Food Chemistry 197 (2016) 1215–1225

Contents lists available at ScienceDirect

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

A novel process for preparing low-fat peanuts: Optimization of the oil

extraction yield with limited structural and organoleptic damage
Joelle Nader ⇑, Nada Fawaz, Charbel Afif, Nicolas Louka ⇑
Laboratoire d’Intensification des Procédés Agro-Industriels (LIPAI), Unité de Recherche: Technologie et Valorisation Agroalimentaire (TVA), Centre d’Analyses et de Recherche,
Faculté des Sciences, Université Saint Joseph, Campus des Sciences et Technologies, Beirut, Lebanon

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

Article history: The main purpose of this study was to extract the maximum amount of oil from peanuts without causing
Received 31 May 2014 major damage and preserving their organoleptic quality after defatting. Accordingly, a successful,
Received in revised form 11 June 2015 healthy, eco-friendly and economic defatting process for peanuts was implemented using mechanical
Accepted 14 November 2015
oil expression, which was optimized by means of Response Surface Methodology. The results
Available online 21 November 2015
demonstrated that maximum extraction yields were obtained at a low initial moisture content (5–7%
d.b.). Defatting and deformation ratios were mostly affected by the pressure and water content with high
Keywords:
correlation coefficients (98.4% and 97.5%, respectively), and overall acceptability decreased following
Defatting
Pressing
higher oil extraction yields. It was concluded that the optimum values for the product moisture content,
Peanuts pressure, and pressing duration were 5% d.b., 9.7 MPa and 4 min, respectively, with a defatting ratio of
Oil yield 70.6%. This resulted in an insignificant irreversible deformation ratio (<1%) and an overall acceptability
Shape preservation of 7.6 over 10.
Response Surface Methodology Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction
The peanut is a plant from the Fabaceae family that is native to
South America and is cultivated in tropical, subtropical, and tem-
perate climate regions (Hammons, 1973). Approximately 29 mil-
lion metric tons of peanuts are produced annually worldwide,
mostly in China, India, and the USA (American Peanut Council,
2014). These oleaginous seeds are concentrated sources of essen-
tial unsaturated fatty acids, with a number of other nutrients and
bioactive components (Kris-Etherton et al., 1999). Several studies
have related the regular consumption of nuts and seeds with
various health benefits, such as lower cholesterol and a decreased
risk of cardiovascular disease and type 2 diabetes (Fraser, 1999).
Peanuts are also an excellent source of protein with a reasonable
amount of carbohydrates, vitamins, and minerals.
Over the past 20 years, peanut processing technologies have
witnessed an outstanding growth in the food industry, notably in
the manufacturing of vegetable oil, peanut butter, and snack
products, such as roasted peanuts. Consumers have become
⇑ Corresponding authors at: Centre d’Analyses et de Recherche, Faculté des
Sciences, Université Saint Joseph, B.P. 11-514, Riad El Solh, Beirut 1107 2050,
Lebanon (J. Nader). Faculté des Sciences, Université Saint Joseph, Campus des
Sciences et Technologies, Mar Roukos, Mkallès, B.P. 11-514, Riad El Solh, Beirut
1107 2050, Lebanon (N. Louka).
E-mail addresses: joelle.nader@net.usj.edu.lb (J. Nader), nicolas.louka@fs.usj.
edu.lb (N. Louka).
increasingly aware of their health, and because the caloric density
of fat is double that of protein and carbohydrates (Holloway &
Wilkins, 1982), an efficient defatting process would produce a
lower calorie product while maintaining the organoleptic qualities
of the product.
Generally, the extraction of oil from oilseeds can be achieved
using a variety of methods that tend to damage the raw materials
and the extracted oil. One of the methods most commonly reported
in the literature is extraction with an organic solvent, such as hex-
ane (Melgarejo Navarro Cerutti, Ulson de Souza, & de Arruda Guelli
Ulson de Souza, 2012; Mani, Jaya, & Vadivambal, 2007). Other
methods include using water for an aqueous extraction
(Campbell & Glatz, 2009) with pressurized gaseous solvents
(Venter, Willems, Kuipers, & de Haan, 2006) or with compressed
solvents, such as propane and ethanol (Jesus et al., 2013). Lately,
supercritical ‘green’ solvents (Salgın & Salgın, 2013) have also been
used. The main drawbacks of solvent extraction are that hexane
(the most used solvent) is expensive and that chronic exposure
to this type of solvent causes neurologic and other disorders
(NIOSH/OSHA/DOE Health Guidelines, 1996), as cited by Russin,
Boye, Arcand, and Rajamohamed (2011).
The second method that is commonly used is mechanical
expression, which leads to a relatively low extraction yield
(40–52%) compared with the solvent extraction method (80–90%)
(Holloway, Finley, & Wheeler, 1991). To increase oil extractability,
several destructive pretreatments need to be performed, such as

http://dx.doi.org/10.1016/j.foodchem.2015.11.079
0308-8146/Ó 2015 Elsevier Ltd. All rights reserved.
1216 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225

