Upload
imenehermassi
View
18
Download
0
Embed Size (px)
Citation preview
5/25/2018 Paddy Dehydration by Adsorption
1/7
Available at www.sciencedirect.com
jo ur na l home pa ge : ww w. el sevi er. com /l oc at e/ is sn /1 53 75 11 0
Research Paper: PHPostharvest Technology
Paddy dehydration by adsorption: Thermo-physical
properties and diffusion model of agriculture residues
S. Tirawanichakula, Y. Tirawanichakulb,, E. Snisob
aDepartment of Chemical Engineering, Faculty of Engineering, Prince of Songkla University, HatYai, 90110 Songkhla, ThailandbDepartment of Physics, Faculty of Science, Prince of Songkla University, HatYai, 90110 Songkhla, Thailand
a r t i c l e i n f o
Article history:
Received 31 January 2007
Received in revised form
6 June 2007
Accepted 1 November 2007
Available online 18 December 2007
Dehydration of high-moisture grain kernels using adsorption is an interesting method,
especially when agricultural residues are used as adsorbents. This dehydration method is a
low-temperature drying process that is simple, requires no mechanical equipment, has low
energy consumption and maintains grain quality. The aim of this research was to
investigate the use of rice husk, coconut husk and sago palm rachis to reduce the moisture
content of fresh paddy. The thermo-physical properties and the drying kinetics of these
agricultural residues acting as moisture adsorbents were studied. The bulk density, specific
heat capacity, void fraction, equilibrium moisture content (EMC) and the effective diffusion
coefficient (Deff) of the adsorbents were determined. The bulk density, specific heat capacity
and void fraction were determined at initial moistures contents ranging from 4.1% to 72.4%
dry-basis (d.b.). The EMC andDeffvalues were measured at surrounding temperatures and
relative humidity varying from of 40 to 701C and from 10% to 90%, respectively. For
evaluating the Deffvalue using thin-layer drying (TLD), an experiment was carried out at
inlet air temperatures of 3070 1C. The bulk density and specific heat capacity of all three
adsorbents were linearly dependent on the moisture content and the void fraction was
inversely related to the moisture content. The four commonly cited EMC equations were
fitted to the experimental data. A modified Henderson equation provided the best fit for
describing the EMC isotherm of all the study cases. The modified Henderson and Pabis
model provided a good fit to the experimental TLD results and theDeffvalues, determined
by non-linear regression, were significantly affected by the inlet drying air temperature.
The Deff values were in the range 1.2 1086.2 108 m2 s1. Sago palm rachis had the
highest Deffvalue compared to coconut husk and rice husk. However, the most suitable
adsorbent material was rice husk because it was easy to prepare and it could be
regenerated by low-temperature heating.
&2007 IAgrE. Published by Elsevier Ltd. All rights reserved.
1. Introduction
Sorption processes are used to remove moisture from
materials (Rouquerol et al., 1999). The application of adsor-
bents for moisture reduction and the dehumidification of
paddy using silica gel were reported by Murata et al. (1993).
The results showed that the silica gel dehumidifier for cereal
grains at ambient air temperature was technically feasible.
Applications of moisture adsorption by agriculture residues
have previously been reported and studied (Inoueet al., 2002).
Thailand as an agro-industrial country has large amounts of
agricultural residues such as coconut shell and coconut husk,
ARTICLE IN PRESS
1537-5110/$ - see front matter & 2007 IAgrE. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.biosystemseng.2007.11.001
Corresponding author.
E-mail addresses:[email protected] (S. Tirawanichakul),[email protected] (Y. Tirawanichakul).
B I O S Y S T E M S E N G I N E E R I N G 9 9 ( 2 0 0 8 ) 2 4 9 2 5 5
http://localhost/var/www/apps/conversion/tmp/scratch_5/dx.doi.org/10.1016/j.biosystemseng.2007.11.001mailto:[email protected]:[email protected]:[email protected]:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_5/dx.doi.org/10.1016/j.biosystemseng.2007.11.0015/25/2018 Paddy Dehydration by Adsorption
2/7
rice husk, starch waste from cassava and sago palm rachis
that can be used in various applications (Malik, 2003).
