown in Table 2.
The statistical optimization procedure applied in the present
study enables the evaluation of the effects caused by both individual independent variables and interactions among them.
Related studies have been reported on the maximization of the
extraction of phenolic compounds from berries [24], evening
primrose meal [25], black currants [13], grape pomace [26] or
wheat [14].
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
MS
AWS
Exp
Cal
Exp
Cal
Exp
Cal
9.86
9.66
9.14
9.10
7.02
9.46
8.32
13.76
6.00
12.25
9.28
11.02
9.32
11.20
12.36
12.54
12.01
12.13
10.22
10.13
10.20
8.93
7.86
8.17
8.11
10.90
9.21
13.58
7.62
10.41
8.73
11.34
10.29
10
11.57
11.57
11.57
11.57
11.57
11.57
11.68
12.68
10.52
11.15
8.16
12.55
10.89
14.97
10.70
12.40
11.00
12.56
11.18
14.39
9.78
10.00
9.75
10.50
9.41
9.32
12.60
12.92
10.38
10.36
9.43
13.17
11.13
14.53
10.79
11.63
9.88
13.01
12.09
12.81
9.81
9.81
9.81
9.81
9.81
9.81
12.16
9.66
12.24
12.72
12.95
18.17
11.01
14.27
10.66
14.17
9.13
16.21
9.36
12.27
12.21
12.34
11.41
10.39
12.44
12.71
10.91
10.73
11.80
12.11
12.81
17.87
9.20
14.77
11.03
14.85
10.93
15.45
10.40
12.27
11.88
11.88
11.88
11.88
11.88
11.88
3.2.1. Solubles extraction yield (SEY)
The maximum extraction yield (18.2%) was reached using
water as a solvent under the conditions of experiment 6 (operation for 90 min at 50 °C using 25 mL AWS/g CsL). The yield was
close to the value reported for extracts from leaves from C. sativa
grown in a near geographical area, obtained with boiling water
for 30 min [7]. Ethanol (ES) and methanol (MS) provided lower
yields than AWS, with maximum values in experiment 8 (performed for 90 min at 50 °C, using 15 mL extracting agent/g CsL).
The lowest yield (6.00 mg soluble solids/mg dry CsL) corresponded to operation with ES for 9.6 min at 37.5 °C using 20 mL
ethanol/g CsL, and for 90 min at 25 °C using 25 mL ethanol/g CsL.
In experiments with ethanol, increasing the contact time while
keeping temperature and liquid to solid ratio constant doubled
the extraction yield (see experiments 15–20, performed in the
central point of the experimental domain). Although a variety of
compounds are soluble in all the liquid phases assayed in this
study, both monomeric sugars (glucose, xylose, arabinose) and sugar oligomers were preferentially solubilized by AWS (data not
shown).
The experimental results in Table 3 were adjusted to Eq. (1) by
minimization of the deviation squares, enabling the calculation of
the results for given operation conditions. For comparative purposes, the data calculated for the considered effects are included
in Tables 3–5 together with the experimental results. Only the
model coefficients significant at a confidence level > 95% (based
on a t-test) were considered for model predictions. The model coefficients are listed in Table 6, as well as the results of the ANOVA.
The equations employed to predict the extraction yield of soluble compounds (SEY), expressed in terms of the dimensional independent variables, and including just the significant terms, were as
follows:
SEYES ¼ 11:6 0:904 t 2
ð2Þ
SEYMS ¼ 9:81 þ 0:932 T þ 0:854 t T 0:981 t L=S
þ 0:931 L=S2
SEYAWS ¼ 11:9 þ 1:14 t þ 1:34 T þ 1:31 t T þ 1:12 t L=S
ð3Þ
ð4Þ
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
values of time and temperature. Response surface for ES showed
a different behavior at low T. The effect of solid to liquid ratio is
showed in Fig. 1A. Models predict different behavior for the solvents used. Ethanol seems not affected by the liquid to solid ratio.
Table 4
Experimental and calculated values for the phenolic extraction yield (PEY).
Exp.
