Heliyon 8 (2022) e09341

Contents lists available at ScienceDirect

Heliyon
journal homepage: www.cell.com/heliyon

Research article

Drying kinetics of two fruits Portuguese cultivars (Bravo de Esmolfe apple

and Madeira banana): An experimental study
M.N.Coelho Pinheiro a, b, d, *, R.O. Madaleno a, Luis M.M.N. Castro a, c, d
a
Departamento de Engenharia Química e Biol
ogica, Instituto Superior de Engenharia do Instituto Polit
ecnico de Coimbra, Rua Pedro Nunes, Quinta da Nora, 3030-199,
Coimbra, Portugal
b
Centro de Estudos de Fen
omenos de Transporte, Faculdade de Engenharia da Universidade do Porto, Rua Dr. Roberto Frias, 4200-465, Porto, Portugal
c
CIEPQPF—Chemical Engineering Processes and Forest Products Research Center, Department of Chemical Engineering, Faculty of Sciences and Technology, University of
Coimbra, Rua Sílvio Lima, 3030-790, Coimbra, Portugal
d
Instituto Polit
ecnico de Coimbra, Instituto de Investigaç~
ao Aplicada, Laborat
orio SiSus, Rua Pedro Nunes, Quinta da Nora, 3030-199, Coimbra, Portugal

A R T I C L E I N F O A B S T R A C T

Keywords: Air convective dehydration was carried out at a laboratory scale using two fruits of cultivars produced in different
Drying kinetics regions of Portugal: Bravo de Esmolfe apple, from Beiras province, and Cavendish banana, from Madeira Island.
Characteristic drying curve Fresh fruits were dried in a tray drier with a hot airstream at different temperatures (35, 40, 45, and 50
C) and
Bravo de Esmolfe apple
velocity of 1.6 m s1. Drying rate curves were obtained using a simple mathematical approach applied to the
Madeira Island banana
moisture content curves adjusting linear and polynomial functions. Different drying rate stages were noticed in
the experiments made with apples (one constant drying rate period followed by two falling drying rate periods),
while in the case of the banana the constant drying rate period was not perceived, being dried entirely during a
unique falling drying rate period. As expected, the constant drying rate value obtained at the beginning of the
experiments with apples is higher when these were conducted at higher temperatures, changing from 8.103 to
14.474 g m2 s1 when the airstream temperature increases from 35 to 50
C. The correspondent critical moisture
contents in the Bravo de Esmolfe apples, at the instant the constant drying rate period stops and the drying rate
starts to fall, slightly decreases from 4.800 to 4.134 kgwater/kgdry solid.
This study explored for the first time the drying behavior of these two important fruits that have been
increasingly used in the food industry in Portugal, giving important information for the industrialization of its
production.

1. Introduction quality dried fruit, it is essential to have a good understanding of the

physico-chemical mechanisms that occur in the dehydration process.
The demand for fruit has increased considerably due to their content Snack food products consumption has increased exponentially
of vitamins, phenols and other antioxidants related to the prevention of worldwide over recent decades, already representing a significant part of
various cancers and degenerative diseases (Guerreiro et al., 2017). the daily intake of nutrients as they are consumed throughout the day
Although having an exceptional taste and notable health benefits, between the traditional three meals to promote satiety and suppress
fresh fruits and vegetables are subject to degradation processes by mi- overconsumption at the subsequent meals (Pires et al., 2021).
croorganisms and chemical and enzymatic reactions, mainly due to their The present research is dedicated to the air-drying of two Portuguese
high water content. Food industries use several strategies to improve the original types of fruits: “Bravo de Esmolfe” apples, a Protected Designa-
fruits stability and availability being the direct dehydration to a dry form tions of Origin (PDO) fruit growing in the enclosed region of north-
a commonly used method to increase the shelf-life of fresh fruit and to central Portugal at Beiras traditional province, and banana from
produce different healthy snacks. The dehydration process during which Madeira Island.
the water activity is reduced considerably defines the final fruit quality, Apples are characterized by their variety and balanced composition,
such as nutritional value, color degradation, texture and rehydration by a moderate energy value and by an adequate balance between sugar
capacity. In order to plan the dehydration process leading to a high and acid contents, giving it a pleasant taste, being one of the most

