See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/346406170

Moisture sorption properties of two varieties of dehydrated mango slices as

determined by gravimetric method using Guggenheim–Anderson–de Boer
model

Article in Journal of Food Processing and Preservation · November 2020

DOI: 10.1111/jfpp.15041

CITATIONS READS

8 278

4 authors:

Isaac Nyangena Owino O. Willis

4 PUBLICATIONS 72 CITATIONS
Jomo Kenyatta University of Agriculture and Technology
157 PUBLICATIONS 1,712 CITATIONS
SEE PROFILE
SEE PROFILE

Jane Ambuko Samuel Imathiu

University of Nairobi Jomo Kenyatta University of Agriculture and Technology
100 PUBLICATIONS 1,241 CITATIONS 79 PUBLICATIONS 1,554 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Isaac Nyangena on 11 December 2020.

The user has requested enhancement of the downloaded file.

Accepted Article
PROF. WILLIS OWINO (Orcid ID : 0000-0003-2198-2595)

Article type : Original Article

1 Moisture sorption properties of two varieties of dehydrated mango slices as determined by

gravimetric method using Guggenheim–Anderson–de Boer model

Nyangena Isaac 1, 2, Willis Owino1, Jane Ambuko3 and Samuel Imathiu 1

1Department of Food Science and Technology, School of Food and Nutrition Sciences, Jomo Kenyatta
University of Agriculture and Technology, P.O. Box 62000-00200 Nairobi, Kenya
2Kenya Bureau of Standards P.O. Box 99376-80100 Mombasa, Kenya
3 Department of Plant Science and Crop Protection, University of Nairobi, P. O. Box 29053-00625, Kangemi,
Nairobi, Kenya
Corresponding author: willis@agr.jkuat.ac.ke

Abstract

The moisture sorption isotherms for oven dried mango were constructed gravimetrically at three
different temperatures of 25°C, 30°C and 35°C, and relative humidity ranging 8.3-97.0%. The
Guggenheim-Anderson-de Boer model was used to fit sorption data and constant equations
determined using non-linear regression analysis. The isotherms had J-shaped curve that represents
type III isotherms characterized by sugar-rich products. At increased temperatures, it was observed
that equilibrium moisture content increased and crossing effect of isotherm curves was observed with
a corresponding increase in water activity. A substantial increase in equilibrium moisture content was

This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/JFPP.15041
This article is protected by copyright. All rights reserved
 observed above water activity of 0.6. Monolayer moisture content of between 8-9.6% was predicted
Accepted Article
for dried slices where the lowest value indicated that moisture content should not be lower than 8.0%
to avoid unnecessary energy consumption. For assured microbiological stability, the safe storage
moisture content for dried mango should be maintained at approximately 15% or less that
corresponded to 0.6 water activity.

Keywords: Equilibrium moisture content; mango; moisture; sorption isotherm; water activity; GAB

Practical Application

Dehydration of fruits has been utilized as one of the most effective value addition procedures for
extending their shelf life. However, exposure to high drying temperature and significant loss of
moisture in fruits cause irreversible stresses in the cellular structure of the fruits and may lead to
robust textural changes especially during storage. This research finds practical application in shelf life
estimation and prediction of storage stability of the mango varieties. Additionally, the results can be
used as a guide in product development and process design for dried mango slices.

Introduction

Mango (Mangifera indica) production and consumption is ranked third and first respectively among
tropical fruits worldwide (Rangel et al., 2011). The fruit is rich in ascorbic acid and beta-carotene
where 100 g of ripe mango consumed provides more than 50% of the recommended daily allowance
for vitamin A and C (Rangel et al. 2011). Besides, mango fruit nutritionally supplies other vitamins
such as niacin and riboflavin, and minerals including calcium, iron and potassium (Gulzar et al.,
2018). Research has shown that high water activity (>0.6) promotes microbial growth and leads to

This article is protected by copyright. All rights reserved

 enhanced quality degradation in foods (Prasantha, 2018). The term water activity denotes free water
Accepted Article
that is available to participate in chemical, physical and microbial reactions (Klewicki et al., 2015)
which influences the shelf-life of foods. Since change in water activity directly affects microbial
deterioration and chemical reactions, the determination of relationship between moisture content and
water activity is highly significant.

One of the most commonly used method of food preservation by most food industries is dehydration
(Bouba et al., 2014). Dehydration of foods has the primary objective of causing reduced microbial
activity, minimizing product deterioration and extending shelf-life of stored products (Prasantha,
2018). Although oven drying has been studied intensively in various fruits, its effect on moisture
sorption isotherms of dried mango slices of Apple and Ngowe mango varieties still remains
undocumented.