PEY (mg GAE/100 mg dry CsL)
ES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
MS
AWS
Exp
Cal
Exp
Cal
Exp
Cal
0.68
0.75
0.35
0.84
0.81
2.24
0.67
1.38
0.24
1.20
0.63
1.43
0.68
0.88
0.98
1.02
0.81
0.95
1.00
0.91
0.50
0.77
0.53
0.65
0.89
1.95
0.55
1.46
0.29
1.30
0.60
1.60
0.66
1.05
0.94
0.94
0.94
0.94
0.94
0.94
0.94
1.88
1.17
1.45
1.70
2.02
1.50
1.26
1.62
2.27
1.95
2.49
1.91
2.03
1.19
1.23
1.67
1.67
1.14
1.13
1.12
2.04
1.42
1.72
1.91
2.25
1.83
1.56
1.35
1.87
1.61
2.15
1.47
1.80
1.36
1.36
1.36
1.36
1.36
1.36
4.14
3.73
3.59
4.45
4.39
9.16
2.36
4.08
4.78
6.10
4.61
6.89
5.08
6.49
3.50
3.23
4.44
4.00
3.75
4.31
4.05
4.57
4.95
4.57
5.28
8.81
2.55
5.19
3.94
5.49
3.70
6.35
3.91
6.21
3.91
3.91
3.91
3.91
3.91
3.91
Table 5
Experimental and calculated values of variable EC50.
Exp.
EC50 (mg extract/mL)
ES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
MS
105
AWS
Exp
Cal
Exp
Cal
Exp
Cal
1.77
1.75
2.91
1.61
0.98
0.45
1.54
1.04
3.71
1.52
1.74
1.27
2.07
1.54
1.56
1.66
1.44
1.30
1.54
1.56
2.18
1.96
2.90
2.05
0.93
0.86
1.73
1.03
3.29
1.38
1.61
0.84
1.90
1.15
1.52
1.52
1.52
1.52
1.52
1.52
1.18
0.87
1.16
0.73
0.44
0.36
1.61
1.14
0.70
0.54
0.48
0.51
0.67
0.91
0.52
0.59
0.47
0.55
0.58
0.45
1.08
0.98
0.88
0.53
0.41
0.40
1.27
1.00
0.86
0.70
0.82
0.51
1.13
0.79
0.52
0.52
0.52
0.52
0.52
0.52
0.37
0.49
0.52
0.40
0.31
0.40
0.68
0.52
0.31
0.39
0.34
0.36
0.34
0.29
0.38
0.41
0.40
0.40
0.40
0.41
0.33
0.47
0.44
0.33
0.28
0.38
0.60
0.45
0.39
0.45
0.42
0.41
0.46
0.31
0.40
0.40
0.40
0.40
0.40
0.40
The three equations were significant at the 95% confidence level. Considering the values of the model coefficients in Table 6, it
can be inferred that extraction time and temperature were the
most influential independent variables, no matter the extraction
agent employed. The effects of the extraction time came mainly
from the interaction terms temperature – L/S; whereas the quadratic effects of time and liquid-to-solid ratio were influential in
the extractions with ES or MS. The response surfaces describing
the dependence of SEY on the selected independent variables are
shown in Fig. 1A for the three liquid phases considered. A related
behavior was observed for SEY in extractions performed with
AWS and MS, for which better yields were obtained at time and
temperature values around 100 min and 55 °C respectively. Comparatively, models for AWS and ES predict maximal SEY values
(around 14% and 20%) for ES and AWS, respectively) for similar
3.2.2. Phenolics extraction yield (PEY)
In the studied operational range, the maximum value of PEY
(9.16 mg GAE/100 mg CsL) achieved with AWS corresponded to
experiment 6 (performed for 90 min at 50 °C using 25 mL acidified
water/g CsL), and was considerably higher than the value obtained
with ES (2.24 mg GAE/100 mg CsL in experiment 6, performed for
90 min at 50 °C using 25 mL aqueous ethanol/g CsL), or with MS
(2.49 mg GAE/100 mg CsL obtained in experiment 12, performed
for 60 min at 58.5 °C using 20 mL methanol/g CsL). In comparison
with results reported in related studies, the phenolic yields
achieved in ES extractions were lower than those reported for
ultrasound-assisted ethanol extraction of leaves from the Italian
‘‘marrone’’ cultivar [9], and when the solvent was methanol the
yields achieved also were lower than those reported for solvent
extraction from a commercial powder plant [3].
