Journal of Food Engineering 114 (2013) 228–234

Contents lists available at SciVerse ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

Optimization and modeling of extraction of solid coconut waste oil

Sarina Sulaiman a,b, A.R. Abdul Aziz a,⇑, Mohamed Kheireddine Aroua a
a
Department of Chemical Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
b
Department of Biotechnology Engineering, International Islamic University Malaysia, P.O. Box 10, Kuala Lumpur 50728, Malaysia

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

Article history: Solid coconut waste was produced after coconut milk extraction process and may still contain up to
Received 27 February 2012 24 wt.% oil content. In this work, extraction of oil from coconut waste using batch and soxhlet extractor
Received in revised form 22 August 2012 was studied. Effect of particle size diameter, type of solvent and solvent to solid ratio on the kinetic and
Accepted 23 August 2012
thermodynamic parameters; entropy, enthalpy and free energy of extraction were investigated. The max-
Available online 31 August 2012
imum oil yields for soxhlet and batch reactor were 23.6% at 80 °C and 21.9% at 65 °C, respectively for par-
ticle size diameter <0.5 mm when hexane was used as solvent. The kinetic of coconut waste oil extraction
Keywords:
was found to be a first order mass transfer model. The DG, DS and DH values were 10.94–13.35 kJ/mol,
Coconut waste
Optimization
33.10–39.57 J/mol K and 0.12–1.25 kJ/mol, respectively shows that the extraction process was spontane-
Mass transfer ous, irreversible and endothermic based on thermodynamic parameters.
Batch extraction Ó 2012 Elsevier Ltd. All rights reserved.
Soxhlet extraction

1. Introduction
Coconut milk is extracted and the flesh is grated before being
pressed to produce milk. The byproduct of this process is coconut
waste. In Malaysia, both coconut oil and milk industry produces
78,000 metric tons of solid coconut waste in 2010 and normally
used as fertilizer, biomass or to feed the animals or left to decay
on the fields (MARDI, 2010; Poarch, 2007; Vetayasuporn, 2007).
Furthermore, coconut is abundant in Malaysia and the area of
coconut plantation is around 115,000 hectares in 2010 (MARDI,
2010). Vetayasuporn, 2007 reported that coconut waste is widely
used as feed for milking cows, pigs and poultry (Epa, 2012; Vetaya-
suporn, 2007). Coconut waste is also used as coconut flour, bio-
mass; firewood or cooking fuel and this is not a good option as
at the world market, coconut oil is sold at a higher price (Guarte
et al., 1996; Poarch, 2007). The oil obtained from coconut waste
can be used as edible oil or for other purposes such as biodiesel
production.
Waste can be extracted either by mechanical pressing or with
solid liquid extraction using solvent (Amin et al., 2010; Sriti
et al., 2011). Extraction using solvent gives a higher yield and less
turbid oil (Amin et al., 2010; Liauw et al., 2008). Hexane and petro-
leum ether are used extensively as the solvent in the oil extraction
from salmon skin, jatropha seed and neem (Aryee and Simpson,
2009; Liauw et al., 2008; Sayyar et al., 2009).
There are many researchers who studied the kinetic model of
the oil extraction (Amarni and Kadi, 2010; Kumoro and Hassan,
2006; Liauw et al., 2008; Meziane et al., 2006; Meziane and Kadi,
⇑ Corresponding author. Tel.: +60 379675300.
E-mail address: azizraman@um.edu.my (A.R. Abdul Aziz).
2008; Perez et al., 2011; Sayyar et al., 2009; Topallar, 2000). Kinetic
study of solid liquid extraction depends on the nature of the oil and
solvent, temperature of the process, particle size of the meal, reac-
tion time and the ratio of the solid to the solvent (Sayyar et al.,
2009). Effect of solvent to solid ratio studies the maximum extract-
ability of oil. Increasing the ratio no longer changes the extraction
yield because particles have been almost extracted (Li et al., 2009).
Therefore, the objective of this paper was to study the oil
extraction process from solid coconut waste using batch and soxh-
let extractor. Effective solvent was chosen based on the yield, sol-
ubility of oil in solvent and density. This study also focuses on
determination of the oil yield, effects of particle size, reaction time
and temperature and solvent to solid. These optimum parameters
were applied to a kinetic model based on mass transfer coefficient.
Mass transfer within the particle was also calculated to determine
that in oil extraction there is no internal diffusion limitation. Ther-
modynamic parameters which include entropy, enthalpy and Gibb
Free Energy of extraction of oil were also determined.
2. Materials and methodology
Coconut waste was collected from a premise that produces
coconut milk at Petaling Jaya, Malaysia. Analytical Grade n-hexane
and petroleum ether were purchased from Merck (Malaysia),
respectively.
2.1. Extraction of oil
The extraction process was conducted using soxhlet and batch
extractors. The soxhlet and batch extractors’ setup includes
100 ml laboratory soxhlet and a 250 ml jacketed glass reactor. Mix-

