Food Measure

DOI 10.1007/s11694-016-9365-5

ORIGINAL PAPER

Investigating shrinkage and moisture diffusivity of melon seed

in a microwave assisted thin layer fluidized bed dryer
Iman Goplour1 • Moein Zarrin Nejad2 • Reza Amiri Chayjan2 •
Ali Mohammad Nikbakht3 • Raquel P. F. Guiné4,5 • Majid Dowlati6

Received: 22 April 2016 / Accepted: 4 July 2016

 Springer Science+Business Media New York 2016

Abstract This work aimed to investigate the drying Keywords Melon seed  Microwave fluidized bed drying 
behavior of melon seed during combined fluidized bed- Moisture diffusivity  Shrinkage
microwave drying system. Three drying air temperatures
(40, 55 and 70 C), three microwave powers (270, 450 and
630 W) and three air velocities (0.8, 1.5 and 2.3 m/s) were Introduction
tested. Five mathematical models were selected to fit the
experimental data for drying kinetics, and the results Melon seed (Colocynthis citrullus L.) is a common Nigerian
revealed that the Aghbashloo et al. model exhibited, in all oil seed containing roughly 50 % by weight of oil, 28.4 %
cases, the best performance in fitting the experimental data protein (60 % of defatted flour), 2.7 % fiber, 3.6 % ash and
(R2 varying from 0.99088 to 0.99998; v2 from 0.00000 to 8.2 % carbohydrate [1]. The seed is an excellent source of
0.00185 and RMSE from 0.02289 to 0.82316). Calculated essential amino acids, particularly arginine, tryptophan and
values of moisture diffusivity for dried melon seed varied methionine, as well as vitamins B1 and B2, and niacin, and
from a minimum of 6.51 9 10-10 to a maximum of minerals, such as sulphur, calcium, magnesium, manganese,
6.59 9 10-9 m2/s under the tested drying conditions. potassium, phosphorus, iron and zinc [2]. The oil obtained
Moisture diffusivity values increased as air temperature from the seeds is utilized for edible purposes [1], while the
and microwave power was increased. Shrinkage values remaining cake is fried and used up as a snack. Containing
were calculated and found to vary in the range from 46.99 mainly linoleic (62.8 %) and oleic (159 %) unsaturated fatty
to 15.09 %. acids, the oil can be an available alternative to maize oil in
diets intended to decrease high levels of blood cholesterol.
Such perception of the product potential has led to an
increase in the production of these seeds.
One major problem that besets melon seeds is that they
& Iman Goplour
Imangolpour@gmail.com
deteriorate quickly during storage due to fungal infection
[3]. The effect of fungal attack on melon seeds includes
1
Young Researchers and Elite Club, Urmia Branch, Islamic decreased nutritive value, change in colour, increase in the
Azad University, Urmia, Iran peroxide value, reduced seed germination and mycotoxin
2
Department of Biosystems Engineering, Faculty of production [2]. The moisture content plays a vital role in
Agriculture, Bu-Ali Sina University, Hamedan, Iran the maintenance of seed shelf-life quality.
3
Department of Mechanics of Biosystems, Faculty of Dehydration refers to the moisture removal process, in
Agriculture, Urmia University, Urmia, Iran which simultaneous heat and mass transfer occur. Moisture
4
CI&DETS, Polytechnic Institute of Viseu, Viseu, Portugal transfer can happen in two forms: surface evaporation and
5
CERNAS, Polytechnic Institute of Coimbra, Coimbra, internal liquid vapor diffusion [4]. Drying is one of the
Portugal most common used methods for food preservation, being
6
Department of Mechanical Engineering of Biosystems, the basic goal when drying food and agricultural products
Faculty of Agriculture, University of Jiroft, Jiroft, Iran the elimination of moisture from the material down to a