grinding (Mrema & Mc Nulty, 1985), extrusion (Evangelista, 2009;
Kartika, Pontalier, & Rigal, 2010), enzymatic (Gaur, Sharma, Khare,
& Gupta, 2007; Jiang, Hua, Wang, & Xu, 2010), ultrasonication
(Abdullah & Koc, 2013), or thermal and microwave pretreatments
(Jiao et al., 2014; Qu et al., 2013).
In addition to solvent extraction and mechanical expression, the
most widely used method in the literature is extraction by super-
critical carbon dioxide or by the combination of enzymatic treat-
ment with any other means of extraction.
The purpose of using enzymes in the enzymatic extraction
method is to damage and/or degrade cell walls, thereby increasing
the permeability of oil in oilseeds (Domínguez, Núnẽz, & Lema,
1995). Although this method has the advantage of a high extrac-
tion rate, it causes the degradation of grains, which is not recom-
mended in the defatting process for snack industries.
On the other hand, the advantages of oil extraction by supercrit-
ical carbon dioxide are its easy removal from food samples without
contamination, the fact that it is environmentally and nutritionally
non-toxic and that the extraction conditions are relatively mild,
thereby leading to a high oxidative stability and low protein denat-
uration. Nevertheless, this extraction operation is very expensive
(Passey & Patil, 1994). Therefore, an alternative method of oil
removal would be more efficient and beneficial to manufacturers.
Consequently, after performing a comparative study of these
various defatting methods, the defatted grains, extracted oil qual-
ity, oil yield and cost effectiveness were the most important fac-
tors, and mechanical expression was adopted to extract oil from
groundnut seeds. This method was chosen because it is the cheap-
est method in the industry and is the healthiest way to remove oil
without polluting the grain samples. Therefore, the highest possi-
ble oil recovery from groundnuts is possible at an economical rate.
After pressing, the seeds are flattened and soaked in water for a
specific time to recover their initial shape and size (Wilkins &
Gannis, 1984; Passey & Patil, 1994).
Several patents have been released with the purpose of
attempting to produce low-fat nuts, but as shown in Table 1, the
inventors were unable to achieve high extraction yields without
damaging or deforming the seeds. Additionally, their processes
consumed high amounts of energy (for heating, long pressing dura-
tions and very high pressures), with relatively low percentages of
extracted oil. As a result, none of these methods have been retained
by the industry because of their inefficiency and low return on
investment. Mechanical pressing, which eliminates oil, is
usually performed at pressures between 6.9 and 10.3 MPa for
15–120 min. At higher pressures (>10.3 MPa), there is a significant
loss of flavor. At lower pressures (<6.9 MPa), the pressing time to
extract a certain percentage of oil has to be extended (Holloway
et al., 1991).
Therefore, studies have been developed to research the most
cost effective methods for defatting. These methods would be use-
ful for industries that produce popular snacks and products from
defatted nuts. The main objective of this study was to optimize
the defatting process parameters of peanuts by means of the
Response Surface Methodology. The aim is to obtain a maximum
oil extraction yield with minimum damage and breakage of the
seeds during pressing, while preserving their organoleptic quali-
ties. The obtained defatted nuts would satisfy consumers because
they would retain their original shape and taste but have a lower
amount of calories.
2. Material and methods
A novel defatting process called MEPPI (Mechanical Extraction
Preserving Product Integrity) was designed, implemented and opti-
mized to prepare low-fat, high fiber, high protein peanuts, thereby
preserving the structural and organoleptic properties of the fin-
ished product.
2.1. Sample preparation
Unshelled peanuts of ‘‘Arachis Hypogaea” botanical origin and
of Virginia type were imported from China (Laixi city shunxiang
peanuts product’s Co. LTD) and shipped to ‘‘El Kazzi”, which is a
local nut roasting facility in Beirut, Lebanon. The peanuts were
manually cleaned and sieved twice using 8.5- and 7.5-mm square
mesh sieves to select those that had approximately the same shape
and dimensions.
2.2. Pretreatments
2.2.1. Light initial roasting and peeling
An initial roasting is often performed to produce the flavor and
color that characterize peanuts. This process also reduces the water
content and denatures proteins to facilitate the subsequent extrac-
tion of oil (Holloway & Wilkins, 1982). Preliminary optimization of
the initial roasting was performed. The optimal conditions
obtained from this study were: roasting at 140 °C for 15 min with
a cake thickness of 3.3 cm. Then, the peanuts were manually
peeled.
2.2.2. Hydration
Hydration before pressing is important to increase the com-
pressibility of the seed and its resistance to disintegration. How-
ever, this process leads to a loss of many important soluble
materials by dissolution in water. Therefore, according to the liter-
ature, the water content before pressing should not exceed 8% d.b.,

Table 1
Process parameters and extraction yields previously obtained by researchers.