0260-8774/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jfoodeng.2012.08.025
 S. Sulaiman et al. / Journal of Food Engineering 114 (2013) 228–234
ing and heating of the solvent was provided by a hot digital mag-
netic stirrer. Cold water circulation was controlled by water bath,
RC (LAUDA). Solid coconut waste was weighted and experiment
was conducted using batch and soxhlet extractor. Solid coconut
waste was grinded for 10 min using a coffee grinder and separated
with sieve plates.
Temperature for batch reactor was varied between 55 and
65 °C, which is below the boiling point of the hexane and for the
soxhlet extractor; the temperature range was between 70 and
80 °C, above the boiling point of the solvent. Hexane and petro-
leum ether were used as solvents for both of the setup to separate
the oil from the waste. The coconut waste was separated by sieving
into three types of particle size, 0.5 mm and below and 0.7 mm and
below and 1.2 mm and below to determine the optimum size and
to obtain the highest oil yield. Solvent to solid mass of coconut
waste were varied from 6:1 to 12:1. The extraction process was
conducted for 15, 30, 45 min and 1, 2, 3, 4 and 5 h. After the equi-
librium was reached, the solvent was recovered using a distillation
system. For batch extractor, the mixture was filtered using Buchi
filter attached to vacuum pump to separate the coconut waste
and solution which contains oil. Solvent was used to wash the meal
for few times to obtain all the oil. The oil yield (oil content) of the
coconut waste was calculated as below:
Oil collected ðgÞ
Oil yield ¼  100%
Mass of coconut waste ðgÞ
Solubility measurement of oil in hexane was conducted in a
closed jacketed reactor and the mixture was mixed continuously.
The reactor was maintained at different temperature, controlled
by a RC (LAUDA) water bath. After 30 min, an aliquot of superna-
tant solution was extracted into a sample bottle, weighed to obtain
the solution withdrawn. After distillation and evaporation of sol-
vent, the weight of oil was determined. Oil solubility in petroleum
ether and hexane was measured.
Density measurement was carried out using DMA 4500 density/
specific gravity meter (Anton Paar, Austria). The adjustment of the
density was measured three times using distilled water at temper-
atures 15, 20, 30 and 40 °C to obtain the mean values and the
uncertainty was ±0.00001 g cm3.
2.2. Kinetic model
Mass transfer kinetic model was used to represent the experi-
ment data because of there is no reaction between the oil and
the solvents, hexane and petroleum ether (Liauw et al., 2008).
Assumption is made that the main mechanism which controls
the rate of extraction of oil is mass transfer of oil from the waste
(solid) to the solvent (liquid). This assumption is in accordance
with other findings (Amin et al., 2010; Saxena et al., 2011). Mass
transfer rate can be written as (Liauw et al., 2008):
dW A
¼ k  AðC Ai  C A Þ
dt
where dWA/dt, mass transfer rate of the coconut waste oil (g/s); CA
and CAi, concentration of coconut waste oil in liquid (solvents) at
time t (g/m3) and at equilibrium (g/m3), respectively; k, mass trans-
fer coefficient, ms1; A, surface area for mass transfer process, m2.
Since the extraction was conducted in a batch process and the
volume was constant throughout the experiment, Eq. (2) can be
written as:
dW A A Y U mS
¼ k  ½W Ai  W A 
dt V
dW A
¼ k  a  ½W Ai  W A 
dt
229
where, ka, is volumetric mass transfer coefficient. To solve Eq. (4)
by integration, following condition was used where the mass of
coconut waste oil is zero in liquid (WA) at the beginning of the
extraction process. Considering this condition, integration of Eq.
(4) resulted as:
W A ¼ W Ai ½1  expðkatÞ  ð5Þ
Rearranging Eq. (5) in terms of yield per mass of coconut waste,
the kinetic model used in this study was:
Y A ¼ Y Ai ½1  expðkatÞ  ð6Þ
where, YA and YAi is yield of coconut waste oil in liquid at time, t, s
and ka-volumetric mass transfer coefficient.
To determine the value of ka (s1), YA and YAi a nonlinear least
square method was used to calculate numerically. ‘‘Origin 8.5’’ Pro-
gram was used to fit the data in order to obtain the mass transfer
value and the yield of oil.
2.3. Mass transfer within the particle
The study of mass transfer within the particle was conducted to
determine that extraction of oil using soxhlet extractor or batch
reactor was not controlled or limited by the internal diffusion. Thi-
ele modulus was used to investigate the mass transfer within the
ð1Þ
particle. To determine the value of Thiele modulus, effective diffu-
sivity (Deff), m2 s1 was calculated. Fick’s second law was used to
determine effective diffusivity by assuming Deff is constant with
the Y, yield at time, t and initial yield of the oil. Pinelo et al.
(2006) calculated effective diffusivity using following equation
(Pinelo et al., 2006):
 
6 pDeff
ln Y ¼ ln  t ð7Þ
p2 r2
where, r is the radius of particles diameter (mm).
By using Eq. (4), a plot ln Y versus time of experiment was plot-
ted. The value of the slope was used to determine the effective
diffusivity.
To determine the effect of mass transfer within the particle on
the extraction of the oil, Thiele Modulus, / was calculated based
on Giri and Sharma, 2000 equation (Giri and Sharma, 2000):
 0:5
dp K qp
/¼ ð8Þ
6 Deff
The value dp represents the particle diameter, K is the extraction
rate, s1, qp is the density of the coconut waste. The system is as-
sumed to have no internal mass transfer limitation if the Thiele
Modulus is <2, and the system suffers from the internal mass trans-
fer limitation if it is above 10 (Giri and Sharma, 2000).
2.4. Thermodynamic parameters
ð2Þ
Thermodynamic parameters (DH, DS and DG) for the oil extrac-
tion of coconut waste using hexane and petroleum ether were esti-
mated using following equation (Liauw et al., 2008):
DG 1 DH 1 DS
ln K ¼  ¼ þ ð9Þ
R T R T R
Y T mL
K¼ ¼ ð10Þ
ð3Þ
where K is equilibrium constant, YT is the yield percent of oil at tem-
perature T, YU is the percent unextracted oil in coconut waste, mL is
ð4Þ the amount of coconut waste oil in liquid at equilibrium tempera-
ture, mS is the amount of coconut waste oil in solid at T, °C equilib-
230 S. Sulaiman et al. / Journal of Food Engineering 114 (2013) 228–234

rium temperature, R, is a gas constant, DH, (kJ/mol) is enthalpy, DS, 0.93

(J/mol K) is entropy and DG, (kJ/mol) is a free energy of extraction.