123
 I. Goplour et al.

safe level to prevent deteriorative reactions and microbial An undesirable shape volumetric change associated
spoilage [5]. Different methods of drying could have dif- with moisture diffusion is known as shrinkage. In general,
ferent bearings on the effective moisture diffusivity, shrinkage takes place as a result of volume reduction due
shrinkage and other physical properties of the drying to evaporation of the moisture contained in the drying
material. One of the most popular methods of drying sample. The changes in dimensions of the solid could be
materials with high moisture content is fluidized bed dry- monitored in most cases. Shrinkage during drying is
ing. This method offers the advantages of good mixing, important not only from the viewpoint of product end-use
high heat and mass transfer coefficients and hence but also for processing simulation [17]. Shrinkage of
increased drying rates resulting in shorter drying time [6]. products such as broad beans [16] and papaya [18] has
However, all fluidized bed dryers, with or without the use been reported.
of inert particles, have a common limit that is independent of In order to design and evaluate modern applicable
the principal mode of heat transfer from the heating medium microwave food processes, mathematical modeling of
to the surface of the material, the transfer from the surface to drying behavior of food stuffs is vital. To predict the drying
the inside of the particle is principally by conduction and behavior of food and agricultural material, many thin-layer
thus the process is slow. This phenomenon causes long drying mathematical models have been developed that fall
drying time during the falling rate period in drying of food onto three main categories: theoretical (Fick’s second law
stuffs [7]. The alternative which has recently been widely of diffusion), semi theoretical (Lewis, Page, modified Page,
employed is the use of rapid transfer of electromagnetic Henderson-Pabis, logarithmic, Two-term, Two-term expo-
energy in the form of microwaves directly to the water nential, approximation of diffusion, Verma, etc.), and
molecules in the material in order to bypass the surface-to- empirical (Wang-Singh). Air temperature and material
center conduction stage. Because the waves can penetrate characteristic dimension are greatly affected by drying
directly into the material, heating is volumetric (from inside kinetics in convective drying. These models are dependent
out) and provides fast and uniform heating throughout the only on drying time and constants, while influence of all
product. The quick absorption of energy by water molecules other factors is negligible [16].
causes rapid evaporation of water, resulting in high drying Any attempt to characterize the drying behavior of the
rates of the food [8]. materials, must inevitably address the physical parameters
The application of microwaves solely can result in of the material such as shrinkage and moisture diffusivity.
uneven heating of certain products depending on their Due to insufficiency of information detailed on melon seed
dielectric and thermophysical properties [9–12]. This in the literature, this research was designed to investigate
problem is more significant when processing at low fre- the drying behavior, the shrinkage properties and moisture
quencies, in which case the dielectric properties are con- diffusivity of melon seed during drying in a fluidized bed
siderably dependent on temperature variations. To dryer combined with microwave at various levels of air
overcome some of the limitations of single microwave or velocity, power and temperature.
fluidized bed driers, one strategy is to combine microwave
with a fluidized bed dryer. Temperature uniformity of the
particles in the bed is provided by good mixing due to Materials and methods
fluidization [13], and drying time is reduced by the use of
microwave energy [8]. This method provides an effective Sample Preparation
way of overcoming the non-uniform heating problems in
conventional pure microwave heating. Some researchers Melon (Colocynthis citrullus L.) seeds used in the drying
have recently focused on the microwave drying kinetics of experiments were purchased from a local market in
foods, such as green pea [14] and chestnut corn [6]. Results Hamedan, Iran. Foreign matter, broken and immature seeds
of such studies elucidated that drying time under micro- were rejected and the good seeds for the work were man-
wave convection would be reduced intensively compared ually unshelled. The seeds were packed in double layered
to common convection methods. low density polyethylene bags and then stored in a refrig-
Effective moisture diffusivity of food and agricultural erator at low temperature (?4±1 C) to avoid the growth
products is strictly required for optimized design and of microorganisms and allowing to uniformity of moisture
construction of a dryer [15] and it represents the effect of distribution [19–21]. The samples were mixed and their
all input parameters on the mass transfer in drying process initial moisture content was determined by oven drying
[16]. Although much information has been reported on the method of whole kernels in triplicate at 103 C for 24 h
modeling of drying kinetics and effective moisture diffu- [22]. Three samples, each weighing 15 g, were placed in
sivity for various food and agricultural products, no this oven set. Initial moisture content of the melon seed
experimental report has been divulged for melon seed. was found to be 67.1 % db.

123
Investigating shrinkage and moisture diffusivity of melon seed in a microwave assisted thin…

The experimental facilities Experimental method and drying conditions

Drying experiments were conducted by means of a labo- A bulk of melon seeds (about 30 ± 1 g) was loaded in the
ratory fluidized bed dryer assisted by a modified domestic dryer chamber and drying experiment was started. The
microwave oven (900 W) as illustrated in Fig. 1. The dryer melon seeds were thoroughly mixed in the dryer system
was designed and fabricated in the department of agricul- and could be easily traced for any moisture content
tural machinery of Bu-Ali Sina University. This device was reductions during the drying processes. All experiments
designed to dry samples while the temperature and velocity were carried out in triplicates.
of air and microwave power were controlled. The drying Measurement of water loss from the samples was off-
chamber consisted of a Pyrex duct of 100 mm diameter and line. Sample weighting (in at most 10 s) was implemented
200 mm height, which was placed within the microwave using a digital balance (AND GF-6000, Japan, ±0.0001 g
oven. A cylindrical Pyrex column, 90 mm in diameter accuracy). The experiments were performed in combined
(100 mm outside diameter) and 280 mm height was uti- fluidized bed dryer at three air temperatures of 40, 55 and
lized as the fluidized bed drying chamber. The chamber 70 C, three air velocity levels (0.8, 1.5 and 2.3 m/s) and
was placed in a domestic microwave oven (Sharp R-I96T, three microwave power levels (270, 450, and 630 W).
Sharp Electronics, Bangkok, Thailand) with outside After turning on the dryer, 30 min was the approximate
dimensions of 540 9 330 9 450 mm. This oven was time required to achieve the steady state. Several trials of
equipped with three power level settings of low (270 W), the experiments were conducted to acquire data for mois-
medium (450 W) and high (630 W), and was calibrated ture contents versus time. At the end of each drying period,
before experimentation. At the first phase of tests, fan when the moisture content of the melon seeds reached the
speed was gradually increased using an inverter (Vincker equilibrium stage (no appreciable changes in three suc-
VSD2, ABB Co., Taipei, Taiwan) and air velocity was cessive sample weighing), the exit air temperature profile
recorded using a multifunction measurement device remained at a nearly constant level. Drying was continued
(Standard ST-8897, Standard instruments Co., Kowloon, until final moisture content was converged to about the
Hong Kong). It consists of a vane-type digital anemometer value of 5 % (db), which is acceptable equilibrium mois-
with ±0.1 m/s1 accuracy. Three air velocities of 0.8, 1.5 ture at room conditions.
and 2.3 m/s were applied in the experiments. An electrical
heating unit equipped with a thermostat (±1)was utilized to Modeling of the moisture ratio
maintain different levels of drying air temperature of 40, 55
and 70 C. A schematic diagram of employed apparatus is The following equation was used to calculate the moisture
shown in Fig. 1. ratio (MR) of melon seed during the drying process:

