The bioactive efficiency of ultrasonic extracts from acorn leaves and green walnut husks against Bacillus cereus: a hybrid approach to PCA with the Taguchi method

Abstract

The objective of this study was to investigate the usefulness of the hybrid approach using the Taguchi method (TM) and principal component analysis (PCA) in determining the optimum conditions for ultrasonic extracts of acorn leaves and green walnut husks, which potentially demonstrate the best bioactive efficiency against Bacillus cereus. First, an L36 (2¹ × 3⁴) mixed level orthogonal array design was implemented, consisting of five factors: extract type, temperature, time, solvent and concentration, respectively. Also, the total phenolic content, antiradical activity, and antibiogram analyses were investigated by design, and signal-to-noise (S/N) ratios were calculated for each trial. An analysis of variance (ANOVA) was performed using S/N ratios to estimate the effects of the measured parameters and their interactions for a single-objective TM. Following that, PCA was used to normalize the S/N ratios for each response to calculate the multi-response performance index for measuring the effects of factors on all responses. In this study, the expected optimal factor levels for each experiment were found to be different for a single-objective TM and inadequate for interpreting all responses simultaneously. The results of the study showed that the optimal conditions for all responses with PCA-based TM was found to be a 5% concentration of acorn leave extract and 50% acetone at 60 °C for 60 min, with 136.96 mg g–1 gallic acid equivalent, 90.75% inhibited 2,2-diphenyl-1-picrylhydrazyl and 14.33 mm inhibition zone against B. cereus.

Full text

Vol.:(0123456789)
1 3
Journal of Food Measurement and Characterization (2019) 13:1257–1268
https://doi.org/10.1007/s11694-019-00041-7
ORIGINAL PAPER
The bioactive efficiency ofultrasonic extracts fromacorn leaves
andgreen walnut husks againstBacillus cereus: ahybrid approach
toPCA withtheTaguchi method
GokturkOzturk1· AhmetE.Yetiman2· MahmutDogan2,3
Received: 10 September 2018 / Accepted: 16 January 2019 / Published online: 24 January 2019
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
The objective of this study was to investigate the usefulness of the hybrid approach using the Taguchi method (TM) and
principal component analysis (PCA) in determining the optimum conditions for ultrasonic extracts of acorn leaves and green
walnut husks, which potentially demonstrate the best bioactive efficiency against Bacillus cereus. First, an L36 (21 × 34)
mixed level orthogonal array design was implemented, consisting of five factors: extract type, temperature, time, solvent and
concentration, respectively. Also, the total phenolic content, antiradical activity, and antibiogram analyses were investigated
by design, and signal-to-noise (S/N) ratios were calculated for each trial. An analysis of variance (ANOVA) was performed
using S/N ratios to estimate the effects of the measured parameters and their interactions for a single-objective TM. Follow-
ing that, PCA was used to normalize the S/N ratios for each response to calculate the multi-response performance index for
measuring the effects of factors on all responses. In this study, the expected optimal factor levels for each experiment were
found to be different for a single-objective TM and inadequate for interpreting all responses simultaneously. The results
of the study showed that the optimal conditions for all responses with PCA-based TM was found to be a 5% concentration
of acorn leave extract and 50% acetone at 60°C for 60min, with 136.96mg g–1 gallic acid equivalent, 90.75% inhibited
2,2-diphenyl-1-picrylhydrazyl and 14.33mm inhibition zone against B. cereus.
Keywords Taguchi method· Principal component analysis· Bacillus cereus· Phenolic· Antimicrobial activity· Hybrid
approach
Introduction
There have been many studies on the relationship between
the intake of synthetic food additives and the development
of diseases such as cancer and obesity [1, 2]. This relation-
ship has brought about concern for the adverse effects of the
synthetic food additives used in food production [3]. In addi-
tion, antimicrobial resistance has become a common prob-
lem for the whole world, and the World Health Organization
has raised attention to this problem [4]. Because of these
concerns, studies investigating the isolation of natural food
additives have been accelerated. Many of the studies have
focused on naturally occurring bioactive compounds, mainly
phenolic substances from agricultural by-products with anti-
microbial activity [5, 6].
It has been reported that some of the agricultural by-
products rich in phenolics possess an antimicrobial effect
on pathogenic microorganisms like Bacillus cereus [7–9].
B. cereus is an aerobic or facultatively anaerobic organism,
a gram-positive, endospore-forming and rod-shaped bacte-
rium that sporulates only in aerobic conditions [10, 11]. The
optimum growth temperature for B. cereus ranges between
28 and 37°C. However, the pathogen can grow between 10
and 50°C, and several strains can even reproduce at 7°C.
The bacterium can multiply in pH conditions between 4.3
and 9.3 and a minimum water activity (aw) of 0.92 [11, 12].
The bacterium is ubiquitous and may frequently spread on
many kinds of foods including meat, dairy products, infant
* Mahmut Dogan
dogan@erciyes.edu.tr
1 Food Technology Program, Kaman Vocational School, Ahi
Evran University, 40300Kırşehir, Turkey
2 Food Engineering Department, Faculty ofEngineering,
Erciyes University, 38030Kayseri, Turkey
3 Tagem Food Analysis Center Co., Erciyes Teknopark Area,
38030Kayseri, Turkey