It is generally agreed that moisture reduction from the grain
kernel, some fruits and food such as rough rice, corn, fish and
starch, is described by unsteady liquid diffusion based on
Ficks law (Henderson & Pabis, 1961). Hill (1949) andCrank
(1975)previously proposed isothermal liquid diffusion models
to describe the drying rate of a grain as a function of
diffusivity. Additionally, they found that the effective diffu-
sion coefficient (Deff) value depended on the inlet drying air
temperature and the initial moisture content. However, a few
previous investigations on application of moisture removal by
adsorption have been reported and studied (Murata et al.,1993; Inoue et al., 2002). The application of moisture deso-
rption for grain using agricultural residues is an interesting
and simple technique which removes the need for mechan-
ical equipment and results in lower energy inputs.
The objectives of this research were to determine the
physical parameters of agricultural residues (rice husk, coco-
nut husk and sago palm rachis) affecting dehydration for
fresh paddy kernel. Thermo-physical properties such as the
specific heat capacity, bulk density and void fraction values of
these agricultural residues need to be determined. In addi-
tion, the equilibrium moisture content (EMC) isotherm, the
effective diffusion coefficient (Deff) and the thin-layer drying
equation for all adsorbents required consideration.
2. Experimental procedures
2.1. Materials
The adsorbents rice husk, raw sago palm rachis and coconut
husk were uniformly rewetted by water spraying to initial
moisture contents of 5.621.5% (d.b.) to determine the EMC
and 14.9%, 20.1% and 27.1% (d.b.) to examine the thin-layer
drying equation, and to 4.172.4% (d.b.) to evaluate specific
heat capacity, bulk density and void fraction.
To determine the equilibrium moisture content, saturatedsalt solutions of potassium nitrate (KNO3), sodium chloride
(NaCl), magnesium nitrate decahydrate (Mg(NO3)2 6H2O),
magnesium chloride decahydrate (MgCl2 6H2O) and lithium
chloride (LiCl) were used.
2.2. Methods
2.2.1. Determination of physical properties
The physical properties of the adsorbents were determined in
terms of the specific heat capacity, bulk density and void
fraction. To determine the specific heat capacity, sago palm
rachis and coconut husk samples were cut into cubes
measuring 23.2mm 23.2mm 23.2mm and rice husk was
formed in a layer with dimensions of 7mm 15mm. Thespecific heat capacity of the agricultural residues was
determined using a calorimeter. Consequently, to calculate
the bulk density, the mass of the samples was weighed
by an electronic balance with an accuracy of 70.01g and
the volume measured using a volumetric flask. The bulk
density was calculated from the mass and volume of the
samples.
2.2.2. Determination of equilibrium moisture content (EMC)
To determine the EMC of the agricultural residues (adsor-
bents), the samples were weighed (2035 g) and kept in
stainless-steel wire nets, hanging in five bottles, whichcontained the five saturated salt solutions (in Section 2.2.1).
The glass bottles were then placed in an incubator at a
constant temperature and a relative humidity (HR) of 4070 1C
and 1090%, respectively. The time to achieve the equilibrium
state between the adsorbent and the saturated salt solution
was approximately 1220 days. The samples were then
removed and weighed to determine the final moisture
content (so-called equilibrium moisture content, Meq) follow-
ing the AOAC (1995) method. Four isotherm models for
predicting the equilibrium moisture content were chosen to
fit the experimental data of EMC, surrounding temperature ( T)
and the HR. They were the modified Brunauer et al. (1938)
(BET), the ChungPfost equation (Chung & Pfost, 1967), themodifiedHenderson (1952)and theHalsey (1948)equations.
ARTICLE IN PRESS
Nomenclature
A,B,C,g,h,k,m and n arbitrary coefficientscp specific heat capacity, kJ kg
11C1
Deff effective diffusion coefficient or effective diffu-sivity, m2 s1
Do Arrhenius factorEa activated energy, kJ kmol
1 K1
HR relative humidity, decimalL characteristic half-thickness, mM moisture content, % (d.b.)Meq equilibrium moisture content, % (d.b.)Mi initial moisture content, decimal (d.b.)Mt moisture content at drying time (t), decimal (d.b.)MR,exp,i experimental data of moisture ratio at the ith
drying time, decimal (d.b.)