Fig. 1 Schematic diagram of

laboratory scale microwave
convection dryer: (1) fan and
electrical motor, (2) electrical
heater, (3) inverter and
thermostat, (4) drying chamber,
(5) thermometer, (6)
hygrometer, (7) precision
balance, (8) air velocity sensor,
(9) computer, (10) chassis

123
 I. Goplour et al.

M  Me Calculation of effective diffusivity

MR ¼ ð1Þ
Mb  Me
It has been accepted that the mass transfer of biological
where MR is the moisture ratio, M is the moisture content
products in the falling rate period can be described by using
at any time (% d.b.); Mb and Me are the initial and equi-
Fick’s second law of diffusion. Simplified equation of
librium moisture contents, respectively (% db).
moisture diffusivity provides an approximate method to
Based on the pretest elucidations, Me values were rela-
present a common quantitative comparison between dif-
tively negligible compared to M and Mb. Therefore, Eq. (1)
ferent products in the aspect of moisture transfer since it
was simplified as follows [16]:
can provide a description analysis for mean diffusion
M coefficient in the entire drying process [16]. To determine
MR ¼ ð2Þ
Mb the effective moisture diffusivity (Deff), was used Eq. (2)
(for an infinite slab), which can be defined as follows [26]:
!
Statistical modeling procedure 8X 1
1 ð2n þ 1Þ2 p2 tDeff
MR ¼ 2 exp  ð6Þ
p n¼0 ð2n þ 1Þ2 4L2
The non-linear regression analysis was performed using
Curve Expert software (Ver. 1.4, Hyams Development, Los where MR is the dimensionless moisture ratio; Deff is the
Angeles, CA) to determine the constants of the model. Five effective diffusivity (m2s-1); n is positive integer, t is
general models shown in Table 1 were fitted to the drying time, and L is the half thickness of the slab in
experimental data. Three criteria were adopted to evaluate samples (m). In practice, only the first term of the series in
the goodness of fit of each model. Correlation coefficient Eq. (6) is used, resulting in the following simplified form
(R2), Chi square (v2) and root mean square error (RMSE) of the equation:
were adopted to determine the quality of the model fitness    2 
8 p Deff t
[13]: lnðMRÞ ¼ exp  ð7Þ
p2 4L2
P
N
½MRexp;i  MRpre;i 2
R2 ¼ 1  i¼12 n 32 ð3Þ Calculation of activation energy
P
MRpre;i
P
N
6 k¼1 7
4 N  MRpre;i 5 Considering that the effective diffusivity varies with tem-
k¼1
perature according to an Arrhenius type function [27, 28]:
 
N 
P 2 Ea
0
MRexp;i  MRpre;i Deff ¼ De exp  ð8Þ
RT
v2 ¼ i¼1 ð4Þ
Nz where D0e is the diffusivity for an infinite temperature
" #12 (m2/s), Ea is the activation energy for moisture diffusion
1X N
RMSE ¼ ðMRpre;i  MRexp;i Þ2 ð5Þ (J/mol), R is the gas constant (R = 8.31451 J/(mol K)) and
N i¼1
T is the drying temperature (expressed in Kelvin) [27, 28],
where MRexp,i is the experimental moisture ratio of ith data, then:
 
MRpre,i is the predicted moisture ratio of ith data, N is the   Ea 1
number of observations and z is the number of drying lnðDe Þ ¼ ln D0e þ  ð9Þ
R T
constants.
Ultimately, the most suitable model to describe of dry- Plotting ln(De) versus (1/T) a straight line is also
ing characteristics of melon seed would be a model with obtained, with:
 
the highest R2 and the lowest v2 and RMSE values [13]. intercept ¼ ln D0e ð10Þ

Table 1 Mathematical models

No. Model name Model equation References
applied to drying curves
1 Aghbashloo et al. MR ¼ expðat=ð1 þ btÞÞ Aghbashloo et al. [23]
2 Midilli et al. MR ¼ a expðktn Þ þ bt Sacilik et al. [8]
3 Page MR ¼ expðktn Þ Amiri chayjan et al. [13]
4 Newton MR ¼ expðktÞ Henderson [24]
5 Logistic MR ¼ a=ð1 þ b expðktÞÞ Cihan et al. [25]