Density,g/cm3
The plot of ln K against 1/T is used to find the value of DH and
0.92
DS from the slope and intercept, respectively.

3. Results and discussion 0.91

3.1. Characterization of coconut waste oil

0.9
0 10 20 30 40 50
Hexane was used as co-solvent throughout the oil extraction
Temperature,°C
process. Hexane improves the mass transfer and diffusivity of oil
Hexane Petroleum ether
with the seed if compared with other solvent such as methanol.
Coconut waste oil is mainly composed of saturated short length Fig. 2. Density data of solid coconut waste oil in hexane and petroleum ether.
fatty acid chains, 2.3% of C8:0, 30.7% of C12:0, 17.91% of C14:0,
17.69% of C16:0, 22.29% of C18:0 and unsaturated fatty acids,
6.10% and 3.01% of C18:1 and C18:2, respectively. The highest com- (Stanisavljević et al., 2007). This shows that oil extracted using
position of methyl ester in the coconut waste oil is lauric acid and hexane has a higher density if compared with petroleum ether.
the acid value and density were 2.3226 mg/g KOH and 0.925 g/cm3,
respectively. 3.3. Effect of solvent

3.2. Solubility and density
Fig. 1 shows the solubility data of hexane and petroleum ether
in solid coconut waste oil. Vegetable oil, hexane and petroleum
ether were non polar and becomes easily miscible. Coconut oil
was most soluble in the non polar solvent because contain high
amount of lauric acid (Kanth Rao and Arnold, 1957). The composi-
tion of lauric acid in this study was 30.70%. Solid coconut waste oil
was partially miscible in both the solvents, hexane and petroleum
ether at temperature of 30 °C. Solubility increases about 10 to
30 w/w% for hexane and petroleum ether when the temperature
of the oil increases (Bera et al., 2006; Franco et al., 2007; Wakelyn,
1997). The solubility of oil in hexane was high because of the
strong solute–solvent interactions (Wakelyn, 1997).
Fig. 2 shows the measurement of density data as a function of
temperature of coconut waste oil. The sample was measured three
times and the uncertainty was ±0.00001 g cm3. The uncertainty of
oil yield was ±0.01 wt.%. For coconut waste oil extracted from hex-
ane, the density decreases from 0.925 g/cm3 to 0.907 g/cm3 when
the temperature increased from 15 to 40 °C. It is reported that
the densities of vegetable oil decrease linearly when temperature
increases because the molecules move apart as their kinetic energy
increases (Rodenbush et al., 1999; Veny et al., 2009). This causes
the volume to increase and the oil become less dense. The density
of oil extracted from both the solvent, hexane and petroleum ether
were almost similar with difference of 0.006 g/cm3 at temperature
15 °C. Stanisavljević et al. (2007) reported that density of tobacco
oil extracted using hexane and petroleum ether were 0.923 g/cm3
and 0.9175 g/cm3 respectively with 0.0055 g/cm3 difference
Fig. 3 illustrates the oil extraction yield of hexane and petro-
leum ether at 80 °C and particle size of 0.5 mm. The oil yield using
soxhlet extractor and hexane showed 1.3 % of oil is extracted if
compared with petroleum ether under same conditions. Sayyar
et al., 2009 reported that extraction of jatropha seed using n-hex-
ane showed that the extraction yield was 1.3% more than petro-
leum ether under the same condition at (temperature of boiling
point, 6:1 solvent to solid ratio) (Sayyar et al., 2009). However
using batch extractor as shown in Fig. 3, the extraction yield using
hexane was 0.9 % more than petroleum ether. Stanisavljević et al.
(2007) also reported that the oil yield from the native Tobacco
seeds were 3.1% and 2.9% by using hexane and petroleum ether,
respectively. It shows hexane extracted 0.2% of oil more than
petroleum ether. Mani et al. (2007) also reported that hexane ex-
tracted 1.3% of oil more than petroleum ether (Bassim et al.,
2003; Kushwaha, 2010; Lawson et al., 2010; Mani et al., 2007;
Nwabueze & Okocha, 2008; Zaher et al., 2002). Hexane is a non po-
lar solvent and able to penetrate into matrix of a seed during
extraction process. This is because they lack of O–H end which if
not would interfere with the extraction process (Nwabueze & Oko-
cha, 2008).
3.4. Effect of temperature and reaction time
Fig. 4a and b shows the effect of temperature and reaction time
on extraction of oil. The oil yield increases from 19.7% to 21.9% and
21.3% to 23.6% at temperature 70 °C and 80 °C for batch extraction
and soxhlet extraction respectively. The increase in temperature
increases the oil yield during extraction process. This is because