Moisture sorption isotherm is a graphic representation of the process wherein water molecules are
progressively and reversibly released from all kinds of hygroscopic forces in food system caused by
colligative effects, capillary effects and direct bonding (Caballero-Ceron et al., 2015). Moisture
sorption isotherms (MSI) curve describes the relationship between equilibrium relative humidity
(ERH)/aw and equilibrium moisture content of the product (Abraham et al., 2017). The shape of a
sorption isotherm is dependent on pressure, temperature, structure and composition of the food
material, as well as the process of humidification/dehydration used; every food product has a unique
isotherm (Caballero-Ceron et al., 2015)(Prasantha, 2018). The moisture sorption isotherms are
yielded by plotting a graph of total moisture content against water activity of a food product over a
varied range and at a fixed temperature (Biswal et al., 2017). They can be determined during
adsorption (begins from dry state) or desorption (begins from wet stage). Knowledge in sorption
isotherms of food products is of importance in prediction of stability in relation to enzymatic,
microbiological and chemical reactions. Additionally, it is used in designing of drying processes,
selection of packaging materials and choosing the appropriate storage conditions (Mariem and
Mabrouk, 2015).

The gravimetric method of determining sorption isotherms uses saturated salt slurry and has been
recommended as the reference method for isotherm determination due to its attributable advantages of

This article is protected by copyright. All rights reserved

 easy handling, lower price and acceptable accuracy (Zhang, Sun and Zhang, 2017) However, typical
Accepted Article
method of saturated salt slurry takes weeks or months for successful moisture sorption isotherm
detection. This reduces its feasibility to determine isotherms for sensitive as well as perishable
samples of biological nature at relatively higher temperatures (Zhang, Sun and Zhang, 2017).

There are over 200 models that are used in the simulation of sorption behaviors aimed at expressing
how equilibrium relative humidity relates with water activity at constant temperature and pressure
(Andrade, 2011). These models are either empirical, theoretical or a simplified mode of the
sophisticated models. Some models are limited for particular types of foods since the chemical
composition and the physical structure have an influence on the amount of water that can be absorbed
which affects the isotherm type that is unique for every product type (Prasantha, 2018). The GAB
model has been reported to provide best fit for adsorption and desorption of foods over a widely
spread range of water activity (Bi et al., 2018). This study aimed to determine the sorption isotherms
for two varieties of dried mango slices at temperatures of 25°C, 30°C and 35°C and relative humidity
of between 8.3-97.0%. The information derived could assist the fruit industry reduce postharvest
losses by establishing optimal drying moisture content to avoid energy wastage during oven drying
and also determine the critical storage moisture content corresponding to 0.6 water activity for
prolonged shelf life.

Materials and methods

Raw material acquisition and preparation

Freshly harvested mango fruits of Apple and Ngowe varieties were obtained at stage 4 of maturity
(skin firmness of 15 to 28.4 N) in the 2017-2018 harvest season from commercial farmers at Masii,
Machakos County, Kenya as previously described in Nyangena et al., (2019). A close similarity for
ripening degree for mangoes with 15-22°Brix was used as a criterion to select sound and uniformly
ripe mangoes for the experiment. Uniformity in size was also considered with weight range of 300-
350g for Apple mango variety and 470-550g for Ngowe variety. The mangoes were hand-picked
during harvesting and arranged into plastic crates cushioned with partially wetted newspapers prior to

This article is protected by copyright. All rights reserved

 transportation to the fruits and vegetable workshop, department of Food Science and Technology,
Accepted Article
Jomo Kenyatta University of Agriculture and Technology for experimentation.

The mango fruits were washed in portable running water and then immersed in 100 ppm sodium
hypochlorite solution contained in stainless steel sink for 20 minutes. The mangoes were subjected to
post-harvest ripening at room temperature conditions (25±2°C) and relative humidity of 50% for 2
days. The samples were rinsed in clean potable running tap water and sliced into uniform thickness of
approximately 4 mm using electric meat slicer (Orion, Japan). Pieces of mango slices were arranged
without overlapping on a meshed tray with perforations prior to placement in a convective dryer
(Memmert, UF 110 108 model DO6836 Germany) at temperatures of 50°C until a moisture content of
approximately 12% was obtained.

Moisture content determination

The moisture content of dried mango slices and that of fresh mango samples was determined by
934.06 AOAC (2016) oven drying method with minor modifications as follows; moisture dishes were
labeled, cleaned and allowed to dry at 105°C, cooled in a desiccator and then weighed to a constant
weight. Five and 2.5 g of fresh and dried mango samples respectively were weighed precisely into the
moisture dishes. The test portions and dishes were oven dried at 105oC for 3 hours to a constant
weight and then transferred quickly into a desiccator for cooling. After cooling, the dishes together
with the samples were weighed while minimizing exposure to atmospheric conditions. The moisture
determination was done in six replicates for each sample. Moisture content was obtained through
differences in weight before and after moisture drying to a constant weight. The values obtained were
expressed as percentage moisture content.