123
Investigating shrinkage and moisture diffusivity of melon seed in a microwave assisted thin…

Ea against drying time for all temperature levels, microwave

slope ¼  ð11Þ
R power levels and air velocities. The initial moisture content
which allows estimating the values of the parameters in of the melon seed was 67.1 % (db.) and they were dried to
Eq. (8): D0e and Ea. about 5 % (db.). As can be seen, with increasing in air
temperature (40–70 C), the drying time was diminished,
Calculation of shrinkage and at higher temperature, due to the quick removal of
moisture, the drying process was shorter. While increasing
Shrinkage is usually defined as the ratio of the final to the microwave energy level (from 270 to 630 W) in the
initial volume of the drying sample. Many researchers have microwave assisted fluidized bed system, the drying time
expressed shrinkage as a function of selected dimension decreased dramatically. To be detailed, at the air velocity
changes of the samples [18]. In many computations it can of 0.8 m/s and 40 C it took 150 min at 270 W, and for the
be assumed that there is no shrinkage just for simplicity; same air velocity but at 70 C it took only 16 min at
nevertheless, shrinkage is never negligible in drying phe- 630 W. The decrease in drying time with increase in drying
nomena. Most of agricultural products shrink during dry- temperature may be due to increase in water vapor pressure
ing; some of them shrink by more than 50 % of their within the melon seed, which increased the migration of
original dimensions depending upon the drying method and moisture, especially as the drying occurs only in falling
degree of drying [16]. Shrinkage of melon seed causes a rate period [30]. Similar observation was reported for
significant variation in mass transfer phenomenon. As a tomato slices [31]. The moisture ratio of melon seed
result, the calculated value of the effective moisture dif- decreased exponentially as the drying time increased.
fusivity will be overestimated. For this reason, dimensional Continuous decline in moisture ratio means that diffusion
changes of the melon seed due to shrinkage should be governed the internal mass transfer [32]. Therefore, the
calculated. Three dimensions of the melon seed in three water vapor concentration on the outer surface of the
directions were measured by means of a digital caliper. drying object reached the equilibrium conditions more
Geometric mean diameter of the melon seed can be cal- rapidly at higher drying air temperature. Similar results
culated as follows noting that the shape is spheroid [29]: were reported by other researchers for drying of apple [33]
1 or green beans [34]. The combined effects of microwave
D ¼ ðA  B  CÞ3 ð12Þ power and drying air temperature on drying time are shown
where D, A, B and C are geometric mean, major, inter- in Fig. 2. The results indicated that by increasing the dry-
mediate and minor diameters (m), respectively. ing air temperature and using microwave energy power as
Melon seed volume before drying (initial volume) was an assisting heat source, the values of drying rate or
computed using the following equation [30]: moisture diffusivity increased, probably due to the pene-
  tration of microwave energy into the sample and also due
4 D 3 to the creation of a large vapor pressure difference between
V0 ¼ p ð13Þ
3 2 the center and the surface of the sample [14, 35]. Similar
where V0 is the initial volume before drying (m3) results were reported by other researchers [14, 35].
Shrinkage of the samples was calculated using the fol-
lowing equation [17]: Mathematical modeling

ðV0  VÞ The drying kinetics of melon seed at different bed condi-

Sb ¼  100 ð14Þ
V0 tions, air temperature and microwave power level was
where Sb is shrinkage (%) and V is the final volume after predicted using five mathematical models shown in
drying (m3). Table 1 (Midilli et al. Newton, Aghbashloo et al. Page and
Logistic), to evaluate their suitability based on the select
criteria such as coefficient of determination (R2), Chi
square (v2) and root mean square error (RMSE). Table 2
Results and discussion shows the quality of the fitting for the five models tested
and the results demonstrated that the Aghbashloo et al.
Drying kinetics model, on average, succeeded comparatively to the others
by corresponding to higher R2 values for all temperature
In order to consider drying behavior, different drying levels, while the v2 and RMSE values were also found to be
curves of a hygroscopic product may be plotted, being the the lowest, thus meaning a good agreement with all
plot of moisture content versus time the more convenient. experimental data. Generally, R2, v2 and RMSE values of
Figure 2 illustrates the variations of moisture content the Aghbashloo et al. model ranged from 0.9908 to 0.9999,

123
 I. Goplour et al.

Fig. 2 Effect of drying air temperature on drying time of melon seed triangle 70 C). The air velocities were 0.8 m/s in the fixed bed,
at different microwave power levels (270, 450, and 630 W) and 1.5 m/s in the semi fluid bead and 2.3 m/s in the fluid bed
various air temperatures (filled circle 40, filled diamond 55 and filled

0.0000 to 0.0018 and 0.0228 to 0.8231, respectively superior for prediction of melon seed drying behavior at
(Table 2). Consequently, Aghbashloo et al. model was 70 C and 1.5 m/s and microwave power level of 630 W
selected as a suitable model to describe the thin layer (Fig. 3). This confirms the suitability of the models to
drying behavior of melon seed in a microwave assisted represent the experimental results.
fluidized bed dryer, and the results of the parameter esti-
mation for this model are shown in Table 3. As the results Effective moisture diffusivity
imply, it was seen that the Aghbashloo et al. model pro-
vided a good fitness to experimental data on drying of The determined values of effective moisture diffusivity
melon seed at 70 C and 1.5 m/s and microwave power (Deff) for all levels of air temperature, velocity and
level of 630 W with the highest R2 of 0.99998, the lowest microwave power are exhibited in Fig. 4. It can be seen
v2 of 0.00000 and RMSE of 0.0228 (Table 2). To validate that Deff values of the melon seed increased progres-
the suitability of the model, the experimental and predicted sively for the same conditions of air temperature as
drying characteristics were compared. Predicted values of applied microwave power was increased. This might be
moisture ratio by Aghbashloo et al. model were plotted described by the enhanced heating energy, which would
against experimental data (Fig. 3). The R2 value (0.9999) increase the activity of the water molecules leading to
of this curve showed that the Aghbashloo et al. model is higher moisture diffusivity when samples were dried at

123
Investigating shrinkage and moisture diffusivity of melon seed in a microwave assisted thin…

Table 2 Statistical criteria obtained from fitting the five models to the experimental data of melon seed drying kinetics
Model name Air Microwave R2 v2 RMSE
velocity power
(m/s) (W) 40 C 55 C 70 C 40 C 55 C 70 C 40 C 55 C 70 C