Grain type Pressure Pressing duration Temperature Water content (weight per Oil extraction yield
(MPa) (min) (°C) cent) (%)
Peanuts (Ammann, 1935) 34.3 45–60 80–90 1–4 60–70
Peanuts (Holloway & Wilkins, 1982) 6.9–10.3 15–120 – 4–8 20–55
8.3a 5a 35–45a
Peanuts (cold pressing) (Vix et al., 1966) 13.8 30–120 0–40 3–8 20–90
Peanuts (hot pressing) (Vix et al., 1966) 13.8 60 82.2–104.4 3–8 20–90
Peanuts (Wilkins & Gannis, 1984) 8.3–10.3 15–120 20–50 4–6 20–60
35–55a
Peanuts (Simelunas, Wilkins, & Gannis, 1985) 17.2 10–20 20–50 5–7 35–55
Peanuts (Wong & Sackenheim, 1992) 10.3–51.7 5–30 15 <6 25–80
35.9a 10a 3–5a 80a
Peanuts (Holloway et al., 1991) 7.6–9 15–20 – – 40–52
Melon seeds (Ajibola, Eniyemo, Fasina, & Adeeko, 25 10 130 9.2 41.6
1990)
a
Optimum values.
 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225
not only to avoid losing flavor (Holloway & Wilkins, 1982) but also
to avoid the restriction and slowing down of oil extractability.
Therefore, the moisture content before pressing was one of the
important variables to optimize during the defatting process.
After several investigations and trials, humidification by soak-
ing at room temperature was found to be the most recommended
method and induced a homogeneous hydration of the product.
This humidification process was followed by a relaxation period
in the refrigerator at 4 °C for 2–3 days. The seeds were packed in
tight PVC bags to ensure homogenization and the uniform distribu-
tion of water throughout the seeds.
The standard hot air oven drying method at 105 ± 1 °C for 24 h
(AOAC, 1990) was used. The moisture content of the raw and ini-
tially roasted peanuts was determined five times and was found
to be 5.56 ± 0.24% d.b. and 2.48 ± 0.12% d.b., respectively. The pea-
nuts were soaked in distilled water with a 1:5 (g/cm3) grain to
water ratio to allow for a sufficient amount of time to reach the
recommended moisture content levels (Eq. (1)), which were used
later in the experimental design:
MiðWf  WiÞ
Q¼
100  Wf
where Q (kg) is the mass of water added, Mi (kg) is the initial mass
of the sample, and Wi and Wf are, respectively, the initial and the
final moisture content of the sample in % d.b.
2.3. Experimental setup
A hydraulic press at the university’s laboratory, which weighs
approximately 1000 kg, was used to provide a maximum pressure
of 26 MPa. This press is mainly composed of: (1) a hydraulic unit,
(2) a steel frame with a plunger for uniaxial compression and (3)
a stainless steel cylinder. These components are operated by (4)
an automated control and measurement system that can be remo-
tely operated (Fig.1). Energy is provided by an electric asyn-
chronous engine of 7.5 kW that drives the hydraulic pump at up
to 12 dm3/min of oil flow.
The press cylinder capacity is 7.7 dm3. The piston stroke is
40 cm, and the internal diameter of the cylinder is 20 cm. The max-
imum downward and upward speed of the piston is 13 mm/s and
Fig. 1. Hydraulic press with details of sieve plates installed at the bottom of the cylinder.
1217
20 mm/s, respectively. Pressure reading is performed by an elec-
tronic pressure sensor with an accuracy of ±0.1 MPa, while the pis-
ton speed is regulated by a speed control valve (±0.5 mm/s). The
press is also fitted with a thermocouple (±0.1 °C) and a position
transducer (±0.01 mm).
2.4. Defatting by hydraulic pressing
The oil extraction technique used in this defatting process was
carried out without using any polluting agents, such as chemicals
or solvents. The finished product, therefore, was natural, with
reduced fat, and was rich in proteins. In addition to the effects of
the physiochemical properties of the material itself (such as the
nature and variety of seeds, oil viscosity, and raw versus pretreated
seeds), hydraulic pressing is generally influenced by five indepen-
dent operating parameters: the moisture content of peanuts before
pressing, applied pressure, displacement speed of the piston or
compression speed, cake thickness expressed as a function of mass
of seeds per pressing unit area (kg/cm2), and pressing duration
(Lanoisellé, 1995). To identify the ranges of the operating parame-
ð1Þ ters, different experiments were carried out to study the effects of
these variables on the defatting ratio and the rate of irreversible
deformation.
2.4.1. Selection of process and response parameters
Based on the preliminary test results, the treatment conditions
were as follows:
(a) After hydration, the samples were allowed to equilibrate in
the refrigerator for 3 days at 4 °C, at which point the water
homogenization inside the seeds was sufficient to resist
deformation during pressing.
(b) The method of pressing with multiple compression cycles
and separation by relaxation periods (Pérez-Gálvez et al.,
2009) was rejected because this method does not improve
the defatting ratio and is economically discouraged in the
industrial field.
(c) The compression speed was set at 8 mm/s for all of the
experiments because it does not significantly affect the oil
extraction yield. Approximately 250 g of peanut seeds were
1218 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225

spread uniformly at the base of the cylinder, with an internal
surface area of 312.6 cm2. The selected variant process
parameters were: pressure (4–10 MPa), pressing duration
(10–30 min), and water content before pressing (7–13% d.b.).
Moreover, it was noted that temperature is a key factor that
positively affects the oil extraction yield. Its effect has been
studied by many authors. Further to several experiments
conducted at the university’s laboratory, the temperature
parameter was omitted to preserve the largest amount of
good quality partially defatted peanuts.
A central composite rotatable design (23 factorial design with 6
repetitions at the central level and 6 axial points where one vari-
able is set at an extreme level (±1.6818) while other variables are
set at their central points) was created. Statistical analyses, as well
as response surfaces, were performed using the analysis design
procedure of Statgraphics Plus (5.1 version, Windows software).
This experimental design included 20 experiments, each repeated
three times.
2.5. Post-treatments

Generally, there were two risks during the pressing process: (a)
breaking the seeds and (b) the loss of flavor because of the
extracted oil. Therefore, the response parameters that will be eval-
uated and optimized in the current study are: the defatting ratio
(%), the rate of irreversible deformation (%), the compression ratio,
and the quality of the defatted peanuts (the overall acceptability,
including the shape, taste, texture, color and odor).
2.4.2. Experimental design and statistical analysis
The effects of the chosen process parameters on oil extraction
yield, aspect and sensory quality of partially defatted peanuts were
studied using Response Surface Methodology (RSM). The levels of
independent variables are shown in Table 2. A nonlinear second
degree polynomial equation (Eq. (2)) was fitted in each response
to describe the process empirically:
X
n X
n1 X
n X
n
Y ¼ a0 þ ai xi þ aij xi xj þ aii x2i
i¼1 i¼1 j¼iþ1 i¼1
At the end of each compression test, the seeds were flattened,
necessitating a reconstitution process. The compressed peanuts
were soaked in 0.6 dm3 of distilled water for a sufficient time
(approximately 30 min) to achieve a full shape recovery. According
to Vix, Spadaro, and Pominski (1966), drying after reconstitution
should be performed in air at 107–260 °C to prevent the kernels
from reverting to their compressed physical dimensions. Further-
more, at these temperatures, the evaporation of water from the
surface of the nutmeats prevents excessive heating and possible
burning. Contrary to what has been mentioned in the literature,
the samples were dried in an oven at 50 °C for an average of 6 h
to reach a water content of 7.5–11% d.b. without affecting their
appearance or causing burning. After drying, the samples were
roasted at 180 °C for 3 min and 30 s (these parameters were iden-
tified after several previous trials and preliminary experiments).
Granular roasting was carried out using thick salt as a heat convec-
tion medium (Mrad, Debs, Maroun, & Louka, 2014).
ð2Þ
2.6. Physicochemical analysis

where a0, ai, aii and aij are the regression coefficients; xi and xj are
the coded levels of independent variables; and n is the number of
operating parameters. The probability (p-value) generated from
analysis of variance (ANOVA) and the coefficient of determination
(R2) represent the significance of independent variables and the
percentage of the experimental results explained by the model as
fitted, respectively.
2.6.1. Lipid content
The oil content was determined in a Soxhlet apparatus accord-
ing to AOAC method 948.22 using petroleum ether as the solvent.
The samples were ground using a mortar and were dried in an oven
at 105 °C until a constant weight was reached. Next, 2 g of the
ground and dried sample were weighed and placed in an extraction
thimble. The oil was extracted continuously with 90 cm3 of

Table 2
Effects of pressing parameters on defatting ratio, aspect, deformation and acceptability of partially defatted finished product.