* Corresponding author.
E-mail address: [email protected] (M.N.Coelho Pinheiro).

https://doi.org/10.1016/j.heliyon.2022.e09341
Received 20 May 2021; Received in revised form 9 November 2021; Accepted 22 April 2022
2405-8440/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

consumed fruits worldwide. In Portugal, in 2018 an average of 24.1 kg of Ws dX

R¼  (1)
apples per capita were consumed, being the most important consumed A dt
fruit, showing its importance in the Portuguese diet (Feliciano et al.,
2010; FAOStat, 2020). where A stands for the drying area (the area exposed to the drying agent),
One of the most important commercialized apples are the Bravo de Ws is the weight of the solid completely dry and t is the time.
Esmolfe apple, a small to medium size Portuguese apple variety with pale- If the solid surface is saturated with water, the evaporation rate de-
yellow skin, white interior and an intense aroma, highly valued by con- pends on the airstream conditions (external mass transfer controlling)
sumers. This apple recognised as a PDO product is a high added value and X decreases linearly with t. During this period, the drying rate is
product with impact in the local and national economy, having an higher constant, Rc. Some products do not exhibit a constant drying rate period
price when compared to others varieties, like Starking and Golden (CDRP), particularly when dealing with food and agricultural products
(Feliciano et al., 2010; Pires et al., 2018; Rocha et al., 2004). Its pro- (Doran, 2013; Guin
e et al., 2014; Aghilinategh et al., 2015).
duction corresponding to 200,000 kg per year, have a growing com- When a solid surface starts to present dry zones, the mechanisms of
mercial demand, owing to its pleasing sensory characteristics, mainly moisture transport through the internal structure of the solid control the
taste and sweetness (Pires et al., 2018; Reis et al., 2009). drying rate (internal mass transfer controlling) and moisture migration to
Bravo de Esmolfe apple is harvested in the middle of September and its the surface is not enough to maintain the maximum drying rate. The solid
commercialization last no longer than May, being more intensive be- moisture content at this instant is designated by critical moisture content
tween December and February (Rocha et al., 2004). The commerciali- (Xc) and the drying rate decrease as the process continues. A falling
zation of dehydrated and snack fruit would be an important way to drying rate period (FDRP) starts and several phenomena are responsible
increase the longevity of this fruit on the market, increasing the envi- for moisture migration limitations through the solid. Moisture transport
ronmental and economic sustainability of this important Portuguese fruit due to capillary forces, induced by pressure gradients and liquid or vapor
production sector. diffusion are some of them, with their magnitudes depending on the solid
The other fruit studied is the Cavendish banana from Madeira Island, structural nature and drying operation conditions.
a small, sweet and tasty fruit. Bananas are cultivated in over 130 coun- Some solids may display more than one critical moisture content. In
tries and grow over a harvested area of approximately 10 million hect- general, this is observed when the principal phenomena associated with
ares (Villaverde et al., 2013; Vilela et al., 2014). The banana fruit has a moisture migration changes due to structural/chemical modifications
relevant nutritional value, providing a significant consumption of car- occurring in the solid (May and Parr
e, 2002). This behavior is reflected in
bohydrates, fiber, vitamins, and minerals, as well of phytochemicals, the drying rate curve with a change of shape. Experiments of drying ki-
including unsaturated fatty acids and sterols (Pontes et al., 2012; Vilela netics are of crucial importance for Xc determination as it is not only a
et al., 2014). solid property but it depends also on drying rate conditions.
The banana plant is a very important crop in the Madeira Island There are several models proposed in the literature to describe drying
economy. The sub-tropical climatic conditions of Madeira Island allowed kinetics. Some were developed from theoretical considerations and