Aghbashloo 0.8 270 0.99294 0.99929 0.99947 0.00174 0.00130 0.00082 0.24199 0.21332 0.16204
et al. 450 0.99886 0.99979 0.99746 0.00112 0.00010 0.00185 0.27721 0.10005 0.32597
630 0.99720 0.99971 0.99998 0.00070 0.00026 0.00001 0.23448 0.11674 0.02562
1.5 270 0.99088 0.99954 0.99931 0.00026 0.00087 0.00038 0.82316 0.14735 0.17060
450 0.99849 0.99997 0.99993 0.00079 0.00000 0.00005 0.30308 0.03139 0.05220
630 0.99977 0.99993 0.99998 0.00089 0.00002 0.00000 0.11373 0.05528 0.02289
2.3 270 0.99368 0.99721 0.99889 0.00172 0.00059 0.00007 0.38338 0.28649 0.21972
450 0.99952 0.99989 0.99639 0.00050 0.00000 0.00031 0.06734 0.06015 0.12579
630 0.99951 0.99996 0.99989 0.00156 0.00001 0.00010 0.11603 0.03449 0.06368
Newton 0.8 270 0.55891 0.79761 0.99530 0.44750 0.16840 0.00300 5.87005 3.60094 0.48062
450 0.77518 0.94127 0.99642 0.19780 0.03717 0.00949 3.90264 1.69195 0.47547
630 0.83297 0.96963 0.00293 0.12963 0.01870 0.00464 3.15935 1.19995 0.14257
1.5 270 0.57059 0.93697 0.00026 0.41460 0.03942 0.08978 5.65015 1.74222 1.03194
450 0.82102 0.98504 0.01383 0.14100 0.00852 0.00650 3.29499 0.81029 0.70762
630 0.87888 0.98693 0.99106 0.08970 0.00774 0.00525 2.62809 0.77239 0.63617
2.3 270 0.81452 0.86616 0.98010 0.12990 0.09334 0.01127 3.16264 2.68089 0.93167
450 0.88567 0.98140 0.98767 0.08567 0.01062 0.00741 2.56838 0.90429 0.75536
630 0.90839 0.99477 0.99842 0.06694 0.00269 0.00089 2.27032 0.45587 0.26192
Logistic 0.8 270 0.89896 0.87520 0.99002 0.00977 0.00919 0.00147 0.86739 0.84120 0.33643
450 0.87087 0.94379 0.99464 0.01321 0.03558 0.00293 1.00854 1.65519 0.47498
630 0.85569 0.94183 0.99996 0.11199 0.00422 0.00001 2.93653 0.57003 0.03842
1.5 270 0.86483 0.94001 0.97564 0.02258 0.00518 0.01340 1.31858 0.63155 1.01577
450 0.86879 0.98104 0.98861 0.02235 0.00117 0.00647 1.31185 0.30040 0.70582
630 0.88485 0.98719 0.99998 0.08520 0.00759 0.00012 2.56132 0.76468 0.09652
2.3 270 0.92784 0.94423 0.94598 0.01144 0.00715 0.00291 0.93855 0.74199 0.47336
450 0.89274 0.98187 0.98758 0.08030 0.01030 0.00730 2.48658 0.89056 0.75014
630 0.85156 0.99998 0.99843 0.01632 0.00009 0.00088 1.12100 0.08731 0.26104
Midili et al. 0.8 270 0.99640 0.99759 0.99960 0.00124 0.00048 0.00025 0.50899 0.19324 0.13957
450 0.99787 0.99951 0.99862 0.00053 0.00003 0.00000 0.20334 0.05508 0.01277
630 0.99753 0.99875 0.99999 0.00041 0.00009 0.00001 0.17961 0.83290 0.29100
1.5 270 0.99740 0.99706 0.99960 0.000826 0.00026 0.00038 0.45219 0.13957 0.17105
450 0.99713 0.99960 0.80174 0.00047 0.00000 0.00234 0.19164 0.02039 0.72447
630 0.99790 0.99980 0.99882 0.00026 0.00001 0.00002 0.14365 0.03052 0.04030
2.3 270 0.99116 0.99966 0.99980 0.00140 0.00004 0.00000 0.42832 0.5760 0.02770
450 0.99805 0.99996 0.99993 0.00027 0.00000 0.00003 0.14551 0.01906 0.05515
630 0.99617 0.99995 0.99998 0.00042 0.00001 0.00007 0.27983 0.0087 0.02770
Page 0.8 270 0.99494 0.99356 0.99872 0.00715 0.00059 0.00034 0.36603 0.31638 0.25127
450 0.99554 0.99868 0.99661 0.00099 0.00013 0.00138 0.29392 0.08775 0.37742
630 0.99587 0.99637 0.99997 0.00371 0.00017 0.00000 0.23249 0.14230 0.03163
1.5 270 0.99723 0.99694 0.99286 0.00880 0.00028 0.00037 0.25882 0.14231 0.17106
450 0.99524 0.99907 0.99998 0.00119 0.00001 0.00003 0.24679 0.06625 0.00277
630 0.99302 0.99993 0.99998 0.00003 0.00016 0.00003 0.26178 0.01618 0.04474
2.3 270 0.99911 0.99944 0.99891 0.00000 0.00442 0.00194 0.36392 0.07393 0.21440
450 0.99644 0.99986 0.99947 0.00035 0.00005 0.00006 0.19699 0.02244 0.15574
630 0.99573 0.99997 0.99980 0.00035 0.00001 0.00450 0.34658 0.03618 0.08775