(𝑊𝑒𝑖𝑔ℎ𝑡 𝑏𝑒𝑓𝑜𝑟𝑒 𝑑𝑟𝑦𝑖𝑛𝑔 ― 𝑊𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒 𝑎𝑓𝑡𝑒𝑟 𝑑𝑟𝑦𝑖𝑛𝑔) 𝑋 100

%𝑀𝐶 = Eq. 1
𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒 𝑏𝑒𝑓𝑜𝑟𝑒 𝑑𝑟𝑦𝑖𝑛𝑔

This article is protected by copyright. All rights reserved

 Sorption isotherms determination procedure
Accepted Article
The sorption isotherms of dried mango slices were determined by static gravimetric technique that has
been reported as the most common method for determination of moisture sorption isotherms for foods
(Bouba et al., 2014). Seven saturated salt solutions (as listed in Table 1) of reagent grade with water
activity ranging 0.08 to 0.97 were prepared to provide respective constant relative humidity. The
saturated salt solutions were made by dissolving appropriate quantities of corresponding salts into
distilled water as recommended by Greenspan (1977). Every saturated salt solution prepared was
placed on a plastic plate held inside the bottom layer of desiccator with a depth of about 2 cm.
Approximately 0.5 g dried samples of each mango variety were weighed in triplicate and placed in
open top aluminium dishes that were subjected to a constant relative humidity environment (enclosed
desiccators). Direct contact between the mango samples and the salt solutions was avoided by placing
the samples in the upper layers of the desiccator. At relatively high water activity levels (above 0.6),
pieces of cotton wool dipped in 40% formalin were placed besides the mango samples inside the
airtight chambers (desiccators) to inhibit mold growth (Akoy and Hörsten, 2013). Sample weights
were evaluated after every 7 days for EMC determination until there was an insignificant change in
sample weight (less than 0.001g) for three consecutive weight measurements. Equilibrium condition
between the environment and the samples was attained after a duration of 28 days. The relative
humidity value and respective water activities of the saturated salt solutions were taken from literature
(Greenspan, 1977), and confirmed using Hygropalm 23-AW-A portable water activity analyzer
(Rotronic AG, Bassersdorf, Switzerland).

Desiccators were put in incubators (Memmert, UF 110 108 model DO6836 Germany) maintained at
25°C, 30°C and 35°C. A data logger (Huato model HE170, China) was placed inside the desiccators
to monitor the actual relative humidity and temperature conditions. Undissolved crystals were
constantly maintained in the salt solution to ensure saturation. Equilibration was monitored through
weighing of the samples on a 7-day interval until the constant weight change of less than 0.001 g (w2)
was obtained. The timing interval for removal of samples from the desiccator, weighing and putting
back in the desiccators was minimized to approximately 30 seconds in order to reduce the effects of
atmospheric moisture desorption or sorption on the results (Akoy et al., 2013).

This article is protected by copyright. All rights reserved

 The EMC (g/100g) was calculated using equation 2 shown below.
Accepted Article MW1 + 100(W3 ― W2)
EMC = Eq. 2
W1 + (W3 ― W2)

Where M=Initial moisture content of dried mango samples (in grams per 100 g dry matter), W1= test
portion weight of dried mango during sorption (g), W3 =initial weight of dried mango sample and
aluminium dish (g), W2=final weight of mango sample and aluminium dish at equilibrium (g) and
EMC=equilibrium moisture content.

The values of EMC were plotted against water activity for construction of moisture sorption
isotherms.

Modelling

A number of mathematical models exist for the description of moisture sorption isotherms of foods.
Some of the most commonly used models include GAB (Guggenheim, Anderson and de Boer), BET
(Brunauer, Emmett and Teller), Halsey Smith, Henderson, and Oswin. Modelling of the experimental
data was done with the GAB equation by Microsoft Office Excel 2013 and analyzed through
polynomial regression. This was done following a detailed literature search that concluded that GAB
equation predicted sorption isotherms most efficiently over wide range of water activity (Bi et al.,
2018). The GAB model is relatively simple and has only three parameters (Mo, C and K) that have
physical significance as presented in equation 3 (Rizvi, 1986).

mCka
M 0 w

(1 kaw )(1 kaw  Ckaw ) Eq. 3

where:

M=Equilibrium moisture content in percent dry basis (d.b).

