Anda di halaman 1dari 7

Journal of Food Engineering 100 (2010) 461467

Contents lists available at ScienceDirect

Journal of Food Engineering


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

Shelf life prediction of packaged cassava-our-based baked product


by using empirical models and activation energy for water vapor
permeability of polyolen lms
Ratchaneewan Kulchan a, Waraporn Boonsupthip b, Panuwat Suppakul c,*
a

Thai Packaging Centre, Thailand Institute of Scientic and Technological Research, 196 Phaholyothin Rd., Ladyao, Chatuchak, Bangkok 10900, Thailand
Department of Food Science and Technology (Food Engineering Major), Faculty of Agro-Industry, Kasetsart University, 50 Phaholyothin Rd., Ladyao,
Chatuchak, Bangkok 10900, Thailand
c
Department of Packaging and Materials Technology, Faculty of Agro-Industry, Kasetsart University, 50 Phaholyothin Rd., Ladyao, Chatuchak, Bangkok 10900, Thailand
b

a r t i c l e

i n f o

Article history:
Received 14 February 2010
Received in revised form 18 April 2010
Accepted 20 April 2010
Available online 15 May 2010
Keywords:
Cassava-our-based baked product
Moisture sorption
Empirical model
Activation energy
Shelf life
Packaging

a b s t r a c t
Moisture sorption kinetics and isotherms of cassava-our-based baked product were investigated. Empirical models were tested to t the experimental data. Textural changes of the product were investigated. In
addition, activation energies (Ep) for water vapor permeability (WVP) of polyolen lms were determined. Finally, the product was packaged in low-density polyethylene (LDPE) or oriented polypropylene
(OPP) pouches, and stored at 30 1 C and 50 2% RH to simulate actual storage conditions and to determine shelf life. This actual shelf life was compared to the predicted shelf life by using empirical models
and Ep for WVP. Moisture sorption kinetics was more rapid during the initial stage, while a lesser amount
of moisture was adsorbed as adsorption time increased. The higher the relative humidity used, the more
pronounced the effect. The sigmoidal moisture sorption isotherms of this product can be classied as type
II. The GAB model was found to be the best-t model for this product. Once the product hardness or work
reached the maximum and began to reduce at moisture content (MC) 6%, the product texture began to
be detected as becoming slightly soft. This implies that hardness and work at the maximum level could be
used to identify the critical MC which causes a loss of crispness to an unacceptable degree. The predicted
shelf lives estimated by employing Ep for WVP of LDPE and OPP, and the GAB model were close to the
actual shelf lives. Therefore, the estimation by empirical models and activation energy was found to be
applicable for rapid and accurate shelf life prediction.
2010 Elsevier Ltd. All rights reserved.

1. Introduction
Cassava (Manihot esculenta Crantz) our is used as a key ingredient in several dry crisp products such as potato chips and puffed
curls. In addition, Asian and Latin American peoples are interested
in its use as a partial substitute for wheat our (Lopez et al., 2004;
Mohamed et al., 2006). To consumers, high crispness of such products indicates not only good quality but also freshness (Rohm,
1990). Unfortunately, few study results have been reported on
the creation and preservation of crispness for cassava-based our
products (Chang et al., 2000). Such research has been especially
rare for multi-component systems.
Sorption characteristics of cassava-our-based baked products
are crucial for the design, modeling and optimization of their
drying, packaging, storage and transport. Knowledge of sorption
isotherms is also important for predicting moisture sorption
* Corresponding author. Tel.: +66 2 562 5058; fax: +66 2 562 5047.
E-mail address: fagipas@ku.ac.th (P. Suppakul).
0260-8774/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jfoodeng.2010.04.031

properties of highly sensitive food products via empirical models.


These isotherms provide information on the moisture-binding
capacity of products at a determined relative humidity, and are a
useful means for analyzing the moisture plasticizing effect and
the effect on textural properties (Bell and Labuza, 2000; Al-Muhtaseb et al., 2002). Chirife and Iglesias (1978) reviewed 23 isotherm
models and their use for tting sorption isotherms of foods and
food products. None of these models accurately described the sorption isotherm over the entire range of relative humidity, since
water is related to the food matrix by different mechanisms in different activity regions. However, these kinetic models are still
important for use in the prediction of moisture sorption properties
of foodstuffs.
In the texture study, crispness was perceived as a combination
of the sound generated and the fracture of the product as it was
bitten completely through with the back molars (Duizer et al.,
1998). Different instrumental and sensory approaches have been
applied to study this quality attribute, and have generated a large
amount of experimental data (Roudaut et al., 2002). Unfortunately,