123
 I. Goplour et al.

Table 3 Model parameters estimated for the different drying conditions

Model name Air velocity Microwave power 40 C 55 C 70 C
(m/s) (W) a b a b a b

Aghbashloo et al. 0.8 270 31.2246 13.1580 19.5767 7.5299 3.6929 0.3889
450 19.1942 7.4470 9.6401 3.0320 4.3371 0.5334
630 16.0276 6.0186 7.8636 2.0553 3.1723 0.1667
1.5 270 27.5069 11.8128 9.2447 2.9614 6.1806 1.5823
450 18.6452 7.0784 5.6811 1.1944 6.8709 1.6566
630 15.8479 5.6647 5.9878 1.2051 6.8755 1.5862
2.3 270 12.4011 5.0175 12.9093 4.8030 5.7796 1.3399
450 14.0307 4.8873 5.6179 1.2918 5.6179 1.2918
630 14.2930 4.8181 4.0336 0.6056 3.4285 0.3174

1.20 Activation energy

Table 4 presents the values of activation energy calculated

1.00
from the values of the effective diffusivity as a function of
Pridicted moisture ratio

temperature for the different drying conditions. The values

0.80 stand in the range from 27.6 to 45.3 kJ/mol, for fluid bead
at 270 W power and fixed bed at 630 W power. It was
observed a trend for the activation energy to be higher for
0.60
fixed bed conditions, followed by semifluid bed and finally
fluid bed. On the other hand, no constant trend was
0.40 R² = 0.9999 observed for the variation of activation energy with
microwave power applied. Lüle and Koyuncu [39] evalu-
0.20 ated the convective and microwave drying characteristics
of Sorbus fruits and found that convective drying air
temperature and microwave oven power levels influenced
0.00 the total drying time as well as the requirements in terms of
0.00 0.20 0.40 0.60 0.80 1.00 1.20
total energy and specific energy. Also Łechtańska et al.
Eperimental moisture ratio [40] studied the microwave- and infrared-assisted convec-
tive drying of green pepper and verified that convective
Fig. 3 Linear fit between predicted moisture ratio and experimental
values (at 70 C and 1.5 m/s and microwave power level of 630 W) drying assisted with both microwave and/or infrared radi-
of Aghbashloo et al. model for thin layer drying of melon seed ation significantly shortened the drying time and saved
energy consumption compared to pure convective drying.
higher microwave power. This effect is more intense for The values for activation energy found in this work are
higher temperature levels. Many researchers have within the same range of other values reported in literature,
reported this phenomenon in their studies, such as [15] although in different drying conditions. The activation
for peaches and [36] for plums. The values of Deff found energy for apples at convective drying in the temperature
in this research fall into the range of 6.51 9 10-10– range 30–60 C was between 32.8 and 35.3 kJ/mol,
6.59 9 10-9 m2/s. Madamba et al. [37] reported that the depending on the variety [28]. The activation energy for
Deff value for food materials is within the range of the convective drying of pumpkin pulp was 33.7 kJ/mol for
10-11–10-9 m2/s. Furthermore, the obtained results were temperatures between 30 and 70 C [41], and for onion was
in agreement with the results of [38] and [4]. The results 26.4 kJ/mol for temperatures in the range 30–60 C [42].
indicate that the maximum value of Deff obtained at
maximum air temperature (70 C) and maximum power Shrinkage effect
level (630 W) for the fixed bed with 0.8 m/s was
6.59 9 10-9 m2/s. Contrarily, the minimum Deff In order to show the effects of various parameters on the
obtained was 6.52 9 10-10 m2/s at the temperature of shrinkage of melon seed, several experiments were carried
40 C, power level of 270 W and also in the fixed bed out under different operating conditions. The percentages
with 0.8 m/s. of shrinkage in melon seed under microwave assisted

123
Investigating shrinkage and moisture diffusivity of melon seed in a microwave assisted thin…

Fig. 4 Deff versus microwave power and air temperature for drying melon seed

fluidized bed dryer at different temperatures of 40, 55 and shrinkage of melon seed. Maximum shrinkage (46.99 %)
70 C and power levels of 270, 450 and 630 W are shown was achieved at air temperature of 70 C, air velocity of
in Fig. 5. It was concluded that drying air temperature and 0.8 m/s and microwave power of 630 W. The lowest
microwave power level both showed significant effect on shrinkage value (15.09 %) was found at air temperature of
the melon seed shrinkage during the drying process. As it 40 C, air velocity of 2.3 m/s and microwave power of
was expected, increasing the heating air temperature and 270 W. More intense shrinkage at higher temperatures and
microwave power resulted in increasing the degree of microwave power is due to faster mass transfer, so the

123
 I. Goplour et al.

Table 4 Calculation of activation energy for different experimental conditions

Drying conditions Microwave power (W) D0e (m2/s) Ea (kJ/mol) Regression coefficient (R)

Fixed bed (0.8 m/s) 270 1.42 9 10-2 44.0 0.9984

450 1.01 9 10-2 42.1 0.9999
630 5.31 9 10-2 45.3 0.9998
-5
Semi fluid bed (1.5 m/s) 270 8.77 9 10 30.2 0.9691
450 5.58 9 10-4 34.2 0.9899
630 1. 9 10-4 28.4 0.9966
Fluid bed (2.3 m/s) 270 4.26 9 10-5 27.6 0.9984
450 4.5 9 10-4 33.1 0.9999
630 1.09 9 10-2 41.1 0.9998