25

100
20
Solubility, w/w %

Yield, %

80 15

60 10

40 5

20 0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
0 Time (hour)
20 30 40 50 60 70 80 Hexane-Soxhlet Extractor Petroleum ether-Soxhlet Extractor
Temperature, °C Hexane-Batch reactor Petroleum ether-Batch reactor
Hexane Petroleum
Fig. 3. Effect of solvent on oil extraction process at 80 °C and particle size of
Fig. 1. Solubility of solid coconut waste oil in hexane and petroleum ether. 0.5 mm.
 S. Sulaiman et al. / Journal of Food Engineering 114 (2013) 228–234
a
30
25
20
Oil yield %
15
10
5
0
0 1 2 3 4 5 6
Time (Hour)
b 25
20
Oil yield %
15
10
5 increases with the decrease in the size of particles during oil
0 Fig. 5 also shows that there was an increase of 0.9% in extraction
0 1 2 3 4 5 6 of oil using batch extractor with hexane and smaller particle size.
Time (Hour)
Fig. 4. Effect of temperature and reaction time on extraction process using (a)
soxhlet extractor and (b) batch extractor.
the rise in temperature increases the diffusion of the oil while
decreasing the viscosity (Eikani et al., 2012; Meziane and Kadi,
2008). The mass transfer coefficient of the extraction process also
increases with temperature thus effecting the diffusion. Sayyar
et al., 2009 reported that using hexane as the solvent, the amount
of oil extracted increased around 1.5% by increasing the extraction
temperature from room temperature to 45 and 45–60 °C (Sayyar
et al., 2009).
The oil yield was slightly higher, 1.7% in the soxhlet extractor if
compared to the batch reactor. This is because soxhlet extractor
introduces continuous extraction whilst batch extractor needs
additional step, washing of the meals after extraction process to in-
crease the oil yield.
Fig. 4a and b also describes that the soxhlet extraction process
reaches optimum at 2 h, with oil content 19% before it reaches to
23.6% at 80 °C. The extraction rate was rapid, 16 % at the beginning
of the process and slows gradually because when the waste was
exposed to the fresh solvent, the free oil on the surface of the waste
was solubilized and oil gets extracted quickly. This causes fast in-
crease in extraction rate. At the initial extraction rate, the oil con-
centration was low in the solvent and mass transfer effect causes
the oil to diffuse quickly from the waste to the solvent. When
the maximum amount of extractable oil was reached, the oil yield
remains the same even after extending the extraction time.
3.5. Effect of particle size
The rate of extraction increases with the decrease in the size of
particle (Sirisompong et al., 2011). Fig. 5 shows the effect of parti-
cle size on oil extraction using soxhlet and batch extractors. Differ-
ent particle size of coconut waste namely 0.5 mm and below; and
0.7 mm and below; and 1.2 mm and below were used to extract
maximum amount of oil. From Fig. 5, oil yield using hexane and
soxhlet extractor was 23.6%, 22.7% and 22.5% with particle size
diameter of 0.5 mm, 0.7 mm and 1.2 mm and below, respectively.
This shows that smaller particle size extracted more than 1.1% of
oil if compared with the larger particles. The extraction yield for
231
25
Oil Yield, %
23
70°C
75°C
80°C 21
19
0.5 0.7 1.2
Particle size, mm
Soxhlet extractor Batch reactor
Fig. 5. Effect of particle size, 0.5 mm, 0.7 mm and 1.2 mm on extraction process
using soxhlet extractor and batch extractor.
55°C 0.7 and 1.2 mm size particles were relatively close to each other.
60°C These findings are in line with those found by Sirisompong et al.
65°C (2011). Sirisompong et al. (2011) found that the rate of extraction
extraction from rambutan kernel.
More oil was extracted from smaller particle size due to the bigger
interfacial area of the solid. The shorter distance the solvent has to
travel to extract the oil from the solid increases the pore diffusion
between solid and solvent. The larger particle has a smaller contact
surface area and is more resistant to solvent entrance and oil diffu-
sion. Smaller amount of oil will be transferred from inside the lar-
ger particle to the surrounding solution (Sayyar et al., 2009).
Sayyar et al. (2009) used three different particle sizes of jatropha
seeds, 0.5 mm and below, 0.5–0.75 mm and 0.75 mm and above.
It was concluded that the intermediate size particle, 0.5 to
0.75 mm produced the highest oil yield, 47.3% using hexane. This
is because when particle was too small, below 0.5 mm, the agglom-
erations of the fine particles reduces the effective surface area
available for the free flow of solvent to solid and prevent the inter-
action between solid and solvent (Sayyar et al., 2009).
3.6. Kinetic of oil extraction
Tables 1 and 2 shows the calculated values of the mass transfer
coefficients using Eq. (3) and the yield of oil, for both soxhlet and
batch extractor respectively at equilibrium at various tempera-
tures, particle size diameter, type of solvent and solvent to solid ra-
tio. Parameters of kinetic model, YAi and ka were estimated by non
linear square fit from equation to experimental data using Origin-
Lab 8.5. It is found that the mass transfer coefficients increase with
the increase in the temperature of extraction. The ka (s1) value
varies from 0.89 to 1.43 for soxhlet extractor and from 1.23 to
2.06 for batch extraction. The plotted graph shows a linear rela-
tionship with R2 values were above 0.9879 for soxhlet and batch
extraction. Increasing the temperature causes the reaction time
to be reduced as reaction occurs faster. The final concentration in-
creases with temperature because of the effect of thermodynamic
on the solubilization of oil inside the solid (Liauw et al., 2008). The
ka also increases around 23% when hexane was used when com-
pared with petroleum ether and extraction occurs significantly.
This proves that higher yield is achieved using hexane in the soxh-
let extractor. This is in agreement with Giri and Sharma, 2000 that
particle size and extracting and solubilising power of the solvent
affects the mass transfer (Giri and Sharma, 2000). Table 3 shows
the comparison value of mass transfer coefficient with previous
work. In this study, the mass transfer coefficient obtained was
0.386  103 s1 using particle size of 0.5 mm. Liauw et al.
232 S. Sulaiman et al. / Journal of Food Engineering 114 (2013) 228–234

Table 1
Parameters fitted for soxhlet extraction of coconut oil.