Mo= GAB monolayer moisture content in percent dry basis (d.b).

This article is protected by copyright. All rights reserved

 C= A dimensionless constant associated with monolayer heat of sorption
Accepted Article
K= A dimensionless factor associated with heat of sorption of the multilayer

aw = water activity

The GAB parameters were determined afterwards using the procedure of transformed form for GAB
isotherm as given by Timmermann et al., (2001). The GAB equation is often rearranged into a
polynomial equation (equation 4):

𝑎𝑤 𝐾𝑎𝑤2 (𝐶 ― 2)𝑎𝑤 1
𝑀 = 1 + 𝑀𝑜𝐶 + 𝑀𝑜𝐶𝐾 (
𝑀𝑜( 𝑐 ― 1)

Eq.4)

If the GAB parameters are substituted with the values:

𝐾 𝐶―2 1
where α= 1 β= 𝑀𝑜𝐶 and ϒ=𝑀𝑜𝐶𝐾
𝑀𝑜( 𝑐 ― 1)

then a different form for the GAB equation is taken as shown in equations 4.

𝑎𝑤
𝑀 = 𝛼.𝑎𝑤2 +𝛽.𝑎𝑤 +𝛾 (Eq.5)

The method of polynomial non-linear regression of aw/EMC against aw was carried out via Microsoft
Excel (2013) to determine the values of constant ϒ, the linear term coefficient (β) and the quadratic
term coefficient.

Least squares regression analysis was used to obtain GAB coefficients (α, β and ϒ) and the
parameters were calculated as shown in equations 5 to 7.

1 1/2
𝑀𝑜 = (𝛽2 ― 4𝛼.𝛾) (Eq.6)

This article is protected by copyright. All rights reserved

 2. ∝ .𝑀𝑜
𝐾 = ― 𝛽.𝑀𝑜 + 1 (
Accepted Article
Eq.7)

1
𝐶 = 𝑀𝑜.𝛾.𝐾𝑔 (

Eq.8)

The monolayer parameter (Mo) is critical since it represents moisture content level where the rate of
any reaction is proved negligible based on the strong binding ability of water. The K parameter values
are used to indicate the GAB equation scope of application and the adjustment of properties of
molecules found in the adsorption monolayer compared to the region of liquid phase (Bi et al., 2018).
The C parameter on the other hand indicates the suitability to which the GAB model is chosen for
empirical data description.

Statistical error for the GAB model was evaluated using the percent of root mean square error
(RMSE) that predicts the deviation between the predicted and experimental moisture content. To
consider the model as acceptable, it is required that the residual should be less than 10% (Caballero-
Ceron et al., 2015, Abraham et al., 2017).

2
(𝑀𝑒𝑥𝑝(𝑖) ― 𝑀𝑒𝑠𝑡(𝑖))
𝑁
( )
[ ]
∑𝑖 = 1
𝑀𝑒𝑥𝑝(𝑖)
RMSE% = 100 ∗ Eq. 9
𝑁

Where N represents the number of observations/experimental data, Mest represents the estimated value
for EMC and Mexp is the experimental value for EMC.

This article is protected by copyright. All rights reserved

 Results and discussion
Accepted Article
Moisture content and equilibration at different humidity.

The moisture content for fresh `Apple` and ‘Ngowe’ mango varieties was found to be 82.49% ± 1.1
and 82.91%±2.5 wet basis (w.b) respectively. After dehydration in a convective dryer at 50°C for 10
hours, the mean moisture content reduced to 12.45% and 11.24% for Apple and Ngowe mango
varieties respectively. The equilibration of moisture for dried mango samples was attained by both
adsorption and desorption processes. Potassium acetate and potassium hydroxide provided water
activity of 0.22 and 0.083 respectively that caused the dried samples to lose moisture content in the
process of equilibration. The rest of the salts had water activity that created higher relative humidity
(above 51) that caused the dried products to gain moisture content during equilibration.

Moisture sorption isotherms of dried mango pieces

The results for moisture sorption isotherms are presented in Fig. 1.and Fig. 2. The sigmoid
characteristic curves for the isotherms obtained for the two mango varieties had fairly similar patterns.
The moisture sorption isotherms followed characteristic shape exhibited by sugar-rich foods
categorized as type III (J-shaped) according to BET classification. This findings were similar to those
reported (Akoy et al., 2013) for sorption characteristics of solar dried Kitchener variety mango slices
and in dried Kent mango variety slices (Akoy and Hörsten, 2013). Dried mango isotherms had gentle
sloping curves at water activity below 0.5 but with a sharp increase in water content above 0.5.
According to Arslan & Tog (2005), the type III isotherms are characteristic of food materials that hold
water at low aw and display a sharp increase in water content held at increased aw. Previous studies
have reported similar results for apple, apricot, raisins, pineapple and sugar (Al-muhtaseb, Mcminn
and Magee, 2002). This might be associated with an increased number of sorption sites. The increased
EMC with corresponding marginal increase in water activity showed that rapid spoilage of mango
slices could be expected at aw levels above 0.70.