The ratio between the yields in phenolics and soluble solids
(PEY/SEY), which measured the phenolic content in terms of mass
ratio respect the total weight of extract, was higher than 50% for
aqueous extracts, doubling the maximal values reached with alcohols. So, under optimal conditions, water was more selective than
ethanol for phenolics extraction than the rest solvents employed in
the present work, as well as in literature studies with the same raw
material [6,11], including extracts obtained with boiling water [7]
and with a sequence of solvents [3]. As a general trend, the phenolic yields obtained with water were between 2–4 or 3–10 times
higher than those obtained with methanol or ethanol, respectively
(Table 4).
Starting from the results listed in Table 6, it can be inferred that
the empirical equations derived from PEY with ES and AWS solvents (considering only the coefficients significant at the 95% confidence level) are as follows:
PEYES ¼ 0:938 þ 0:300 t þ 0:296 T þ 0:116 L=S þ 0:197 t T
ð5Þ
PEYAWS ¼ 3:91 þ 0:788 T þ 0:682 L=S þ 0:908 t L=S
ð6Þ
Both equations were significant at the 95% confidence level,
whereas no significant dependence of PEY on the independent
variables was observed in extractions performed with methanol.
Besides the significant terms shown in Eqs. (5) and (6), the contribution of the interaction term time to temperature was significant at the 90% level in ethanol extractions. Fig. 1B shows the
response surfaces calculated for PEY as a function of the extraction
time and temperature (at the optimal L/S) and as a function of temperature and L/S (at the optimal contact time) for ES and AWS
extractions. The amount of solubilized phenolics increased with
temperature in both solvents, reaching maximal values (around
6 mg GAE/100 mg dry CsL) operating with AWS at 58.5 °C for
9.54 min. The solubilized phenolics with AWS decreased with time
in such a way that prolonged contact times provided the lowest
PEY values. This behavior is possibly due to a thermal decomposition of phenolic compounds in acidified water at high temperatures. A similar behavior during conventional solvent extraction
was reported for compounds from wheat [14], black currants
[13] or from berries [23].
ES showed an inverse behavior, with maximal values of solubilized phenolics at temperatures higher than 50 °C and 90 min. In
this case, as expected, increased temperatures resulted in higher
PEY, particularly, where the enhanced solubility and diffusivity of
solutes play a major role.
1.57
0.079
3.2.3. DPPH radical scavenging activity
Extracts with high radical scavenging capacity (DPPH RS), measured as EC50, were obtained in AWS extractions performed under
the conditions of experiments 5 (in which the phases were in contact for 90 min at 25 °C using 25 mL of acidified water/g CsL), 9
(9.6 min, 37.5 °C, 20 mL of acidified water/g CsL) and 14 (60 min,
37.5 °C, 28.4 mL of acidified water/g CsL) with an half maximal
Effective Concentration of 0.31 and 0.29 mg aqueous extract/mL
(Table 5). These experiments resulted in extracts with the highest
phenolic concentrations (0.338, 0.448 and 0.529 mg GAE/mg soluble solids, respectively). AWS extraction under the conditions of
experiment 14 led to the extract with the highest concentration
of phenolics, which also showed the highest antioxidant activity.