Hexane Petroleum ether

T 0.7 mm 0.5 mm 1.2 mm T 0.7 mm 0.5 mm 1.2 mm
(K) (K)
Yai, ka, R2 Yai, ka, R2 Yai, ka, R2 Yai, ka, R2 Yai, ka, R2 Yai, ka, R2
wt.% s1 wt.% s1 wt.% s1 wt.% s1 wt.% s1 wt.% s1
10:1 Solvent to solid ratio
343 19.45 1.17 0.9988 21.58 1.12 0.9982 19.43 1.23 0.9977 343 18.23 1.16 0.9986 19.9 1.11 0.9986 18.17 1.15 0.9983
348 21.73 1.25 0.9937 22.71 1.34 0.9938 21.63 1.24 0.9952 348 19.65 1.2 0.9972 20.58 1.26 0.9945 19.66 1.18 0.9971
353 22.8 1.39 0.9939 23.82 1.38 0.9931 22.49 1.39 0.9926 353 21.06 1.22 0.9964 21.33 1.43 0.9945 20.92 1.23 0.9962
12:1 Solvent to solid ratio
343 19.41 1.22 0.9978 21.61 1.1 0.9982 19.24 1.22 0.9978 343 18.29 1.15 0.9985 20.18 1.09 0.9988 18.17 1.14 0.9981
348 21.38 1.25 0.9935 22.58 1.33 0.9942 21.22 1.24 0.9951 348 19.57 1.23 0.9966 20.62 1.26 0.9945 19.59 1.18 0.9971
353 22.34 1.39 0.9921 23.35 1.41 0.9923 22.21 1.4 0.9924 353 21.08 1.23 0.9965 21.33 1.43 0.9946 20.87 1.22 0.9961
6:1 Solvent to solid ratio
343 20.3 0.92 0.9974 21.31 1.09 0.9975 20.3 0.92 0.9974 343 15.82 0.91 0.995 17.85 1.1 0.9939 15.76 0.91 0.995
348 20.89 1.19 0.9956 22.27 1.22 0.9925 20.89 1.19 0.9956 348 17.92 0.92 0.9971 19.53 1.33 0.9927 17.93 0.89 0.9968
353 21.4 1.4 0.9918 22.74 1.42 0.9934 21.4 1.4 0.9918 353 19.68 1.09 0.9948 20.5 1.36 0.9949 19.68 1.07 0.9951

Table 2
Parameters fitted for batch extraction of coconut oil.

Hexane Petroleum ether

T 0.7 mm 0.5 mm 1.2 mm T 0.7 mm 0.5 mm 1.2 mm
(K) (K)
Yai, ka, R2 Yai, ka, R2 Yai, ka, R2 Yai, ka, R2 Yai, ka, R2 Yai, ka, R2
wt.% s1 wt.% s1 wt.% s1 wt.% s1 wt.% s1 wt.% s1
10:1 Solvent to solid ratio
343 19.64 1.34 0.9927 19.96 1.37 0.9924 19.69 1.31 0.9906 343 18.2 1.26 0.9924 18.68 1.34 0.9905 18.09 1.24 0.9933
348 20.6 1.37 0.9936 21.06 1.37 0.9959 20.55 1.36 0.9931 348 18.81 1.36 0.9924 19.58 1.38 0.9915 18.79 1.34 0.992
353 21.16 1.41 0.9926 22.02 1.38 0.9947 21.16 1.39 0.9931 353 19.25 1.39 0.9919 20.25 1.42 0.9909 19.23 1.37 0.9919
12:1 Solvent to solid ratio
343 19.61 1.34 0.993 19.99 1.37 0.9926 19.81 1.28 0.9884 343 18.24 1.26 0.9918 18.71 1.35 0.9907 17.83 1.23 0.9919
348 20.59 1.38 0.9931 21.11 1.37 0.9957 20.6 1.37 0.9927 348 18.69 1.37 0.9935 19.56 1.38 0.9919 18.63 1.32 0.9912
353 21.16 1.41 0.9926 22.02 1.39 0.9946 21.16 1.4 0.993 353 19.28 1.4 0.9926 20.3 1.42 0.9908 19.36 1.36 0.991
6:1 Solvent to solid ratio
343 18.23 1.4 0.9968 18.76 1.41 0.9977 18.25 1.4 0.9969 343 14.99 1.46 0.9952 15.71 1.45 0.9959 14.51 1.51 0.9879
348 18.55 1.59 0.999 19.09 1.67 0.9987 18.57 1.59 0.999 348 15.69 1.62 0.9955 16.32 1.64 0.9934 15.62 1.61 0.9953
353 18.87 2.05 0.9965 19.34 2.44 0.9975 18.87 2.06 0.9966 353 16.15 1.77 0.9948 16.97 1.75 0.9914 16.07 1.75 0.9952

Table 3
Comparison of mass transfer coefficient with previous work.

