This article was downloaded by: [Eindhoven Technical University]

On: 17 November 2014, At: 07:58

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK

Drying Technology: An
International Journal
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/ldrt20

Experimental Determination
and Modeling of Sorption
Isotherms of Tropical Fruits:
Banana, Mango, and Pineapple
a c b
André Talla , Yves Jannot , George Elambo
c b
Nkeng & Jean-Rodolphe Puiggali
a
LAEN-ENSP , Yaoundé, Cameroun
b
Laboratoire TREFLE , Talence Cedex, France
c
ENSTP , Yaoundé, Cameroun
Published online: 06 Feb 2007.

To cite this article: André Talla , Yves Jannot , George Elambo Nkeng & Jean-
Rodolphe Puiggali (2005) Experimental Determination and Modeling of Sorption
Isotherms of Tropical Fruits: Banana, Mango, and Pineapple, Drying Technology: An
International Journal, 23:7, 1477-1498, DOI: 10.1081/DRT-200063530

To link to this article: http://dx.doi.org/10.1081/DRT-200063530

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
 losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the
Content.

This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014
 Drying Technology, 23: 1477–1498, 2005
Copyright Q 2005 Taylor & Francis, Inc.
ISSN: 0737-3937 print/1532-2300 online
DOI: 10.1081/DRT-200063530

Experimental Determination and Modeling

of Sorption Isotherms of Tropical Fruits:
Banana, Mango, and Pineapple
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

André Talla1,3, Yves Jannot
2, George Elambo Nkeng3,

and Jean-Rodolphe Puiggali2
1
LAEN-ENSP, Yaoundé, Cameroun
2
Laboratoire TREFLE, Talence Cedex, France
3
ENSTP, Yaoundé, Cameroun

Abstract: Sorption isotherms of banana, mango, and pineapple have been

experimentally found at 40, 50, and 60C by use of the salt method for a range
of water activities from 0.056 to 0.85. The sorption capacity of these fruits
increases with the temperature for a given water activity. The experimental curves
have been simulated by the GAB and BET models. The BET model ensures a
better representation of the experimental results for water activity lower than
0.35 with maximum deviation of 1.0, 1.5, and 2.0 kgwkgdm 1 for respectively
banana, mango and pineapple. On the contrary, the GAB model enables the
representation of the whole desorption isotherms and the estimation of the water
content corresponding to monolayer saturation and of the isosteric heat of sorp-
tion. It also takes into account the temperature effect. The isosteric sorption heat
of the three studied products are deduced from experimental results and empirical
correlations are proposed leading to a satisfactorily representation.

Keywords: Banana; Mango; Ananas; Sorption isotherm; Isosteric heat; GAB;

BET; Specific area

INTRODUCTION

Water activity aw of a product essentially depends on its water content X

and on its temperature h. When the water in a product is in balance with
the water vapor present in the surrounding air, the product temperature
is equal to the air temperature (no heat transfer) and the water vapor


Correspondence: Yves Jannot, Laboratoire TREFLE, Esplanade des Arts et
Métiers, 33405 Talence Cedex, France; Tel.: þ33 (0) 556 845 422; Fax: þ33(0)
556 845 436; E-mail: [email protected]
 1478 Talla et al.

pressure in the product is equal to the water vapor pressure of the air (no
mass tranfer): Water activity of the product is thus equal to air relative
humidity. The curve representing, for a temperature h, the water content
Xeq of a product in equilibrium as a function of its water activity aw (or of
air relative humidity HR) is called the sorption isotherm. The desorption
isotherms are obtained when the initial product is wet and reaches the
equilibrium with air by losing water; the (ad)sorption isotherms are
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

obtained when the initial product is dry and reaches the equilibrium with
air by adsorbing water.
The sorption isotherm is particularly important when considering
convective drying of a product by a hot air flow. It enables the calcu-
lation of the equilibrium water content Xeq, the limit of the product water
content at the end of the drying. The equilibrium water content Xeq is a
parameter that appears in models predicting the product water content
evolution during its drying.
Furthermore, the analysis of the processes influencing physical,
biochemical, and microbiological stability of a product (that determines
its quality) is mostly dependent on the sorption isotherm. The sorption
isotherm also gives information on the sorption mechanism and on the
interaction between adsorbate and adsorbent as pointed out by Bizot
and coworkers.[1] The equilibrium water content of a product is one of
the most important parameter to predict its behavior during storage.[2,3]
Nevertheless, it has been shown by Yang and Atallah[4] that the drying
method could have an effect on the characteristics of a dried product such
as: sorption and desorption isotherms, porosity, specific area, and color.
Several authors[5–18] have published results about sorption isotherms
studies of various products, but very little information is available about
sorption isotherms of tropical fruits, except that presented by Talla
et al.[19]
This article presents a study on the determination of sorption
and desorption isotherms of three tropical fruits: banana, mango, and
pineapple. The variety of mango used in our tests is Mangifera indica;
there exist around 300 varieties of mango, which can be quite different
from one to another.