MR,pre,i predicted value of moisture ratio at theith dryingtime, decimal (d.b.)
MR moisture ratio at the ith drying time which isdefined as (MtMeq)/(MiMeq), decimal
N number of samplesn number of termsR universal gas constant, 8.314 kJ kmol1 K1
R2 coefficient of determination, decimalT absolute temperature, KTin inlet drying air temperature, 1Ct drying time, sln root of the Bessel function of the first kind of
order zero
r bulk density, kg m3
e void fraction, %
B I O S Y S T E M S E N G I N E E R I N G 9 9 ( 2 0 0 8 ) 2 4 9 2 5 5250
5/25/2018 Paddy Dehydration by Adsorption
3/7
The equations used were
Modified BET equation (Brunaueret al., 1938)
HR1 HRMeq
1AB
HRB 1
AB , (1)
where HR is the relative humidity in decimals; A and B arearbitrary coefficients; and Meq is the equilibrium moisture
content in decimals (d.b.).
ChungPfost equation (Chung& Pfost, 1967)
lnHR A
RTeBMeq, (2)
whereRis the universal gas constant, 8.314kJ kmol1 K1;Tis
the absolute temperature in K.
Modified Hendersons equation (Henderson, 1952)
1 HR eATMB
eq
. (3)
Halseys equation (Halsey, 1948)
HR eA=RTMBeq. (4)
The constants in these models were determined by the
non-linear regression analysis from the experimental
data. The coefficient of determination (R2) value was used as
the criterion for determining the best fit of the experimental
data.
2.2.3. Development of a thin-layer drying equation and the
effective diffusion coefficient
The samples were dried in a thin layer (thickness of 15 mm)
to determine the effective diffusion coefficient. Drying
was performed at hot-air inlet temperatures of 3070 1C,
which were suitable for the dehydration of adsorbents.
Above drying temperatures of 701C the adsorbents
may burn and decompose. An air velocity for determining
the thin-layer drying kinetics of materials was fixed at
1.8ms1.
Samples were maintained at initial moisture contents of
14.9%, 20.1% and 27.1% (d.b.). The inlet, outlet and ambient air
(wet and dry bulb) temperatures were recorded using a
datalogger with an accuracy of 71 1C. Air velocity was
measured by a hot-wire anemometer with an accuracy of
70.1ms1.
(a)Empirical model:
Empirical models were developed using non-linear regres-
sion. The four conventional models were formulated as
follows:
Newton (Mujumdar, 1987)
MR ekt. (5)
Page (Page, 1949)
MR MtMeqMiMeq ektm. (6)
Two-term model (Sharaf-Eldeenet al., 1980)
MR Aektn Begt. (7)
Modified Henderson and Pabis (Karathanos, 1999)
MR Aekt Begt ceht, (8)
where MR is the moisture ratio at the ith drying time. It isdimensionless and is defined as (MtMeq)/(MiMeq) whereMt,
MiandMeqare the predicted, initial and equilibrium moisture
contents, respectively, decimal (d.b.); A, B, k, g,h, m and n are
arbitrary coefficients; and t is drying time in s.
The empirical equations were evaluated for best fit using
the coefficient of determination (R2). The experiments were
replicated.
(b)Theoretical model:
It is assumed that moisture is transferred by liquid
diffusion and that shrinkage of materials is negligible duringdrying. In the partial differential equations for moisture
diffusion to a single piece of adsorbent, the rice husk is
assumed to be an infinite slab and both the coconut husk and
the sago palm rachis are assumed to be cubic in shape. The
general solution for moisture ratio can be expressed by the
following equation:
For an infinite slab
MRMtMeq
Mi Meq
8p2
X1n1
1
2n 12ep
22n12=4LDefft; (9)
whereL is the characteristic half-thickness of the adsorbent
in m; t is the drying time in s.
For the cubic shape
MRMtMeq
Mi Meq
8p2
3 X1n0
1
2n12
3e2n1
23p2Defft=L2.