At higher water activities, EMC increased sharply with a corresponding increase in moisture content
which resulted in crossing effect of isotherm curves. The crossing points of sorption isotherms

This article is protected by copyright. All rights reserved

 depends on structure, type (Lutovska et al., 2017) and sugar content of the food material (Amankwah
Accepted Article
et al., 2018). The crossing effect indicated a variation in water activity versus moisture content, and
dissolution of sugar around that water activity range. The inversion which shows dependence of
sorption isotherms on temperature can be associated with endothermic dissolution of sugar at high
moisture content and water activities, so that food products hold more water at higher temperatures.
The inversion is relevant as it opposes the normal trend of increased microbial activity with
temperature. At lower water activities of below 0.70, the EMC for dried mango slices were lower at
higher temperatures (35°C) compared to lower temperature (25°C). An increased EMC was observed
for every increased water activity for all isotherms at each temperature. The increased EMC has been
associated with increased relative humidity that results from increased vapor pressure (Rosalizan et
al., 2018). According to Rosalizan et al., (2018), the phenomenon of decreased EMC at constant
water activity for increased temperature is related to excitation state experienced by molecules. With
an increase in temperature, water molecules are imparted with an elevated state of excitation that in
turn decreases attractive forces between them and increases their distance resulting in decreased
degree of water sorption with increased temperature at a given state of relative humidity. This has an
implication that increased temperatures results in decreased hygroscopic nature for the product
(Ocheme, Ariahu and Ingbian, 2013). According to Ariahu et al., (2006), at a fixed moisture content,
increased aw at elevated temperatures rendered products more easly susceptable to microbial
deterioration since the water activity occur at a point above critical levels estimated for safe storage.
However, the dried mango slices dried to a moisture content of approximately 12% attained water
activity of below 0.6 which provides stability during storage. The GAB equation estimated safe
storage moisture content at approximately 25% which corresponded to water activity 0.6 for assured
microbiological stability.

Evaluation of GAB model fitting ability

The GAB equation fitted the moisture sorption isotherms for dried mango slices reasonably well as
demonstrated by values of percentage root mean square (%RMSE) given in Table 2. All the RMSE
values were less than 10 which has been recommended for models that give best description for
sorption data. A loss and gain in moisture was observed concurrently for the dried mango slices as it

This article is protected by copyright. All rights reserved

 depended on sample-air interaction at the storage environment. According to Haque et al., (2006),
Accepted Article
biological materials either lose moisture or gain in same depending on the interaction of materials and
air during storage. Low water activities of 0.22 and 0.083 provided by potassium acetate and
potassium hydroxide salts respectively resulted in moisture loss. Water activities of the other salts
used created higher relative humidity (above 51) that caused the dried products to gain moisture
content.

Fig. 3. and 4. shows the GAP fit predicted EMCs compared with the experimental EMCs for the
adsorption isotherms. The EMCs for experimental and predicted data were plotted against water
activity which resulted in a close estimate for the two EMCs using GAP model. The GAP model can
therefore be recommended to characterize moisture sorption features for mango fruit since it produced
reasonably good fit.

Monolayer moisture adsorption

The equilibrium data results were reasonably modeled using GAB model and the value for monolayer
moisture content (Mo) was determined at ranges between 8.0-9.15% as shown in Table 2. The Mo
parameter is important since it represents the final moisture content value for the product that
enhances better preservation (Coşkun and Pekmez, 2012). The Mo values displayed a decreased trend
with increased temperature range of 25°C to 35°C for Apple mango variety and 25°C to 30°C for
Ngowe variety. The trend of this results was in agreement with that recorded for moisture sorption
isotherms of solar dried Kitchener mango variety (Akoy et al., 2013) with Mo values range of 5.79 to
7.68 and for Jamun fruit (Biswal et al., 2017). The highest Mo recorded for Apple mango variety at
25°C was 9.15% while the lowest value recorded was 8.0% at 35°C. Dried Ngowe slices had highest
Mo value of 8.80 at 25°C and lowest Mo value of 8.15 at 30°C. Slightly higher Mo values (9.97%-
11.14%) were recorded for mango at temperature range of 30-50°C (Prasantha and Amunogoda,
2013). The monolayer moisture differences for dehydrated mango can be attributed to the different
physical-chemical composition of the varieties used and the difference in choice of techniques of
drying used. According to Zuo et al., (2015), there is also a reduced number for sorption sites that
actively bind water with increased temperature while water molecules movement increases resulting
in reduced Mo values. The least Mo value obtained (8.00%) indicated that during dehydration,

