ppi 201502ZU4659
Esta publicación cientíca en formato digital es
continuidad de la revista impresa
ISSN 0254-0770 / Depósito legal pp 197802ZU38
UNIVERSIDAD DEL ZULIA
Una Revista Internacional Arbitrada
que está indizada en las publicaciones
de referencia y comentarios:
• SCOPUS
• Compendex
• Chemical Abstracts
• Metal Abstracts
• World Aluminium Abstracts
• Mathematical Reviews
• Petroleum Abstracts
• Current Mathematical Publications
• MathSci
• Revencyt
• Materials Information
• Periódica
• Actualidad Iberoamericana
DE LA FACULTAD DE INGENIERÍA
REVISTA TÉCNICAREVISTA TÉCNICA
“Buscar la verdad y aan-
zar los valores transcen-
dentales”, misión de las
universidades en su artículo
primero, inspirado en los
principios humanísticos.
Ley de Universidades 8 de
septiembre de 1970.
“Buscar la verdad y aan-
zar los valores transcen-
dentales”, misión de las
universidades en su artículo
primero, inspirado en los
principios humanísticos.
Ley de Universidades 8 de
septiembre de 1970.
VOL.43 ENERO - ABRIL 2020 No.1
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, Enero-Abril, pp. 03-56
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, 49-55
Sorption characteristics of peeled beans and shells of
fermented and dry Trinitario cocoa beans (Theobroma
cacao L.)
Aleida J. Sandoval* , José A. Barreiro , Andrea De Sousa , Daniela Blanco and
César Giménez
Laboratorio de Procesamiento de Alimentos. Depto. de Tecnología de Procesos Biológicos y Bioquímicos,
Universidad Simón Bolívar, Aptdo. 89000, Caracas 1080-A, Venezuela.
*Autor de contacto: asandova@usb.ve
https://doi.org/10.22209/rt.v43n1a07
Recepción: 03/12/2019 | Aceptación: 12/12/2019 | Publicación: 20/12/2019
Abstract
Water sorption data of the peeled bean and shell of Trinitario fermented and dry cocoa beans were determined
at 25 °C and adjusted to the isotherm models of Brunauer-Emmet-Teller-(BET) and Guggenheim-Anderson-de Boer-(GAB).
The difference in the moisture sorption capacity of the shell and peeled beans was clearly established. The monolayer
water contents estimated using the BET and GAB models for the peeled bean was 3.53 and 3.60 g water/100 g dry solids,
respectively; while for the shell were 8.42 and 8.50 g water/100 g dry solids. The energy constants (C) obtained applying
the BET model for the peeled bean and shell were 154.3 and 163.2, respectively. Those estimated from the GAB model (C´)
were 21.63 and 83.56, respectively. A larger C´ value for the shell indicates stronger water binding to the active sites in the
monolayer as compared with the peeled beans that could indicate the presence of more active polar sites in the former. The
value of the K constant in the GAB model was 0.871 and 0.942 for the peeled bean and shell respectively that could suggest
a larger water adsorption by proteinaceous material in the peeled bean.
Keywords: Sorption isotherms; BET and GAB; cocoa beans; cocoa bean shell; peeled cocoa bean.
Características de sorción del haba pelada y la cascarilla de
cacao Trinitario (Theobroma cacao L.) fermentado y seco
Resumen
Se determinaron los datos de sorción de humedad de la cascarilla y habas peladas de cacao Trinitario, fermentado y seco a
una temperatura de 25 °C y se ajustaron a los modelos de Brunauer-Emmet-Teller-(BET) y Guggenheim-Anderson-de Boer-
(GAB). El contenido de agua de la monocapa para el haba pelada se estimó utilizando los modelos BET y GAB en 3,53 y 3,60
g de agua/100 g de sólidos secos, respectivamente; mientras que para la cascarilla fueron de 8,42 y 8,50 g de agua/100 g
de sólidos secos. La constante de energía (C) obtenida al aplicar el modelo BET para el haba pelada y la cascarilla fueron
154,3 y 163,2, respectivamente. Aquéllas calculadas con el modelo GAB (C´) fueron de 21,63 y 83,56. Un valor mayor de C’
para la cascarilla indicó una mayor unión del agua a los sitios activos en la monocapa en comparación con el haba pelada, lo
que podría evidenciar la presencia de sitios polares más activos en el primero. El valor de la constante K en el modelo GAB
fue 0,871 y 0,942 para el haba pelada y la cascarilla, respectivamente, pudiendo sugerir una mayor adsorción de agua por
el material proteico en el haba pelada.