462

R. Kulchan et al. / Journal of Food Engineering 100 (2010) 461467

Nomenclature
A
test area (m2)
aw
water activity
a, b, c, d constants of the Peleg modelCB
constant of the BET model
CG, k
constants of the GAB model
Ep
apparent activation energy of water vapor permeability
F, G, H constants of the Lewicki modelG
weight change (g)
k, c
constants of the Oswin model
k1, k2
parameters of Peleg kinetic model
l
thickness (mil)
M0
initial moisture content

no conclusion can be soundly drawn for the relationship between


instrumental and sensory results. This is due to the fact that many
denitions of crisp were applied (Roudaut et al., 2002), and only
a few studies of sensory data have been reported to the public
(Hecke et al., 1995; Roudaut et al., 2002). The crispness of dry crisp
products is controlled by product composition and structure
(Roudaut et al., 2002). Process conditions affect the nal moisture
content which governs crispness of the nished product (Roudaut
et al., 2002). During storage, water adsorption from the atmosphere or by mass diffusion from neighboring components can also
cause a loss of crispness (Nicholls et al., 1995).
Moisture-sensitive products may absorb moisture during longterm storage, as the commonly used packaging materials are permeable to moisture. Moisture content can be used as the critical
data for judging the quality of products that have been degraded
by moisture. Water vapor permeability of packaging materials is
one of the important criteria for predicting the rate of moisture uptake (Chen and Li, 2003). Recently there has been increased interest in the development of a mathematical model for optimization
of exible lm packaging of moisture-sensitive foods (Del Nobile
et al., 2003; Azanha and Faria, 2005; Araromi et al., 2008; Siripatrawan, 2009).
This study is aimed at: (1) investigating the moisture sorption
kinetics and empirically modeling the moisture sorption isotherm
of cassava-our-based baked product; (2) determining a critical
water activity of cassava-our-based baked product based on
mechanical and sensory approaches; and (3) determining the activation energy for water vapor permeability of polyolen lms, and
applying this to the predicted shelf life of moisture-sensitive food
products.

2. Materials and methods


2.1. Sample preparation
A cassava-our-based baked sample was prepared using cassava our (55.2%) (Cho Heng Rice Vermicelli Co., Ltd., Nakhon
Pathom, Thailand); coconut milk (18.4%); egg yolk (1.1%); and sucrose (23%), obtained from various commercial retailers. Firstly, a
mixture of coconut milk and sucrose was heated at 90 C until
40% sample weight loss was reached. The obtained mixture, with
egg yolk and cassava our then added, was kneaded into dough
using a domestic mixer (KM 410, Kenwood Limited, UK) at a minimum speed. The dough was stored in a tightly sealed container at
room temperature overnight. Then the dough, after adding water
(2.3%), was kneaded to obtain homogeneous distribution before
being divided roll dough into small balls (1 cm dia) using a

Mc
Mt
P
psat

Dp
RH0
RH
T
Tg,m
t

critical moisture content


moisture after time (t)
water vapor permeability coefcient (g mil m2 d1
mmHg1)
saturated vapor pressure at constant temperature,
mmHg
vapor pressure difference (mmHg)
relative humidity in test dish
relative humidity in desiccator
temperature (K)
mid-point glass transition temperature (C)
time (d)

1 cm plain biscuit cutter. The balls were placed on a greased pan


and baked at 150 C for 20 min. After baking, they became porous
and expanded to 1.5 cm dia. The baked products were left to cool,
and kept in a tightly sealed container for further use.
2.2. Proximate analysis
The sample was analyzed for moisture, protein, carbohydrate,
starch, fat, ash and ber using AOAC methods (Lane, 1998). All
determinations were carried out in triplicate.
2.3. Moisture sorption kinetics and isotherm
A standard gravimetric methodology (weighing samples equilibrated in thermally stabilized desiccators) was used for determination of the adsorption kinetics. The baked product was crushed,
and completely dried in a vacuum oven at 70 C and 76 mm Hg
for 48 h, and then in a desiccator over P2O5 for 2 weeks. The dried
samples (in triplicate) were placed into desiccators with saturated
salt solutions at 30 C. The salt solutions included LiCl, MgCl2,
Mg(NO3)2, NaCl, and K2NO3 of known relative humidity (% RH):
11.3, 32.4, 51.4, 75.1, and 92.5% RH, respectively (Greenspan,
1977). Weights of samples as a function of time were measured;
moisture content was then measured by drying in an oven at
105 C for 3 h (Lane, 1998). Set of experiments was performed in
two replications. This was expressed on a dry-weight basis as g
H2O/100 g dry sample. Water activity (aw) was determined using
a water activity instrument (Testo 650, Testo, Inc., Germany). Moisture adsorption curves of the samples were tted to a mathematical model suggested by Peleg (1988):