MATHEMATICAL MODELS

Sorption isotherms of a material have been described by various math-

ematical models with two or more parameters as presented by Van der
Berg and Bruin.[20] Nevertheless, models with more than three parameters
lead to uneasy use and physical interpretation. The model presented by
Brunauer et al.[5] known as BET model, and its modified version from
Guggenhein et al. (GAB) as described by Bizot[1] have been successfully
used by many authors[10–14,16,17] for modeling sorption isotherms. These
 Tropical Fruit Isotherm Modeling 1479

models include parameters that have physical meaning. Moreover, the

GAB model takes into account the temperature effect.

BET Model

This model allows the study of the phenomenon of water molecules

Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

adsorption and desorption. Its mathematical expression is as follows:

" #
Xm baw 1  ðn þ 1Þanw þ nanþ1
w
Xeq ¼ ð1Þ
1  aw 1 þ ðb  1Þaw  banþ1w

where

Xm ¼ water content corresponding to monolayer kgw  kgdm 1);

b ¼ constant; and
n ¼ number of water molecules layers.

If the monolayer is not saturated, a simplified expression is:

" #
Xm baw 1  2aw þ a2w
Xeq ¼ ð2Þ
1  aw 1 þ ðb  1Þaw  ba2w

GAB Model

The GAB model enables a representation of sorption isotherms for

the whole values of water activity aw. Its mathematical expression is as
follows:
Xm CKaw
Xeq ¼ ð3Þ
ð1  Kaw Þ½1 þ ðC  1ÞKaw 
where

Xeq ¼ equilibrium water content (kgwkgdm 1);

Xm ¼ water content corresponding to monolayer saturation (kgw 
kgdm 1);
aw ¼ product water activity;
C ¼ constant linked to monolayer sorption heat; and
K ¼ constant linked to monolayer sorption heat.



H1  Hm H1  Hq
C ¼ C0 exp and K ¼ K0 exp ð4Þ
RT RT
 1480 Talla et al.

with H1, Hm, Hq, respectively, condensation heat of pure water, total
sorption heat of the monolayer, and total sorption heat of the other
layers
where

T ¼ absolute temperature (K); and

R ¼ perfect gas constant (8,314 J mol 1K 1).
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

With the hypothesis of an initial homogenous monolayer filling of

the product surface, one can calculate its area. Supposing that the mono-
layer is recovered with aligned water molecules, it could easily be shown
that its area can be calculated by a formula.[5] The value of the specific
area of a product may give information about its rehydration and its
internal structure.
pffiffiffip ffiffiffi

332 N 1=3
Sm ¼ qdm Xm ð5Þ
2 Mq2w
where

Sm ¼ monolayer area or specific area (m2  m 3);

N ¼ Avogadro number ¼ 6.023  1023;
M ¼ absorbing material molecular mass (kg  mol 1);
qw ¼ water density (kg  m 3); and
qdm ¼ solid matrix density (kgdm  m 3).

Isosteric Sorption Heat

A heat that equals to the sum of phase change latent heat Lv and of sorp-
tion heat Qsorp is necessary to evaporate one kilogram of water. The
desorption heat may be calculated from the curves representing, for a
constant water activity, product activity as a function of temperature,
these curves are called the (water) sorption isosters. The equation of these
curves is given by the Clausius-Clapeyron formula:



Qsorp 1
lnðaw Þ ¼  þ Const ð6Þ
R T
In a two-dimensional diagram where ln(aw) is represented as a function of
1=T, the sorption isosters are straight lines according to the relation:
!
@ lnðaw Þ Qsorp
  ¼ ð7Þ
@ T1 R
X¼Const

The isosteric sorption heat can be deduced from the slope of the
sorption isosters representing ln(aw) as a function of 1=T for each product
 Tropical Fruit Isotherm Modeling 1481

water content X, and then the curve Qsorp ¼ f(X) can be plotted. It is an
important parameter whose values show that at the end of the drying it
cannot be neglected in energy balance since it has the same order of mag-
nitude as the vaporization latent heat.