Palabras clave: datos de sorción; BET y GAB; habas de cacao; cascarillas; habas peladas.
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, Enero-Abril, pp. 03-56
50 Sandoval y col.
Introduction
Cocoa beans are a tropical product obtained
from the pods of the cocoa tree (Theobroma cacao L.). It
is an important commodity used in the manufacture of
chocolate and chocolate products with applications in
other industries. Once harvested, the pods are opened
and the beans that are covered by mucilage rich in
carbohydrates, allow to ferment in fermentation boxes at
ambient temperature. The fermentation process involves
multiple microorganisms and enzymatic transformations
dried and packed in new permeable jute or similar sacks.
The stability and preservation of cocoa beans depends on
their moisture content, which is affected by temperature
and air relative humidity, and ultimately by their water
activity. Moisture sorption isotherms correlate the
moisture content of a product with its water activity at
constant temperature when it is in equilibrium with an
ambient of known relative humidity. When the product
with a determined moisture content is exposed to the
ambient during storage or transportation, it can gain or
lose moisture depending on the psychrometric conditions
of air. Therefore, the knowledge of the moisture sorption
isotherms is important to predict the stability of cocoa
beans and carry out engineering calculations.
Water sorption isotherms of cocoa beans at
elevated temperatures have been determined for the
desorption isotherms for fermented cocoa beans at 30,
40 and 60 °C. The data for 30 and 40 °C was coincident
and there were no differences in sorption data with
temperature in this range.
Sorption moisture isotherms for cocoa beans
at ambient temperatures found in tropical zones (i.e., up
obtained moisture sorption data for cocoa beans for water
activities below 0.70 using the static jar method at 15 and
isotherms for non-fermented cocoa beans at temperatures
studied and all data was adjusted to a single isotherm. The
water sorption characteristics of cocoa powder obtained
from fermented cocoa beans in at temperatures of 5, 15
< 0.05) between the moisture sorption characteristics in
the temperature range studied and the data was adjusted
to a single isotherm in that range. In both papers, the
monolayer water content was estimated using the BET
model.
All works regarding the sorption isotherms
for fermented and non-fermented cocoa beans at
temperatures from 5 to 40 °C indicated that there
characteristics in this temperature range, and the
moisture sorption isotherms could be assimilated to a
properties and the composition and thermogravimetric
characteristics of well fermented and dry Trinitario cocoa
beans. The weight fraction estimated for the peeled bean
was 0.845 and that of the shell 0.155 in relation to the
found between the moisture content of the shells (17.30 %,
w.b.) and that of the peeled bean (6.22 %, w.b.), evidencing
different water sorption capacity. It was pointed out that
the shell with 15.5 % of the weight and average thickness
of 0.310 mm, was able to retain 39.7 % of the total water
present in the whole bean. This evidenced that the
moisture distribution in cocoa beans was not uniform
and it could have incidence in the water vapor diffusion
and adsorption processes within the cocoa bean. These
work, particularly to study mass transfer and stability
problems of cocoa beans and to better understand the
phenomenon of moisture migration in dry containers and
ship holds or while stored in warehouses that could result
in condensation damage and eventual deterioration of the
cocoa beans.
literature revised regarding the sorption isotherms of
cocoa beans components (shells and peeled beans). The
objective of this research work was to determine the
moisture sorption isotherm of shells and peeled fermented
and dry Trinitario cocoa beans at 25 °C and determine the
Materials and methods
Sample collection
A composite sample of about 5 kg of fermented
and dry cocoa beans (Theobroma cacao L.) Trinitario type,
selected at random from beans packed in new and clean
jute sacks was provided by Cacao de Origen, Hacienda
La Trinidad, Caracas, Venezuela. The cocoa beans had
been grown and harvested at Cúpira, Miranda state,
Venezuela, in 2014. The cocoa beans had been graded as
Fino de Primera) by the same,
according to the Venezuelan standard for grading cocoa
random for peeling to obtain the cocoa bean shells and the
peeled cocoa beans used in this work.