References This work Liauw (2008) Topallar (2000) Amin et al. (2010)
Raw material Solid coconut waste Neem Sunflower seed Jatropha
K, s1 0.386  103 7.29  103 0.107  103 0.13  103

(2008), Topallar (2000) and Amin et al. (2010) reported that the
particle size of neem, sunflower and jatropha seeds were 0.425–
0.71 mm, 0.125 mm and 0.425 mm, respectively. The mass transfer
coefficient obtained for neem, sunflower and jatropha seeds were
7.29  103 s1, 0.107  103 s1 and 0.13  103 s1, respectively.
These shows that characteristic of solid matrix were almost similar
and the value of mass transfer coefficient lies within the reported
value from 0.107  103 to 7.29  103 s1.
3.6.1. Mass transfer within the particle
Effect of mass transfer within the particle was studied to deter-
mine if the diffusion to the surface controls and solvent diffuses
well into the particles. The following data was used in Eqs. (7
and 8) in order to calculate the Thiele Modulus, density of coconut
waste: 2.413 g/cm3, effective diffusivity (Deff) = 7.26  106 cm2 -
s1 and average value of particle diameter, dp = 0.5 mm.
Table 4 represents the Thiele Modulus calculate for the soxhlet
and batch reactors at 10:1 hexane to solid ratio at temperature 70–
80 °C and particle size of 1.2, 0.7 and 0.5 mm. Thiele Modulus
ranges between 0.1645 to 0.4339, 0.2315 to 0.3211 and 0.3025
to 0.4310 when particle size of 0.5 mm, 0.7 mm and 1.2 mm were
used respectively. The results shows that the Thiele modulus value
decreases when smaller size particles were used (Scott, 2006).
Small value of Thiele Modulus indicate surface reaction controls
and the solvent diffuses well without reacting (Scott, 2006). The
reaction was surface reaction limited and decreases the internal
mass transfer diffusion limitation (Giri and Sharma, 2000). The
batch reactor shows a higher value of Thiele modulus if compared
to the soxhlet extractor. However, overall both the system was not
affected by the mass transfer within the particle. This was due to
the small value of the Thiele modulus, which was <2. Giri and Shar-
ma (2000) reported that the Thiele Modulus values of coal were
0.1057 and 0.016 for particle size of 0.1278 mm and 0.016 cm,
respectively. This shows that the process was not affected by the
internal mass transfer limitation.
3.7. Thermodynamic parameters
Tables 5 and 6 show the values of equilibrium constant and
thermodynamic parameters for the oil extraction process of coco-
nut waste. Fig. 6 shows the plot of ln K vs. 1/T which was used to
determine the value of thermodynamic parameters. The values of
 S. Sulaiman et al. / Journal of Food Engineering 114 (2013) 228–234 233

Table 4
Thiele modulus on the internal mass transfer.

Particle size diameter /2 Thiele modulas

0.5 mm 0.7 mm 1.2 mm
Temperature 70 °C 75 °C 80 °C 70 °C 75 °C 80 °C 70 °C 75 °C 80 °C
Soxhlet extractor 0.1645 0.1850 0.1954 0.2424 0.2643 0.2681 0.4310 0.3321 0.3110
Batch reactor 0.4339 0.3260 0.3286 0.3211 0.2315 0.2333 0.4065 0.2978 0.3025

Table 5
Thermodynamic parameters for soxhlet extraction of coconut oil.

Soxhlet extractor
T (K) 1.2 mm 0.7 mm 0.5 mm
K DH, kJ/mol DS, J/mol K DG, kJ/mol K DH, kJ/mol DS, J/mol K DG, kJ/mol K DH, kJ/mol DS, J/mol K DG, kJ/mol
Hexane – 10:1 solvent to solid ratio
343 1.48 10.94 35.1 1.1 1.21 11.29 34.39 0.51 1.34 11.43 35.62 0.79
348 1.59 1.28 1.29 0.68 1.4 0.97
353 1.65 1.45 1.35 0.85 1.5 1.15
Petroleum ether – 10:1 solvent to solid ratio
343 1.26 11.19 33.31 0.24 1.19 11.73 35.61 0.48 1.33 12.37 38.34 0.78
348 1.33 0.4 1.3 0.66 1.41 0.97
353 1.43 0.57 1.33 0.84 1.51 1.16

Table 6
Thermodynamic parameters for batch extraction of coconut oil.

Batch extractor
T (K) 1.2 mm 0.7 mm 0.5 mm
K DH, kJ/mol DS, J/mol K DG, kJ/mol K DH, kJ/mol DS, J/mol K DG, kJ/mol K DH, kJ/mol DS, J/mol K DG, kJ/mol
Hexane – 10:1 solvent to solid ratio
343 1.19 12.16 36.73 0.44 1.18 12.82 38.59 0.41 1.06 13 38.24 0.12
348 1.24 0.62 1.24 0.61 1.13 0.31
353 1.34 0.81 1.34 0.8 1.2 0.5
Petroleum ether – 10:1 solvent to solid ratio
343 0.94 12.01 37.55 0.87 1.08 12.77 37.76 0.18 1.1 13.35 39.57 0.23
348 0.99 1.06 1.16 0.37 1.16 0.42
353 1.06 1.25 1.22 0.56 1.26 0.62

the enthalpy were in the range of 10.94–12.37 kJ/mol for soxhlet y = -3.0108x + 9.9644
1.60
R2 = 0.9999
extractor and 12.01–13.35 kJ/mol for batch extractor using hexane.
The value obtained was in the range (4–13.5 kJ/mol) obtained by 1.20
other researches for extraction of melon, rubber seed and olive
cake oil (Ibemesi and Attah, 1990; Meziane and Kadi, 2008). These
Ln K