MATERIAL AND METHOD

Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Sorption Method

The static method of the saturated salts solutions has been used for
sorption isotherms determination. It is a method where diffusion is the
only way of mass transfer between the tested product and the surround-
ing air. The relative air humidity is fixed by contact with saturated salts
solution whose water vapor pressure at a given temperature is perfectly
known. This method is commonly used for product sorption isotherms
determination.[11,12,17,21,22]
The time necessary to reach equilibrium is quite long compared to
the dynamic method since the diffusion rate is a limitating factor.

Experimental Process

The first step is the choice of the salts so that a large interval of water
activities could be obtained. For our experiments, nine salts have been
used that covered the range 0.056 to 0.85 for water activities; these
include: KOH, LiCl, KCH3CO2, MgCl2, K2CO3, NaBr, CuCl2, NaCl,
and KCl. The corresponding values of water activities for all these salts
at the various experimental temperatures used are presented in Table 1
according to Bizot and Multon.[23]
Three samples of the tested product (leading to an average value) are
then set in each of the nine recipients containing the saturated salts solu-
tions. Figure 1 shows the experimental facility. The recipients are set in a
temperature regulated chamber. For adsorption isotherms determination,
the samples were previously dehydrated at 60C. For desorption
isotherms, fresh products without any preparation were used.
Then, the samples were weighed at regular time intervals until the
mass becomes constant (variation less than 1 mg by 48 h) and they could
be considered as in equilibrium with air at (h, HR). The mass of a sample
when equilibrium is reached with its bone dry mass measured after dehy-
dration for 48 h at 102C enables the calculation of the equilibrium water
content by applying the following formula:


meq
Xeq ¼ 100 1 ð8Þ
md
 1482 Talla et al.

Table 1. Air relative humidities imposed by saturated salt solutions

for several temperatures[23]

Relative humidity

Salts 25C 40C 50C 60C

KOH 8.2 6.3 5.7 5.6

LiCl 11.3 11.2 11.4 11.0
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

KCH3CO2 22.5 21.6 21.6 21.6

MgCl2 32.8 31.6 30.5 29.3
K2CO3 43.2 42.2 45.6 45.0
NaBr 57.6 53.2 50.9 49.7
CuCl2 68.5 68.0 65.5 63.3
NaCl 75.3 74.7 74.4 74.5
KCl 84.3 82.3 81.2 80.3

where

meq ¼ sample equilibrium mass (kg); and

md ¼ sample bone dry mass (kg  dm).

The sorption isotherm is given by the experimental couples (aw, Xeq).

The experimental procedure consists of introducing the recipients in the
chamber regulated to the lowest temperature value of the series (40C),

Figure 1. Experimental device for sorption isotherm determination.

 Tropical Fruit Isotherm Modeling 1483

weighing all the samples when the equilibrium was reached, and then
raising the temperature to the next value (50C). This procedure was
repeated for 60C. During the experiments, a partial vacuum was applied
inside the recipients to increase the dehydration rate.

RESULTS AND DISCUSSIONS

Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figures 2a to 4a represent the experimental desorption isotherms of

banana, mango, and pineapple obtained at different temperatures. The
isotherms are sygmoidal as usually presented for foodstuff as done by
Iglesias and Chirife.[7] Meanwhile, it can be observed for banana that
these ‘‘static’’ equilibrium water contents are greater by a few percent
than the ‘‘dynamic’’ equilibrium water contents reported by Talla
et al.[19] This result is in agreement with the conclusions of Nadeau and
Puiggali[24] related to the comparison between the static and dynamic
methods for desorption isotherms determination. For a given value of
the water activity, the equilibrium water content decreases with the
temperature. Furthermore, it is not possible to obtain experimental
couples (Xeq, aw) for aw > 0.9 with the static saturated salts method
since the produts (fruits) are decomposed after a few days before the
thermodynamic equilibrium being reached.
The constants Xm, C, and K of the GAB model depends on product
characteristics and temperature, similar to the constants Xm and b of the
BET model. All these constants are estimated from experimental results.
For each temperature, these parameters are estimated by an optimization
method minimizing the sum S of the quadratic errors calculated by
formula[9] between the experimental equilibrium water contents and the
values calculated with relation.[3]
!2
X n
Xexp  Xmod
S¼ ð9Þ
i¼1 Xexp
i