Mt M 0 t=k1 k2 t;

where, Mt = moisture after time t; M0 = initial moisture; and k1 and


k2 = parameters.
A standard gravimetric methodology was used for determination of the adsorption isotherms. The baked product was prepared
and conditioned, as described in Section 2.3. The dried samples in
triplicate were equilibrated over saturated salt solutions inside
desiccators at 30 C for 4 weeks. The salt solutions included LiCl,
CH3COOK, MgCl2, K2CO3, Mg(NO3)2, KI, NaCl, KCl and K2NO3 of
known relative humidity (% RH): 11.3, 21.6, 32.4, 43.2, 51.4, 67.9,
75.1, 83.6 and 92.5, respectively (Greenspan, 1977). Moisture content was then measured by drying in an oven at 105 C for 3 h
(Lane, 1998). Set of experiments was performed in 4 replications.
This was expressed on a dry-weight basis as g H2O/100 g dry
sample. Water activity was determined using a water activity
instrument.

R. Kulchan et al. / Journal of Food Engineering 100 (2010) 461467

2.4. Moisture sorption isotherm modeling


Isotherm models from the literature (Berg and Bruin, 1981; Toledo, 1991) were selected for modeling the experimental data of
adsorption isotherms of cassava-our-based baked samples. Those
models are expressed and rearranged as given in Table 1. The
parameters of the equations were estimated using Kyplot 2.0 for
Windows (Kyence Inc., Japan). The value of the root mean square
percentage error (%RMS) represents the tting ability of a model
in association with the number of data points.
2.5. Determination of critical water activity of cassava-our-based
baked product based on mechanical and sensory approaches
2.5.1. Sensory evaluation
Twelve panelists were trained to ensure the same perception of
crispness attribute, as dened by Duizer et al. (1998). Crispness
was rated on a nine-point category scale (1 = not crisp/soggy,
9 = very crisp). The panelists evaluated the samples in a random order, three times over three sessions. A three-way variance analysis
products, panelists and replications (block factor) ensured no
interaction between products and panelists. This ensured a greater
understandability and homogeneity of panelists results regarding
the crispness evaluation.
2.5.2. Mechanical measurement
Mechanical measurement was performed with a texture analyzer (LLOYD Instrument TM LRX S/N 10313, Lloyd Instruments
Ltd., UK). A sample was placed on top of the lower hollow cylinder.
A at cylindrical plunger (4.77 mm dia) was set to a crosshead
speed of 10 cm/min and a load of 50 kgf. Force and deformation
data were recorded. Each sample was measured in 1520 replicates. Hardness (kgf) was dened as the maximum force at the
breaking point of the product, and work (kgf.mm) as the integral
area under the force and deformation curve (Li et al., 1998).
2.6. Determination of activation energy for water vapor permeability
of LDPE and OPP lms
Water vapor permeability (WVP) was measured gravimetrically, according to ASTM Standard Method E 9519. The test dish
was lled with desiccant within 6 mm of the specimen. The specimen was then attached to the dish, and the edges of the specimen
sealed with melted parafn wax to prevent the passage of vapor
into, out of, or around the specimen edges. Three test dishes were
used per sample. Each was weighed at once, placed in a separate
desiccator over saturated salt solution having known relative
humidity of 90 2% RH and conditioned in a temperaturecontrolled chamber at 5 different temperatures (20 1 C,
25 1 C, 30 1 C, 35 1 C and 38 1 C). Test dishes were

Table 1
Models describing the moisture sorption isotherms of cassava-our-based baked
product.
Model

Mathematical expression

BET (BrunauerEmmettTeller)
Brunauer et al. (1938)
GAB (GuggenheimAndersonde
Boer) Berg and Bruin (1981)
Lewicki Lewicki (1998)

me mo C B aw =1  aw 1 C B  1aw 

Oswin Oswin (1946)


Peleg Peleg (1993)

me mo C G kaw =1  aw 1 C G  1kaw 
me F=1  aw G   F=1 aH
w 
me kaw =1  aw c
me aabw cadw

me, Equilibrium moisture content; mo, monolayer moisture content; a, b, c, CB, CG, d,
F, G, H, k, constants specic to individual mathematical expression.