the successful introduction of this crop in the middle of the sixteenth others are entirely supported by experimental results.
century (Oliveira et al., 2005). According to official statistics, the banana Page model is a popular and successful model in describing the drying
plantations produce annually about 30,000 tonnes of fruit, representing behavior of fruits and legumes. It is a semi-theoretical model approved by
20% of the agricultural production and about 30% of the exportations ASABE (American Society of Agricultural and Biological Engineers) for
from the Island (Oliveira et al., 2006; Vilela et al., 2014; DREM, 2020). modelling thin-layer drying of crops (ANSI/ASAE, 2014) and is repre-
“Dwarf Cavendish” (Musa acuminata Colla var cavendish) specie is sented by an exponential function with two parameters (k and n),
economically the most important cultivar in Madeira Island, and is also
X  Xe
the most relevant in productivity (ca. 50–60% of the total banana pro- ¼ expð  k t n Þ (2)
Xi  Xe
duction) and occupied area (Oliveira et al., 2005; Vilela et al., 2014).
The application of a drying step in the manufacture of foodstuffs, and where, Xi is the initial solid moisture content (dry basis) and Xe is the
in particular, of vegetables and fruits has the following advantages: i) to equilibrium solid moisture content.
increase the product stability in terms of conservation and consequently In a recent review study (Onwude et al., 2016), the thin-layer the-
the rise of its expiration dates due to its lower water activity; ii) to pre- ories, models, and applications, particularly to fruits and legumes drying
vent the development of microorganisms; iii) to facilitate the manipu- behavior, were presented. In this study, Page model (Equation 2) is
lation of the dried food by the customer; iv) to reduce the solid weight, referred to as one of the most suitable models to describe the drying
facilitating the transport and v) to reduce costs of dehydration compared kinetics of the fruits and legumes, particularly to represent the banana
with other drying processes such as freeze drying (Mujumdar, 2006). drying. In the study carried out by Kaleta et al. (2013), the drying
Moisture content of a solid is usually expressed in a dry basis, mass of characteristics of apple cubes (variety Ligol) were evaluated in a fluidized
water per mass of dry solid, and is designated by X. The drying curve of a bed drier, and Page model (Equation 2) was considered as one of the most
specific material is the plot of X versus t, representing the moisture con- appropriate models.
tent evolution in the solid along the drying process. Very often, by in- Empirical models are also widely used to obtain the relationship be-
spection of that curve, different behaviors in the solid drying are tween moisture content and the drying time and they adequately
observed due to the influence of the product internal structure in the describe the drying kinetics of various products, and some fruits and
mass transfer phenomena. Frequently, after a transitory period, at the legumes in particular (Onwude et al., 2016). The model of Wang and
beginning, where the solid temperature adjusts to the air drying condi- Singh (1978) considers a quadratic equation and was originally proposed
tions, there is an initial period where drying is controlled by the external to fit the data registered in a thin-layer of medium-grain rough rice
moisture transport through the airstream, after evaporation at the solid drying experiment,
surface, being the solid dried at the maximum drying rate. Afterwards,
the drying rate starts to decrease when the moisture internal transport X  Xe
¼ 1 þ a t þ b t2 (3)
becomes the dominant mechanism and is conditioned by the nature of Xi  Xe
the product being dried. In this phase one, or two, periods with different
behaviors in the drying curve is observed. where a and b are constants.
The instantaneous drying rate at which the moisture is removed from This empirical model is referred by Onwude and coauthors (2016) to
the solid can be obtained from drying curves (X versus t) as the slope dX/ explain adequately the drying behavior of bananas, supported in the
dt. If R represents the drying rate at any time t, thus: conclusions made by Omolola et al. (2014) and Fadhel et al. (2011).