1258 G.Ozturk et al.
1 3
food, spices, cereals and vegetables [10, 13]. B. cereus and
members of its bacterial group can cause food poisoning
that may arise from an intoxication (emetic syndrome) or an
infection (diarrheal syndrome). An emetic syndrome occurs
through the intake of a heat-stable toxin (cereulide), while
a diarrheal syndrome grows from ingesting viable cells that
generate enterotoxins in the small intestine [14]. The inhibi-
tion of both the microbial growth and the toxin formation
of B. cereus is possible with the phytochemicals acquired
from medicinal and non-medicinal herbal extracts [15–17].
One of the phytochemicals, phenolics, can be extracted
by various techniques: either conventional techniques like
soxhlet extraction, infusion and maceration, or by modern
methods such as ultrasound-assisted extraction, microwave-
assisted extraction, pressurized solvents, enzyme-assisted
extraction and pulsed electrical fields [18]. Because there
are many limitations in classical extraction techniques,
which can be time-consuming, produce low yields and
involve higher solvent consumption, the ultrasound extrac-
tion technique, or a combination of it with other extraction
techniques, has received considerable attention in recent
years. The ultrasound technique can be applied successfully
in extracting phenolic compounds from plants, agricultural
by-products, species, and so more yield can be achieved in a
shorter time, using less solvent than conventional techniques
[19].
Free radicals result from metabolic activity in aerobic
systems (endogenous sources) or exogenous sources like
pollution, heavy metals, tobacco and so on [20]. These radi-
cals are deactivated by the antioxidative defense systems
(enzymatic and non-enzymatic antioxidants). If there is no
balance between the free radicals and the defense systems,
the former is capable of causing damage to DNA, proteins,
lipids and carbohydrates, thereby leading to oxidative stress,
which is the most significant phenomenon in a variety of
diseases [20, 21]. Phenolics can scavenge the free radicals,
which act as natural antioxidants. This phenolic effect is
closely associated with its antioxidant activity, depending
on the phenolic type or phenolic structure–activity relation-
ship [22]. In addition to their antioxidant activity, phenolics
may exhibit antimicrobial activity against gram-positive and
gram-negative bacteria [23].
Among agricultural by-products, acorn leaves (genus
Quercus) and the green husk of walnuts (genus Juglans)
are rich in phenolic compounds. Ferulic acid, chlorogenic
acid, ellagitannins, hydrolysable tannins, gallic acid and its
derivatives, and quercetin or kaempferol glycosides, have
been detected in acorn leaves [24, 25]. The methanol extracts
of acorn leaves (Quercus ilex L.) have been shown to exhibit
inhibitory activity against Brucella, Bacillus, Enterobacter,
Neisseria, Pseudomonas and Escherichia bacteria, and
also Candida albicans while they do not exhibit antifungal
activity [26]. Walnut and its by-products, such as leaves and
green husks, contain phytochemicals with antioxidant, anti-
microbial and anticarcinogenic activity [27–29]. Thirteen
phenolics, including gallic acid, catechin, chlorogenic acid,
vanillic acid, caffeic acid, epicatechin, syringic acid, syrin-
galdehyde, p-coumaric acid, ferulic acid, rutin, myricetin,
and juglone were identified, and juglone was found to be
the major phenolic in the green walnut husk [30, 31]. As
a result, it has been suggested that both acorn leaves and
the green walnut husk can be used as a cheap antioxidant
resource, antimicrobial agent and chemopreventive agent
in both the pharmaceutical and food industry due to their
phytochemical contents [26, 28, 32–34].
Using a response surface design of experiments (DoE)
methodology [24, 29, 32, 35–37] along with classical and
ultrasonic extraction methods, it has been reported that their
some parts show antioxidant, antimicrobial and anticarcino-
genic activity. The DoE consists of mathematical methods
which are used to specify the relationships between the dif-
ferent factors that affect the process/trial and the output of
the process. The DoE can be used for comparison, charac-
terization (screening), modeling and optimization (estima-
tion) of factors and responses [38].
One of the DoE methodologies, the Taguchi method
(TM), is an extremely effective statistical approach that was
developed by Dr. Genichi Taguchi, who studied the improve-
ment of both the process and product quality of manufactur-
ing in Japan [39]. Although the Taguchi Method has mainly
been applied to various fields of industrial production, its
has only been implemented in a limited number of specific
case studies in the biotechnology field (e.g., food process-
ing, fermentation, bioremediation, molecular biology and
wastewater treatment) [40]. The prime objective of TM is
that it can be performed with minimal experimental runs
for a determination of the effects of factors on characteristic
features and the determination of optimal conditions. TM
has definite benefits compared to conventional statistical
methodologies regarding the specification of optimal experi-
mental conditions. The TM can specify an optimum state
that has the least experimental variable conditions. The vari-
ability of the conditions is expressed by using the S/N ratio
to identify control factors that minimize the variability in an
experiment by decreasing the effects of noncontrollable (also
called noise) factors. The maximum value of the S/N ratio
corresponds to the optimal settings of an experiment that is
useful for single-response cases [41, 42]. However, classi-
cal TM is not sufficient for the estimation or optimization
of multi-response, problem-based studies [43]. Hence, the
TM combined with PCA has been used by several authors
in the engineering field to overcome the problem [43–46].
In the PCA-based Taguchi hybrid method, PCA is used to
transform highly correlated independent variables into a
linear combination of uncorrelated variables called princi-
pal components (PCs) that are used to select PCs whose

1259
The bioactive efficiency ofultrasonic extracts fromacorn leaves andgreen walnut husks against…
1 3
eigenvalue is higher than 1, and which can then be employed
as a MRPI. The optimal conditions for multi-response stud-
ies can also be determined by the maximum value of MRPI
that corresponds to its respective experimental run in the
design matrix. The use of PCA with TM is not only to ensure
objectivity but also to contribute to the development of the
robustness of designs [43, 46].
As far as we are concerned, there has been no research
about the estimation or optimization of the effects of ultra-
sonically obtained herbal extracts on microorganisms, anti-
radical activity, or total phenolic content (TPC) by using the
PCA-TM hybrid approach. It seems that most of the pub-
lished papers on TM have been interested in single response
cases, and multi-response studies are not prevalent among
Taguchi practitioners. Thus, the suitability of using TM
with PCA for the determination of the bioactive efficiency
of ultrasonic extracts of acorn (Quercus L.) leaves and green
walnut (Juglans regia L.) husks against B. cereus was tested
by considering the effects of extract type, temperature, time,
solvent type and extract concentration.
Materials andmethods
Samples
Acorn (Quercus L.) leaves and green husks of walnut
(Juglans regia L.) were collected in Kayseri, Turkey, after
harvesting them in the season of 2016. They were dried in
the laboratory until 5% moisture remained [47], and then
they were ground into a fine powder with a blender (War-
ing Blender, HGB2WTS3, Waring Commercial, USA). The
powder was passed through a 0.45µm sieve to extract a
uniform particle size. The samples were stored at − 20°C
after grinding.
Chemicals
Sodium carbonate, Folin–Ciocalteu reagent, ethanol, metha-
nol, acetone, nutrient broth and agar were purchased from
Merck (Germany) while the 2,2-diphenyl-1-picrylhydrazyl
(DPPH) and gallic acid were obtained from Fluka (Ger-
many) and Sigma-Aldrich (Germany), respectively.
Taguchi experimental design andstatistical
analyses
In this study, a mixed level orthogonal array design (L36:
21 × 34) was applied, which comprised five factors in two,
three, three, three and three levels. The experimental factors
and levels of the design are indicated in Table1. For ana-
lyzing the Taguchi design, 36 experimental runs were per-
formed, and S/N ratios were computed in compliance with a
larger-the-better equation. The results of each trial were ana-
lyzed separately using the mean values. Later, data analysis
was carried out to estimate the effects of the factors and their
interactions by using ANOVA with a significance level of
0.05. All computations were performed using Minitab v18.1
(Minitab Ltd., Coventry, United Kingdom) statistical software.
Principal component analysis (PCA)
PCA is a multivariate method that enables data to be presented
in an alternative way using a smaller number of uncorrelated
variables called principal components [44, 48]. By using PCA,
all the response data obtained from the trials were converted
into principal components (PCs) that were a linear combina-
tion of the original multi-response dataset. Before PCA, S/N
ratios for each response were normalized to values between 0
and 1 by the following equation:
where
XiK
represents the normalized S/N ratio of the Kth
response data from the ith experimental run,
Xi(K)
is the
S/N ratio of the Kth data response from the ith experimental
run,
Xi
(K)
+
and
Xi(K)−
are the maximum and minimum S/N
ratios of all the experimental runs for the Kth data response.
Afterwards, the eigenvalues and eigenvectors of the correla-
tion matrix were calculated using the normalized S/N ratios
with PCA. The PC scores were computed as follows:
where
Pi(K)
is the Kth PC equating to the ith experimental
run,
vk(j)
is the jth constituent of the Kth eigenvector. In addi-
tion, e(K) values were calculated by dividing each eigen-
value into the sum of eigenvalues [45, 46]. MRPI values
for each experimental run were calculated according to the
following equation:
(1)
X
iK =
X
i
(K)−X
i
(K)−
X
i
(K)+−X
i
(K)−
(2)
P
i(K)=
m
∑
k=1
Xik ×vk(j
)
(3)
MRPI
=
m
∑
k=1
Pi(K)×e(K
)
Table 1 Factors and levels used in Taguchi mixed level array design
Factors Levels
1 2 3
(A) Extract type Acorn leaves Walnut green husk –
(B) Temperature (°C) 40 50 60
(C) Time (min) 30 45 60
(D) Solvent (50%) Methanol Ethanol Acetone
(E) Extract concentra-
tion (%)
1.25 2.5 5