463

weighed periodically. The relationship of gain weight and time


were plotted, with WVTR calculated as follows:

WVTR

 
G
1

t
A

The water vapor permeability coefcient (P) can be expressed


as:

WVTR
 l;
Dp

where: G = weight change (from the straight line), g; t = time, d; G/


t = slope of the straight line, g d1; A = test area (cup mouth area),
m2; WVTR = rate of water vapor transmission, g d1 m2; P = water
vapor permeability, g mil d1 m2 mmHg1; l = thickness, mil; and
Dp = vapor pressure difference, mmHg.
Plotting ln P versus 1/T gives a straight line. The slope of the
straight line represents Ep/R. The Ep of the reaction can now be calculated by multiplying the slope by the gas constant (R).
2.7. Shelf life simulation of moisture-sensitive products
The shelf life simulation of moisture-sensitive products was
developed based on Eq. (10):

Gl
;
AP Dp

where: t = time, d; G = mass of products (dry)  [critical moisture


(Mc)  initial moisture (M0)], g; A = area, m2; l = thickness, mil;
P = permeability coefcient, g mil d1 m2 mmHg1;

Dp psat 

RH0  RH
; mmHg:
100

Shelf life simulation was rendered into two cases. In the rst
case, the water vapor permeability coefcient was in accordance
with a standard condition of storage at 38 C, as a worst-case scenario. In the second case, the water vapor permeability coefcient
was in accordance with the actual condition of storage at 30 C.
This coefcient can be calculated by employing the activation energy for WVP.
2.8. Shelf life determination of moisture-sensitive products
The shelf life of cassava-our-based baked products can be
determined experimentally. About 50 g of samples were packed
in 0.103  0.156 m of 50 lm LDPE and 50 lm OPP pouches. Storage conditions were 30 1 C and 50% RH. The pouches containing
the products were evaluated for moisture content, sensory hardness, and work every 34 days, until products reached their critical
moisture content. The results obtained from analytical and experimental shelf life predictions were compared.
3. Results and discussion
3.1. Food properties
The chemical composition of the cassava-our-based baked
sample was 84.78% 0.09% carbohydrate (approximately 60.08%
starch), 0.63% 0.07% protein, 10.24% 0.09% fat, 3.74% 0.05%
water, 0.32% 0.01% ash and 0.29% 0.03% ber, on a wet basis.
This dry crisp sample was high in starch and fat contents. After
baking and cooling down, the sample was analyzed for aw and
moisture content. It was found that the sample properties were
0.38 and 3.9%, respectively.

464

R. Kulchan et al. / Journal of Food Engineering 100 (2010) 461467

3.2. Moisture sorption kinetics and isotherm


Moisture sorption kinetic curves of cassava-our-based baked
product are depicted in Fig. 1. Moisture adsorption was more rapid
in the initial stages, and declined with increasing time. Then, moisture content of the product reached a plateau, indicating that it became equilibrated with relative humidity in each condition. At
relative humidity of 11.3%, 32.4%, 51.4%, 75.1% and 92.5%, the required times to reach each equilibrium were 34, 60, 76, 100 and
384 h, respectively. Baucour and Daudin (2000) reported that, at
high relative humidity, the mass transfer is very slow, making it
difcult to reach equilibrium in the range 0.91.0 aw. Measured
sorption kinetic curve data were tted to Eq. (1). The constants
k1 and k2, which were derived from the linear t, are shown in
Table 2. The coefcients of determination were found to be high
in all cases (r2 > 0.90); this is an indication of a good t to the
experimental data. Generally, food products stored at a higher
relative humidity tended to have lower k1 and k2 values, and vice
versa. As constants associated with mass transfer and maximum
moisture adsorption capacity, the lower the k1, the higher the initial moisture adsorption rate, and the lower the k2, the higher the
moisture adsorption capacity (Turhan et al., 2002). However in this
case, at 92.5% RH, k1 showed a higher value, which means a lower
degree of initial moisture adsorption rate. This is due to a slow
mass transfer at high relative humidity.
The moisture sorption isotherm curve of cassava-our-based
baked product, represented in Fig. 2, can be classied as a type II
sigmoidal isotherm, which is obtained for soluble materials and
shows an asymptotic trend as water activity tends toward 1 (Bell
and Labuza, 2000). Moisture sorption was more rapid in the initial
stages, and a lesser amount of moisture was adsorbed as adsorption time increased. The higher the relative humidity used, the
more pronounced the effect. The equilibrium moisture content of
the product dramatically soared above aw = 0.73.

Fig. 1. Moisture sorption curves of cassava-our-based baked product at various


relative humidity as a function of time.

Table 2
Sorption kinetic model constants and coefcient of determination for cassava-ourbased baked product at selected relative humidity.
Relative humidity (%)

11.3
32.4
51.4
75.7
92.5

Fig. 2. Moisture sorption isotherm of cassava-our-based baked product.