The lowest EC50 of the extracts, indicative of optimal antioxidant activity, was below the one determined for BHA
(EC50 = 0.25 mg/mL). The most active alcoholic extracts were produced in experiment 6 (performed for 90 min at 50 °C using
25 mL of alcohol/g CsL), with EC50 = 0.45 mg ethanolic extract/mL
and 0.36 mg methanolic extract/mL. It can be noted that the temperature needed to obtain extracts with high antioxidant activity
was lower for water extractions (25–37.5 °C) than for alcohols
(50 °C). The dependence of the radical scavenger capacity of extracts obtained with the three solvents studied on the operational
variables are given by the following equations:
2.18
0.261
6.49
0.376
4.00
0.973
0.586
0.662
0.854
0.783
0.011
0.052a
0.062a
0.037
0.019
0.158
0.754a
0.908b
0.223
0.023
0.264b
0.064
0.094
0.052
0.156b
0.288b
0.105
0.002
0.285
0.394
0.406
0.009
0.008
0.004
0.048
0.092
0.101
0.566c
0.230b
0.223b
0.462
0.788b
0.682b
0.018
0.003
0.046a
0.517
MS
1.52
3.91
EC50 (mg extract/mL)
ES
AWS
0.396
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
AWS
106
DPPH RSES ¼ 1:52 0:566 t 0:230 T 0:223 L=S þ 0:288 t2
1.37
0.390
0.552
0.143
0.094
0.153
0.087
0.185
0.096
0.156
0.162
0.099
1.36
ES
MS
PEY (mg GAE/100 mg dry CsL)
ES
SEY (mg/100 mg dry CsL)
DPPH RSMS ¼ 0:517 0:264 t L=S þ 0:155 L=S2
ð8Þ
DPPH RSAWS ¼ 0:396 0:046 L=S 0:052 t L=S þ 0:062 T L=S
11.1
0.177
5.25
1.270
c
Coefficients significant at p P 0.90.
Coefficients significant at p P 0.95.
Coefficients significant at p P 0.99.
a
b
2.53
1.450
F
Error
4.92
1.005
0.909
0.825
0.816
0.700
Model
0.197b
0.095
0.037
1.31b
1.12b
0.127
1.02a
0.860
0.395
Interaction
Time T (b12)
Time L/S (b13)
T L/S (b23)
0.854b
0.981b
0.084
0.052
0.057
0.030
0.373
0.461
0.193
0.904b
0.542
0.503
Quadratic
Time (b11)
T (b22)
L/S (b33)
0.495a
0.577a
0.931c
0.300c
0.296c
0.116b
1.14c
1.34c
0.555
0.828a
0.775a
0.086
Linear
Time (b1)
T (b2)
L/S (b3)
0.249
0.932c
0.215
11.9
9.81
11.6
Intercept (b0)
MS
0.938
ð9Þ
AWS
Model parameters
Table 6
Regression coefficients and ANOVA of the adjusted models for the objective functions (SEY, PEY, radical scavenging, EC50).
ð7Þ
The equations developed for methanol and ethanol extracts
were significant at the 95% confidence level, whereas the one deduced for water extracts was significant at the 90% confidence level. The major contributions of the dependent variables (and
interactions among them) can be identified in Eqs. (7)–(9).
Activities of commercial antioxidants, such as ascorbic acid
(EC50 = 0.12 mg/mL) and rutin (EC50 = 0.16 mg/mL) determined in
our laboratory, were slightly higher than the best extracts obtained
in this work. Higher DPPH radical scavenging capacities were reported for an ethanolic extract produced by a series of five extraction stages of 10 min (EC50 = 0.0126 mg/mL) [11], for crude
extracts and fractions obtained with a sequence of solvents
(EC50 = 0.017–0.071 mg/mL) [3], and slightly higher for a boiling
water extract from chestnut leaves (IC50 = 0.17 mg/mL) [7].
The effects of the most influential variables on the extract concentration needed to scavenge 50% of DPPH (EC50) are shown in
Fig. 1c. Optimal conditions for antioxidant activity of the aqueous
extracts are defined by contact times in the range of 50–90 min
and temperatures below 30 °C, whereas alcoholic extracts manufactured at increased temperatures (up to 50 °C) showed increased
activity.
The ABTSþ radical scavenging capacity of extracts produced under selected conditions showed similar potency regardless the
extracting solvent. The most active extracts were obtained with
ES (EC50 = 0.34 mg/mL). This activity was near the ones determined
for AWS (0.38 mg/mL) and MS (0.37 mg/mL) extracts.
The phenolic content of extracts, measured by the ratio PEY/SEY
showed a close dependence on the DPPH radical scavenging capacity of the corresponding extracts (see Fig. 2). This finding suggests
that the total phenolic content of different extracts was a major
factor to explain their antioxidant activity, as it has been proposed
for polyphenols both in polar solvents [3] and in hot water extracts,
but not for flavonoids [7].