2
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

Among the models tested by Omolola et al. (2014), the quadratic equa- 2. Materials and methods
tion (Equation 3) proposed by Wang and Singh was the one to be suitable
in describing the microwave drying kinetics of a thin-layer of Mabonde 2.1. Materials and samples preparation
banana variety. Fadhel and coauthors (2011) found the same conclusion
when essayed to describe the thin-layer behavior of banana in a forced Bravo de Esmolfe PDO apples, a fruit of the cultivar derived from the
convection indirect solar drier. Among the eight models tested, the Wang species Malus domestica Borkh, growing in the enclosed region of north-
and Singh model presented the higher values of the correlation central of Portugal at Beiras traditional province, and Cavendish ba-
parameters. nana cultivar from Madeira Island were selected to be studied in the
Kemp et al. (2001) proposed some methods for processing drying drying experiments. Both fruits were obtained fresh, in a local super-
kinetics raw data, but a simple and expeditious approach has been market and immediately dried or stored in a fresh local for a short period.
applying by the authors with success. This approach was presented in Care was taken to choose apples and bananas with the indication of
detail using the cork powder drying as a case study (Castro and Pinheiro, products marketed under-protected designations of origin and with the
2016) and afterwards, was used in processing drying data recorded in same category and caliber.
experiments with granular cork products with different sizes (Madaleno Apples and bananas were peeled and cut into 3 mm thickness slices.
et al., 2017) and foodstuffs as courgette (Pinheiro et al., 2015) and po- Each slice was cut in a (almost) square shape and placed in one tray dryer
tatoes (Madaleno et al., 2018). The data acquired during experiments, to obtain a uniform single layer covering all the tray surface (Figure 1). A
after being expressed in terms of moisture content is fitted using appro- clean metallic tray with dimensions 0.276 m  0.184 m  0.015 m was
priate functions. A linear function is firstly essayed to fit data corre- used. To prevent enzymatic browning reactions in the fruits, this phase of
sponding to the process beginning (after neglecting the initial transitory sample preparation and disposal in the tray was made as rapidly as
instants) possible (as recommended by Guin
e and Barroca (2014) for this banana
cultivar).
X ¼a t þb (4)
2.2. Convective hot air-drying experiments
and a second-degree (or higher) polynomial function is used to fit the
remaining data,
A laboratory tray dryer (Armfield Ltd, model UOP8) equipped with an
X ¼ c t 2 þ d t þ e or X ¼ f t 3 þ g t 2 þ h t þ i (5) axial flow fan and heating elements near the tunnel entrance was used to
perform the hot air convective drying kinetics experiments. In all the
where a, b, c, d, e, f, g, h and i are constants. experiments, the fan speed was maintained to obtain nearly the same hot
Therefore, the drying rates correspondent to CDRP and FDRP ac- airstream flow rate passing through the dryer tunnel. The heater power
cording to Castro and Pinheiro (2016) are obtained differentiating the was modified to change the air temperature downstream of the drying
functions fitted (Equation 4 and Equation 5), as described by Eq. (1). The chamber from 35 to 50
C.
procedure is repeated with new iterations until continuous functions of A humidity/temperature sensor (HigroClip2 from Rotronic) con-
drying rate are attained when the transition between the two periods nected to an external probe of the data logger HygroLog HL-NT3 was
occurs. Hence, at this step, the critical moisture content is determined. used to measure the temperature of the airstream downstream and up-