(10)
The effective diffusion coefficient (Deff) is conventionally
described by the Arrhenius-type equation as follows:
Deff DoeEa=RT, (11)
whereDois an Arrhenius factor and Eais an activated energy
in kJkmol1
K1
.Many studies have reported that Do depends on the
absolute drying air temperature (T) and the moisture content.
However, this Do value can be determined as a constant
(Muletet al., 1989;Rovedo et al., 1998). The experiments were
replicated three times.
3. Results and discussion
3.1. Determination of physical property
Fig. 1shows the variability of the specific heat capacity of rice
husk, coconut husk and sago palm rachis with the moisturecontent. The experimental data of each adsorbent were fitted
ARTICLE IN PRESS
B I O S Y S T E M S E N G I N E E R I N G 9 9 ( 2 0 0 8 ) 2 4 9 2 5 5 251
5/25/2018 Paddy Dehydration by Adsorption
4/7
to linear regression. The results show that the specific heat
capacity (cp) in kJkg1 K1 of all adsorbents has a linear
relation to the moisture content (M) in % (d.b.). These specific
heat capacities are determined corresponding to the pervious
works (Muletet al., 1989). The equations are as follows:
rice husk cp 0:0287M 1:6686, (12)
coconut husk cp 0:0232M 2:2325, (13)
sago palm rachis cp 0:0343M 2:2986 (14)
with values for R2 of 0.953, 0.948 and 0.935, respectively.
The results showed that rice husk had the lowest specific
heat capacity, while at lower moisture contents coconut husk
and sago palm rachis had similar specific heat capacities.
However, at higher moisture contents, the rice husk had a
relatively low heat adsorption compared to the others.
Fig. 2shows the relationship between the bulk density and
the moisture content of the three adsorbents. The bulk
density of the adsorbents was linearly dependent on the
moisture content. Moreover, the bulk density of rice husk hada relatively high value compared to the coconut husk and
sago palm rachis. The bulk densities (r) in kgm3 could be
expressed as follows:
rice husk r 1:27M 80:65, (15)
coconut husk r 0:26M 34:41, (16)
sago palm rachis r 0:26M 56:13 (17)
with values ofR2 of 0.973, 0.918 and 0.932, respectively.
The experimental and fitted void fraction data for all
adsorbents are shown in Fig. 3. The results indicated that
the percentage of void fraction (e) slightly decreases with anincrease in the moisture content. The least-squares regres-
sion analysis indicated that the fitted regression equations
were in close agreement with the experimental data. More-
over, the percentage of void fraction of rice husk was higher
than that for coconut husk and sago palm rachis, respectively
rice husk 0:0922M 98:75, (18)
coconut husk 0:0887M 81:19, (19)
sago palm rachis 0:0483M 69:75 (20)
with values ofR2 of 0.952, 0.974 and 0.973, respectively.
However, rice husk is a more suitable adsorbent than
coconut husk and sago palm rachis, especially in Thailand.
This is because rice husk is a direct by-product of dehusking
of paddy and it is not necessary to cut it into a in rectangular
shape like the others.
3.2. Determination of the equilibrium moisture content
The four EMC desorption models for rice husk, coconut husk
and sago palm rachis were evaluated as shown inTables 13,
respectively. The results indicated that the predicted values
using the modified Henderson model were in close agreementwith the experimental results for the temperature and
ARTICLE IN PRESS
0
1
2
3
4
5
0 10 20 30 40 50 60
Moisture content,% (d.b.)
Specificheatcapacity,
kJkg
1
K
1
Experiment Simulation
Fig. 1 Relationship between specific heat capacity and
moisture content of three Absorbents; , sago palm rachis;, coconut husk; , rice husk; , regression.
0
20
40
60
80
100
120
0 5 10 15 20 25 30
Moisture content, % (d.b.)
Bulkdensity
,kg
(m3)1
Fig. 2 Change in bulk density of rice husk, coconut husk
and sago palm rachis at various moisture contents; , sago
palm rachis; , coconut husk; , rice husk; , regression.
60
70
80
90
100
0 5 10 15 20 25 30
Moisture content, % (d.b.)
