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
462
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
Mc
Mt
P
psat
Dp
RH0
RH
T
Tg,m
t
Mt M 0 t=k1 k2 t;
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
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
WVTR
G
1
t
A
WVTR
l;
Dp
Gl
;
AP Dp
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
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
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
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
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,
465
Fig. 4. Relationships between water activity and texture properties and moisture content of the cassava-our-based baked product.
466
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
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
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.