where n is the number of measurements for a temperature (nine in the

present case).
Contrarily to the GAB model which represents the sorption isotherm
on a wide range of aw values, the BET model enables the representation
of the sorption isotherms only in the range of activities from 0 to 0.35
according to Brunauer et al.[5] Table 2 lists the whole results of the esti-
mation. Furthermore, Figs. 2a to 4a and 5a to 7a present a comparison
between the experimental curves and the curves simulated by the GAB
model and by the BET model, respectively.
The representation of the residues on Figs. 2b to 4b shows higher
deviation for high values of activities with the GAB model. The
maximum deviation for the whole products is lower than 8% for a water
 1484 Talla et al.
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 2. (a) Desorption isotherms of banana at several temperatures: simulation

by the GAB model and experimental results; (b) Residues between experimental
values of Xeq and values calculated by the GAB model for banana.

activity of 0.85. Nevertheless, the deviation is quite low for water activi-
ties lower than 0.35 and are thus interesting for drying modeling since the
literature recommends a final water content between 0.15 and 0.20 for a
 Tropical Fruit Isotherm Modeling 1485
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 3. (a) Desorption isotherms of mango at several temperatures: simulation

by the GAB model and experimental results; (b) Residues between experimental
values of Xeq and values calculated by the GAB model for mango.
 1486 Talla et al.
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 4. (a) Desorption isotherms of pineapple at several temperatures: simula-

tion by the GAB model and experimental results; (b) Residues between experi-
mental values of Xeq and values calculated by the GAB model for pineapple.
 Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Table 2. Estimated values of the parameters of the GAB and BET models

GAB model BET model

Product h (C) Xm (kgw  kgdm 1) C K Xm (kgw  kgdm 1) b Sm (m2  m 3) X 0 qs=Sm (kgw  m 2)

Banana X 0 ¼ 3:77 25 0.127 9295.7 0.979 0.181 74.49 8.91E þ 08 5.92E06

(kgw  kgdm 1) 40 0.108 6531.4 0.993 0.160 49.23 7.88E þ 08 6.70E06
50 0.095 3420.9 1.009 0.147 35.66 7.24E þ 08 7.29E06
60 0.083 1828.3 1.011 0.128 34.76 6.30E þ 08 8.38E06
Mango X 0 ¼ 6:83 40 0.096 4798.6 1.054 0.129 94.30 6.35E þ 08 1.51E05
(kgw  kgdm 1) 50 0.080 2954.5 1.082 0.110 66.08 5.41E þ 08 1.77E05
60 0.065 1104.0 1.111 0.094 45.87 4.63E þ 08 2.07E05
Pineapple X 0 ¼ 6:15 40 0.185 83.292 0.934 0.266 24.45 1.31E þ 09 6.58E06
(kgw  kgdm 1) 50 0.164 57.285 0.937 0.234 21.94 1.15E þ 09 7.47E06
60 0.143 45.769 0.938 0.199 20.64 9.80E þ 08 8.79E06

1487
 1488 Talla et al.
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 5. (a) Desorption isotherms of banana at several temperatures: simulation

by the BET model and experimental results; (b) Residues between experimental
values of Xeq and values calculated by the BET model for banana.
 Tropical Fruit Isotherm Modeling 1489
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 6. (a) Desorption isotherms of mango at several temperatures: simulation

by the BET model and experimental results; (b) Residues between experimental
values of Xeq and values calculated by the BET model for mango.

good conservation.[24] For water activities lower than 0.35, the BET
model fits better of the experimental results as shown on Figs. 5b to 7b
representing the residues. In this range of water activities, the maximum
 1490 Talla et al.
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 7. (a) Desorption isotherms of pineapple at several temperatures:

simulation by the BET model and experimental results; (b) Residues between
experimental values of Xeq and values calculated by the BET model for pineapple.

observed deviation is 1.0, 1.5, and 2.0%, respectively, for banana, mango,
and pineapple. The relatively low deviation leads to the conclusion that
these two models represent the desorption isotherm quite satisfactorily.
 Tropical Fruit Isotherm Modeling 1491