Calculated model constants, coefcient of determination (r2),


and %RMS for each model for the product are represented in
Table 3. The GAB model is a semi-theoretical multilayer sorption
model with a physical meaning for each constant. In general, it is
the most accepted model for foods or edible materials. The product
presented a monolayer moisture content of 2.46% (dry basis). This
value indicates the maximum amount of water that can be adsorbed in a single layer of the dry product, and is a measure of
the number of sorbing sites. This monolayer moisture content denes the physical and chemical stability of foods. It has an effect on
lipid oxidation, enzyme activity, non-enzymatic browning, avor
preservation, and product structure (Menkov, 2000). This value is
in the range of acceptability because the maximum monolayer
moisture content should not be more than 10% dry basis for food
products (Labuza et al., 1985). Araromi et al. (2008) also reported
that monolayer moisture content decreases with increasing temperature. This is due to a decrease in the number of active sites
for water binding, which may be caused by changes in physical
or chemical structures in the food products as a result of changes
in temperature (Geankoplis, 1993). The trends are in line for
high-carbohydrate foods, as reported by Labuza (1968).
The BET model can be also used to determine the monolayer
water content of a product (2.39%, dry basis). However, this model
is applicable only between aw values of 0 and 0.5 (Bell and Labuza,
2000). The Lewicki model was developed for applicability to a high
range of aw. It ts well with the moisture sorption data at high
humidity, and predicts that water content tends to innity when
aw reaches 1.0. The Oswin model provides good descriptions of
the moisture isotherms throughout the entire range of water

Table 3
Sorption isotherm model constants, coefcient of determination and percentage of
root mean square error for cassava-our-based baked product.
Sorption isotherm model

Constant

r2

BET

m0 = 2.3867
CB = 4.6705
m0 = 2.4598
CG = 49.0599
k = 1.0904
F = 1.6765
G = 1.5727
H = 3.9982
k = 4.3764
c = 1.1409
a = 335.0922
b = 16.6709
c = 13.6947
d = 1.1218

0.9481

6.6812

0.9976

3.1309

0.9888

4.7429

0.9668

13.9072

0.9991

3.6487

GAB

Lewicki

Cassava-our-based baked product


k1

k2

r2

3.32
1.86
1.09
0.64
1.79

0.76
0.30
0.16
0.09
0.02

0.9286
0.9429
0.9391
0.9634
0.9650

Oswin
Peleg

%RMS

R. Kulchan et al. / Journal of Food Engineering 100 (2010) 461467

activity. However in this case, the maximum %RMS value was obtained for the Oswin model. The Peleg model can predict both sigmoid and non-sigmoid isotherms. This model might be tted as
well or better than the GAB model, but unfortunately its constants
had no physical meaning. In the case of the Peleg model, the value
of r2 was highest and was similar to the GAB model. Nevertheless,
the %RMS value from the Peleg model produced a higher result
than that of the GAB model. Thus, the GAB model was found to
be the best estimator for predicting the equilibrium moisture content of the product, followed by the Peleg and Lewicki models. This
is in agreement with Rohvein et al. (2004) and Siripatrawan and
Jantawat (2006) who reported that the GAB model is considered
to be the most versatile sorption model available in the literature,
since it has been shown to t the experimental sorption data for
nearly all products and over the whole water activity range.
Fig. 3 reveals the experimental vs. predicted moisture content of
the product. The obtained points lie on the diagonal for low and
intermediate aw levels, indicating low interaction between components in accordance with their separation in independent phases,

Fig. 3. Comparison between experimental moisture content and those predicted by


various sorption isotherm models for cassava-our-based baked product.

465

as observed during the product baking. At a high level of water


activity, it can also be observed that points fall on the diagonal,
as a result of the interaction between water molecules and the polar groups of the product.
3.3. Relationship between texture and water activity
Crispness of the baked samples was determined using a sensory
approach. Fresh samples were highly crispy, with a score of 7.8,
and were very moisture-sensitive. As the samples adsorbed more
water, the crispness acceptance sharply declined in a linear manner with an increase in water activity (Fig. 4). It was noticed that
the product crispness was preserved to a satisfactory degree
(score P 5) when containing a small amount of water (aw < 0.54)
or moisture content (dry basis < 6%). These specic values of aw
and moisture content (0.54 or 6%) could be considered as critical
points of crispness loss. These critical water activity values corresponded to those of other dry crisp starchprotein-matrix products, reported as approximately 0.5 (Roos et al., 1998; Hough
et al., 2001). The cassava baked sample also contained protein
and starch, which would be anticipated to form such a crispy matrix during baking. This matrix was signicantly softened by the
plasticization effect of water adsorbed above the critical level
(Martinez-Navarrete et al., 2004). It is interesting to point out that
the water activity at the critical level (0.54) is much higher than
that at the monolayer water content (2.5% dry basis or
aw  0.12). As demonstrated in previous work, at the critical water
activity the product was still in a glassy stage (Tg,m  132.8 C)
(Kulchan et al., 2010).
The product texture was also examined using a mechanical approach. The results showed that product hardness and work were
changed in a concave manner with increasing water activity
(Fig. 4). Each of them increased to a maximum point at
aw  0.54. As pointed out above, aw at this value is the critical point
of crispness loss (score = 5). This suggested that the mechanical
hardness and work data at the maximum level could be used to
identify the critical aw of crispness loss in sensory data. The same
incidence was also reported for extruded at bread (Roudaut
et al., 1998).
An increase in water content in the product resulted in a decrease of the glass transition temperature (Kulchan et al., 2010)
as well as crispness over the whole range of aw, but for hardness
and work it occurred only at aw higher than 0.54. Such reductions
are caused by molecular mobility which is facilitated by water