1260 G.Ozturk et al.
1 3
General linear ANOVA was employed for the determi-
nation of the factors that affect the levels of each parame-
ter. All calculations were carried out using Minitab v18.1
statistical software.
Extraction ofphenolics fromthesamples
The extraction of phenolics from the samples was per-
formed with respect to the Taguchi design matrix con-
sisting of thirty-six points obtained from the Minitab
software (Table2). At first, a sonification process was
conducted with an ultrasonic bath (WiseClean, WUC-
A03H, 296W/1AMPS, Daihan Scientific Co. Ltd., Korea)
Table 2 L36 (21 × 34) mixed
level orthogonal array
design table with factors
and, experimental results
that are consist of measured
responses and computed S/N
ratios of TPC, DPPH radical
scavenging activity (RSA) and
antimicrobial activity (AA)
assays
a TPC: Total phenolic content as mg g−1 GAE
b RSA: DPPH radical scavenging activity (%)
c Diameters of inhibition zones in mm for B. cereus
Experi-
mental
run
Factors Measured responses Computed S/N ratios
A B C D E TPCaRSAbAAcTPC RSA AA
1 1 1 1 1 1 27.56 93.55 9.52 29.29106 39.39962 18.31177
2 1 2 2 2 2 55.81 92.54 10.83 34.63962 39.33631 17.77296
3 1 3 3 3 3 136.96 90.75 14.33 42.6063 39.14712 22.68712
4 1 1 1 1 1 31.04 93.10 7.36 – – –
5 1 2 2 2 2 52.26 92.74 6.34 – – –
6 1 3 3 3 3 133.11 90.54 13.01 – – –
7 1 1 1 2 3 109.42 90.02 13.87 40.78201 39.0872 22.84362
8 1 2 2 3 1 37.49 93.50 9.48 31.4782 39.4158 19.53617
9 1 3 3 1 2 62.43 92.26 11.46 35.90805 39.29989 21.18622
10 1 1 1 3 2 69.91 91.22 11.96 36.89028 39.20193 21.55704
11 1 2 2 1 3 120.97 88.57 15.20 41.6534 38.94551 23.63687
12 1 3 3 2 1 30.53 91.09 8.21 29.69585 39.18958 18.28334
13 1 1 2 3 1 34.41 91.18 10.48 30.73309 39.19764 20.40723
14 1 2 3 1 2 62.08 90.19 12.07 35.85865 39.10314 21.63415
15 1 3 1 2 3 117.77 88.40 13.14 41.42046 38.92891 22.37411
16 1 1 2 3 2 68.81 90.52 12.34 36.75362 39.13523 21.8263
17 1 2 3 1 3 120.69 88.17 14.19 41.63341 38.9067 23.04169
18 1 3 1 2 1 31.02 90.58 8.61 29.83367 39.14032 18.70006
19 2 1 2 1 3 19.98 77.34 7.63 26.01138 37.7685 17.6467
20 2 2 3 2 1 2.73 88.97 4.00 8.732973 38.98527 12.0412
21 2 3 1 3 2 10.64 83.98 7.98 20.53921 38.48336 18.03643
22 2 1 2 2 3 17.36 79.38 7.99 24.79024 37.9939 18.05094
23 2 2 3 3 1 3.58 88.27 4.00 11.07167 38.91623 12.0412
24 2 3 1 1 2 11.12 84.59 4.00 20.92561 38.54632 12.0412
25 2 1 3 2 1 2.62 89.14 4.00 8.36177 39.00153 12.0412
26 2 2 1 3 2 13.59 84.00 6.21 22.6628 38.48521 15.85717
27 2 3 2 1 3 20.33 78.08 8.80 26.1612 37.85043 18.88636
28 2 1 3 2 2 9.31 85.29 6.33 19.37886 38.61826 16.0235
29 2 2 1 3 3 23.44 77.09 6.96 27.39868 37.73969 16.85634
30 2 3 2 1 1 4.29 88.29 4.00 12.6547 38.91791 12.0412
31 2 1 3 3 3 21.06 77.64 7.34 26.46737 37.80198 17.31392
32 2 2 1 1 1 3.79 88.62 4.00 11.56387 38.9508 12.0412
33 2 3 2 2 2 8.71 85.59 5.35 18.80053 38.64894 14.56708
34 2 1 3 1 2 9.53 85.01 5.61 19.57876 38.58956 14.97409
35 2 2 1 2 3 18.49 79.51 8.05 25.34038 38.00866 18.11592
36 2 3 2 3 1 4.85 87.88 4.00 13.71592 38.87796 12.0412

1261
The bioactive efficiency ofultrasonic extracts fromacorn leaves andgreen walnut husks against…
1 3
at 50Hz at a sample to solvent ratio of 1:10 in 50mL
glass flasks (Schott, Germany). Next, the supernatant
was obtained by centrifuging (Sigma 3-30K, Germany)
at 13,000rpm for 10min at 5°C and lyophilized with a
laboratory type freeze-dryer (Christ, Alpha 2–4 LSCplus,
Germany). The lyophilized extracts were used for further
analyses (Fig.1).
Determination ofTPC
The TPC of the herbal extracts was determined by using
the Folin–Ciocalteu method, according to Hayta and
Iscimen [49]. Briefly, 30µL of the extract, 150µL of
10-fold diluted Folin–Ciocalteu reagent and 120µL of
sodium carbonate (70g L−1) were placed in a 96 well
polyethylene microplate. The absorbance of the mixtures
were measured with a microplate-reader (Multiscan
FC, Thermo-Fisher Scientific, USA) at a wavelength of
750nm against pure water as a blank after 1h of incuba-
tion at room temperature, and the results were calculated
as mg of gallic acid equivalents, ranging from 7.81 to
125mgkg−1.
Radical scavenging activity (RSA)
Antiradical activity analysis was performed by using the
2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical in keep-
ing with the method of Hayta and Iscimen [49]. A 30 µL vol-
ume of the extract from the microplate well was mixed with
270 µL of DPPH solution (300µM prepared in 80% ethanol).
The mixture was stirred vigorously for 5min and allowed
to stand at room temperature for 55min in the microplate
reader. At the end of the incubation period, the absorbance
of the mix was measured at a wavelength of 520nm with the
microplate reader against 80% ethanol as a blank. The radi-
cal scavenging activity was expressed as the percent inhibi-
tion by using the following equation:
Antimicrobial activity (AA) assay
The strain of Bacillus cereus (ATCC 11778) used as a test
microorganism was obtained from the Food Engineering
%
Inhibition
=
(Absorbance of the control −Absorbance of the extract
Absorbance of the control
)
×100
Fig. 1 Phenolic extraction methodology from acorn leaves and walnut green husk