Voidfraction,
%
Fig. 3 Change in void fraction of rice husk, coconut husk
and sago palm rachis at various moisture contents; , sago
palm rachis; , coconut husk; , rice husk; , regression.
B I O S Y S T E M S E N G I N E E R I N G 9 9 ( 2 0 0 8 ) 2 4 9 2 5 5252
5/25/2018 Paddy Dehydration by Adsorption
5/7
relative humidity ranges. TheR2 values of the EMC absorption
isotherm of these three adsorbents simulated using the
modified Henderson equation were 0.942, 0.773 and 0.942
for rice husk, coconut husk and sago palm rachis, respec-
tively. Additionally, the modified EMC equations are as
follows:
rice husk 1HR exp3:03 104TM0:91eq , (21)
coconut husk 1 HR exp2:57 104TM0:92eq , (22)
ARTICLE IN PRESS
Table 1 Equilibrium moisture content constants andcoefficients of different desorption isotherm models forrice husk
Model Arbitrary constants inmodel
R2
A B
Modified BET 0.998 1.638 0.815
Chung and Pfost 7.426 103 0.199 0.996
Modified Henderson 2.750 104 1.130 0.988
Halsey 8.673 103 1.017 0.995
R2, coefficient of determination.
Table 2 Equilibrium moisture content constants andcoefficients of different desorption isotherm models forcoconut husk
Model Arbitrary constants inmodel
R2
A B
Modified BET 0.613 1.557 0.902
Chung and Pfost 3.899 104 0.393 0.731
Modified Henderson 1.608 106 2.374 0.773
Halsey 1.010 106 3.103 0.842
R2,coefficient of determination.
Table 3 Equilibrium moisture content constants andcoefficients of different desorption isotherm models forsago palm rachis
Model Arbitrary constants inmodel
R2
A B
Modified BET 0.725 1.415 0.944
Chung and Pfost 1.858 104 0.240 0.996
Modified Henderson 1.289 105 2.257 0.999
Halsey 5.071 105 2.528 0.999
R2, coefficient of determination.
0.0
0.2
0.4
0.6
0.8
1.0
0 20 40 60
Drying time, min
0 20 40 60
Drying time, min
Drying time, min
Moistureratio,
decimal
0.0
0.2
0.4
0.6
0.8
1.0
Moistureratio,
decimal
0.0
0.2
0.4
0.6
0.8
1.0
Moistureratio,
decimal
0 10 20 30 40 50 60
(a)
(b)
(c)
Fig. 4 Comparison between experimental data and
simulated data from the modified Henderson and Pabis
empirical model at various drying times, drying
temperatures of 3070 1C and a constant airflow rate of
1.8ms1; , drying temperature of 30 1C; , drying
temperature of 50 1C;K, drying temperature of 70 1C; ,
simulation. (a) Rice husk with initial moisture content of
14.9% (d.b.). (b) Coconut husk with initial moisture content
of 27.1% (d.b.). (c) Sago palm rachis with initial moisture
content of 14.9% (d.b.).
B I O S Y S T E M S E N G I N E E R I N G 9 9 ( 2 0 0 8 ) 2 4 9 2 5 5 253
5/25/2018 Paddy Dehydration by Adsorption
6/7
sago palm rachis 1 HR exp1:02 104TM1:30eq (23)
with the values ofR2 of 0.942, 0.773 and 0.942, respectively.
The EMC values of all adsorbents will be used for evaluating
the empirical and semi-theoritical moisture ratio in the next
section.
3.3. Establishment of a thin-layer drying equation and an
effective diffusion coefficient
To validate the moisture ratio (MR) equation, including the
EMC value for adsorbents, the thin-layer drying of rice husk,
coconut husk and sago palm rachis was conducted at the
drying temperatures of 30, 50 and 70 1C. The inlet drying air
velocity was fixed at 1.8m s1. The experimental results were
used to formulate the thin-layer drying equation by curve
fitting [Eqs. (5)(8)]. The rate of measured and predicted
moisture transfer was significantly related to the inlet drying
air temperature. Additionally, drying at high temperatures
was more rapid.Fig. 4compares the measured and predicted
values for moisture content at various drying times. The
results also showed that the proposed modified Henderson
and Pabis model could predict the moisture content in close
agreement with the measurements. The R2 value is higher
than 0.99 as shown inTable 4.