Fig. 4. Relationships between water activity and texture properties and moisture content of the cassava-our-based baked product.

466

R. Kulchan et al. / Journal of Food Engineering 100 (2010) 461467

molecules (Roudaut et al., 2002). However at lower water activity


(from 0.23 to 0.54), hardness and work, on the contrary, sharply increased with an increase in water activity. This is due to the fact
that at such low water activity (such as 0.23, which is close to
0.12 of the monolayer water) the product had only a small number
of water molecules. An addition of water could be just enough to
ll the free volume (at a microscopic level), leading to an increase
in product density (Benczedi 1999; Seow et al., 1999) and interactions among water and other component molecules (Roudaut et al.,
2002). These phenomena cause strong increases in hardness and
work (a stronger puncture force was required). However, such an
increase in water had only a slight impact on glass transition temperature, corresponding to a slight increase in molecular mobility
(Kulchan et al., 2010).
3.4. Activation energy of water vapor permeability of LDPE and OPP
lms
Arrhenius plots of WVPs of LDPE and OPP lms are depicted in
Fig. 5. It was found that WVPs of LDPE and OPP lms are not highly
temperature-dependent. This resulted in low values of activation
energies for WVPs of LDPE and OPP lms: 22.33 and
21.26 kJ mol1, respectively. This low Ep implies that temperature
uctuation during storage does not have a signicant impact on
WVP, and in turn shelf life estimation (and vice versa). Xiong
(2002) reported that activation energy of WVP of LDPE was
22.86 kJ mol1. In addition, George et al. (1997) reported that activation energies for WVPs of LDPE composite lms reinforced with
20% (w/w) untreated and treated dicumyl peroxide (05% by
weight of polymer) pineapple-leaf ber were 23.64 and
22.23 kJ mol1, respectively.
3.5. Shelf life simulation of moisture-sensitive products
Predicted shelf life by the GAB model was simulated using the
WVPs of LDPE and OPP lms at a standard condition (38 C), which
were 0.2785 and 0.0861 g mil d1 m2 mmHg1, respectively.
Predicted shelf life by the GAB model and Ep for WVP were simulated using the WVPs of LDPE and OPP lms at an actual storage
condition (30 C), which were 0.2288 and 0.0687 g mil d1
m2 mmHg1, respectively. The simulated shelf life values of cassava-our-based baked products in all packages are shown in Table
4. It was found that predicted shelf life by the GAB model and Ep for
WVP yields better shelf life estimation, closer to the actual shelf life
as predicted by the GAB model. The differences between the experimental and the predicted shelf life of the product are minute,
especially in the case of the predicted shelf life by the GAB model
and Ep for WVP, which is based solely on the relationship between

Fig. 5. Arrhenius plots of water vapor permeability of LDPE and OPP lms.

Table 4
Comparison between experimental shelf life and predicted shelf life of cassava-ourbased baked product.
Shelf life

Experimental shelf life


Predicted shelf life by GAB model
Predicted shelf life by GAB model
and Ep for WVP