1262 G.Ozturk et al.
1 3
department’s culture collection of Erciyes University in Kay-
seri, Turkey. The determination of the antimicrobial activity
of the extracts was carried out by the agar well diffusion
method. The test strain was suspended in sterile nutrient
broth and incubated at 35 ± 0.1°C. The final cell concentra-
tion of the culture suspensions was adjusted by comparing
it to 0.4–0.5 McFarland turbidity standards [50]. Afterward,
1mL of the culture suspension was inoculated into 100mL
of molten nutrient agar, and approximately 15mL of the
medium was poured into 90 × 15mm Petri dishes. After
solidification of the agar plates (at 4°C, for 1h), four equi-
distant wells of 4mm in diameter were created on the agar
plates using a sterile cork borer. Fifty microliters of 1.25%,
2.5% and 5% concentrations of extracts were transferred into
these wells using a sterile micropipette according to Table2.
The Petri dishes were incubated at 35 ± 0.1°C for 18 to 24h.
After that period, the inhibition zones were measured in mil-
limeters. All experiments were performed in triplicate [51].
Results anddiscussion
Results ofthesingle objective Taguchi experiments
In this study, the aim was to estimate the most effective con-
ditions for the maximum level of total phenolic content, radi-
cal scavenging activity and antimicrobial activity. Therefore,
the usage of a larger-the-better objective function was pre-
ferred to evaluate the response variables. The mean values
of the results of the response trials and their calculated S/N
ratios are presented by a mixed level orthogonal array design
matrix (L36: 21 × 34) in Table2. The numbers in each col-
umn from A to E in the design matrix indicate the environ-
ment level that implemented the factors in the experiments.
The average S/N ratios with delta and rank values for each
trial are also shown in Table3. The lowest rank and highest
delta value corresponded to the most effective factor for each
response. Additionally, as a result of ANOVA, the degrees
of freedom (DF), the sequential sum of squares (Seq SS),
the adjusted sums of squares (Adj SS), the adjusted mean
squares (Adj MS), and the F-value and P-value were set to
interpret the factors and interactions that have statistically
significant effects on each response. Tables4 and 5 displays
the ANOVA results of S/N ratios for TPC, RSA and AA,
respectively. The results of the single objective TM for each
experiment are discussed in separate sections below.
Total phenolic content
The TPC of extracts were significantly affected (p < 0.05) by
the following factors; extract type (A), solvent (D) and con-
centration (E), and extract type–concentration (A*E) interac-
tion (Table4.). The ANOVA results for TPC demonstrated
that the temperature (B) and time (C) factors, and A*B,
A*C, A*D, B*D, B*E and C*D interactions, were not found
to be significant (p > 0.05). The response table of S/N ratio
means (Table3) revealed that the extract type was the most
efficient factor for TPC. Also, the following conditions: 5%
concentration of acorn leaf extract, 50% acetone, 60°C and
60min, were found to be the most suitable for TPC that cor-
responded to the 3rd experimental run in Table2.
Phenolic compounds in plants have an essential role in
the formation of color and taste, and as a defence mecha-
nism against pathogen microorganisms. Also, phenolic
compounds have been reported to have a positive effect on
human health; it is known that there is an inverse correlation
between the consumption of a diet rich in phenols and the
risk of developing disease. This relationship is due to the
antioxidant properties of the phenols present in plants, thus
contributing to a reduction in oxidative stress, and creat-
ing this effect at the appropriate amount of phenolics [52].
Therefore, the extraction stage is of great importance in the
recovery of phenolics from agricultural by-products. Several
extraction methods have been used to extract the phenolics
from agro-waste. They could be influenced by the extrac-
tion parameters such as the sample type, time, temperature,
solvent and its polarity. Another factor that may affect the
yield and structure of phenolics is the extraction method.
The ultrasound-assisted extraction method, one of the
extraction techniques, has some advantages over the others.
In this process, cavitation bubbles occur, causing the mate-
rial used in the extraction to be destroyed, which accelerates
Table 3 Response table for means of signal-to-noise (S/N) ratios for
each experiment
Factors Level 1 Level 2 Level 3 Delta Rank
Total phenolic content
(A) Extract 35.95 19.12 – 16.83 1
(B) Temperature 27.19 26.55 26.57 0.64 5
(C) Time 27.88 27.04 25.39 2.49 3
(D) Solvent 27.39 25.62 27.30 1.77 4
(E) Concentration 19.74 27.45 33.11 13.38 2
DPPH radical scavenging activity
(A) Extract 39.16 38.45 – 0.71 2
(B) Temperature 38.71 38.80 38.82 0.11 4
(C) Time 38.72 38.74 38.87 0.14 3
(D) Solvent 38.75 38.81 38.76 0.06 5
(E) Concentration 39.09 38.86 38.38 0.71 1
Diameters of inhibition zones of B. cereus
(A) Extract 20.92 15.03 – 5.89 1
(B) Temperature 18.27 17.51 17.35 0.92 3
(C) Time 17.88 17.86 17.39 0.50 5
(D) Solvent 17.77 17.35 18.01 0.67 4
(E) Concentration 15.23 17.77 20.13 4.91 2

1263
The bioactive efficiency ofultrasonic extracts fromacorn leaves andgreen walnut husks against…
1 3
Table 4 Analysis of variance
(ANOVA) results for S/N
ratios of total phenolic content
and DPPH radical scavenging
activity analyses
*Significant
Factors DF Seq SS Adj SS Adj MS F p
Total phenolic content
A 1 2316.23 1318.18 1318.18 6738.38 0.000*
B 2 2.88 0.26 0.13 0.66 0.579
C 2 36.88 2.30 1.15 5.87 0.092
D 2 22.97 9.97 4.99 25.49 0.013*
E 2 967.04 131.36 65.68 335.75 0.000*
A*B 2 4.10 0.88 0.44 2.25 0.253
A*C 2 2.61 0.72 0.36 1.83 0.302
A*D 2 0.59 0.14 0.07 0.36 0.725
A*E 2 14.15 7.09 3.54 18.11 0.021*
B*D 4 2.16 1.73 0.43 2.21 0.270
B*E 4 3.71 2.03 0.51 2.59 0.230
C*D 4 1.22 1.22 0.30 1.56 0.373
Error 3 0.59 0.59 0.20
Total 32 3375.12
DPPH radical scavenging activity
A 1 4.09699 2.77279 2.77279 228.80 0.001*
B 2 0.07813 0.00334 0.00167 0.14 0.877
C 2 0.15765 0.00285 0.00142 0.12 0.893
D 2 0.02761 0.01737 0.00868 0.72 0.557
E 2 2.74915 0.41254 0.20627 17.02 0.023*
A*B 2 0.08926 0.00270 0.00135 0.11 0.898
A*C 2 0.05099 0.00348 0.00174 0.14 0.872
A*D 2 0.06221 0.00830 0.00415 0.34 0.735
A*E 2 0.67698 0.28431 0.14216 11.73 0.038*
B*D 4 0.07579 0.07791 0.01948 1.61 0.363
B*E 4 0.01757 0.00732 0.00183 0.15 0.950
C*D 4 0.03979 0.03979 0.00995 0.82 0.589
Error 3 0.03636 0.03636 0.01212
Total 32 8.15848
Table 5 Analysis of variance
(ANOVA) results for S/N ratios
of inhibition zone diameters of a
tested strain of Bacillus cereus
*Significant
Factors DF Seq SS Adj SS Adj MS F p
B. cereus
A 1 283.424 147.901 147.901 4499.40 0.000*
B 2 5.363 2.967 1.483 45.13 0.006*
C 2 2.143 0.945 0.473 14.38 0.029*
D 2 2.344 0.194 0.097 2.96 0.195
E 2 129.372 5.256 2.628 79.95 0.002*
A*B 2 0.038 0.275 0.137 4.18 0.136
A*C 2 1.188 2.261 1.130 34.39 0.009*
A*D 2 4.392 4.133 2.066 62.86 0.004*
A*E 2 3.576 4.112 2.056 62.55 0.004*
B*D 4 6.935 2.136 0.534 16.25 0.023*
B*E 4 0.696 7.046 1.762 53.59 0.004*
C*D 4 19.625 19.625 4.906 149.25 0.001*
Error 3 0.099 0.099 0.033
Total 32 459.195