Cocoa beans were carefully peeled by hand in
the laboratory using a knife to separate the cocoa bean
shell and obtain the peeled bean. The samples so prepared
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, Enero-Abril, pp. 03-56
51
Sorption isotherms of peeled bean and shells of Trinitario cocoa beans
were kept at an average temperature of 25 °C in a hermetic
until they were used for analysis and further testing.
Physical and chemical characterization
The moisture content was determined using the
atmospheric drying method at 100-102 °C for 16 hours
until constant weight, following the procedure indicated
determined using a Decagon CX-1 equipment previously
The fraction of peeled cocoa bean and shell
was determined following the procedure presented by
of whole cocoa beans, shells and peeled beans was
determined for each sub-sample using an analytical
balance Ohaus Adventurer (± 0.0001 g); and the shell
and the peeled bean fractions calculated. Also, the shell
thickness was determined by selecting at random 100
pieces of shells and measuring their thickness in triplicate
with a digital Mitutoyo micrometer (± 0.001 mm).
Chemical analyses for this product were done
reference, the proximate composition is presented here
(g/100 g ± standard deviation): Moisture = 6.51 ± 0.05;
fat = 44.23 ± 0.30; protein = 12.65 ± 0.22; ash = 3.34 ±
Moisture sorption isotherms
In order to determine the moisture sorption
isotherms, the peeled cocoa beans and shells were
separately ground in a motorized mill (Wiley N°4), using
a 2-mm sieve. A static gravimetric method was followed
weighing of about 2 g of the ground samples (peeled
bean or shells). The samples were placed in open shallow
0.6 cm. The containers with the samples were placed
by triplicate inside desiccators, each one containing an
oversaturated salt solution of known equilibrium relative
humidity at the temperature studied. Eleven oversaturated
salt solutions were used. These solutions covered a water
activity range from 0.08 to 0.97 at 25 °C. The following
salts and the corresponding water activities at 25 °C in
3CO2 (0.225), MgCl22CO3 (0.432), Mg(NO3)2
(0.529), CoCl2 (0.649), NaCl (0.753), (NH4)2SO4 (0.810),
3 2SO4 (0.973). Thymol was used as
antifungal agent in the desiccators with ambient with
relative humidity above 81% to avoid alterations by mold
growth that could affect the results of the experiments
samples were kept in an ambient at 25 °C.
In order to determine when equilibrium was
reached, an identical setup was prepared in parallel for
each salt. In this experiment, the samples were periodically
weighed using an analytical balance (Ohaus Adventurer; ±
0.0001g) until reaching constant weight at equilibrium.
Also, after equilibrium was reached, the water activity
of the samples was measured using a Decagon CX-1
equipment, in order to compare the water activity of
the samples with those given by the oversaturated salt
solutions.
The moisture content of the equilibrated samples
determined by the oven method described before were
correlated with the water activities of the corresponding
salt solutions. The data obtained were adjusted to the
equations (1) and (2), respectively.
Where:
V: moisture content (g water/100 g dry solids)
aw: water activity (dimensionless)
Vm: moisture content corresponding to the monomolecular
layer (g water/100 g dry solids)
C: energy constant related to the free energy of sorption
(dimensionless)
In order to select the data for the adjustment to
were followed. According to these recommendations, the
application of the BET equation should be restricted to a
water activity range where the term V.(1-aw) continuously
increases when plotted as a function of aw. The selection
of a wrong range of water activity could result in negative
values of the term C which is physically unsound. This
problem arises by the impossibility of separating the
monolayer adsorption given by the BET equation from the
the C value.