The proposed methodology could be also useful in detecting more than stream the drying compartment. The airstream velocity at the outlet of
one falling drying rate period if present. For more details of the meth- the tunnel was measured with an anemometer from Airflow De-
odology implemented, the description in Castro and Pinheiro (2016) is velopments (LCA6000).
recommended. One stainless steel tray with the fruit sample was placed in a metallic
Although several studies have been reported on the drying process of structure connected to a digital balance OHAUS, Adventurer Pro AV8101
various food products, there is still a lack of information regarding the (accuracy 0.1 g). The solid weight reduction was recorded in a com-
kinetics of the drying process of the two Portuguese fruits used in the puter with an acquisition frequency of 0.025 Hz using the HyperTerminal
present study. The typical aroma and texture of a fruit is characterized by software. When the weight recorded remains approximately constant,
the presence of a wide range of volatile metabolites, with different vol- revealing that the equilibrium with the air humidity prevailing in the air-
atilities and concentrations that can vary among the different cultivars drying stream is achieved, the experiment was interrupted.
from different regions (Pontes et al., 2012). Thus, the specificity of the During the drying experiment, the airstream conditions were peri-
characteristics of the two studied fruits also determines their behavior odically verified by measuring temperature, humidity, and velocity. The
during the drying process. Given the importance that these two fruits parameters used to characterize the drying conditions were determined
present and the relevance that dehydrated and snack fruit assume in the averaging the values registered during the experiments and are sum-
diet of modern societies, the increase in the knowledge of the drying marized in Table 1 with the corresponding standard deviation. Although
process of these typical Portuguese fruits is essential to allow the devel- the humidity in the airstream was measured downstream of the drying
opment of the food industry that improves the fruit sector supply chain compartment, the equipment used does not allow humidity control, and
and consequently the sustainability of this important economic activity in the humidity in the inlet air is the one from the laboratory environment.
Portugal. In the experiments performed with apple samples, the average value of
Two studies (Guin
e and Barroca, 2014; Guin
e et al., 2015) were the relative humidity in ambient air ranged from 49 to 56%, and for the
published on the topic of convective drying using the same Portuguese experiments with bananas, the variation was similar (43–51%).
variety of banana with origin in Madeira Island, but with different At the end of the experiments, the initial moisture content of the
objectives of the present study. Guin
e and Barroca (2014) dried the humid solid was determined by putting the entire contents of dried fruit
banana with hot air at different temperatures to evaluate the kinetics of in an oven at (104  1)
C and weighting until constant weight in a high
colour change during the drying process. Guin
e et al. (2015) dried the precision digital balance (Mettler Toledo PB3002, accuracy 0.0001 g).
banana with hot air at 50
C and 70
C in order to investigate the
impact on the phenolic content and antioxidant activity. The present 2.3. Drying kinetics data processing
work constitutes, to our knowledge, the first complete study concern-
ing the experimental characterization of the drying kinetics behavior at Monitoring the moisture content evolution during the drying process
different temperatures, of Bravo de Esmolfe apples and banana from of both fruits and further application of the mathematical approach
Madeira Island. proposed by Castro and Pinheiro (2016) allowed to determine some