RN Coded levels Actual Response parameters

X1 X2 X3 X1 X2 X3 Y1a Y2a Y3b Y4a
1 1 1 +1 7 4 30 40.85 2.33 7.78 1.0363
2 +1 1 1 13 4 10 13.92 8.13 8.49 1.0154
3 1.6818 0 0 4.95 7 20 70.62 3.28 6.44 1.0335
4 1 1 1 7 4 10 34.15 1.76 8.23 1.0337
5 +1 +1 1 13 10 10 24.98 36.02 5.51 1.0122
6 +1 +1 +1 13 10 30 27.07 37.94 5.78 1.0511
7 +1.6818 0 0 15 7 20 17.79 37.85 7.58 1.0528
8 0 0 1.6818 10 7 3.18 22.97 8.88 7.33 1.0053
9 1 +1 +1 7 10 30 57.44 11.1 5.5 1.0856
10 0 +1.6818 0 10 12.05 20 32.68 18.11 5.69 1.0862
11 1 +1 1 7 10 10 48.55 9.95 7.86 1.0606
12 0 0 +1.6818 10 7 36.82 24.77 12.55 6.26 1.0437
13 0 1.6818 0 10 1.95 20 9.6 0.03 8.37 1.0321
14 +1 1 +1 13 4 30 14.29 11.6 8.19 1.0236
15 0 0 0 10 7 20 23.6 15.76 7.42 1.0345
16 0 0 0 10 7 20 24.61 15.69 7.81 1.0342
17 0 0 0 10 7 20 23.34 15.6 7.85 1.0379
18 0 0 0 10 7 20 24.6 16.35 7.97 1.0355
19 0 0 0 10 7 20 21.34 15.32 7.89 1.0373
20 0 0 0 10 7 20 22.94 15.41 7.87 1.0357

RN: Run Number.

X1: Water content (% d.b.); X2: Pressure (MPa); X3: Pressing duration (min); Y1: Defatting ratio (%); Y2: Unrecoverable deformation ratio (%); Y3: Acceptability (/10); Y4:
Compression ratio.
a
Values are represented as means of three repeated experiments with a mean standard deviation of 1.22 for Y1, 0.36 for Y2 and 0.0015 for Y4.
b
Values are represented as means of five tasting results with a mean standard deviation of 0.19.
 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225
petroleum ether, which gradually dissolved the fat at 130 °C for
16 h. The lipid extract and solvent were then dried in an oven at
95 °C for a sufficient time to evaporate the remaining solvent.
The oil content was calculated in terms of the percentage of dry
basis.
2.6.2. Defatting ratio
The defatting ratio is the oil extraction yield, which is the ratio
of extracted oil to the total oil that is initially present in peanut
seeds before pressing. The mass of extracted oil is the difference
between the initial oil content and the residual oil in the cake,
which was determined by Soxhlet extraction (as shown in
Section 2.6.1).
2.6.3. Deformation ratio
The deformation ratio is the percentage of broken or irre-
versibly deformed seeds. After pressing, reconstitution and drying,
each batch was weighed and the total mass was noted (Wt in g).
Then, all of the seeds with an unrecoverable deformation were
set aside and weighed (Wd in g). The deformation ratio was calcu-
lated as follows (Eq. (3)):
Wd
Deformation ratio ð%Þ ¼  100 ð3Þ
Wt
2.6.4. Compression ratio
The method used for measuring the compression ratio was
identical to that described by Louka (1996). The main difference
is that the compression ratio is the opposite of the expansion ratio
– it reflects the decrease in volume of the product after pressing.
This was calculated from the ratio of the bulk density of the pro-
duct after defatting by pressing (qa) to the bulk density of the con-
trol product before pressing (qb), as shown in (Eq. (4)).
qa
Compression ratio ¼ ð4Þ
qb
The bulk density of the peanuts (q1) was calculated five times
using Eqs. (5) and (6), as reported by Louka and Allaf (2004), with
slight modifications:
M1 M1
q1 ¼ ¼ ð5Þ
Vs ðV t  V i Þ
M2 Mi
q2 ¼ ¼ ð6Þ
Vt Vi
where q1 (g/cm3) and q2 (g/cm3) are the peanuts and inert powder
bulk densities, respectively; M1 (g), M2 (g) and Mi (g) are the pea-
nuts mass, powder mass filling the total volume and powder mass
filling the interparticle space, respectively; and Vs (cm3), Vt (cm3)
and Vi (cm3) are the actual volume occupied by the sample, the vol-
ume of the whole cylinder and the interparticle space, respectively.
2.6.5. Moisture content
The moisture content was determined gravimetrically by mea-
suring weight loss on 2 g of sample by drying in an oven at 105 °C
until a stable weight was reached for a 60 min time frame (AOAC,
1990). The percentage of the dry basis (% d.b.) result was expressed
as the ratio of the water content in the product to the mass of the
same dry product.
2.7. Sensory analysis
Generally, consumers make their decisions when purchasing
and consuming certain foods by relying on many sensory specifica-
tions, including appearance, flavor, aroma, color, and texture.
1219
Therefore, our sensory panel consisted of five expert assessors from
the El-Kazzi industry who are specialized in the evaluation of the
snack products. Twenty peanut samples that were treated differ-
ently according to the experimental design were packed in blind
labeled tight transparent PVC bags and randomly subjected to sen-
sory evaluation by each panelist. A scale was adopted to evaluate
each characteristic, from 1 (very low/dislike) to 10 (very high/like).
The values were averaged to calculate the overall acceptability
because the 5 sensory properties, as previously mentioned, have
the same impact on the acceptability, and each quality attribute
was evaluated on a 10-point basis.
3. Results and discussion
3.1. Effect of different treatments on the response parameters
The defatting ratio, percentage of deformation, and acceptabil-
ity of partially defatted peanuts are highly influenced by the
applied treatment and its variant parameters. As is shown in
Table 2, the maximum defatting ratio (70.62%) was achieved by
pressing peanuts (5% d.b. water content) at 7 MPa for 20 min.
Under these pressing conditions, the percentage of deformed seeds
was low (3.28%) and the overall acceptability of the product was
found to be relatively low as well (6.44/10) during tasting.
Ammann (1935) reported a 60–70% oil yield when peanuts were
pressed at 34.3 MPa for 45–60 min, while an 80% oil removal over
10 min at 35.1 MPa was reported by Wong and Sackenheim (1992).
Compared to the literature, the largest advantage observed from
our experimental results is that high extraction yields are being
achieved at the lowest cost due to a relatively low applied pressure
and pressing duration.
3.2. Effects of independent variables on the response parameters
3.2.1. Defatting ratio
The water content (linear and quadratic effect) and pressure
(linear effect) appeared to have a significant influence (p-
value < 0.01%) on the defatting ratio (Fig. 2a). The high water con-
tent tended to reduce the extraction of oil, while the application of
an increased pressure led to an improved extraction yield (p-
value < 0.01%).
The defatting ratio of peanut seeds decreased linearly with an
increasing water content because the water decreased the internal
and intra-particle mobility of oil (Holloway & Wilkins, 1982). This
means that the energy required to initiate the movement of the oil
from the oleosomes to the grain surface became higher (Lanoisellé,
1995). A water content of 5% d.b. proved to be adequate for effi-
cient oil extraction. This rate increased linearly with increasing
pressure on the peanut seeds during mechanical pressing
(Fig. 2e). Generally, the higher the pressure, the higher the propor-
tion of extracted liquids. Previous studies have shown that an
increase in the oil extraction yield follows a pressure increase.
However, the performance tends be asymptotic. In the current
experimental design, linearity persisted in the range of pressures
applied to peanuts (4–10 MPa). Below these values, the plateau
was not reached. The pressing time had no significant influence
(p-value = 5.64%) on the defatting ratio, but once it was extended,
there was a slight increase in this rate.At the beginning of pressing,
the seeds were essentially rearranged and inter-particle air was
eliminated. The compression arose continuously (at this stage, oil
was removed through the pores by filtration under pressure) until
the peanut cake became uncompressible. In fact, because the
matrix was generally composed of seeds, air between particles
and oil, the process occurred in two phases (Lanoisellé, 1995):
1220 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225