Eqs. (9) and (10) were used to determine the effective
diffusion coefficient (Deff) and the diffusion coefficient at air
temperatures of 3070 1C was formulated using Eq. (11). The
equations were written as follows:
rice husk Deff 0:815e9:69103Tin, (24)
coconut husk Deff 0:914e2:01102Tin, (25)
sago palm rachis Deff 1:260e1:59102Tin , (26)
whereTinwas the inlet drying air temperature in 1C and with
values ofR2 of 0.997, 0.997 and 0.998, respectively.
Figs. 57 show the effective diffusion coefficient versus
drying air temperature. The results show that the Deffvalue
was exponentially related to the drying temperature. At the
drying air temperature range between 30 and 70 1C, the range
of values of theDeffvalue for rice husk [(110) 1010 m2 s1] is
lower than that obtained of coconut husk [(16) 108 m2 s1]
and sago palm rachis [(25) 108
m2
s1
]. This indicates thatcoconut husk and sago palm rachis more rapidly transfer
moisture compared to rice husk. TheseDeffvalues are useful
for predicting the drying kinetic of adsorbents. All of these
physical parameters are necessary to develop a mathematical
model for predicting paddy drying by adsorption. Moreover,
some physical parameters will be useful for studying thecombustion and adsorption of these agricultural residues.
ARTICLE IN PRESS
Table 4 Illustration of arbitrary constants of modified Henderson and Pabis drying model for three different absorbents
Type Absorbent Drying model coefficient of Modified Henderson and Pabis model R2
Mi, % d.b. Inlet drying temperature,1C
A B C g h k
Rice husk 20.1 50 0.4400 0.4393 0.1267 0.2258 0.0130 0.2258 0.999Coconut husk 27.1 50 0.5372 0.4217 0.0412 0.0017 1.7761 0.0526 0.999
Sago palm
rachis
14.9 50 0.1369 0.1776 0.1327 0.0226 0.7173 0.0226 0.991
Mi, initial moisture content;R2, coefficient of determination.
0.0
1.0
2.0
3.0
4.0
0 30 60 90 120
Temperature, C
Diffusioncoefficient,1
0
10,m
2s
1
Fig. 5 Relationship between the effective diffusion
coefficient and the temperature of rice husk; ,
experimental data of rice husk; , non-linear regression.
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
0 30 60 90 120
Temperature, C
Diffusioncoefficient,
1
0
8m
2s
1
Fig. 6 Relationship between the effective diffusion
coefficient and the temperature of coconut husk; ,
experimental data of coconut husk; , non-linear
regression.
B I O S Y S T E M S E N G I N E E R I N G 9 9 ( 2 0 0 8 ) 2 4 9 2 5 5254
5/25/2018 Paddy Dehydration by Adsorption
7/7
4. Conclusions
The physical properties of three agriculture residues (rice
husk, coconut husk and sago palm rachis) using initial
moisture contents of 4.172.4% (d.b.) were studied. The
important parameters for drying kinetics such as equilibrium
moisture content, thin-layer drying equation and effective
diffusion coefficient were determined at drying temperatures
of 3070 1C and initial moisture contents of 14.9%, 20.1% and
27.1% (d.b.). In a future paper these physical properties will be
used for developing a mathematical model of paddy drying
using these agriculture residues. The following conclusions
are drawn:
1. The physical property of adsorbents in terms of the
specific heat capacity, bulk density and void fraction was
found to be linearly dependent on the moisture content.
The rice husk was found to have a high specific heat
capacity, bulk density and void fraction compared to
coconut husk and sago palm rachis.
2. The modified Hendersons equation developed to predict
the equilibrium moisture content for all adsorbents was
valid for ambient relative humidity in the range between
10% and 90% and temperature in the range of 4070 1C.
3. The empirical equation for thin-layer drying using themodified Henderson and Pabis model provided close
agreement between predicted and measured results for
all three adsorbents.