This article is protected by copyright. All rights reserved

 moisture content should not be allowed to go below 8% in order to prevent unnecessary energy and
Accepted Article
time consumption. Water activity levels of less than 0.3 are attained at moisture contents in this level
and microbiological stability is conferred to the product. Chemical reactions are also expected to be
slowed down depending on solvation (Coşkun and Pekmez, 2012).

The mango sorption isotherm curves presented three different regions that corresponded to different
means by which water is bound to the system of the mango slices. The first region (monolayer)
occurred at low water activity (0.1 to ~ 0.25) and represented firmly bound water that is strongly
bonded via the hydrogen bonds as explained by Mariem & Mabrouk (2015). It is impossible to
remove this type of water and is water that is therefore not available for reaction as a solvent nor for
use as a reagent in chemical reactions due to increased adsorption energy. The second region
(multilayer) occurred at the intermediate zone of water activity (~0.25 -0.75) with moderately active
water that is readily available as a reagent or solvent for chemical reactions. According to Mariem and
Mabrouk (2015), this region contains multilayer water that is less firmly held by capillary forces. The
third region occurred at higher water activities (above 0.75) where the water`s reactive capacity as a
reagent or solvent for any chemical reaction is high. The water present at this region occurs as liquid
phase and as a free water fraction that supports microorganism growth as well as enzymatic reactions
that accelerate food spoilage.

Conclusion

Conventional gravimetric method was applied in the determination of moisture sorption isotherms for
dried mango slices at temperatures of 25°C, 30°C and 35°C. Dried mangoes of Apple and Ngowe
varieties presented sorption isotherms with J-shaped sigmoid curves characteristic of high sugar
content foods. Moisture sorption isotherms of dried mango slices showed dependence on temperature.
The obtained data was satisfactorily described by the GAB model that gave reasonably good %
RMSE values. An increased EMC was observed for decreased trend in temperature when water
activity was kept constant. Moisture sorption isotherms for both Apple and Ngowe variety displayed
tendency of crossing over as a result of crystalline sugar dissolution at raised temperature and
moisture content. Dried mango slices can be better preserved for longer periods when safe storage

This article is protected by copyright. All rights reserved

 moisture content is maintained at approximately 15% and below with corresponding water activity of
Accepted Article
0.6 for inhibition of micro-organisms. The data established from this study may be applied in shelf
life estimation and prediction of storage stability of the tested mango varieties. Additionally, the
results can be used as a guide in product development and process design for dried mango slices.

Acknowledgement
The authors express sincere thanks to the Rockefeller Foundation Yieldwise Initiative that supported
this research through the University of Nairobi Postharvest project. The first author wishes to
acknowledge the German Academic Exchange Service (DAAD) for scholarship award (Personal ref.
no: 91636877).

References
Abraham, S. I. et al. (2017) ‘Moisture Adsorption and Thermodynamic Properties of Sorghum-Based
Complementary Foods’, International Journal of Food Engineering and Technology, 1(1), pp. 1–8.
doi: 10.11648/J.IJFET.20160201.13.
Akoy, E. O. M. and von Hörsten, D. (2013) ‘Moisture sorption isotherms of mango slices’, International
Journal of Agricultural and Food Science, 3(2), pp. 164–170.
Akoy, E. O. M., Von Hörsten, D. and Ismail (2013) ‘Moisture adsorption characteristics of solar-dried
mango slices’, International Food Research Journal, 20(2), pp. 883–890.
Al-muhtaseb, A. H., Mcminn, W. A. M. and Magee, T. R. A. (2002) ‘Moisture sorption isotherm
characteristics of food products: A review’, Food and Bioproducts Processing, 80(June), pp. 118–
128.
Amankwah, E. A. et al. (2018) ‘Modeling the equilibrium moisture content of desorption and adsorption
of yam ( Dente )’, Agricultural Engineering International: CIGR Journal, 20(1), pp. 184–192.
Andrade, R. (2011) ‘Models of sorption isotherms for food:uses and limitations’, (September 2011), pp.
325–334.