Where:
: moisture content corresponding to the monomolecular
layer (g water/100 g dry matter)
C´: energy constant related to the free energy of
sorption (dimensionless)
K: constant related to the properties of the multilayer
molecules relative to the bulk liquid
(dimensionless)
(1)
(2)
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, Enero-Abril, pp. 03-56
52 Sandoval y col.
The constants K and C´
relations: 0 < KC´ C´
GAB equation turns into a Type II sigmoidal shape curve.
However, for 0 < C´
C´
is a constant related to the difference in enthalpy between
monolayer (H1) and multilayer (Hm) sorption, and is
expected to be positive due to the exothermic nature of the
interaction of water with the active sites in the surface. The
constant K is related to the difference between the heat of
condensation H of water and the heat of sorption of the
multimolecular layer (Hm). This difference is expected to
be negative and smaller since the multilayer molecules are
Statistical analysis
Comparison of central tendency of the
equilibrium mositure contents of shell and peeled beans
was done by means of one-factor analysis of variance
sorption models of BET and GAB. In all cases, OriginPro
8.5.0 SR1 (Northampton, MA, U.S.A.) software was used.
linear regressions.
Results and discussion
Physical Characterization
The moisture content obtained for the cocoa
bean samples was 6.51 ± 0.05% (wet basis) with a range
from 6.48 to 6.59% (wet basis). Its water activity at 25-
27°C was 0.657. The ± sign indicates standard deviation.
The weight fraction determined for the shells
in relation to the whole bean was 0.147 while the weight
Trinitario fermented and dry cocoa beans who reported a
weight fraction for the shell of 0.155 ± 0.004 with a 95 %
determined was 0.21 ± 0.03 mm.
The moisture content of the peeled beans was
5.21 ± 0.11 g water/100 g dry solids and the moisture
determined for the shells was 18.76 ± 0.67 g water/100
g dry solids. Based on these results it was calculated that
35.6 % of the water in the whole bean was contained in
the shell and the rest in the peeled bean. These results
are indicative of the higher water adsorption capacity of
the cocoa bean shell as compared with that of the peeled
bean. The difference in the water adsorption capacity of
the shell and the peeled been and the possible causes were
the physicochemical changes that take place during cocoa
activity taking place during fermentation results in the
formation of compounds capable of adsorbing water
such as reducing sugars, peptides, free aminoacids,
organic acids and minerals. These compounds are mainly
associated to the external area of the bean (shell) where
most of the fermentation occurs.
Moisture sorption isotherms
In order to clarify the difference in the moisture
sorption capacity of shells and peeled cocoa beans,
the moisture sorption isotherms of these components
of Trinitario fermented and dried cocoa beans were
determined at 25 °C. Since the initial moisture content of
the peeled beans and shells was 5.21 and 18.76 % (dry
basis) respectively, the moisture data in the isotherms
below the initial moisture content were obtained by
desorption and those above by adsorption. The sorption
data obtained for the shell and peeled bean are presented
in Table 1.
Table 1. Sorption data of shell and peeled cocoa beans
at 25 °C (average moisture content ± standard deviation)
Moisture content (g/100 g of dry solids)
awPeeled bean Shell
0.082 3.39 ± 0.07 8.50 ± 0.46
0.113 4.10 ± 0.14 9.14 ± 0.05
0.225 4.34 ± 0.27 10.79 ± 0.50
0.328 4.67 ± 0.03 12.32 ± 0.11
0.432 4.75 ± 0.05 14.65 ± 0.67
0.529 5.13 ± 0.09 15.28 ± 0.93
0.649 5.50 ± 0.30 19.26 ± 2.59
0.753 7.39 ± 0.02 27.16 ± 1.99
0.810 9.13 ± 0.13 32.37 ± 4.26
0.936 19.68 ± 0.18 72.70 ± 9.20
0.973 24.55 ± 0.07 n/d*
* n/d: non determined
< 0.05) between the average moisture sorption capacity
(p < 0.05) than that of the peeled bean. This increased
moisture adsorption capacity of shells can be explained
by the water-binding compounds formed during bean
fermentation by microbiological and enzymatic action.