3
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

Figure 1. Fruit samples arrangement in the tray before drying in the dryer: (a) apple samples and (b) banana samples.

ii) a linear equation (X ¼ a t þ b) followed by a third-degree poly-

Table 1. Hot air conditions used in drying experiments performed at different nomial function (X ¼ f t3 þ g t2 þ h t þ i);
temperatures with apple and banana. The mean values of the airstream tem- iii) a linear equation (X ¼ a t þ b) followed by an exponential function
perature and velocity followed by the associated standard deviation are depicted.
(X ¼ Xi exp (-k t n), Page model type).
apple banana


temperature ( C) velocity (m s 1
) temperature (
C) velocity (m s1) The linear function in the three scenarios tested was used to represent
35.13  0.05 1.60  0.00 35.0  0.2 1.59  0.02 the CDRP and the exponential and polynomial functions to represent the
40.01  0.08 1.59  0.01 40.0  0.1 1.59  0.01 unique FDRP or the two FDRP. The methodology proposed by Castro and
45.0  0.2 1.60  0.01 45.1  0.1 1.59  0.01 Pinheiro (2016) was used to detect the transition between the different
50.1  0.2 1.61  0.01 49.9  0.4 1.62  0.00
drying rate periods.
Concerning the processing of data obtained from the banana drying
experiments, the Page model and an empirical model were used as op-
tions to describe the fruit dehydration behavior, according to the state of
important aspects to be accounted for in potential industrial-scale the art presented in the Introduction. Hence, the two options tested were:
operations.
As discussed before, Page model is widely used in describing the i) a linear equation (X ¼ a t þ b) followed by a third-degree polynomial
dehydration behavior of fruit and legumes. Empirical models, like the function (X ¼ f t3 þ g t2 þ h t þ i);
quadratic equation proposed by Wang and Singh, are also successfully ii) a linear equation (X ¼ a t þ b) followed by an exponential function
essayed relating moisture content in foodstuffs and drying times as (X ¼ Xi exp (-k tn), Page model type).
already enhanced. Moreover, in the literature, some studies concerning
the kinetics of convective drying of apples suggest the existence of two Also in the banana drying modelling case, the linear function in the
FDRP (Feng et al., 2000; Veli
c et al., 2004; Nowak and Lewicki, 2005; two scenarios tested was used to represent the CDRP. According to the
Golestani et al., 2013; Guin
e et al., 2014). For example, Veli
c et al. (2004) bibliography, evidence exists for a unique FDRP when drying banana that
refer two well-defined FDRP and a short CDRP when drying Jonagold is represented by the polynomial and exponential functions in the two
apples in a convection tray dryer at 60
C and airflow rate velocities of possibilities tested.
0.64 and 1.00 m s1. The authors used an exponential model (like Page The methodology proposed by Castro and Pinheiro (2016) was used
model) for an approximation of data and drying curves calculation. to detect the transition between the different drying rate periods when
Nowak and Lewicki (2005) using infrared energy and convective drying the functions indicated are used. The correlation performance in
referred that apple slices dried in most cases in the FDRP. Only at the describing data was evaluated by statistical analysis for each scenario
lowest air velocity (0.5 m s1) a CDRP was observed. tested in both fruits, and the best fit was used to obtain the drying rate
To understand the best option in processing the data obtained in the curves at the different temperatures used in the experiments.
experiments performed in the present study, three different equation In general, a unique drying rate curve can be used for a specific ma-
types to model drying kinetics of Bravo de Esmolfe apples were tested: terial dried at different air conditions over a narrow range (Mujumdar,
2006). This curve is designated by characteristic drying curve (or
i) a linear equation (X ¼ a t þ b) followed by two different functions, normalized drying curve) and is obtained plotting R/Rc as a function of
a quadratic (X ¼ c t2 þ d t þ e) and a cubic equation (X ¼ f t3 þ g t2 the characteristic moisture ratio. The characteristic moisture ratio is a
þ h t þ i); dimensionless parameter calculated as the ratio of the free moisture

4
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

content at the instant t, ðX  Xe Þ, and the free moisture content at the 3. Results and discussion
critical time tc, ðXc  Xe Þ.
Applying this approach to food products is not always straightforward A set of experiments were planned to study the effect of the airstream
as in some cases the constant drying rate period is absent and it is not temperature on the drying performance of Bravo de Esmolfe apple and
available the value of the critical moisture content. When the existence of Madeira banana. The airstream flowing tangentially over the tray surface
a constant drying rate is not clear, Baini and Langrish (2007) and Jannot with the wet (continuous) fruit layer was forced to pass with virtually the
et al. (2004) suggested that Xc can be replaced by Xi (the initial value of same velocity and its temperature was modified from 35 to 50
C, with
X) and Rc can be taken as the higher value of R.

Figure 2. Dimensionless moisture content evolution during drying carried out with the air-drying stream (1.6 m s1) at different temperatures for: (a) Bravo de Esmolfe
apple samples and (b) Madeira banana samples.