Fig. 2. Standardized effects on defatting ratio, deformability and acceptability of defatted peanuts and trends of the estimated response surfaces.

(1) instantaneous compression, where the gases in the cake are cake. Therefore, there is a threshold pressure (oil-point pressure)
compressed, and (2) the consolidation phase, which begins with at which the extraction phase begins and the oil emerges from
the simultaneous removal of the gas and liquid content from the the seed kernels (Faborode & Favier, 1996). Extending the pressing
 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225
duration and increasing the pressure intensity increased and stabi-
lized the defatting ratio. Studies have shown a correlation between
the oil extraction yield and the pressing time, with a gradual slow-
ing of the flow at higher pressures for longer durations due to the
clogging of the pores that allowed this flow (Adeeko & Ajibola,
1990).
3.2.2. Unrecoverable deformation ratio
The unrecoverable deformation ratio seemed to be highly and
positively affected by the water content (linear and quadratic
effect with p-values of less than 0.01% and 1%, respectively) and
by the linear (positive) and quadratic (negative) effects of the pres-
sure (p-value < 0.01% and = 1.1%, respectively) (Fig. 2b).The per-
centage of the deformation of peanut seeds increased linearly
with an increase in the water content. According to the literature
(Holloway & Wilkins, 1982), there has been a need to raise the
water content of the seeds to more than 4% d.b. to avoid the irre-
versible deformation during the pressing phase. With reference
to the results of the current experimental design, it has been found
that this light moistening procedure led to an increase in com-
pressibility and elasticity. However, by exceeding a water content
of 8% d.b., the irreversible deformation rate increased due to a
reduction in the elasticity of the seeds, which was followed by a
modification of the cell structure – from an elastic to a more plastic
behavior. Thus, the seeds could no longer recover their initial form
when subjected to high pressures (Fig. 2f).
As shown by Vix et al. (1966), at water contents below 3% d.b.,
the product is brittle, friable and more likely to suffer physical
damage. However, at water contents greater than 8% d.b., the pro-
duct enters a viscoelastic state and induces an irreversible plastic
deformation. Furthermore, at a high moisture content, the seeds
tend to become ‘‘squashy” during pressing, the efficiency of oil
removal is decreased, the growth of molds during storage (prior
to pressing) is highly promoted, and the individual kernels are also
reduced in size (approximately one-fourth to one-half) after press-
ing. In conclusion, for a water content ranging between 3 and 8% d.
b., the product had an elastic behavior, which limited the percent-
age of deformation. Moreover, the increase in the pressure exerted
on the peanut seeds during pressing resulted in increased friction
forces between the seeds and the shear stresses applied subse-
quently contributed to an increase in the percentage of irreversible
deformation (Fig. 2f). Ultimately, the pressing duration had no sig-
nificant influence (p-value = 17.2%) on the deformation ratio.
The previously described results were confirmed by laboratory
observations, as is shown in Fig. 3. At low water contents and
low pressures (1 and a), the grains had very low fractures and
were effortlessly disintegrated and easily reconstituted afterwards
(Fig. 3.1, 3.3, 3.4 and 3.13). As shown in figures 3.2 (+1, 1, 1), 3.8
(0, 0, a), 3.9 (1, +1, +1), 3.11, 3.12, and 3.14 through 3.20, these
pressed seeds had a more flattened appearance with a moderate
unrecoverable deformation and a mostly reasonable disintegration
when either the pressure or the water content used were
increased. On the other hand, the most distorted and paste-like
aspect, as well as the highest percentage of cracks, is clearly shown
in figure 3.6, where all of the operating parameters were set at a
high value (+1, +1, +1). This undesirable grain appearance was also
obtained in runs 5 (+1, +1, 1) and 7 (+a, 0, 0), where high water
contents were used (Fig. 3.5 and 3.7).
3.2.3. Overall acceptability
The overall acceptability of the finished product is shown in
Fig. 2c and was significantly and negatively affected by the pres-
sure (p-value < 0.01%) and pressing duration (linear and quadratic
effects with p-values of 2.4% and 3.7%, respectively). Because the
oil contains flavor precursors, its removal resulted in a loss of the
organoleptic properties of the partially defatted peanuts. The pro-
1221
duct became more acceptable to the consumer when seeds with a
higher water content were used. In fact, the high water content
prevented oil from draining from the seeds, which contributed to
the preservation of the natural flavors of the seeds due to a low
oil extraction yield (Fig. 2g).
Fig. 3 also shows the progression of the appearance of the pea-
nuts at different stages of the process. After pressing for oil extrac-