Fig. 5 Variation of shrinkage

with air temperature and
microwave power at different
air velocities for melon seed

created free space causes tension in the hawthorn tissue and the drying kinetics for all the conditions tested. The
therefore the product becomes more wrinkled [16, 18]. Moisture diffusivity values of the melon seed were inten-
Similar results have been obtained for broad beans [16], sively increased as higher microwave power was applied.
and papaya [18]. When water is gradually removed from a The minimum value of Deff obtained was 6.59 9 10-9 m2/
foodstuff, a pressure unbalance is generated between the s, respectively for the fixed bed with 0.8 m/s, 70 C and
inner section of the material and the external pressure 630 W while the minimum value was 6.51 9 10-10 m2/s
creates contracting stresses that lead to material shrinkage for the fixed bed with 0.8 m/s, 40 C and 270 W. The
or collapse, changes in shape and occasionally cracking of activation energy varied from 27.6 to 45.3 kJ/mol, showing
the product [17]. At higher microwave power, much severe a trend for increasing as the air velocity diminished.
shrinkage may occur owing to the fact that high radiation Shrinkage percentage increased with microwave power
power engenders more heat in melon seed samples. Similar and air temperature, so that the highest shrinkage was
behavior has been observed on potato drying [43] and observed at air temperature of 70 C, air velocity of 0.8 m/s
drying of carrot and garlic [44]. and microwave power of 630 W.

Acknowledgments The authors would like to thank Bu-Ali Sina

University for financial support of this study.
Conclusion

Five empirical models were utilized to prognosticate the

drying kinetics of melon seed, from which the Aghbashloo
References
model was the best, showing high values of R2 (from 1. O.O. Ajibola, S.E. Eniyemo, O.O. Fasina, K.A. Adeeko,
0.99294 to 0.99998) in addition with low values of RMSE Mechanical expression of oil from melon seeds. J. Agr. Eng. Res.
(from 0.02289 to 0.82316), thus being indicated to predict 45, 45–53 (1990)