5
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

increments of 5
C. A resume of the operating conditions used in the species of apples, was evaluated by Cruz et al. (2015) for several tem-
experiments can be seen in Table 1. perature conditions, in drying experiments with a halogen moisture
analyzer. Regarding the moisture of fresh apples, the authors reported
values that correspond to 4.56 kg/kg and 5.25 kg/kg (dry basis) for ap-
3.1. Drying curves
ples from cultivar Golden Delicious and Granny Smith, respectively, that
are somewhat lower than the initial moisture content determined for
Figure 2a shows the evolution of the Bravo de Esmolfe apples moisture
Bravo de Esmolfe in the present study.
content during the drying experiments carried out at the different tem-
The moisture content in bananas when equilibrium with air-drying
peratures. An equivalent representation is shown in Figure 2b for
conditions was reached (Xe ) does not change significantly (less than
Madeira banana. The moisture content of the fruits was divided by the
9.5%) in the experiments performed at different temperatures (Table 2).
correspondent Xi , and is represented as a dimensionless parameter (X=Xi )
In opposition, in the drying experiments carried out with apples, the
to avoid the effect introduced by the variability in Xi at the used samples.
residual moisture at higher temperatures (45 and 50
C) is about 25% less
An evident influence of the airstream temperature on the solid
than the values obtained in the experiments at 35 and 40
C.
moisture reduction can be observed from Figure 2, being the rate of
moisture depletion more intense when the air temperature is higher.
Consequently, the equilibrium conditions between the solid and the 3.2. Drying rate curves
drying agent were reached faster for the highest temperature used. For
apples, the equilibrium was reached in 390 min (23 400 s) and 600 Understanding the migration of water through the cellular structure
min (36 000 s) when the drying was at 50 and 35
C, respectively. The of food, and particularly in fruits like apples and bananas, is a complex
same trend was observed for Madeira banana, although the moisture issue. Water can be found in different locations within the food tissue in
reduction with time is shallower than in the apples drying, and to achieve different proportions: in the intracellular spaces; in the spaces between
a certain (dimensionless) moisture content the time required is greater. cells (intercellular) and in the cell walls (Khan et al., 2018). In opposition
The nature of the fruit conditioned the drying kinetics, the initial to the water residing inside the cells, the intercellular water behaves like
moisture content, and also the value of residual moisture when equilib- free water, and frequently is designated by capillary water. In its turn, the
rium was reached with the drying air conditions. Information about water in the cell wall is bounded more strongly than intracellular water
initial and equilibrium moisture contents in the fresh fruit samples used (Khan and Karim, 2017).
in all the drying experiments is given in Table 2. The average value ob- Apples and bananas present different microstructural characteristics,
tained for the initial moisture content in apples was (6.0  0.6) kg/kg dry the former with high porous tissue (Khan et al., 2016a) and the latter
solid and for bananas was (3.2  0.2) kg/kg dry solid. with a high starch content and similar to the low porous tissue of potato.
Initial moisture contents found in the literature for fresh bananas are This diversity in microstructure definitely will be associated with
similar to the average value encountered for Madeira bananas in this different behaviors in cell level water transport, certainly visible in the
study (see Table 2). For example, Baini and Langrish (2007) performed drying rate curves.
experiments where peeled ripe bananas were dried continuously for 72 h In order to select the best functions to adjust the moisture content
at temperatures of 60 and 80
C in a pilot-scale kiln dryer. The initial evolution in the fruits during drying and obtain the correspondent drying
moisture content obtained in four different banana samples shown a rate curves, some correlation parameters were calculated to evaluate the
coefficient of variation of 4.8%, being the average value of 3.617 kg/kg description performance of data by the functions tested. The root mean