1264 G.Ozturk et al.
1 3
mass transfer from the material into the solvent. Hence, a
higher yield and higher amount of bioactive compound can
be obtained compared to conventional extraction techniques
[49, 53].
The total amount of phenolics in the acorn leaves in this
study was higher than those found by Moreno-Jimenez
etal. [34] who applied the infusion-classical technique
(80°C/10min) to Quercus sideroxyla, Quercus eduardii
and Quercus durifolia leaves, and they found that the phe-
nolic content of the leaves varied from 15.42 to 35.15mg of
GAE/g. In a study conducted by Sanchez-Burgos etal. [54],
it was determined that the TPC for different Quercus species
leaves ranged between 6.98 and 17.26mg of GAE/g, with
the infusion method (boiling water/10min). Similar results
(7.44–35.52mg of catechin/g) were obtained by Popovic
etal. [55], but less than those reported by Rivas-Arreola
etal. [56]. The ultrasound process used during the extraction
process can lead to an increase in the amount of phenolics
from the samples in this current study with the cavitation
phenomena.
Another factor in extracting the bioactive substances from
herbs is the solvent and its polarity. The effect of differ-
ent solvents on the phenolic and antimicrobial activities of
the herbs was investigated. As a result of these investiga-
tions, non-polar compounds dissolve in non-polar solvents,
and polar ones dissolve in polar solvents, and thus the total
amount of phenolics will vary depending on the type of plant
used in the extraction. A 50% aqueous solution was preferred
in this study since pure organic solvents extracted less phe-
nolics than aqueous solutions [57, 58].
The highest total amount of phenolic substances in the
acorn and green walnut husks was found in experiments 3
and 29, respectively. Previous studies emphasized that acorn
leaves are rich in total phenolic content. The combination
of acetone with water can be widely used in the extraction
of tannins from plants and agro-waste. Ahmed etal. [59]
investigated the effect of different solvents on phytochemical
compounds and the biological activity of Quercus dilatata,
and the aqueous acetone was found to be the best solvent
system compared to single or binary solvent systems. Simi-
lar results have been reported by Alasalvar etal. [60] and
Mokhtarpour etal. [61].
Tabaraki and Rastgoo [62] obtained natural antioxidants
from green walnut husks by both conventional and ultrasonic
extraction methods. Compared to the traditional technique,
higher total phenolic content was provided by the ultrasonic
method with a RSM, varying from 6.28 to 7.23mg GAg−1,
with 50% ethanol as the extraction solvent. In another study,
the amount of phenolics in the methanol extract from green
walnut husks was found to be higher than those in petro-
leum ether [28]. They reported that the difference in the
total amount of phenolic content in each extract could be
attributed to the polarity of the solvent used in the extraction.
Another reason for this condition might be attributed to the
environmental conditions. Ghasemi etal. [63] stated that
ecological conditions, including altitude, geography and cli-
mate affected the phytochemical compounds obtained from
a methanol-based extract of green walnut husk in Iran (TPC
varied from 15.15 to 108.11), which is in agreement with
our results.
Antiradical activity
The analysis of variance on the DPPH radical scavenging
activity (Table4) demonstrated that the main effects of
extract type (A), concentration (E) and extract type–con-
centration (A*E) interaction were affected by the percent-
age inhibition of DPPH (p < 0.05). In addition to this, the
observed factors of temperature (B), time (C) and solvent
(D), and the interactions of A*B, A*C, A*D, B*D, B*E
and C*D were not found to be influential (p > 0.05). The
concentration was found to be the most effective factor
on the percentage inhibition of DPPH in comparison with
Table3. According to the single objective Taguchi Method,
the 1.25% concentration of acorn leaf extract, 50% acetone,
50°C and 45min that was shown in the 8th experimental
run of Table2, was found to be suitable for DPPH inhibition.
DPPH is the most commonly applied antioxidant method
as it is a simple and effective method for evaluating the
antioxidant capacity of substances from plants and agro
by-products [64]. In general, there is a definite correlation
between the total phenolics and DPPH [65], and aqueous
solvents have been reported to have higher DPPH activity
than those of pure solvents [66]. Trabelsi etal. [58] exam-
ined the effect of certain solvents on antioxidant compounds
of L. monopetalum leaves, and mixing pure solvents with
water resulted in higher radical-scavenging activity than
pure solvents. Also, they found that 80% acetone had the
highest antiradical activity among solvents used in the study.
In contrast, a study by Tabaraki etal. [67] was conducted
to extract tannin from different acorn tissues, and they opti-
mized the conditions of ultrasonic-assisted extraction by
using a response surface methodology, which was 60°C,
45min and a methanol concentration of 74–82%. Both
researchers reported a positive correlation between the total
phenolics and DPPH. On the other hand, in a study on anti-
oxidants and the cardioprotective potential of leaves of dif-
ferent acorn species, including Quercus sideroxyla, Quercus
eduardii and Quercus resinosa, it was found that the highest
total amount of phenolic material and DPPH activity was
in the extract obtained with the acetone–water mixture (at
3:1) [56], however, they didn’t report a positive correlation
between the total phenolic content and DPPH. A similar
result has also been found for the hazelnut kernels of dif-
ferent cultivars [68]. The absence of this relationship in the
results obtained in the study could be related to the different

1265
The bioactive efficiency ofultrasonic extracts fromacorn leaves andgreen walnut husks against…
1 3
Table 6 Normalized S/N
ratios for TPC, RSA, AA and
principal component scores and
their integrated MRPI
Experi-
mental run
Normalized S/N ratios Principal component scores MRPI
TPC RSA AA PC1 PC2 PC3
1 0.611172 0.990347 0.540768 1.066263 0.713755 − 0.01959 0.944142
2 0.767359 0.952575 0.494302 1.130275 0.660275 − 0.16459 0.969667
3 1 0.8397 0.918094 1.541374 0.413338 − 0.0362 1.178088
4 – – – – – – –
5 – – – – – – –
6 – – – – – – –
7 0.946728 0.803951 0.931591 1.503899 0.385594 0.009882 1.144761
8 0.67504 1 0.646359 1.183627 0.68731 0.009529 1.016829
9 0.804399 0.930846 0.788658 1.347484 0.565306 0.01511 1.092125
10 0.833082 0.872401 0.820637 1.371642 0.496914 0.015127 1.088003
11 0.972174 0.719416 1 1.542855 0.284463 0.036873 1.141269
12 0.622992 0.865033 0.538316 1.036428 0.592315 − 0.03414 0.886884
13 0.653282 0.869841 0.721479 1.18181 0.549458 0.073177 0.974908
14 0.802957 0.813461 0.827287 1.338595 0.444404 0.039133 1.04991
15 0.965371 0.709512 0.891101 1.462004 0.301195 − 0.03515 1.089994
16 0.829092 0.832606 0.843858 1.373068 0.454232 0.032873 1.076355
17 0.97159 0.696261 0.948672 1.501187 0.274194 0.000412 1.109152
18 0.627017 0.835643 0.574254 1.05488 0.555232 − 0.01279 0.888619
19 0.515399 0.017189 0.483413 0.681182 − 0.18689 − 0.02613 0.408963
20 0.01084 0.743137 0 0.222186 0.708987 0.018501 0.366137
21 0.355602 0.443688 0.517023 0.71842 0.242064 0.126104 0.56608
22 0.47974 0.151667 0.518274 0.719285 − 0.05985 0.028449 0.474145
23 0.079134 0.701947 0 0.25673 0.657335 − 0.03149 0.373444
24 0.366886 0.481251 0 0.388653 0.394606 − 0.24381 0.38128
25 0 0.752838 0 0.217617 0.720211 0.026548 0.366518
26 0.417615 0.444792 0.329086 0.634311 0.275122 − 0.04996 0.515996
27 0.519774 0.06607 0.59032 0.770306 − 0.16545 0.047588 0.477441
28 0.321718 0.524172 0.34343 0.601695 0.364966 0.031075 0.521983
29 0.555911 0 0.415253 0.657851 − 0.19496 − 0.10341 0.389455
30 0.125361 0.702949 0 0.288462 0.650005 − 0.06432 0.392452
31 0.528715 0.037163 0.454715 0.676679 − 0.16359 − 0.05505 0.412498
32 0.093507 0.722572 0 0.272468 0.674489 − 0.04098 0.389244
33 0.30483 0.542476 0.21783 0.510887 0.414325 − 0.04451 0.473695
34 0.327556 0.507049 0.25293 0.539749 0.368304 − 0.03726 0.479646
35 0.495805 0.160473 0.523878 0.736532 − 0.05559 0.021278 0.487119
36 0.156351 0.679114 0 0.30265 0.621646 − 0.08718 0.393263
Table 7 The results of eigen analysis of the correlation matrix
Principal compo-
nent (PC)
Eigenvalue Proportion
explained (%)
Cumula-
tive total
(%)
1 2.0512 68.4 68.4
2 0.9053 30.2 98.6
3 0.0435 1.4 100
Table 8 The eigenvectors of the correlation matrix
Variable Principal component
PC1 PC2 PC3
TPC 0.680 − 0.179 − 0.711
RSA 0.289 0.957 0.035
AA 0.674 − 0.229 0.703