This article is protected by copyright. All rights reserved

 AOAC (2005) ‘Official Methods of Analysis of the Association of Official Analytical Chemists.’,
Accepted Article
Arlington, VA: Association of Official Agricultural Chemists., 18th edn(February).
Ariahu, C. C., Kaze, S. A. and Achem, C. D. (2006) ‘Moisture sorption characteristics of tropical fresh
water crayfish (Procambarus clarkii)’, Journal of Food Engineering, 75(3), pp. 355–363. doi:
10.1016/j.jfoodeng.2005.03.062.
Arslan, N. and Tog, H. (2005) ‘Moisture Sorption Isotherms for Crushed Chillies’, Biosystems
Engineering, 90(1), pp. 47–61. doi: 10.1016/j.biosystemseng.2004.10.008.
Bi, Y. et al. (2018) ‘The prediction of moisture adsorption isotherm for sucrose powder in Côte d ’ Ivoire’,
06(04), pp. 1–9. doi: 10.18535/ijsrm/v6i4.ft01.
Biswal, S. et al. (2017) ‘Moisture desorption isotherms of fresh Jamun ( Syzygium cumini ) fruit’, Indian
Journal Of Agricultural Research, 51(3), pp. 267–271. doi: 10.18805/ijare.v51i03.7917.
Bouba, A. A. et al. (2014) ‘Desorption isotherms, net isosteric heat and the effect of temperature and
water activity on the antioxidant activity of two varieties of onion (Allium cepa L)’, International
Journal of Food Science and Technology, 49(2), pp. 444–452. doi: 10.1111/ijfs.12321.
Caballero-Ceron, C. et al. (2015) ‘Moisture Sorption Isotherms of Foods: Experimental Methodology,
Mathematical Analysis, and Practical Applications’, Water Stress in Biological, Chemical,
Pharmaceutical and Food Systems, pp. 187–214. doi: 10.1007/978-1-4939-2578-0.
Coşkun, A. and Pekmez, H. (2012) ‘Effect of drying methods on the moisture sorption isotherms and
thermodynamic properties of mint leaves’, Journal of Food Science and Technology, 49(August), pp.
439–449. doi: 10.1007/s13197-011-0302-7.
Greenspan, L. (1977) ‘Humidity Fixed Points of Binary Saturated Aqueous Solutions’, Journal of
research of the national bureau of standards, 81(1), pp. 89–96.
Gulzar, A. et al. (2018) ‘Effect of Blanching Techniques and Treatments on Nutritional Quality of Dried
Mango Slices During Storage’, 68(1). doi: 10.1515/pjfns-2017-0012.
Haque, A. H. et al. (2006) ‘Performance of an Accelerated Method for the Determination of Equilibrium
Moisture Content’, Food science and technology research, 12(1), pp. 1–7.
Klewicki, R. et al. (2015) ‘Sorption isotherms for osmo-convectively-dried and osmo-freeze-dried apple,
sour cherry, and blackcurrant’, Journal of Horticultural science and Biotechnology, 0316(75–79). doi:
10.1080/14620316.2009.11512599.

This article is protected by copyright. All rights reserved

 Lutovska, M. et al. (2017) ‘Different methods of equilibrium moisture content determination’, Journal on
Accepted Article
Processing and Energy in Agriculture, 4487(21), pp. 91–96.
Mariem, S. Ben and Mabrouk, S. Ben (2015) ‘Moisture Sorption Isotherms and Isosteric Heats of Sorption
of Tomato Slices’, American Journal of Renewable and Sustainable Energy, 1(3), pp. 140–155.
Nyangena, I. et al. (2019) ‘Effect of selected pretreatments prior to drying on physical quality attributes of
dried mango chips’, Journal of Food Science and Technology. Springer India. doi: 10.1007/s13197-
019-03857-9.
Ocheme, O. B., Ariahu, C. C. and Ingbian, E. K. (2013) ‘Moisture Sorption Characteristics of Dakuwa (
Nigerian Cereal / Groundnut Snack )’, 9(4), pp. 499–504. doi: 10.1515/ijfe-2012-0242.
Prasantha (2018) ‘Prediction of Moisture Adsorption Characteristics of Dehydrated Fruits Using the GAB
Isotherm Model’, Annals of Agricultural & Crop Sciences, 3(1), pp. 1–4.
Prasantha, B. D. R. and Amunogoda, P. N. R. J. (2013) ‘Moisture Adsorption Characteristics of Solar-
Dehydrated Mango and Jackfruit’, Food and Bioprocess Technology, 6(7), pp. 1720–1728. doi:
10.1007/s11947-012-0832-7.
Rangel, M. et al. (2011) ‘Estimation of Moisture Sorption Isotherms of Mango Pulp Freeze- dried .’,
International Journal of Biology and Biomedical Engineering, 5(1), pp. 18–23.
Rizvi, S. S. . (1986) Thermodynamic properties of foods in dehydration in engineering properties of foods.
Edited by MA Rao & SSH Rizvi,.
Rosalizan, M. S. et al. (2018) ‘Mathematical modelling of adsorption isotherms of Malaysian variety of
purple flesh sweet potato at different temperatures’, Thermal Science and Engineering Progress. doi:
10.1016/j.tsep.2018.07.007.
Timmermann, E. O., Chirife, J. and Iglesias, H. A. (2001) ‘Water sorption isotherms of foods and
foodstuffs : BET or GAB parameters ?’, Journal of Food Engineering, 48, pp. 19–31.
Zhang, L., Sun, D. and Zhang, Z. (2017) ‘Methods for measuring water activity ( a w ) of foods and its
applications to moisture sorption isotherm studies’, Critical Reviews in Food Science and Nutrition.
Taylor & Francis, 57(5), pp. 1052–1058. doi: 10.1080/10408398.2015.1108282.
Zuo, L., Rhim, J. and Lee, J. H. (2015) ‘Moisture Sorption and Thermodynamic Properties of Vacuum-
Dried Capsosiphon fulvescens Powder’, Preventive nutrition and food science, 20(3), pp. 215–220.