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
107
Fig. 1. Predicted dependence of (a) the solubles extraction yield (SEY), (b) phenolic extraction yield (PEY) and (c) antioxidant activity (EC50) of extracts on the contact time,
temperature and liquid solid ratio.
It can be noted that temperature caused different effects on the
PEY and on the antioxidant activity: PEY increased with temperature, whereas an intermediate temperature led to maximal antioxidant activity. It can be noted that increased temperatures enhance
extraction by improving both solubilities and diffusion coefficients.
However, intermediate temperatures can be more favorable for
extracting active phenolics, due to the limited degradation.
3.3. Antioxidant properties characterization
CsL aqueous extracts obtained under optimal conditions
described above were assayed for their ability to scavenge ROS
1
(O
2 , H2O2, ROO , HOCl, and O2) and RNS ( NO and ONOO , in the
presence and absence of HCO
).
Limited
effects
were
observed
in
3
experiments with H2O2 and HOCl, whereas the scavenging activi-
108
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
3.4. Composition of the CsL extracts
3.5
Water
3
Ethanol
EC50 (g/L)
The literature reported on the phenolic compounds present in
CsL lists a number of phenolic acids (p-hydroxyl, vanillyl, syringil,
gallic, ellagic, p-coumaric, chlorogenic and ferulic acids) [1,22] and
flavonoids (rutin, hesperidin, quercetin, hyperoside, isoquercitrin,
apigenin, morin, naringin, narcissin, astragalin, galangin and
kaempferol) [6,8,10]. In the aqueous extracts produced according
to the optimized process of the present work, the major phenolic
compounds identified by HPLC were gallic acid, protocatechuic
acid, 4-hydroxybenzoic acid, vanillic acid, rutin, quercetin and
apigenin.
Methanol
2.5
2
1.5
1
0.5
0
0
0.1
0.2
0.3
0.4
0.5
0.6
4. Conclusions
PEY/SEY (g GAE / g solid)
Fig. 2. Relationship between phenolic content (calculated as the ratio PEY/SEY) and
DPPH radical scavenging capacity of extracts.
O2.-
ROS and RNS scavenging activity (%)
100
1
HOCl
-
ONOO
O2
.
NO
ONOO- with HCO3-
75
50
25
An experimental design was employed to develop mathematical equations suitable for assessing the yield in soluble compounds,
the yield in phenolics and the antioxidant activity of extracts produced from C. sativa leaves (CsL) with water, ethanol and methanol. This feedstock is bio-renewable material of huge availability,
employed in traditional medicine, and a potential source of radical
scavengers. Both extraction time and temperature affected the
yields of soluble solids and phenolics strongly. Prolonged extraction periods (>90 min) at high temperatures (50 °C) led to optimal
extraction of phenolics with water and ethanol; whereas limited
effects were associated to the contact time in extractions with
methanol at high temperatures. Acidified water was the best solvent for phenolic extractions, which yield extracts of high purity.
Under selected conditions, extracts containing up to 54 wt% of phenolic acids and flavonoids were produced, with a radical scavenging capacity similar to that of BHA. The process proposed in this
work is simple, flexible, and causes low environmental impact.
Acknowledgements
0
0
5
10
15
20
Conc (µg/mL)
Fig. 3. ROS and RNS scavenging activity of extracts obtained under optimized
conditions.
Table 7
ROS and RNS scavenging effects (IC50, mean ± standard deviation) and ORAC ROO. ±
Standard deviation) of the aqueous extract obtained under optimized conditions.
IC50 (mg/mL) 103
O
2
1
O2
HOCl
H2O2
NO
ONOO without NaHCO3
ONOO with NaHCO3
15.1 ± 2.5
11.6 ± 0.9
63.8 ± 5.2
32.3 ± 9.4*50 lg/mL
7.15 ± 0.18
1.56 ± 0.23
3.24 ± 0.23
ROO
ORAC (lmol trolox equiv./mg extract)
1.09 ± 0.11
*
Scavenging effect (mean%) at the highest tested concentration (in superscript).