According to experimental results, the estimated parameters Xm and

C decrease when the temperature rise since the parameter K increases
with temperature (Table 2). This is in agreement with the reported beha-
vior of other foodstuff.[26–29] The sensitivity of the GAB model to the
parameters Xm, C, and K is deduced from Eqs. (10) to (12) and repre-
sented in Fig. 8 for banana. This figure shows that the GAB model is
quite insensitive to the parameter C: a relative variation of 10% on the
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

parameter C has no effect on the simulated values of Xeq, which is also

verified for mango and pineapple. This confirms, as pointed out by
Rouquerol et al.,[30] that the simplified theoretical expression of the
parameter C given by formula (5) cannot lead to a reliable estimation
of the total sorption heat Hm of the monolayer.
e @Xeq e
Xm ¼ Xeq ð10Þ
100 @Xm 100
 
e @Xeq e CKaw
C ¼ Xeq 1  ð11Þ
100 @C 100 1 þ ðC  1ÞKaw

 
e @Xeq e Kaw 1
K ¼ Xeq þ ð12Þ
100 @K 100 1  Kaw 1 þ ðC  1ÞKaw

Figure 8. Variation of Xeq induced by a relative variation of 10% of each para-

meter.
 1492 Talla et al.

These three formulas represent the variation of the estimated value of Xeq
indices by a relative variation of e% of respectively the parameters Xm, C,
and K from their nominal values.
The parameter Xm is the water content corresponding to the mono-
layer saturation. Table 2 shows that the values of Xm estimated from the
BET model are always greater than the values obtained from the GAB
model. This difference is linked to the fitting of the model with the experi-
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

mental curve specially in the area corresponding to the monolayer

saturation (water activity lower than 0.2). The values obtained from
the BET model which better fits the experimental values in this area must
be more reliable than those obtained with the GAB model.
Furthermore, it can be observed that the specific area (calculated
from the BET model) is not linked to the initial product water content,
which is rather linked to the porosity (Table 2). The specific area is
directly proportional to the water content at the saturation of the mono-
layer. Table 2 also indicates the values of X0qs=Sm, which represents, at
the initial state, the water content filling the pores by unit of specific area
and so has the dimension of a water thickness. The density of the solid
phase is deduced from the results presented by Lozano,[31] who has mea-
sured the density of several fruits and vegetables as a function of their
water content. For a zero water content, the density varies from 1300
to 1550 kg  m 3; a mean value of 1400 kg  m 3 has been used in our
calculation. Table 2 shows a lower value for mango than for banana
and pineapple, which may indicate that mango will be easier to dry; this
prediction is in agreement with the results of Talla et al.[19] On the other-
hand, according to the values in Table 2 for banana and pineapple, these
two products might have a rather similar behavior during drying.
The parameters C0 and H1  Hm have not been estimated since it will
not be reliable as previously pointed out. Table 3 presents the estimated
values of K0 and Hq  H1. The values of Hq  H1 represent the mean
value of sorption heat of the studied fruits, which is greater than the
vaporization heat of water.
Figures 9 to 11 represent the sorption isosters ln(aw) vs. 1=T that
must be straightlines according to formula (6), for, respectively banana,
mango, and pineapple. For each water content X, a sorption isoster has
been plotted and the isosteric sorption heat was calculated from its

Table 3. Estimated parameters for K calculation

Product K0 H1  Hq (kJ  mol 1)

Banana 1.359 97.74

Mango 2.533 274.63
Pineapple 1.003 22.04
 Tropical Fruit Isotherm Modeling 1493
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 9. Sorption isosters for banana.

slope. For each product, the representation of Qsorp vs. X has then
been plotted on Fig. 12. It can be observed that the isosteric sorption
heat is high for the low water content, indicating the strong link between
the adsorbate (water) and the adsorbent, but it is quite negligible
compared to vaporization latent heat for high water content. The same

Figure 10. Sorption isosters for mango.

 1494 Talla et al.
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Figure 11. Sorption isosters for pineapple.

procedure has been repeated for mango and pineapple and all the
experimental results have been correlated satisfactorily by relations
(13) to (15).