This article is protected by copyright. All rights reserved

Accepted Article
Table 1. Saturated salt solutions with respective standard relative humidity values for EMC
determination for dried mango at 25°C, 30°C and 35°C.

Standard relative humidity (%)

Saturated salt solution Temperatures
25°C 30°C 35°C
Potassium Hydroxide (KOH) 8.23 8.23 8.23
Potassium Acetate (KC2H3O2) 23 22.6 22
Magnesium Nitrate Mg (NO3)2 6H2O 52.89 52.8 51
Potassium Iodide (KI) 68.86 68 68
Sodium Chloride (NaCl) 75.25 75.2 75
Potassium Chloride (KCl) 83.34 84 84
Potassium Sulphate (K2SO4) 97.30 97 97
Source: Greenspan, (1977)

Table 2. Fitting parameters and % root mean square (RMS) for GAB model tested at different
temperatures for dried Apple and Ngowe variety mango slices.

Sample Temperature (°C) Mo(%d.b) C K % RMSE

Apple 25 9.15 33.96 0.85 4.21
30 8.43 36.17 0.86 5.36
35 8.00 27.61 0.87 8.52
Ngowe 25 8.80 23.40 0.87 7.27
30 8.15 21.21 0.88 6.73
35 8.17 8.20 0.87 5.82

C and K = Dimensionless constants associated with heat of sorption; Mo= monolayer moisture
content; RMSE= root mean square error.

This article is protected by copyright. All rights reserved

 jfpp_15041_f1-3.docx

55
Accepted Article
Equilibrium moisture content g/100g
50
EMC 25°C
45
EMC 30°C
40
EMC 35°C
35
30
25
20
15
10
5
0
0 0.2 0.4 0.6 0.8 1
Water activity, aw
Fig. 1. Temperature effects on the adsorption isotherms for dried Apple

55
50 EMC 25°C
Equilibrium moisture content g/100g

45 EMC 30°C
40 EMC 35°C
35
30
25
20
15
10
5
0
0 0.2 0.4 0.6 0.8 1
Water activity, aw
Fig. 2. Temperature effects on the adsorption isotherms for dried Ngowe

This article is protected by copyright. All rights reserved

 55 55
Experimental EMC (25 °C)

Equilibrium moisture content g/100g

50 50 Experimental EMC (25°C)
Predicted EMC (25 °C)
Equilibrium moisture content g/100g
Accepted Article 45 45 Predicted EMC (25)
40 40 (a)
(a)
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Water activity, aw Water activity, aw

55
55 Experimental EMC (30°C)

Equilibrium moisture content g/100g

Experimental EMC (30°C) 50
Equilibrium moisture content g/100g

50 Predicted EMC (30°C)

Predicted EMC (30°C) 45
45
(b)
40 (b) 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Water activity, aw Water activity, aw
55 55
Equilibrium moisture content g/100g

50 Experimental EMC (35°C) 50 Experimental EMC (35°C)

Predicted EMC (35°C)
Equilibrium moisture content g/100g

45 45
Predicted EMC (35°C)
40 (c) 40 (c)
35 35
30 30
25 25
20 20
15 15
10 10

5 5

0 0
0 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 0
Water activity, aw Water activity, aw
Fig. 4. Adsorption isotherms for Apple mango at Fig. 3. Adsorption isotherms for Ngowe mango at
25°C, 30°C and 35°C. 25°C, 30°C and 35°C
This article is protected by copyright. All rights reserved

View publication stats