1266 G.Ozturk et al.
1 3
polarity solvents and the extract was taken under different
conditions (Table2) so that the different phytochemicals
could pass into the solvents. Thus, there may not be a differ-
ence in DPPH between extracts at different points.
Antimicrobial activity
The ANOVA results for the S/N ratios of the inhibition
zones shown in Table5 indicated that four factors and five
interactions were found to be influential on the antimicrobial
effects of the herbal extracts against B. cereus (p < 0.05).
No significant differences were observed for solvent (D)
and extract type–temperature (A*B) interaction (p > 0.05).
Pursuant to the delta and rank values in Table3, the extract
type is the most significant factor, followed by concentra-
tion, temperature, solvent and time, respectively. As a result,
the optimum conditions for antimicrobial activity are speci-
fied as follows: 5% concentration of acorn leaf extract, 50%
methanol, 50°C and 45min, which corresponds to the 11th
experimental run in Table2. According to the results of the
analysis, although the maximum value for the antimicrobial
activity on B. cereus was obtained on the 11th run, when
the variables including TPC, RSA and AA were evaluated
simultaneously by PCA, it was determined that the third run
was the optimum value.
Compared to green walnut husk, oak leaves have higher
antimicrobial activity against B. cereus. The most critical
factor in this antimicrobial activity may be attributed to the
tannin compounds. Oak leaves are tannin-rich plant tissues
[69]. The tannins have two forms: condensed and hydro-
lyzable tannins. With these compounds being secondary
compounds, tannins can taste bitter [70]. Tannins have a pro-
tective effect against plant herbivores, and can grow on oak
leaves as well as tasting bitter [24]. They also exhibit radical
scavenging, antioxidant and antimicrobial activity [71]. Tian
etal. [72] investigated the polarity effect of certain solvents
on the antimicrobial and antioxidant activities of the extract
obtained from Galla chinensis. As a result of this research,
they noted that tannins are abundant in Galina chinensis,
and the majority of the tannins from Galina chinensis were
found to be non-polar or weak polar, and they also found
that ethyl acetate, ethanol and 80% acetone extracts of it
were effective against Staphylococcus aureus, Bacillus sub-
tilis, and Bacillus cereus at 0.25mgmL−1, 0.5mgmL−1
and 0.5mgmL−1, respectively. Moreno-Jimenez etal. [34]
examined the antioxidant, anti-inflammatory and anticar-
cinogenic activity of red oak (Quercus spp.) leaves. They
determined that the extract from Quercus sideroxyla leaves
have a higher content of condensed tannin than the other
two leaves, Quercus durifolia and Quercus eduardii. Moreo-
ver, the results of the study indicated that the extract from
Quercus sideroxyla leaves had anti-inflammatory and anti-
proliferative effects. Gan etal. [73] investigated the identi-
fication and bioactivity of the basic gallotannins extracted
from red beans, one of the hydrolyzable tannins. The extract
was divided into nine different fractions which consisted of
the I–V fractions eluted by 0–100% ethanol solutions and
the others eluted by 40–100% acetone solutions. From these
fractions, fraction VIII with acetone had the highest antibac-
terial activity against all tested bacteria, including Staphylo-
coccus aureus (ATCC 25923), Bacillus cereus (QAP D15),
Shigella flexneri (QC 5820) and Salmonella enterica serovar
Typhimurium (ATCC 14028), with inhibition zones of 11.9,
19.8, 12.9 and 12.6mm, respectively.
Principal component analysis
The results of the single objective TM for each response
obtained from the trials revealed that the method above is
inadequate for detection of the joint optimum level for all
responses. For this purpose, the PCA-based hybrid Tagu-
chi approach was preferred for multi-response optimization/
Table 9 Response table for means of MRPI values
Factors Level 1 Level 2 Level 3 Max–Min Rank
(A) Extract type 1.0434 0.4372 – 0.6062 1
(B) Temperature 0.7371 0.7390 0.7448 0.0077 5
(C) Time 0.7364 0.7382 0.7462 0.0098 4
(D) Solvent 0.7349 0.7274 0.7585 0.0311 3
(E) Concentration 0.6621 0.7729 0.7859 0.1238 2
Table 10 General linear
ANOVA results for MRPI
values
*Significant
Factors DF Seq SS Contribution (%) Adj SS Adj MS F-value p-Value
Extract 1 3.00658 94.21 3.00658 3.00658 859.24 0.000*
Concentration 2 0.09872 3.09 0.09872 0.04936 14.11 0.000*
Solvent 2 0.00492 0.15 0.00563 0.00281 0.80 0.460
Time 2 0.00040 0.01 0.00057 0.00029 0.08 0.922
Temperature 2 0.00037 0.01 0.00035 0.00017 0.05 0.952
Error 23 0.08048 2.52 0.08048 0.00350
Total 32 3.19147 100.00