1
ties against O
2 , O2 and RNS were remarkable (Fig. 3), presenting
IC50s in low mg/mL range (Table 7). The scavenging effects observed for the studied RNS were much higher than those obtained
for ROS, a finding in agreement with literature [6].
Sonication-assisted ethanol extraction of chestnut leaves was
reported by Zˇivkovic´ et al. [9]. In the present work, this strategy enhanced SEY and the phenolic content of extracts by a factor of 1.2,
without compromising the extraction selectivity significantly
(EC50 = 0.36 mg/mL).
The authors greatly acknowledge MEC and FEDER for the financial support of the Project AGL2006-05387. The authors wish to
thank Ms. Noelia González (who held a ‘‘Lucas Labrada’’ contract
from ‘‘Xunta de Galicia’’). Beatriz Díaz Reinoso thanks the Spanish
MEC for her research grant (BES-2005-7525). Andrés Moure thanks
‘‘Xunta de Galicia’’ for his research contract (‘‘Isidro Parga Pondal’’
Program). Diana Couto acknowledges the Fundação para a Ciência
e Tecnologia financial support for her PhD Grant (reference SFRH/
BD/72856/2010).
References
[1] J. Zˇivkovic´, Z. Zekovic´, I. Mujic´, D. Gocrossed, D. Signevac, M. Mojovic´, A. Mujic´,
I. Spasojevic´, EPR spin-trapping and spin-probing spectroscopy in assessing
antioxidant properties: example on extracts of catkin, leaves, and spiny burs of
Castanea sativa, Food Biophys. 4 (2009) 126–133.
[2] C.A. Calliste, P. Trouillas, D.P. Allais, A. Simon, J.L. Duroux, Free radical
scavenging activities measured by electron spin resonance spectroscopy and
B16 cell antiproliferative behaviors of seven plants, J. Agric. Food Chem. 49
(2001) 3321–3327.
[3] C.A. Calliste, P. Trouillas, D.P. Allais, J.L. Duroux, Castanea sativa Mill. leaves as
new sources of natural antioxidant: an electronic spin resonance study, J.
Agric. Food Chem. 53 (2005) 282–288.
[4] J.C.M. Barreira, I.C.F.R. Ferreira, M.B.P.P. Oliveira, J.A. Pereira, Antioxidant
potential of chestnut (Castanea sativa L.) and almond (Prunus dulcis L.) byproducts, Food Sci. Technol. Int. 16 (2010) 209–216.
[5] A. Mujic´, N. Grdovic, I. Mujic´, M. Mihailovic´, J. Zˇivkovic´, G. Poznanovic´, M.
Vidakovic´, Antioxidative effects of phenolic extracts from chestnut leaves,
catkins and spiny burs in streptozotocin-treated rat pancreatic.beta.-cells,
Food Chem. 125 (2010) 841–849.
[6] I.F. Almeida, E. Fernandes, J.L.F.C. Lima, P.C. Costa, M.F. Bahia, Protective effect
of Castanea sativa and Quercus robur leaf extracts against oxygen and nitrogen
reactive species, J. Photochem. Photobiol. B: Biol. 91 (2008) 87–95.
[7] J.C.M. Barreira, I.C.F.R. Ferreira, M.B.P.P. Oliveira, J.A. Pereira, Antioxidant
activities of the extracts from chestnut flower, leaf, skins and fruit, Food Chem.
107 (2008) 1106–1113.
B. Díaz Reinoso et al. / Chemical Engineering Journal 203 (2012) 101–109
[8] A. Basile, S. Sorbo, S. Giordano, L. Ricciardi, S. Ferrara, D. Montesano, R. Castaldo
Cobianchi, M.L. Vuotto, L. Ferraration, Antibacterial and allelopathic activity of
extract from Castanea sativa leaves, Fitoterapia 71 (2000) S110–S116.
[9] J. Zˇivkovic´, I. Mujic´, Z. Zekovic´, S. Vidovic, A. Mujic´, S. Jokic´, Radical scavenging,
antimicrobial activity and phenolic content of Castanea sativa extracts, J. Cent.