tion, the raw peanuts (Fig. 3a) were flattened and lost their original
shape. Moreover, several shown cracks on the surfaces reduced
their overall acceptability (Fig. 3b). At the third stage, the reconsti-
tuted peanut seeds regained their volume and became slightly lar-
ger than the original ones. However, their color1 lightened due to
the loss of oil (Fig. 3c). Further drying and roasting resulted in the
disappearance of cracks, and a desirable balance between the calorie
content and final product aroma, color, taste and texture were
achieved (Fig. 3d).
3.2.4. Compression ratio
As expected, the compression ratio was significantly and posi-
tively affected by pressure (linear and quadratic effect) and the
pressing time (linear effect) (Fig. 2d). In fact, high pressures and
long pressing durations contributed to a more effective defatting.
The lipid cells were emptied of their contents and the seeds were
flattened (Figs. 2h and 3.6). A larger decrease in volume occurred
and the compression ratio was greater. Moreover, this ratio
decreased with an increasing water content. In fact, water inhib-
ited the elimination of oil during mechanical compression and
the seeds were slightly compacted (Fig. 2h).
3.3. Interaction between operating parameters
As per the Pareto charts and according to the response surfaces
shown in Fig. 2, only the significant interactions are explained. The
interaction between the water content and pressure positively
affected the rate of irreversible deformation. In fact, at low pres-
sures and with a gradual increase of the water content, the rate
of deformation increased slightly. However, when higher pressures
were applied, the rise of the deformation ratio rapidly amplified
due to the increased water content (Fig. 2f).
The acceptability parameter was especially affected by the
interaction between the water content and pressure. At low pres-
sures, the defatting ratio was lower and the product was more rec-
ommended by the customer compared to the peanuts subjected to
high pressures. At low pressures and an increasing water content,
the overall acceptability, as well as the taste, increased due to the
small percentage of oil removed when the seed moisture content
was increasing. However, by applying high pressures and while
increasing the water content, the acceptability decreased instead.
This acceptability reduction can be explained by the fact that high
levels of water contributed to the loss of organoleptic qualities dur-
ing the process. Flavors may have been lost by dissolution during
soaking (before pressing or during reconstitution) or by volatiliza-
tion during drying and roasting (Fig. 2g).
3.4. Model fitting
Based on the responses generated by the 20 experiments, a
model was fitted to give a suitable regression coefficient to the
effect of each linear and quadratic parameter, along with the effect
of the interaction between the operating parameters. In fact, the
developed model equations were able to effectively predict the
1
For interpretation of color in Fig. 3, the reader is referred to the web version of
this article.
1222 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225
Fig. 3. Grain appearance of the 20 experiments and after different stages of the defatting process: (a) raw peanuts, (b) after pressing, (c) after reconstitution, (d) after drying
and roasting.
peanut oil yield and quality at any given initial water content, pres-
sure and pressing duration.
The equations of the fitted models for each response parameter,
as well as the correlation coefficients (R2), are as follows:
Defatting Ratio ðDRÞ ¼ 123:895  20:37  W þ 0:34  P þ 0:31  T
þ 0:88  W ^ 2  0:01  W  P  0:05  W  T
 0:0003  P^ 2 þ 0:002  P  T
þ 0:007  T ^ 2 with R2 ¼ 98:38%;
Unrecoverable Deformation Ratio ðUDRÞ
¼ 10:2026  5:39  W þ 0:052  P þ 0:55  T þ 0:23  W ^ 2
þ 0:052  W  P þ 0:015  W  T  0:002  P^ 2
 0:0004  P  T  0:014  T ^ 2 with R2 ¼ 97:45%;
Acceptability ðACÞ ¼ 4:53243 þ 0:54  W þ 0:05  P þ 0:008  T
 0:025  W ^ 2  0:004  W  P þ 0:012  W  T
 0:00024  P^ 2  0:00056  P  T
 0:003  T ^ 2 with R2 ¼ 88:73%;
Compression Ratio ðCRÞ ¼ 1:04462  0:003  W  0:0004  P
þ 0:0005  T þ 0:0002  W ^ 2
 0:00007  W  P þ 0:00008  W  T
þ 0:000008  P^ 2 þ 0:00002  P  T
 0:00005  T ^ 2 with R2
¼ 82:33%:
For all of the studied response parameters, the coefficient of deter-
mination, R2, was high (>82%), which indicated that second order
models of multiple regressions could be used to explain the func-
tional relationship between process and response parameters.
3.5. Response surfaces
As shown in Fig. 2, the trends of the estimated response surfaces
varied according to each response parameter. The pressing dura-
tion was found to have an insignificant effect on the defatting ratio
and the percentage of unrecoverable deformations. Therefore, this
process parameter was fixed accordingly. The opposite effect was
noticed between the following response parameters: at a low ini-
tial water content and for a pressure ranging from 4 to 10 MPa, a
relatively high defatting ratio was observed while the deformation
 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225 1223