1267
The bioactive efficiency ofultrasonic extracts fromacorn leaves andgreen walnut husks against…
1 3
estimation. The normalized S/N ratios for each response were
computed with Eq.(1) and displayed in Table6. The calcu-
lated eigenvalues and eigenvectors of the correlation matrix
are given in Tables7 and 8, respectively. The principal com-
ponent (PC) scores (PC1, PC2, and PC3) and their combined
multi-response performance index (MRPI) values for each
experimental run were acquired with Eqs.(2, 3) and shown in
Table6. The prediction of optimum levels for all factors was
possible with a maximum MRPI value that corresponded to the
factor levels of the third experimental run (5% concentration of
acorn leaf extract, 50% acetone, 60°C and 60min) in Table2.
The factors affecting the levels of each parameter were ranked
by means of all the MRPI values for every level that was indi-
cated in Table9. The results of the general linear ANOVA
in Table10 demonstrated the contribution of each factor in
descending order as extract type, concentration, solvent, time
and temperature, and the extract type and concentration were
found to be the most significant (p < 0.05) of all the factors.
Conclusions
In this study, the use of the PCA-TM hybrid method has been
attempted to find the solution to a multi-response estimation/
optimization problem, in conjunction with a case investi-
gation in both the fields of food and agricultural sciences.
Furthermore, a comparison has been performed between
single objective TM and PCA-TM hybrid methods for this
problem. According to this study, the following conclusions
can be declared:
(1) In single objective estimation with TM, different factors
levels were found for each response. The predicted opti-
mum factor settings for TPC, RSA and AA are 13333,
12231 and 12213, respectively. Also, the results of tri-
als were confirmed by comparison with studies in pub-
lished literature based on phytochemicals and herbal
extracts.
(2) The results of the RSA analysis with a single objec-
tive TM showed that the inhibition percentage is inad-
equate to express the RSA of the studied herbal extracts
with the mixed level orthogonal array design. It could
be more suitable to use more reliable methods (e.g.,
Trolox equivalent antioxidant capacity assay [TEAC] or
ascorbic acid equivalent, EC50) instead of the percent-
age of inhibition.
(3) From the comparison of significant factors identified
by ANOVA, only the extract type and concentration
were found to be significant among the MRPI factors,
while various factors and their interactions that are effi-
cient, as identified by the single-objective TM method,
are mentioned in the preceding sections. This kind of
loss of expression or quality of some characteristics
in a PCA-TM hybrid method is always probable when
compared to a single objective TM [45]. However, the
overall reliability was improved.
(4) In some cases, the use of PCA can be disadvantageous
if there is more than one principal component whose
eigenvalue is > 1, and the solution of this shortcom-
ing is unknown [43]. However, there was only one
eigenvalue found, which was used for the calculation
of MRPI values, to be higher than 1 (Table7).
(5) In a PCA-based Taguchi hybrid method, an optimal
factor setting was determined with the MRPI for all
responses corresponding to the factor levels (13333)
of the third experimental run in the design matrix. This
method has proved itself from the point of the estima-
tion of optimal levels in this study. However, it is neces-
sary to probe the usage of the PCA-TM hybrid method
in the field of both food and agricultural sciences by
considering different conditions and factors.
Acknowledgements We would like to thank ERU-TAUM (Erciyes
University, Technology Research and Application Center) for letting
us use their laboratory-type freeze-dryer.
Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflict of
interest.
References
1. K.W. Wahle, I. Brown, D. Rotondo, S.D. Heys, Adv. Exp. Med.
Biol. 698, 36 (2010)
2. S. Jafari, S. Saeidnia, M. Abdollahi, Curr. Pharm. Biotechnol.
15(4), 409 (2014)
3. M. Carocho, M.F. Barreiro, P. Morales, I.C.F.R. Ferreira, Compr.
Rev. Food Sci. Food Saf. 13(4), 377 (2014)
4. WHO, Antibiotic resistance. http://www.who.int/media centr e/
facts heets /antib iotic -resis tance /en/. Accessed 5 Feb 2018
5. A. Moure, J.M. Cruz, D. Franco, J.M. Dominguez, J. Sineiro,
H. Dominguez, M.J. Nunez, J.C. Parajo, Food Chem. 72(2), 145
(2001)
6. N. Balasundram, K. Sundram, S. Samman, Food Chem. 99(1),
191 (2006)
7. F. Kallel, D. Driss, F. Chaari, L. Belghith, F. Bouaziz, R. Ghorbel,
S.E. Chaabouni, Ind. Crops Prod. (2014) https ://doi.org/10.1016/j.
indcr op.2014.07.047
8. O.H. Lee, B.Y. Lee, Bioresour. Technol. 101(10), 3751 (2010)
9. B. Melgar, M.I. Dias, A. Ciric, M. Sokovic, E.M. Garcia-Castello,
A.D. Rodriguez-Lopez, L. Barros, I. Ferreira, Ind. Crops Prod.
107, 353 (2017)
10. J.L. McKillip, Antonie Van Leeuwenhoek 77(4), 393 (2000)
11. A. Osimani, L. Aquilanti, F. Clementi, Int. J. Hosp. Manag. 72,
145 (2018)
12. A. Tewari, S. Abdullah, J. Food Sci. Technol. 52(5), 2500 (2015)
13. E.J. Bottone, Clin. Microbiol. Rev. 23(2), 382 (2010)
14. P.E. Granum, in Applications and systematics of Bacillus and rela-
tives, ed. by R. Berkeley, M.N. Heyndrickx, P. de Vos (Blackwell,
Oxford, 2002), p.37