123
Investigating shrinkage and moisture diffusivity of melon seed in a microwave assisted thin…

2. S.A. Bankole, O.A. Lawal, A. Adebanjo, Storage practises and 24. S.M. Henderson, Progress in developing the thin-layer drying
aflatoxin B1 contamination of ‘egusi’ melon seed in Nig. Trop equation. Trans. ASAE. 17, 1167–1168 (1974)
Sci 44, 150–153 (2004) 25. A. Cihan, K. Kahveci, O. Hacihafizoglu, Modeling of intermittent
3. S.A. Bankole, Moisture content, mould invasion and seed ger- drying of thin layer rough rice. J. Food. Eng. 79, 293–298 (2007)
minability of stored melon. Mycopathologia 122, 123–126 (1993) 26. E.K. Akpinar, Determination of suitable thin layer drying curve
4. I. Doymaz, Air-drying characteristics of tomatoes. J. Food. Eng. model for some vegetables and fruits. J. Food. Eng. 73, 75–84
78, 1291–1297 (2007) (2006)
5. I. Golpour, R.A. Chayjan, J.A. Parian, J. Khazaei, Prediction of 27. N.R. Swami Hulle, P.S. Rao, Effect of high pressure pre-treat-
paddy moisture content during thin layer drying using machine ments on structural and dehydration characteristics of Aloe Vera
vision and artificial neural networks. Jast 17, 287–298 (2015) (Aloe barbadensis Miller) cubes. Dry. Tech. 34(1), 105–118
6. L. Momenzadeh, A. Zomorodian, D. Mowla, Experimental and (2016)
theoretical investigation of shelled corn drying in a microwave- 28. A.C. Cruz, R.P.F. Guiné, J.C. Gonçalves, Drying kinetics and
assisted fluidized bed dryer using artificial neural network. Food. product quality for convective drying of apples (cvs. Golden
Bioprod. Process. 89, 15–21 (2011) Delicious and Granny Smith). Int. J. Fruit. Sci. 15(1), 54–78
7. M.S. Hatamipour, D. Mowla, Experimental and theoretical (2015)
investigation of drying of carrots in a fluidized bed with energy 29. N.N. Mohsenin, Physical characteristics: physical properties of
carrier. Dry. Technol. 21, 83–101 (2003) plant and animal materials (Gordon and Breach Science Pub-
8. G.P. Sharma, S. Prasad, Specific energy consumption in micro- lisher, New York, 1986)
wave drying of garlic cloves. Energy. 31, 1921–1926 (2006) 30. M. Kaveh, R.A. Chayjan, M. Esna-Ashari, Thermal and physical
9. M.E.C. Oliveira, A.S. Franca, Finite element analysis of micro- properties modelling of terebinth fruit (Pistacia atlantica L.)
wave heating of solid. Int. Comm. Heat. Mass. Transf. 27(4), under solar drying. Res. Agr. Eng. 61(4), 150–161 (2015)
527–536 (2000) 31. J. Khazaei, G.R. Chegini, M. Bakhshiani, A novel alternative
10. G.R. Askari, Z. Emam-Djomeh, S.M. Mousavi, Heat and mass method for modeling the effects of air temperature and slice
transfer in apple cubes in a microwave-assisted fluidized bed thickness on quality and drying kinetics of tomato slices: super-
drier. Food. Bioprod. Process. 91(3), 207–215 (2013) position technique. Dry. Technol. 26, 759–775 (2008)
11. M.O. Fakhouri, H.S. Ramaswamy, Temperature uniformity of 32. A.K. Haghi, N. Amanifard, Analysis of heat and mass transfer
microwave heated foods as influenced by product type and during microwave drying of food products. Braz. J. Chem. Eng.
composition. Food. Res. Int. 26(1), 89–95 (1993) 25, 491–501 (2008)
12. K.G. Ayappa, H.T. Davis, E.A. Davis, J. Gordon, Two-dimen- 33. H. Feng, J. Tang, R.P. Cavalieri, Dielectric properties of dehy-
sional finite element analysis of microwave heating. AIChE. J. drated apples as affected by moisture and temperature. Trans.
38(10), 1577–1592 (1992) ASAE 45, 129–135 (2001)
13. R.A. Chayjan, M. Kaveh, S. Khayati, Modeling drying charac- 34. B.A. Souraki, D. Mowla, Axial and radial moisture diffusivity in
teristics of hawthorn fruit under microwave-convective condi- bed dryer with energy carrier: modeling with and without
tions. J. Food. Process. Pres. 39(3), 239–253 (2015) shrinkage. J. Food. Eng. 88, 9–19 (2008)
14. L. Momenzadeh, A. Zomorodian, D. Mowla, Applying artificial 35. B.A. Souraki, D. Mowla, Simulation of drying behavior of a
neural network for drying time prediction of green pea in a small spherical foodstuff in a microwave assisted fluidized bed of
microwave assisted fluidized bed dryer. JAST 14, 513–522 inert particles. Food. Res. Int. 41, 255–265 (2008)
(2012) 36. R.K. Goyal, A.R.P. Kingsly, M.R. Manikantan, S.M. Ilyas,
15. A.R.P. Kingsly, R.K. Goyal, M.R. Manikantan, S.M. Ilyas, Mathematical modeling of thin layer drying kinetics of plum in a
Effects of pretreatments and drying air temperature on drying tunnel dryer. J. Food. Eng. 79, 176–180 (2007)
behavior of peach slice. Int. J. Food. Sci. Technol. 42, 65–69 37. P.S. Madamba, R.H. Driscoll, K.A. Buckle, Enthalpy—entropy
(2007) compensation models for sorption and browning of garlic.
16. G. Hashemi, D. Mowla, M. Kazemeini, Moisture diffusivity and J. Food. Eng. 28, 109–119 (1996)
shrinkage of broad beans during bulk drying in an inert medium 38. K. Sacilik, R. Keskin, A.K. Elicin, Mathematical modeling of
fluidized bed dryer assisted by dielectric heating. J. Food Eng. 92, solar tunnel drying of thin layer organic tomato. J. Food. Eng. 73,
331–338 (2009) 231–238 (2006)
17. L. Mayor, A.M. Sereno, Modelling shrinkage during convective 39. F. Lüle, T. Koyuncu, Convective and microwave drying char-
drying of food materials: a review. J. Food. Eng. 61, 373–386 acteristics of sorbus fruits (Sorbus domestica L.). Proced. Soc.
(2004) Behav. Sci. 195(3), 2634–2643 (2015)
18. L.E. Kurozawa, M.D. Hubinger, K.J. Park, Glass transition phe- 40. J.M. Łechtańska, J. Szadzińska, S.J. Kowalski, Microwave- and
nomenon on shrinkage of papaya during convective drying. infrared-assisted convective drying of green pepper: Quality and
J. Food. Eng. 108, 43–50 (2012) energy considerations. Chem. Eng. Proc. Process. Intens. 98,
19. B. Suresh, A. Kaur, M.R. Manikantan, M R, Moisture dependant 155–164 (2015)
physical properties of sunflower seed (Psh 569). Int. J. Eng. Sci. 41. R.P.F. Guiné, S. Pinho, M.J. Barroca, Study of the convective
2(8), 23–27 (2013) drying of pumpkin (Cucurbita maxima). Food. Bioprod. Process.
20. M. Shakeri, R. Khodabakhshian, The physical attributes of saf- 89, 422–428 (2011)
flower seed as a function of moisture content, variety and size. 42. C.L. Mota, C. Luciano, A. Dias, M.J. Barroca, R.P.F. Guiné,
Electron. J. Polish. Agric. Univ. 14(3), 1–10 (2011) Convective drying of onion: kinetics and nutritional evaluation.
21. R. Khodabkhshian, B. Emadi, M.H.A. Fard, Gravimetrical Food. Bioprod. Process. 88, 115–123 (2010)
properties of sunflower seeds and kernels. World. Appl. Sci. J. 43. N. Wang, J.G. Brennan, Changes in structure, density and
8(1), 119–128 (2010) porosity of potato during dehydration. J. Food. Eng. 24, 61–76
22. AOAC, Official method of analysis, 13th edn. (Association of (1995)
Official Analytical Chemists, Washington, 1980) 44. B. Taner, I. Filiz, E. Seda, Y. Hasan, Effects of microwave and
23. M. Aghabashlo, M.H. Kianmehr, A. Arabhosseini, Performance infrared drying on the quality of carrot and garlic. Eur. Food. Res.
analysis of drying of carrot slices in a semi-industrial continuous Technol. 218, 68–73 (2003)
band dryer. J. Food. Eng. 91, 99–108 (2009)

123