Fig. 4. Contours of estimated response surfaces, optimum values and contours plots generated from multiple response analysis.
1224 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225
ratio was low (0–12%), which is convenient in this case study. On
the other hand, as long as the pressure increased and the moisture
content was high, the oil yield slightly increased, whereas the
deformation ratio was highly intensified (Fig. 2e and f). As a result,
despite the applied pressure, it was clearly observed that a low
moisture content (<8% d.b.) is recommended to maximize the oil
yield and to lower as well the waste of the deformed product.
Due to its insignificant effect on the overall acceptability and
the compression ratio, the water content was fixed at the central
level. The overall acceptability of the partially defatted peanuts
was mostly reduced with the increase in pressure (Fig. 2g). How-
ever, it is highly recommended that the pressure remains below
9 MPa during mechanical oil expression, especially when the initial
moisture content used was approximately 8% d.b. In fact, the
decreasing trend of the response surface of the overall acceptabil-
ity was mainly due to 3 subsequent phenomena: (1) at high pres-
sures, the flavor precursors were lost with a high percentage of
extracted oil; (2) the high defatting ratios induced higher com-
pressibility ratios and therefore required a longer soaking time to
reconstitute the flattened seeds, thereby leading to a loss of flavors
by dissolution; and (3) after excessive soaking, a longer drying pro-
cess was required before roasting, during which flavors were
reduced by volatilization. The shape of the response surface in
(Fig. 2h) shows an obvious increase in the compression ratio while
increasing the pressure and pressing duration. This is due to the
high oil extraction yield, leading to an intensification of the com-
pressibility of the grains.
3.6. Multiple response optimization
The optimum values for each response were chosen differently.
To compare the variation of the response parameters based on
operating parameters, ‘‘Pressure” and ‘‘Water content” were cho-
sen to be variables, while the ‘‘Duration of pressing” was fixed at
its optimum value (Fig. 4a–d). The main purpose of this study
was to maximize the extraction yield (defatting ratio) and the com-
pression ratio, while minimizing the percentage of deformed seeds
and keeping the organoleptic properties satisfactory (highest over-
all acceptability). The different optimum values are indicated by a
cross in Fig. 4 and are displayed at the bottom left of each contours
of the estimated response surface plot. The maximum degree of
defatting and the maximum compression ratio were reached at a
water content of 5% d.b. by applying approximately 12 MPa of
pressure. On the other hand, an optimum rate of irreversible defor-
mation was seen at low water contents (approximately 5.4% d.b.)
while applying low pressures, approximately below 2.5 MPa
(2 MPa in average). Regarding the acceptability, the best values
were obtained by hydrating the peanut seeds at approximately
15% d.b. before pressing and by applying a low pressure of approx-
imately 2 MPa.
Because the optimum values contradicted each other, a com-
promise was found. To satisfy consumers who are aware of their
caloric intake, it was imperative to reasonably maximize the defat-
ting ratio and minimize the deformed products. Thus, the highest
impact factor (IF) was assigned to the defatting ratio. An average
IF factor was assigned to the deformation ratio and the overall
acceptability. In this case, the optimum results were generated
from a multiple response analysis and were aimed at maximizing
the desirability (80%). Applying a hydraulic pressure of 9.7 MPa
for 4 min on peanuts previously hydrated to 5% d.b. resulted in a
70% defatted product with less than 1% of deformed seeds. The
overall consumer acceptability was 7.6/10 (Fig. 4e). The application
of higher pressures in this defatting method guaranteed over 70%
oil extraction yields. However, this would cause higher percent-
ages of irreversible deformations and a loss of flavor, leading to
lower consumer acceptability. The current results are promising
for industrial applications because a high defatting ratio could be
reached with a minimum loss of deteriorated peanuts. Meanwhile,
the acceptability of defatted snacks could be easily recovered by a
simple addition of flavoring agents.
4. Conclusion
An efficient defatting process of peanut seeds has been opti-
mized using RSM. Many conventional methods of oil extraction
(a) cause oilseed damage (cracking, grinding, or flaking), (b) leave
solvent residues in the defatted peanuts and in the extracted oil,
and (c) require high energy consumption. This novel process has
solved all of the above drawbacks and preserves the structure
and the physicochemical properties of the defatted seeds, and it
also provides extracted oil for daily human consumption. It is the
simplest, most economical and most environmentally friendly
method among all of the other industrial applications.
Conflict of interest
The authors declare that there are no conflicts of interest.
Acknowledgements
The authors gratefully acknowledge the Lebanese company ‘‘El-
Kazzi” and the Council for Scientific Research of Saint Joseph
University, Lebanon (Project FS34) for providing financial support
to carry out this research work.
References
Abdullah, M., & Koc, A. B. (2013). Oil removal from waste coffee grounds using two-
phase solvent extraction enhanced with ultrasonication. Renewable Energy, 50,
965–970.
Adeeko, K. A., & Ajibola, O. O. (1990). Processing factors affecting yield and quality of
mechanically expressed groundnut oil. Journal of Agricultural Engineering
Research, 45, 31–43.
Ajibola, O. O., Eniyemo, S. E., Fasina, O. O., & Adeeko, K. A. (1990). Mechanical
expression of oil from melon seeds. Journal of Agricultural Engineering Research,
45, 45–53.
American Peanut Council (2014). <http://www.peanutsusa.com/> Accessed
10.02.14.
Ammann, P. (1935). Process of treating earth-nut kernels. U.S. Pat., no 2,003,415.