Figure 12. Sorption heat of banana, mango, and pineapple vs. water content.
 Tropical Fruit Isotherm Modeling 1495

6:822
722:3 þ 0:021X
Banana : Qsorp ¼ 1:028
ðkJ  kgw1 Þ ð13Þ
6:14 þ X
0:370
44:6 þ 90:8X
Mango : Qsorp ¼ 1:807
ðkJ  kgw1 Þ ð14Þ
0:03 þ X
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

3:653
82:49  0:320X
Pineapple : Qsorp ¼ 0:542
ðkJ  kgw 1Þ ð15Þ
0:433 þ X

CONCLUSION

In this study, the desorption isotherms of banana, mango and pineapple

have been experimentally established and then modeled by BET and
GAB models. The BET model enables the calculation of the monolayer
water content leading to the value of the specific area for different tem-
peratures. A comparaison between BET and GAB models shows that
the latter one gives a better fitting with experimental results for water
activities lower than 0.35. Nevertheless, the maximal deviation between
experimental results and calculated values of the water content X by
the GAB model is around 0.08 kgw  kgdm 1 for X ¼ 0.85, which remains
acceptable. This model also leads to the calculation of the isosteric
desorption heat of each fruit.
Furthermore, the curves giving the isosteric desorption Qsorp heat vs.
water content X have been deduced from the experimental desorption
isotherms. The high values observed for low water content indicates
the strong link between the adsorbate (water) and the adsorbent. Empiri-
cal relations Qsorp ¼ f(X) have been established for each product.

NOMENCLATURE

aw Product water activity

C Parameter linked to monolayer heat of sorption
K Parameter linked to multilayers heat of sorption
m Product mass (kg)
Q Sorption heat (kJ  mol 1)
S Specific area (m2  m 3)
X Dry basis water content (kgw  kgdm 1)
X Mean dry basis water content (kgw  kgdm 1)

Greek Symbols
q Density (kg  m 3)
h Air temperature (C)
 1496 Talla et al.

Subscripts
0 Initial
w Water
eq Equilibrium
i Sample number
m Monolayer
n Number of layers of water molecules
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

p Number of measurements for a sorption isotherm

dm Dry matter
sorp Sorption
exp Experimental
mod Model

ACKNOWLEDGMENTS

The authors gratefully acknowledge the National High School of Public

Works (NHSPW) of Yaounde, the Agence Universitaire de la Franco-
phonie (AUF), and the Laboratoire Énergétique et Phénomènes de
Transfert (LEPT-ENSAM) of the École Nationale Supérieure des Arts
et Métiers of Bordeaux for their financial support.

REFERENCES

1. Bizot, H.; Riou, N.; Multon, J.L. Guide pratique pour la détermination des
isothermes de sorption et de l’activité de l’eau. Science des Aliments Hors
Série, 1987,
2. Chirife, J.; Buera, M.P. Water activity, glass transition and micribial stability
in concentrated semi-moist food systems. Journal of Food Science 1994,
59, 921–927.
3. Vilades, S.L.; Malee, L.F.; Gerchenson, L.N.; Alzamora, S.M. Water
sorption characteristics of sugar impregnated strawberries. Drying
Technology 1995, 13, 1993–2010.
4. Yang, C.S.T.; Atallah, W.A. Effect of four drying methods on the quality of
intermediate moisture lowbush blueberies. Journal of Food Science 1985, 50,
1233–1237.
5. Brunauer, S.; Emmet, P.H.; Teller, E. Adsorption of gases in multimolecular
layers. Journal of the American Chemical Society 1938, 60, 309–319.
6. Henderson, S.M. A basic concept of equilibrium moisture. Agricultural
Engineering St Joseph 1952, 32, 29–32.
7. Iglesias, H.A.; Chirife, J. Handbook of Food Isotherms Water Sorption. Para-
meters for Food and Food Components; Academic Press: New York, 1982.
8. Ajibola, O.O.; Adams, B.A. Desorption equilibrium moisture content of
cassava at several temperatures. International Agrophysics 1986, 2, 235–239.
9. Bizot, H. Using the GAB model to construct sorption isotherms. In Physical
Properties of Food; Jowitt, R., Esther, F., Hallstrom, B., Meffert, H.F.T.,
 Tropical Fruit Isotherm Modeling 1497

Spiess, W.E.L., Vos, G., Eds.; Applied Science Publishers: London=New

York, 1983; 43.
10. Mir, M.A.; Nath, N. Sorption isotherms of fortified mango bars. Journal of
Food Engineering 1995, 25, 141–150.
11. Mercier, F. Séchage de gel d’alumine: maı̂trise de la texture de supports de
catalyseur. Thèse, Université de Pau et des Pays de l’Adour, Pau, 1996.
12. Pillard, W. Evolution des propriétés hydriques et rhéologiques d’un sel
d’alumine au cours d’un processus de séchage isotherme. Thèse, Université
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

de Montpellier II, Montpellier, 1997.