1268 G.Ozturk et al.
1 3
15. H. Jun, J. Kim, J. Bang, H. Kim, L.R. Beuchat, J.H. Ryu, Int. J.
Food Microbiol. 160(3), 260 (2013)
16. L.R. Beuchat, in Microbial Food Contamination, ed. by C.L. By,
S. Wilson, Droby (CRC Press, Boca Raton, 2000), p.149
17. T.K. Lim, Edible Medicinal and Non-medicinal Plants, vol.3
(Springer, New York, 2012), pp.70–72
18. J. Azmir, I.S.M. Zaidul, M.M. Rahman, K.M. Sharif, A.
Mohamed, F. Sahena, M.H.A. Jahurul, K. Ghafoor, N.A.N. Noru-
laini, A.K.M. Omar, J. Food Eng. 117(4), 426 (2013)
19. B.K. Tiwari, Trac-Trend Anal. Chem. 71, 100 (2015)
20. A. Phaniendra, D.B. Jestadi, L. Periyasamy, Indian J. Clin. Bio-
chem. 30(1), 11 (2015)
21. V. Lobo, A. Patil, A. Phatak, N. Chandra, Pharmacogn. Rev. 4(8),
118 (2010)
22. H.M. Ali, A. Abo-Shady, H.A.S. Eldeen, H.A. Soror, W.G.
Shousha, O.A. Abdel-Barry, A.M. Saleh, Chem. Cent. J. 7(1), 53
(2013)
23. M.J. Alves, I.C. Ferreira, H.J. Froufe, R. Abreu, A. Martins, M.
Pintado, J. Appl. Microbiol. 115(2), 346 (2013)
24. J.P. Salminen, T. Roslin, M. Karonen, J. Sinkkonen, K. Pihlaja, P.
Pulkkinen, J. Chem. Ecol. 30(9), 1693 (2004)
25. C. Moctezuma, A. Hammerbacher, M. Heil, J. Gershenzon, R.
Méndez-Alonzo, K. Oyama, J. Chem. Ecol. 40(5), 458 (2014)
26. M. Gulluce, A. Adiguzel, H. Ogutcu, M. Sengul, I. Karaman, F.
Sahin, Phytother. Res. 18(3), 208 (2004)
27. V. Nour, I. Trandafir, S. Cosmulescu, J. Chromatogr. Sci. 51(9),
883 (2013)
28. M. Carvalho, P.J. Ferreira, V.S. Mendes, R. Silva, J.A. Pereira, C.
Jeronimo, B.M. Silva, Food Chem. Toxicol. 48(1), 441 (2010)
29. A. Fernandez-Agullo, E. Pereira, M.S. Freire, P. Valentao, P.B.
Andrade, J. Gonzalez-Alvarez, J.A. Pereira, Ind. Crops Prod. 42,
126 (2013)
30. F. Stampar, A. Solar, M. Hudina, R. Veberic, M. Colaric, Food
Chem. 95(4), 627 (2006)
31. B.B. Shi, W.N. Zhang, X. Li, X.J. Pan, Int. J. Food Prop. 20,
S2635 (2018)
32. I. Oliveira, A. Sousa, I.C. Ferreira, A. Bento, L. Estevinho, J.A.
Pereira, Food Chem. Toxicol. 46(7), 2326 (2008)
33. N.E. Rocha-Guzman, J.A. Gallegos-Infante, R.F. Gonzalez-
Laredo, R. Reynoso-Camacho, M. Ramos-Gomez, T. Garcia-
Gasca, M.E. Rodriguez-Munoz, S.H. Guzman-Maldonado, L.
Medina-Torres, B.A. Lujan-Garcia, Food Chem. 115(4), 1320
(2009)
34. M.R. Moreno-Jimenez, F. Trujillo-Esquivel, M.A. Gallegos-
Corona, R. Reynoso-Camacho, R.F. Gonzalez-Laredo, J.A.
Gallegos-Infante, N.E. Rocha-Guzman, M. Ramos-Gomez, Food
Chem. Toxicol. 80, 144 (2015)
35. M.H. Xu, X.Y. Yang, M.R. Fu, Waste Biomass Valoriz. 7(5), 1159
(2016)
36. P.S. Makkar, B. Singh, Bioresour. Technol. 37(2), 185 (1991)
37. M.T. Moradi, A. Karimi, S. Alidadi, Chin. J. Nat. Med. 14(3), 196
(2016)
38. A. Bevilacqua, M. Corbo, M. Sinigaglia, Current Research, Tech-
nology and Education Topics in Applied Microbiology and Micro-
bial Biotechnology (Formatex Research Center, Badajoz, 2010),
p.1419
39. R.K. Roy, A Primer on the Taguchi Method, 2ndedn. (Society of
Manufacturing Engineers, USA, 2010), pp.9–24
40. R.S. Rao, C.G. Kumar, R.S. Prakasham, P.J. Hobbs, Biotechnol.
J. 3(4), 510 (2008)
41. M.H. Oztop, S. Sahin, G. Sumnu, J. Food Eng. 79(1), 83 (2007)
42. Minitab Getting Started with Minitab 18. https ://www.minit
ab.com/uploa dedFi les/Docum ents/getti ng-start ed/Minit abGet
tingS tarte d_EN.pdf. Accessed 5 Feb 2017
43. R. Jeyapaul, P. Shahabudeen, K. Krishnaiah, Int. J. Adv. Manuf.
Technol. 26(11–12), 1331 (2005)
44. J. Antony, Qual. Reliab. Eng. Int. 16(1), 3 (2000)
45. A.K. Dubey, V. Yadava, Opt. Lasers Eng. 46(2), 124 (2008)
46. T. Yih-Fong, C. Fu-Chen, Int. J. Mach. Mach. Mater. 1(1), 76
(2006)
47. R.L. Bradley, in Food Analysis, ed. by S.S. By, Nielsen (Springer,
Boston, 2010), p.85
48. H. Abdi, L.J. Williams, Wiley Interdiscip. Rev. Comput. Stat. 2(4),
433 (2010)
49. M. Hayta, E.M. Iscimen, LWT-Food Sci. Technol. 77, 208 (2017)
50. K. Gezer, M. Duru, I. Kivrak, A. Turkoglu, N. Mercan, H. Turko-
glu, S. Gulcan, Afr. J. Biotechnol. 5 (20), 1924 (2006)
51. H. Baydar, O. Sağdiç, G. Özkan, T. Karadoğan, Food Control
15(3), 169 (2004)
52. A. Soto-Vaca, A. Gutierrez, J.N. Losso, Z.M. Xu, J.W. Finley, J.
Agric. Food Chem. 60(27), 665 (2012)
53. K.M. Hammi, A. Jdey, C. Abdelly, H. Majdoub, R. Ksouri, Food
Chem. 184, 80 (2015)
54. J.A. Sanchez-Burgos, M.V. Ramirez-Mares, M.M. Larrosa, J.A.
Gallegos-Infante, R.F. Gonzalez-Laredo, L. Medina-Torres, N.E.
Rocha-Guzman, Ind. Crops Prod. 42, 57 (2013)
55. B.M. Popovic, D. Stajner, R. Zdero, S. Orlovic, Z. Galic, Sci.
World J. (2013) https ://doi.org/10.1155/2013/13465 6
56. M.J. Rivas-Arreola, N.E. Rocha-Guzman, J.A. Gallegos-Infante,
R.F. Gonzalez-Laredo, M. Rosales-Castro, J.R. Bacon, R. Cao, A.
Proulx, P. Intriago-Ortega, Pak. J. Biol. Sci. 13(11), 537 (2010)
57. M. Markom, M. Hasan, W.R.W. Daud, H. Singh, J.M. Jahim, Sep.
Purif. Technol. 52(3), 487 (2007)
58. N. Trabelsi, W. Megdiche, R. Ksouri, H. Falleh, S. Oueslati, B.
Soumaya, H. Hajlaoui, C. Abdelly, LWT-Food Sci. Technol.
43(4), 632 (2010)
59. M. Ahmed, H. Fatima, M. Qasim, B. Gul, BMC Complement
Altern. Med. 17, 386 (2017)
60. C. Alasalvar, M. Karamac, R. Amarowicz, F. Shahidi, J. Agric.
Food Chem. 54(13), 4826 (2006)
61. A. Mokhtarpour, A.A. Naserian, R. Valizadeh, M. Danesh Mes-
garan, Annu. Rev. Res. Biol. 4(8), 1330 (2014)
62. R. Tabaraki, S. Rastgoo, Korean J. Chem. Eng. 31(4), 676 (2014)
63. K. Ghasemi, Y. Ghasemi, A. Ehteshamnia, S.M. Nabavi, S.F.
Nabavi, M.A. Ebrahimzadeh, F. Pourmorad, J. Med. Plant Res.
5(7), 1128 (2011)
64. S.B. Kedare, R. Singh, J. Food Sci. Technol. 48(4), 412 (2011)
65. N. Pawar, S. Pai, M. Nimbalkar, G. Dixit, Food Chem. 126(3),
1330 (2011)
66. I.F. Almeida, M.H. Amaral, P.C. Costa, M.F. Bahia, P. Valentão,
P.B. Andrade, R.M. Seabra, T.M. Pereira, Biofactors 33(4), 267
(2008)
67. R. Tabaraki, A. Safari, F.A. Yeganeh, J. Appl. Chem. Res. 7(3),
67 (2013)
68. J. Jakopic, M. Mikulic-Petkovsek, A. Likozar, A. Solar, F. Stam-
par, R. Veberic, Food Chem. 124(3), 1100 (2011)
69. M.Y. Elahi, Y. Rouzbehan, Anim. Feed Sci. Technol. 140(1–2),
78 (2008)
70. R. Amarowicz, A. Troszynska, N. Barylko-Pikielna, F. Shahidi,
J. Food Lipids 11(4), 278 (2004)
71. K.T. Chung, T.Y. Wong, C.I. Wei, Y.W. Huang, Y. Lin, Crit. Rev.
Food Sci. Nutr. 38(6), 421 (1998)
72. F. Tian, B. Li, B. Ji, G. Zhang, Y. Luo, LWT-Food Sci. Technol.
42(7), 1289 (2009)
73. R.Y. Gan, K.W. Kong, H.B. Li, K. Wu, Y.Y. Ge, C.L. Chan, X.M.
Shi, H. Corke, Front. Chem. 6, 39 (2018)
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.