AOAC (1990). Official methods of analysis (15th ed., Vol. 2). Washington, DC:
Association of Official Analytical Chemists.
Campbell, K. A., & Glatz, C. E. (2009). Mechanisms of aqueous extraction of soybean
oil. Journal of Agriculture and Food Chemistry, 57, 10904–10912.
Domínguez, H., Núnẽz, M. J., & Lema, J. M. (1995). Aqueous processing of sunflower
kernels with enzymatic technology. Food Chemistry, 53(4), 427–434.
Evangelista, R. L. (2009). Oil extraction from lesquerella seeds by dry extrusion and
expelling. Industrial Crops and Products, 29(1), 189–196.
Faborode, M. O., & Favier, J. F. (1996). Identification and significance of the oil-point
in seed-oil expression. Journal of Agricultural Engineering Research, 65(4),
335–345.
Fraser, G. E. (1999). Nut consumption, lipids, and risk of a coronary event. Clinical
Cardiology, 22(S3), 11–15.
Gaur, R., Sharma, A., Khare, S. K., & Gupta, M. N. (2007). A novel process for
extraction of edible oils: Enzyme assisted three phase partitioning (EATPP).
Bioresource Technology, 98(3), 696–699.
Hammons, R. O. (1973). Early history and origin of the peanut. In C. T. Wilson (Ed.),
Peanuts—Culture and uses (pp. 17–45). Stillwater, Oklahoma: American Peanut
Research and Education Association, Inc..
Holloway, Jr., & Wilkins, H. (1982). Low-fat nuts with improved natural flavor. U.S.
Pat., no 4,329,375.
Holloway, O. E., Finley, J. W., & Wheeler, E. L. (1991). Low calorie nuts with the
organoleptic character of full fat nuts. Eur. Pat., no WO 91/10372.
Jesus, A. A., Almeida, L. C., Silva, E. A., Filho, L. C., Egues, S. M. S., Franceschi, E., ...
Dariva, C. (2013). Extraction of palm oil using propane, ethanol and its mixtures
as compressed solvent. The Journal of Supercritical Fluids, 81, 245–253.
Jiang, L., Hua, D., Wang, Z., & Xu, S. (2010). Aqueous enzymatic extraction of peanut
oil and protein hydrolysates. Food and Bioproducts Processing, 88(2–3), 233–238.
Jiao, J., Li, Z. G., Gai, Q. Y., Li, X. J., Wei, F. Y., Fu, Y. J., & Ma, W. (2014). Microwave-
assisted aqueous enzymatic extraction of oil from pumpkin seeds and
evaluation of its physicochemical properties, fatty acid compositions and
antioxidant activities. Food Chemistry, 147, 17–24.
 J. Nader et al. / Food Chemistry 197 (2016) 1215–1225
Kartika, I. A., Pontalier, P. Y., & Rigal, L. (2010). Twin-screw extruder for oil
processing of sunflower seeds: Thermo-mechanical pressing and solvent
extraction in a single step. Industrial Crops and Products, 32(3), 297–304.
Kris-Etherton, P. M., Pearson, T. A., Wan, Y., Hargrove, R. L., Moriarty, K., Fishell, V., &
Etherton, T. D. (1999). High-monounsaturated fatty acid diets lower both
plasma cholesterol and triacylglycerol concentrations. American Journal of
Clinical Nutrition, 70, 1009–1015.
Lanoisellé, J.-L. (1995). Contribution à l’étude du Pressage des Graines
Oléoprotéagineuses. PhD thesis submitted at Paris XII University Val-de-Marne,
France.
Louka, N. (1996). Maîtrise de la qualité des produits agro-alimentaires séchés;
modification texturale et réduction du coût énergétique par Détente Instantanée
Contrôlée ‘‘DIC” vers le vide. PhD thesis submitted at Université de Technologie
de Compiègne, France.
Louka, N., & Allaf, K. (2004). Expansion ratio and color improvement of dried
vegetables texturized by a new process ‘‘controlled sudden decompression to
the vacuum”: Application to potatoes, carrots and onions. Journal of Food
Engineering, 65, 233–243.
Mani, S., Jaya, S., & Vadivambal, R. (2007). Optimization of solvent extraction of
Moringa (Moringa oleifera) seed kernel oil using response surface methodology.
Food and Bioproducts Processing, 85(C4), 328–335.
Melgarejo Navarro Cerutti, M. L., Ulson de Souza, A. A. U., & de Arruda Guelli Ulson
de Souza, S. M. (2012). Solvent extraction of vegetable oils: Numerical and
experimental study. Food and Bioproducts Processing, 90(2), 199–204.
Mrad, R., Debs, E., Maroun, R., & Louka, N. (2014). Multiple optimization of chemical
components and texture of purple maize expanded by IVDV treatment using
the response surface methodology. Food Chemistry, 165, 60–69.
Mrema, G. C., & Mc Nulty, P. B. (1985). Mathematical model of mechanical oil
expression from oilseeds. Journal of Agricultural Engineering Research, 31,
361–370.
1225
NIOSH/OSHA/DOE Health Guidelines (1996). Occupational safety and health guideline
for n-hexane. Occupational Safety and Health Administration (OSHA), United
States Department of Labour, Washington DC, USA. Available at: <http://www.
osha.gov/SLTC/healthguidelines/n-hexane/recognition.html)> Accessed
12.07.09.
Passey, C. A., & Patil, N. D. (1994). Process for preparing low-calorie nuts. U.S. Pat., no
5,290,578.
Pérez-Gálvez, R., Chopin, C., Mastail, M., Ragon, J. Y., Guadix, A., & Bergé, J. P. (2009).
Optimisation of liquor yield during the hydraulic pressing of sardine (Sardina
pilchardus) discards. Journal of Food Engineering, 93, 66–71.
Qu, X. J., Fu, Y. J., Luo, M., Zhao, C. J., Zu, Y. G., Li, C. Y., ... Wei, Z. F. (2013). Acidic pH
based microwave-assisted aqueous extraction of seed oil from yellow horn
(Xanthoceras sorbifolia Bunge.). Industrial Crops and Products, 43, 420–426.
Russin, T. A., Boye, J. I., Arcand, Y., & Rajamohamed, S. H. (2011). Alternative
techniques for defatting soy: A practical review. Food and Bioprocess Technology,
4(2), 200–223.
Salgın, U., & Salgın, S. (2013). Effect of main process parameters on extraction of
pine kernel lipid using supercritical green solvents: Solubility models and lipid
profiles. The Journal of Supercritical Fluids, 73, 18–27.
Simelunas, W., Wilkins, H., & Gannis, P. (1985). Process for preparing low-fat nuts.
Eur. Pat., no 0,168,210 A2.
Venter, M. J., Willems, P., Kuipers, N. J. M., & de Haan, A. B. (2006). Gas assisted
mechanical expression of cocoa butter from cocoa nibs and edible oils from
oilseeds. The Journal of Supercritical Fluids, 37, 350–358.
Vix, H., Spadaro, J., & Pominski, J. (1966). Partially-defatted nut meats and process. U.
S. Pat., no 3,294,549.
Wilkins, H., & Gannis, P. (1984). Process for preparing low-fat nuts. Eur. Pat., no
0,139,459 A1.
Wong, V., & Sackenheim, R. (1992). Process of making low fat nuts. U.S. Pat., no
5,164,217.