13. Myhara, R.M.; Taylor, M.S.; Slominski, B.A.; Al-Bulushi, I. Moisture sorp-
tion isotherms and chemical composition of omani dates. Journal of Food
Engineering 1998, 37, 471–479.
14. Tsami, E.; Krokida, M.K.; Drouzas, A.E. Effect of drying on the sorption
characteristics of model fruit powders. Journal of Food Engineering 1999,
38, 381–392.
15. Vijayanand, P.; Yadav, A.R.; Balasubramanyam, N.; Narasimham, P.
Storage stability of guava fruit bar prepared using a new process.
Lebensm.-Wiss. u.-Technol. 2000, 33, 132–137.
16. Arogba, S.S. Effect of temperature on the moisture sorption isotherm of a
biscuit containing processed mango (Mangifera indica) kernel flour. Journal
of Food Engineering 2001, 48, 121–125.
17. Lahsani, S.; Kouhila, M.; Mahrouz, M.; Fliyou, M. Moisture adsorption-
desorption isotherms of prickly pear cladode (Opuntia ficus Indica) at
different temperatures. Energy Conversion and Management 2003, 44,
923–936.
18. Arlabosse, P.; Rodier, E.; Ferasse, J.H.; Chavez, S.; Lecomte, D. Comparison
between static and dynamic methods for sorption isotherm measurements.
Drying Technology 2003, 21 (3), 479–497.
19. Talla, A.; Jannot, Y.; Kapseu, C.; Nganhou, J. Experimental study and
modelling of the kinetics of drying of tropical fruits. Application to banana
and mango. Sciences des Aliments 2001, 21, 499–518.
20. Van der Berg, C.; Bruin, S. Water activity and its estimation in food systems:
theorical aspects. In Water Activity: Influence on Food Quality; Rockland,
L.B., Stewart, G.E., Eds.; Academic Press: New York, 1981; 1–61.
21. Belahmidi, M.; Belghit, A.; Mrani, A.; Mir, A.; Kaoua, M. Approche
expérimentale de la cinétique de séchage des produits agro-alimentaires—
Application aux peaux d’orange et à la pulpe de betterave. Revue Générale
de Thermique 1993, 380, 444–453.
22. Kouhila, M.; Belghit, A.; Daguenet, M.; Boutaleb, B.C. Experimental
determination of the sorption isotherms of mint (Mentha viridis), sage (Salvia
officinalis) and verbena (Lippia citriodora). Journal of Food Engineering
2001, 47, 281–287.
23. Bizot, H.; Multon, J.L. Méthode de reference pour la mesure de l’activité de
l’eau dans les produits alimentaires. Anales de Technologie Agricole 1978, 27
(2), 441–449.
24. Nadeau, J.P.; Puiggali, J.R. Séchage: des procédés physiques aux procédés
industriels. Lavoisier Tec. et Doc., Paris, 1995; 67–75.
25. GRET. Le point sur le séchage solaire des produits alimentaires; GRET:
Paris, 1986.
 1498 Talla et al.

26. Labuza, T.P.; Kaanane, A.; Chen, J.Y. Effect of temperature on the moisture
sorption isotherms and water activity shift to two dehydrated foods. Journal
Food Science 1985, 50, 385–391.
27. Maroulis, Z.B.; Tsami, E.; Marinos-Kouris, D. Application of the GAB
model to the moisture sorption isotherms for dried fruits. Journal of Food
Engineering 1988, 7, 63–78.
28. Okos, M.R.; Narsimhan, G.; Singh, R.K.; Weitnaeur, A.C. Food
dehydration. In Handbook of Food Engineering; Hedman, D.R., Lund,
Downloaded by [Eindhoven Technical University] at 07:58 17 November 2014

Marcel Dekker, New York, 1992; 437–562.

29. Sopade, P.A.; Ajisegiri, E.S. Moisture sorption study on Nigerian foods:
Maize and sorghum. Journal of Food Process Engineering 1994, 17, 33–56.
30. Rouquerol, F.; Rouquerol, J.; Sing, K. Adsorption by Powders and Porous
Solids: Principles, Methodology and Applications; Academic Press: San Diego,
1999; 93–115.
31. Lozano, J.E.; Rostein, E.; Urbicain, M.J. Shrinkage, porosity and density of
foodstuffs at changing moisture contents. Journal of Food Science 1983, 48,
1497–1502, 1553.