0.80
values was also similar with Topallar (2000) finding which was
11.2 kJ/mol for oil extraction from sun flower seed. Positive enthal-
y = -8.0771x + 23.943
py change indicates the endothermic nature of the extraction pro- 0.40
R2 = 0.9999
cess and requires energy during the process with agreement with
previous studies (Amin et al., 2010; Topallar, 2000). 0.00
The process initially includes coconut waste in hexane. Entropy 2.82 2.84 2.86 2.88 2.90 2.92
1/T
of the mixture increases due to the oil molecules extraction. The Petroleum Ether: 6:1 Hexane 6:1
positive value of entropy change describes the process is irrevers- Linear (Petroleum Ether: 6:1) Linear (Hexane 6:1)
ible which is in agreement with other findings (Amin et al., 2010;
Fig. 6. Plot of ln K (equilibrium constant) vs. 1/T (temperature, °C) – particle size of
Meziane and Kadi, 2008; Topallar, 2000). The free energy and en-
0.7 mm.
tropy values lie between 0.12–1.16 kJ/mol and 33.10–38.24 J/
mol K, respectively for both of the extractor using hexane. Free en-
ergy and entropy of extraction using petroleum ether varies be-
tween 0.23–1.25 kJ/mol and 33.31–39.57 J/mol K for 1.2, 0.7 and nut waste using hexane and petroleum ether with R2 above 0.99.
0.5 mm of particle size respectively. This shows that the energy re- The rate of extraction increases by 1.9% for soxhlet extraction using
quired to break the solute–solute and solvent–solvent interactions hexane when the temperature (80 °C) and solvent to solid ratio
are less than the energy released in solute–solvent interactions. (10:1) was increased. Increasing the temperature from 70 to
80 °C; increases the yield by 2.3% for soxhlet extractor. Solubility
and density data showed hexane as a better solvent if compared
4. Conclusion
with the petroleum ether with more 1.3% oil yield for soxhlet
extractor. The mass transfer coefficients also increase with solvent
This study demonstrated that the mass transfer model showed
to solid ratio and temperature by 0.1% and 0.2%, respectively. Thi-
a good fit with the experimental data for the oil extraction of coco-
234 S. Sulaiman et al. / Journal of Food Engineering 114 (2013) 228–234
ele modulus value shows that the batch and soxhlet extraction
were not affected by the mass transfer within the particle. The
DG, DS and DH values were 10.94–13.35 kJ/mol, 33.10–39.57 J/
mol K and 0.12–1.25 kJ/mol, respectively proves that the extraction
process was spontaneous, irreversible and endothermic, respec-
tively. The maximum yields obtained from extraction of coconut
waste oil with particle size diameter of 0.5 mm were 23.6% at
80 °C for soxhlet extractor and 21.9% at 65 °C for batch extractor.
References
Amarni, F., Kadi, H., 2010. Kinetics study of microwave-assisted solvent extraction
of oil from olive cake using hexane: comparison with the conventional
extraction. Innovative Food Science & Emerging Technologies 11 (2), 322–327.
Amin, S., Hawash, G., El Diwani, El Rafei, S., 2010. Kinetics and thermodynamics of
oil extraction from jatropha curcas in aqueous acidic hexane solutions. Journal
of American Science 6 (11), 8.
Aryee, A.N.A., Simpson, B.K., 2009. Comparative studies on the yield and quality of
solvent-extracted oil from salmon skin. Journal of Food Engineering 92 (3), 353–
358.
Bassim, H.H., Ahmad, A.L., Ng, A.H., 2003. Removal of residual oil from palm oil mill
effluent using solvent extraction method. Jurnal Teknologi 38, 33–42.
Bera, D., Lahiri, D., De Leonardis, A., De, K., Nag, A., 2006. A novel azeotropic mixture
for solvent extraction of edible oils. Agricultural Engineering International: The
CIGR Ejournal VIII, 1–6.
Eikani, M.H., Golmohammad, F., Homami, S.S., 2012. Extraction of pomegranate
(Punica granatum L.) seed oil using superheated hexane. Food and Bioproducts
Processing 90 (1), 32–36.
Epa, 2012. How to Make Cookies and Macaroons from Coconut Residue (Sapal).
Franco, D., Sineiro, J., Pinelo, M., Núñez, M.J., 2007. Ethanolic extraction of Rosa
rubiginosa soluble substances: oil solubility equilibria and kinetic studies.
Journal of Food Engineering 79 (1), 150–157.
Giri, C.C., Sharma, D.K., 2000. Mass-transfer studies of solvent extraction of coals in
N-methyl-2-pyrrolidone. Fuel 79 (5), 577–585.
Guarte, R.C., Mühlbauer, W., Kellert, M., 1996. Drying characteristics of copra and
quality of copra and coconut oil. Postharvest Biology and Technology 9 (3), 361–
372.
Ibemesi, J., Attah, J., 1990. Temperature effects on the extraction of rubber and