Most of the microbiological action during fermentation
occurs in the surface of the beans where most of these
The data in Table 1 was adjusted to the models of
The bars in each point indicate the standard
error.
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, Enero-Abril, pp. 03-56
53
Sorption isotherms of peeled bean and shells of Trinitario cocoa beans
Figure 1. BET model moisture sorption isotherms for the
shell and peeled beans at 25 °C
Table 2. Constant parameters obtained for the BET
and GAB sorption models
Cocoa component
Model Parameter Peeled bean Shell
BET
Vm 3.53 8.42
C154.3 163.2
R2 0.987 0.999
GAB
3.60 8.50
C’ 21.64 83.56
K0.871 0.942
R2 0.949 0.993
Table 2 shows that in all cases elevated
2) were obtained for the
statistical adjustment of the data to the BET and GAB
models. The analysis of variance carried out for both
linear regressions (p < 0.001).
The monolayer water content (Vm and ) of the
shell was higher than that obtained for the peeled bean in
both models, evidencing the presence of more active sites
capable of binding water in the shell, as discussed before.
When using the BET model, C values of 154.3 and 163.2
were obtained for the peeled bean and shell, respectively.
The values obtained were in the same order magnitude to
2.94 and 3.56 g/100 g dry solids, respectively.
The estimated GAB constants were: monolayer
water content ( ) of 3.60 and 8.50 g/100 g of dry solids;
C´ values of 21.64 and 83.56; and K values of 0.871 and
0.942, for the peeled bean and shell, respectively. The
results obtained comply with the inequalities summarized
sense that Vm < and C > C´ with K
physical soundness of the results.
The larger C´ value obtained for the shell
indicates stronger water binding to the active sites in
the monolayer as compared with the peeled beans,
evidencing the presence of more active polar sites in the
former. C´ is related to the difference between the free
enthalpy of water molecules adsorbed in the multilayer
Figure 2. GAB moisture sorption isotherms for the shell
and peeled bean at 25 °C
and the peeled bean used to adjust the BET model, were
selected according to the guidelines given by Thommes
beans and six for the shell, including the origin.
The adjustment for the GAB equation included
data points for water activity ranging from 0 to 0.973
for the peeled bean and from 0 to 0.936 for the shell.
The constants obtained from the non-linear adjustment:
Vm and C for the BET model and , C´ and K for the GAB
model are presented in Table 2.
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, Enero-Abril, pp. 03-56
54 Sandoval y col.
and that of the molecules bound to the monolayer. The
K value obtained for the shell was close to one (0.942),
indicating small difference between the molecules in the
multilayer and non-bound molecules in the bulk of liquid.
As K approaches to one, the water molecules beyond the
monolayer are not structured in a multilayer and have
similar characteristics as those in the bulk liquid and the
. A smaller
K value was obtained for peeled beans (0.871) indicating
structuration of a sorption multilayer. According to Chirife
K for proteins falls in a range of 0.82
0.77. This could indicate predominant water adsorption
by proteinaceous material in the peeled bean.
Conclusions
The water sorption data of shell and peeled
cocoa beans were determined and adjusted using non-
GAB. The analysis of variance for both models showed
the peeled bean and the shell and elevated correlation
The difference in the moisture sorption capacity
of the shell and peeled beans was clearly established. The
monolayer water content calculated using the BET (Vm)
and GAB ( ) models for the peeled bean was 3.53 and
3.60 g water/100 g dry solids, respectively; while for the
shell were 8.42 and 8.50 g water/100 g dry solids. The
C values obtained using the BET model for the peeled
bean and shell were 154.3 and 163.2, respectively and C´
values of 21.63 and 83.56 were calculated when the GAB
model was applied. A larger C´ value for the shell indicates
stronger water binding to the active sites in the monolayer
as compared with the peeled beans that could evidence
the presence of more active polar sites in the former.