Days

% Difference in
shelf life

LDPE
pouch

OPP
pouch

LDPE
pouch

OPP
pouch

35
30.47
36.83

119
97.85
122.77

12.94
5.23

17.77
3.17

moisture content of the product and the barrier property of the


packaging material, as well as the storage condition. However,
Roca et al. (2006) stated that shelf life of a moisture-sensitive food
product is affected not merely by moisture adsorption but also by
moisture migration in the food product, which is greatly affected
by the complexity of the food structure.
4. Conclusion
The moisture sorption kinetics of cassava-our-based baked
products was more rapid in the initial stages; a lesser amount of
moisture was adsorbed as the adsorption time increased. GAB, Peleg and Lewicki models were useful to t moisture sorption isotherm data of the products. The product hardness or work
reached the maximum and began to reduce at moisture content
(MC) 6%, when the product texture began to be detected as
becoming slightly soft. The predicted shelf lives estimated by
employing Ep for WVP of LDPE and OPP, and the GAB model were
close to the actual shelf lives. Therefore, the estimation by empirical models and activation energy was found to be applicable for
rapid and accurate shelf life prediction.
Acknowledgments
This work was fully supported by a fund for the promotion of
research at the Center of Advanced Studies for Agriculture and
Food (CASAF), Kasetsart University. The authors express their
thanks and appreciation for this support.
References
Al-Muhtaseb, A.H., McMinn, W.A.M., Magee, T.R.A., 2002. Moisture sorption
isotherm characteristics of food products: a review. Transaction of Institution
of Chemical Engineers 80 (C), 118128.
Araromi, D.O., Olu-Arotiowa, O.A., Olajide, J.O., Afolabi, T.J., 2008. Neuro fuzzy
modeling approach for prediction of equilibrium moisture characteristics and
shelf-life of corn our. International Journal of Soft Computing 3 (2), 159166.
Azanha, A.B., Faria, J.A.F., 2005. Use of mathematical models for estimating the
shelf-life of cornakes in exible packaging. Packaging Technology and Science
18 (4), 171178.
Baucour, P., Daudin, J.D., 2000. Development of a new method for fast measurement
of water sorption isotherms in the high humidity range: validation on gelatine
gel. Journal of Food Engineering 44 (2), 97107.
Bell, L.N., Labuza, T.P., 2000. Moisture Sorption: Practical Aspects of Isotherm
Measurement and Use, second ed. American Association of Cereal Chemists,
Inc., St. Paul.
Benczedi, D., 1999. Estimation of the free volume of starchwater barriers. Trends in
Food Science and Technology 10 (1), 2124.
Berg, C., Bruin, S., 1981. Water activity and its estimation in food systems:
theoretical aspects. In: Rockland, L.B., Steward, G.F. (Eds.), Water Activity:
Inuences on Food Quality. Academic Press, New York, pp. 147177.
Brunauer, S., Emmett, P.H., Teller, E., 1938. Adsorption of gases in multimolecular
layers. Journal of the American Chemical Society 60 (2), 309319.
Chang, Y.P., Cilean, P.B., Seow, C.C., 2000. Variations in exural and compressive
fracture behavior of a brittle cellular food (dried bread) in response to moisture
sorption. Journal of Texture Studies 31 (5), 525540.
Chen, Y., Li, Y., 2003. A new model for predicting moisture uptake by packaged solid
pharmaceuticals. International Journal of Pharmaceutics 255 (12), 217225.

R. Kulchan et al. / Journal of Food Engineering 100 (2010) 461467


Chirife, J., Iglesias, H.A., 1978. Equations for tting water sorption isotherms of
foods. Journal of Food Technology 13 (2), 159174.
Del Nobile, M.A., Buonocore, G.G., Limbo, S., Fava, P., 2003. Shelf life prediction of
cereal-based dry foods packed in moisture-sensitive lms. Journal of Food
Science 68 (4), 12921300.
Duizer, L.M., Campanella, O.H., Barnes, G.R.G., 1998. Sensory, instrumental and
acoustic characteristics of extruded snack food products. Journal of Texture
Studies 29 (4), 397411.
Geankoplis, C.J., 1993. Transport Process and Unit Operations, third ed. PrenticeHall, Inc.
George, J., Bhagawan, S.S., Thomas, S., 1997. Effects of environment on the
properties of low-density polyethylene composites reinforced with pineappleleaf bre. Composites Science and Technology 58 (9), 14711485.
Greenspan, L., 1977. Humidity xed points of binary saturated aqueous solutions.
Journal of Research of the National Bureau of Standards 81A (1), 89102.
Hecke, E.V., Allaf, K., Bouvier, J.M., 1995. Texture and structure of crispy-puffed food
products I: mechanical properties in bending. Journal of Texture Studies 26 (1),
1125.
Hough, B., Buera, M.P., Chirife, J., Moro, O., 2001. Sensory texture of commercial
biscuits as a function of water activity. Journal of Texture Studies 32 (1),
5774.
Kulchan, R., Suppakul, P., Boonsupthip, W., 2010. Texture of glassy tapioca-ourbased baked product as a function of moisture content. In: Reid, D.S.,
Sajjaanantakul, T., Lillford, P.J., Charoenrein, S. (Eds.), Water Properties in
Food, Health, Pharmaceutical and Biological Systems: ISOPOW 10. WileyBlackwell, New York.
Labuza, T.P., 1968. Sorption phenomena in foods. Food Technology 23 (1),
1519.
Labuza, T.P., Kaanane, A., Chen, J.Y., 1985. Effect of temperature on the moisture
sorption isotherms and water activity shift of two dehydrated foods. Journal of
Food Science 50 (2), 385389.
Lane, R.H., 1998. Cereal foods. In: AOAC, Ofcial Methods Analysis of AOAC
International, 16th ed. AOAC International, Gaithersburg, pp. 137.
Lewicki, P.P., 1998. A three parameter equation for food moisture sorption
isotherms. Journal of Food Process Engineering 21 (2), 127144.
Li, Y., Kloeppel, M.K., Hsieh, F., 1998. Texture of glassy corn cakes as a function of
moisture content. Journal of Food Science 63 (5), 869872.
Lopez, A.C.B., Pereira, A.J.G., Junqueira, R.G., 2004. Flour mixture of rice our, corn
and cassava starch in the production of gluten-free white bread. Brazilian
Archives of Biology and Technology 47 (1), 6370.
Martinez-Navarrete, N., Moragu, G., Talens, P., Chiralt, A., 2004. Water sorption and
the plasticization effect in wafers. International Journal of Food Science and
Technology 39 (5), 555562.
Menkov, N.D., 2000. Moisture sorption isotherms of vetch seeds at four
temperatures. Journal of Agricultural Engineering Research 76 (4), 373380.