melon seed oils. Journal of the American Oil Chemists’ Society 67 (7), 443–445.
Kanth Rao, R., Arnold, L., 1957. Alcoholic extraction of vegetable oils. Part IV.
Solubilities of vegetable oils in aqueous 2-propanol. Journal of the American Oil
Chemists’ Society 34 (8), 401–404.
Kumoro, A.C., Hassan, M., 2006. Modelling of andrographolide extraction from
Andrographis paniculata. Leaves in a Soxhlet Extractor, INRET 2006, Putrajaya,
Malaysia.
Kushwaha, P., 2010. Need to discontinue the use of solvents for the extraction of
edible oils for ensuring safety of public health. International Journal of
Pharmaceutical Research 2 (4), 7–14.
Lawson, O.S., Adefemi, O., Ologunagba, F.O., Ojomo, A.O., 2010. Evaluation of the
parameters affecting the solvent extraction of soybean oil. ARPN Journal of
Engineering and Applied Sciences 5 (10), 51–55.
Li, X.M., Tian, S.L., Pang, Z.C., Shi, J.Y., Feng, Z.S., Zhang, Y.M., 2009. Extraction of
Cuminum cyminum essential oil by combination technology of organic solvent
with low boiling point and steam distillation. Food Chemistry 115 (3), 1114–
1119.
Liauw, M.Y., Natan, F.A., Widiyanti, P., Ikasari, D., Indraswati, N., Soetaredjo, F.E.,
2008. Extraction of neem oil (Azadirachta indica A. Juss) using n-hexane and
ethanol: studies of oil quality, kinetic and thermodynamic. ARPN Journal of
Engineering and Applied Sciences 3 (No. 3), 6.
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 (4C), 328–335.
MARDI, (2010). Asian and Pacific Coconut Community.
Meziane, S., Kadi, H., Lamrous, A.O., 2006. Kinetic study of oil extraction from olive
foot cake. Grasas Y Aceites 57 (2), 175–179.
Meziane, S., Kadi, H., 2008. Kinetics and thermodynamics of oil extraction from olive
cake. Journal of the American Oil Chemists’ Society 85 (4), 391–396.
Nwabueze, T.U., Okocha, K.S., 2008. Extraction performances of polar and non-polar
solvents on the physical and chemical indices of African breadfruit (Treculia
africana) seed oil. African Journal of Food Science 2, 119–125.
Perez, E.E., Carelli, A.A., Crapiste, G.H., 2011. Temperature-dependent diffusion
coefficient of oil from different sunflower seeds during extraction with hexane.
Journal of Food Engineering 105 (1), 180–185.
Pinelo, M., Sineiro, J., Núñez, M.J., 2006. Mass transfer during continuous solid–
liquid extraction of antioxidants from grape byproducts. Journal of Food
Engineering 77 (1), 57–63.
Poarch, D., (2007). Coconut flesh or sapal uses as fuel and flour.
Rodenbush, C., Hsieh, F., Viswanath, D., 1999. Density and viscosity of vegetable oils.
Journal of the American Oil Chemists’ Society 76 (12), 1415–1419.
Saxena, D.K., Sharma, S.K., Sambi, a.S.S., 2011. Kinetics and thermodynamics of
cottonseed oil extraction. Grasas Y Aceites 62 (2), 198–205.
Sayyar, S., Abidin, Z.Z., Yunus, R., Muhammad, A., 2009. Extraction of oil from
jatropha seeds-optimization and kinetics. American Journal of Applied Sciences
6 (7), 1390–1395.
Scott, F.H., 2006. Elements of Chemical Reaction Engineering, forth ed. Prentice Hall
Professional Technical Reference.
Sirisompong, W., Jirapakkul, W., Klinkesorn, U., 2011. Response surface
optimization and characteristics of rambutan (Nephelium lappaceum L.) kernel
fat by hexane extraction. LWT – Food Science and Technology 44 (9), 1946–
1951.
Sriti, J., Talou, T., Faye, M., Vilarem, G., Marzouk, B., 2011. Oil extraction from
coriander fruits by extrusion and comparison with solvent extraction processes.
Industrial Crops and Products 33 (3), 659–664.
Stanisavljević, I.T., Lazić, M.L., Veljković, V.B., 2007. Ultrasonic extraction of oil from
tobacco (Nicotiana tabacum L.) seeds. Ultrasonics Sonochemistry 14 (5), 646–
652.
Topallar, Ü.G., 2000. Kinetics and thermodynamics of oil extraction from sun ower
seeds in the presence of aqueous acidic. Turkish Journal of Chemistry 24 (3),
247–254.
Veny, H., Baroutian, S., Aroua, M., Hasan, M., Raman, A., Sulaiman, N., 2009. Density
of Jatropha curcas seed oil and its methyl esters: measurement and estimations.
International Journal of Thermophysics 30 (2), 529–541.
Vetayasuporn, S., 2007. The feasibility of using coconut residue as a substrate for
oyster mushroom cultivation. Biotechnology 6, 5.
Wakelyn, P.J., (1977). Technology and Solvents for Extracting Oilseeds and
Nonpetroleum Oils. Amer Oil Chemists Society (November 1997).
Zaher, F.A., El Kinawy, O.S., Haron, a.D.E.E., 2002. Solvent extraction of jojoba oil
from pre-pressed jojoba meal. Grasas y Aceites 55 (2), 129–134.