The value of the K constant in the GAB model was 0.871
and 0.942 for the peeled bean and shell respectively that
could suggest a largest water adsorption by proteinaceous
material in the peeled bean.
References
del agua en semilla de café y cacao. Acta Científica
Venez, Vol. 33, No. 1 (1982), 398.
desorption isotherms of cocoa beans. Trans.
ASAE
Sorption isotherms and drying characteristic
curve offermented cocoa. Int. J. Sci. Technol., Vol.
and isosteric heats of fermented cocoa beans
(Theobroma cocoa). J. Food Res., Vol. 4, No 3
(2015) 138-147.
vegetables and plant products. J. Sci. Food Agric.,
Vol. 1, (1950) 42, 1950.
isotherms of non-fermented cocoa beans
(Theobroma cacao). J. Food Eng., Vol. 51, No. 2
M.: Sorption characteristics of fermented cocoa
powder (Theobroma cacao). Rev. Tec. la Fac. Ing.
Univ. del Zulia
and thermogravimetric characterization of
components of Venezuelan fermented and dry
Trinitario cocoa beans (Theobroma cacao L.):
Whole beans, peeled beans and shells. Rev.
Técnica la Fac. Ing. Univ. del Zulia, Vol. 42, No. 1
(2019) 39-46.
properties of fermented and dried beans of
Venezuelan Trinitario cocoa (Theobroma cacao
L.). Rev. Técnica la Fac. Ing. Univ. del Zulia, Vol. 42,
No. 2 (2019) 50-56.
Caracas, Venezuela, 1995.
ed. Arlington, VA, 1996.
Decagon Devices, Inc. USA, 1984.
saturated aqueous solutions. J. Res. Natl. Bur.
Stand. Sect. A Phys. Chem., Vol. 81A, No. 1 (1977)
89-96.
dynamic techniques with and without thymol as
an antimicrobial agent. Food Biophys., Vol. 6, No. 3
Rev. Téc. Ing. Univ. Zulia. Vol. 43, No. 1, 2020, Enero-Abril, pp. 03-56
55
Sorption isotherms of peeled bean and shells of Trinitario cocoa beans
of gases in multimolecular layers. J. Am. Chem.
Soc., Vol. 60, No. 2 (1938) 309-319.
J. Am. Chem. Soc.
1946.
1953.
:
Physisorption of gases, with special reference
to the evaluation of surface area and pore size
distribution (IUPAC Technical Report). Pure Appl.
Chem.
parameters: BET or GAB values?. Colloids Surfaces
A Physicochem. Eng. Asp.
235-260.
Technology. Food Bioprocess Technol., Vol. 1, No. 1
(2008) 82-90.
Strate, G.: Sorption isotherms, GAB parameters
and isosteric heat of sorption. J. Sci. Food Agric.,
Vol. 85, No. 11 (2005) 1805-1814.
and biochemistry of cocoa fermentation in the
Dominican Republic. Int. J. Food Microbiol., Vol.
114, No. 1 (2007) 124-130.
composition during fermentation of pulp pre-
conditioned cocoa (Theobroma cacao) beans. Int.
Food Res. J., Vol. 20, No. 4 (2013) 1843-1853.
Boquet, R.: Some features of the parameter k of
the GAB equation as applied to sorption isotherms
of selected food materials. J. Food Eng., Vol. 15, No.
1 (1992) 75-82.
REVISTA TECNICA
DE LA FACULTAD DE INGENIERIA
UNIVERSIDAD DEL ZULIA
www.luz.edu.ve
www.serbi.luz.edu.ve
www.produccioncientica.luz.edu.ve
Esta revista fue editada en formato digital y publicada
en Diciembre de 2019, por el Fondo Editorial Serbiluz,
Universidad del Zulia. Maracaibo-Venezuela
Vol. 43. N°1, Enero - Abril 2020, pp. 03 - 56__________________