4. The drying kinetics of adsorbents was well explained by
the diffusion model while an effective diffusion coefficient
can be formed as an Arrhenius function of drying air
temperature.
Finally, it can be concluded that rice husk is the most
suitable agriculture residue for dehydrating fresh paddy. This
is because it is a direct by-product of paddy dehusking and
does not need to be prepared before being used in the drying
process. However, in the absence of rice husk, both sago palm
rachis and coconut husk can be used for moisture removal
from fresh paddy.
Acknowledgements
The authors wish to express their sincere thanks to Graduate
School, Department of Chemical Engineering, Faculty ofEngineering and Department of Physics, Faculty of Science,
Prince of Songkla University, Thailand for their financial
support and facilities. Finally, the authors would like to
gratefully acknowledge Dr. K. Inoue, Agricultural Engineering
Division, the National Agricultural Research Center for
Hokkaido Region (NARCH) Japan for his warmly advice.
R E F E R E N C E S
AOAC (1995). Official Methods of Analysis. Association of OfficialAnalytical Chemists, 15th edn. Washington, DC, USA
Brunauer S; Emmett P H; Teller E (1938). Adsorption of gasesin multi-molecular layers. Journal of American ChemistrySociety, 60, 309319
Chung D S; Pfost H B (1967). Adsorption and desorption of watervapor by cereal grain and their products. Transactions of theASAE,10, 549557
Crank J (1975). The Mathematics of Diffusion. London, OxfordHalsey G (1948). Physical adsorption on non-uniform surface.
Journal of Chemical Physics, 16, 8392Henderson S M(1952). A basic concept of equilibrium moisture.
Agricultural Engineering Journal, 3, 2932Henderson S M; Pabis S (1961). Grain drying theory, l. Tempera-
ture effect on drying coefficients. Journal of AgriculturalEngineering Research,6, 169174
Hill T L(1949). Physical adsorption and the free volume model for
liquids. Journal of Chemical Physics,17, 590Inoue K; Tirawanichakul Y; Prachayawarakorn S; Soponronnarit
S(2002). Analysis of dry processing of paddy in Thailand.Annual Meeting of Japanese Society of Agricultural Machinery,pp 469470, Tokyo, Japan
Karathanos V T (1999). Determination of water content of driedfruits by drying kinetics. Journal of Food Engineering,39,337344
Malik P K(2003). Use of activated carbons prepared from sawdustand rice-husk for adsorption of acid dyes: a case study of acidyellow 36. Dyes and Pigments, 56, 239249
Mulet A; Berna A; Rossello C (1989). Drying of carrots, I: dryingmodel. International Journal of Drying Technology,7(3),536557
Mujumdar A S (1987). Handbook of Industrial Drying. Marcel-
Dekker Inc., New YorkMurata S; Enomoto T; Miyauchi K (1993). Adsorption character-
istics of thin silica-gel layer and simulation in the fixed layer.Journal of the Jap anese of Agricultural Machinery (Japanese),55(3), 4149
Page G E (1949). Factors influencing the maximum rate of dryingshell corn in layers. MSc Thesis, Purdue University, WestLafeyette, Indiana, USA
Rovedo C O; Suarez C; Viollaz P (1998). Analysis of moistureprofiles, mass Biot numbers and driving forces duringdrying of potato slabs. Journal of Food Engineering, 36(2),211311
Rouquerol F; Rouquerol L; Sing K S W (1999). Adsorption byPowders and Porous Solids. Academic Press, San Diego, USA
Sharaf-Eldeen Y L; Blaisdell J L; Hamdy M Y (1980). A model forear corn drying. Transactions of the ASAE, 23, 12611271
ARTICLE IN PRESS
0.0
2.0
4.0
6.0
8.0
10.0
0 30 60 90 120
Temperature, C
Diffusioncoefficient,1
0
8m
2s
1
Fig. 7 Relationship between the effective diffusion
coefficient and the temperature of sago palm rachis; ,
experimental data of sago palm rachis; , non-linear
regression.
B I O S Y S T E M S E N G I N E E R I N G 9 9 ( 2 0 0 8 ) 2 4 9 2 5 5 255