467

Mohamed, S., Abdullah, N., Muthu, M.K., 2006. Physical properties of keropok (fried
crisps) in relation to the amylopectin content of the starch ours. Journal of the
Science of Food and Agriculture 49 (3), 369377.
Nicholls, R.J., Appelqvist, I.A.M., Davies, A.P., Ingman, S.J., Lillford, P.J., 1995. Glass
transition and the fracture behavior of gluten and starches within the glassy
state. Journal of Cereal Science 21 (1), 2536.
Oswin, C.R., 1946. The kinetics of package life. III. The isotherm. Journal of the
Society of Chemical Industry 65 (4), 419423.
Peleg, M., 1988. An empirical model for the description of moisture sorption curves.
Journal of Food Science 53 (4). 12161217 and 1219.
Peleg, M., 1993. Assessment of a semi-empirical 4 parameter general model for
sigmoid moisture sorption isotherms. Journal of Food Process Engineering 16
(1), 2137.
Roca, E., Guillard, V., Guilbert, S., Gontard, N., 2006. Moisture migration in a cereal
composite food at high water activity: effects of initial porosity and fat content.
Journal of Cereal Science 43 (2), 144151.
Rohm, H., 1990. Consumer awareness of food texture in Austria. Journal of Texture
Studies 21 (3), 363373.
Rohvein, C., Santalla, E., Gely, M.C., 2004. Estimation of sorption isotherm and the
heat of sorption of quinoa (Chenopodium quinoa Wild.) seeds. Food Science and
Technology International 10 (6), 409413.
Roos, Y., Roininen, K., Jouppila, K., Tuorila, H., 1998. Glass transition and water
plasticization effects on crispness of a snack food extrudate. International
Journal of Food Properties 1 (2), 163180.
Roudaut, G., Dacremont, C., Le Meste, M., 1998. Inuence of water on the crispness
of cereal-based foods: acoustic, mechanical, and sensory studies. Journal of
Texture Studies 29 (2), 199213.
Roudaut, G., Dacremont, C., Valles Pamies, B., Colas, B., Le Meste, M., 2002.
Crispness: a critical review on sensory and material science approaches. Trends
in Food Science and Technology 13 (67), 217227.
Seow, C., Cheah, P.B., Chang, Y.P., 1999. Antiplasticization by water in reducedmoisture food systems. Journal of Food Science 64 (4), 576581.
Siripatrawan, U., 2009. Shelf-life simulation of packaged rice crackers. Journal of
Food Quality 32 (2), 224239.
Siripatrawan, U., Jantawat, P., 2006. Determination of moisture sorption isotherms
of jasmine rice crackers using BET and GAB models. Food Science and
Technology International 12 (6), 459465.
Toledo, R.T., 1991. Fundamentals of Food Processing Engineering, second ed. Van
Nostrand Reinhold, New York.
Turhan, M., Sayar, S., Gunasekaran, S., 2002. Application of Peleg model to study
water absorption in chickpea during soaking. Journal of Food Engineering 53
(2), 153159.
Xiong, L., 2002. Determination and prediction of shelf life of moisture-sensitive

Post shredded wheat cereal. School of Packaging, Michigan State University 12,
7794.

Anda mungkin juga menyukai