Eur. Agric. 10 (2009) 175–182.
[10] F. Henry, L. Danoux, G. Pauly, Cosmetic composition comprising an extract of
the leaves of Castanea sativa, PCT Int. Appl. WO 079741 (2005).
[11] I.F. Almeida, P. Valentão, P.B. Andrade, R.M. Seabra, T.M. Pereira, M.H. Amaral,
P.C. Costa, M.F. Bahia, In vivo skin irritation potential of a Castanea sativa
(chestnut) leaf extract, a putative natural antioxidant for topical application,
Basic Clin. Pharm. Toxicol. 103 (2008) 461–467.
[12] F.P. Liebel, Use of flavonoid glycosides. Tanning agents and microorganisms
for the therapy and prophylaxis of diabetes mellitus, Eur. Pat. EP0956867
(1999).
[13] J.E. Cacace, G. Mazza, Extraction of anthocyanins and other phenolics
from black currants with sulfured water, J. Agric. Food Chem. 50 (2002)
5939–5946.
[14] C. Liyana-Pathirana, F. Shahidi, Optimization of extraction of phenolic
compounds from wheat using response surface methodology, Food Chem. 93
(2005) 47–56.
[15] B.L. Browning, Methods of Wood Chemistry, John Wiley & Sons, New York,
1967. p. 356.
[16] N. Blumenkrantz, G. Asboe-Hansen, New method for quantitative
determination of uronic acids, Anal. Biochem. 54 (1973) 484–489.
[17] E. Conde, C. Cara, A. Moure, E. Ruiz, E. Castro, H. Domínguez, Antioxidant
activity of the phenolic compounds released by hydrothermal treatments of
olive tree pruning, Food Chem. 114 (2009) 806–812.
109
[18] A. Gomes, E. Fernandes, A.M. Silva, C.M. Santos, D.C. Pinto, J.A. Cavaleiro, J.L.
Lima, 2-Styrylchromones: novel strong scavengers of reactive oxygen and
nitrogen species, Bioorg. Med. Chem. 15 (2007) 6027–6036.
[19] D. Costa, E. Fernandes, J.L.M. Santos, D.C.G.A. Pinto, A.M.S. Silva, J.L.F.C. Lima,
New noncellular fluorescence microplate screening assay for scavenging
activity against singlet oxygen, Anal. Bioanal. Chem. 387 (2007) 2071–
2081.
[20] M. Whiteman, U. Ketsawatsakul, B. Halliwell, A reassessment of the
peroxynitrite scavenging activity of uric acid, Ann. NY Acad. Sci. 962 (2002)
242.
[21] E. Fernandes, A. Gomes, D. Costa, J.L.F.C. Lima, Pindolol is a potent scavenger of
reactive nitrogen species, Life Sci. 77 (2005) 1983–1992.
[22] T. Sariyildiz, J.M. Anderson, Variation in the chemical composition of green
leaves and leaf litters from three deciduous tree species growing on different
soil types, Forest Ecol. Manage. 210 (2005) 303–319.
[23] F. Adjé, Y.F. Lozano, P. Lozano, A. Adima, F. Chemat, E.M. Gaydou, Optimization
of anthocyanin, flavonol and phenolic acid extractions from Delonix regia tree
flowers using ultrasound-assisted water extraction, Ind. Crops Prod. 32 (2010)
439–444.
[24] J.E. Cacace, G. Mazza, Mass transfer process during extraction of phenolic
compounds from milled berries, J. Food Eng. 59 (2003) 379–389.
[25] M. Wettasinghe, F. Shahidi, Evening primrose meal: a source of natural
antioxidants and scavenger of hydrogen peroxide and oxygen-derived free
radicals, J. Agric. Food Chem. 47 (1999) 1801–1812.
~ ez, Effect of solvent,
[26] M. Pinelo, M. Rubilar, M. Jerez, J. Sineiro, M.J. Nu´n
temperature, and solvent-to-solid ratio on the total phenolic content and
antiradical activity of extracts from different components of grape pomace, J.
Agric. Food Chem. 53 (2005) 2111–2117.