(dry basis). A reference of the initial moisture content of banana fruits square error (RMSE), the determination coefficient (r2), and the chi-
from Madeira Island was found in the work of Barroca and Guin
e (2013). square test (χ 2) were used as correlation indicators of the functions
The value indicated by the authors is 2.06 kg/kg (dry basis), which is tested. Those criteria are traditionally accepted to assess model robust-
lower than the value encountered in this study for the Cavendish banana ness and the equations for their calculation can be found easily in the
from Madeira Island. It should be noted that the banana moisture content bibliography (Lee and Kim, 2009; Roberts et al., 2008; Vega-G
alvez et al.,
depends on the ripe stage (Corr^ea et al., 2012) as on the harvesting 2009). The values obtained for the functions tested to represent all the
season of the fruit (Nguyen and Price, 2007). experiments performed with apples are shown in Table 3. Values of r2
Concerning the value of moisture content for Bravo de Esmolfe apples, near the unity, and values of χ 2 and RMSE approaching zero indicate that
it seems to be slightly higher than other cultivars as Golden Delicious and the prediction is close to the experimental data. Analyzing those popular
Granny Smith, when comparing to values published for these fruit vari- parameters, it can be concluded that the best fit was obtained when a
eties. Product quality before and after convective drying, for those linear equation followed by two polynomial functions (i.e., considering
the existence of two FDRP) are used in data approximation, once this
scenario presented the smallest values of RMSE and χ 2, and the highest
value of r2. For that, this option was considered the most appropriated to
Table 2. Moisture content in both fruits (Bravo de Esmolfe apple and Madeira
banana) for fresh samples (before drying) and equilibrium conditions with the
fit drying kinetics data for Bravo de Esmolfe apple. However, it should be
airstream used in the experiments performed with different temperatures. For the enhanced that when a unique FDRP was assumed a good fit to data was
apple, the values obtained at the constant drying rate period and the moisture also obtained, probably due to a weak changing behavior between the
content at the transitions to the falling drying rate periods (Xc ) are also depicted. two FDRP. In opposition, when Page model was used the correlation
T (
C) 35 40 45 50
indicators are clearly worse, as can be observed in Figure 3. In fact,
considering the histograms of the deviations (Figure 4) it is evident that
Bravo de Xi (kgwater/kgdry solid) 6.838 6.045 5.682 5.239
Esmolfe apple in Page model the frequency of occurrence of high deviations is signifi-
Xe (kgwater/kgdry solid) 0.196 0.201 0.147 0.146
cant and in the other two options, the occurrence of deviations is
Rc (gwater m2 min1) 8.103 10.088 11.845 14.474
concentrated in the small ranges.
Xc * (kgwater/kgdry solid) 4.800 4.258 4.155 4.134
When applying the methodology proposed by Castro and Pinheiro
1.426 1.140 2.025 2.018
(2016) to detect the transition between the different drying rate periods
tc * (min) 85.26 67.20 53.51 34.32
271.39 229.37 137.82 104.53 in both options considered for processing data from drying experiments
Madeira Xi (kgwater/kgdry
with banana, no convergence of the method had been attained, indi-
solid) 3.039 3.023 3.535 3.255
banana Xe (kgwater/kgdry
cating the inexistence of a CDRP in this case. As was found in the liter-
solid) 0.210 0.199 0.240 0.214
ature, drying kinetic studies with bananas do not show the evidence of
*
transition between constant and first falling drying rate periods/transition such a period in the drying rate curve, referring to a unique FDRP. For
between first and second drying rate periods.
that reason, in Table 3, only the model constants f, g, h and i, and k and n

6
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

Table 3. Model constants and parameters used as criteria for selecting the best model to define drying curves for Bravo de Esmolfe apples and banana from Madeira
Island: root mean square error (RMSE), determination coefficient (r2) and, reduced chi-square (χ 2). The parameters determining the best quality of the fit are in bold.
apple 35
C 40
C

linear+empirical linear+empirical linear+Page linear+empirical linear+empirical linear+Page

model model model model model model
(two FDRP*) (one FDRP*) (two FDRP*) (one FDRP*)
model constants a 102 -2.436407 -2.402638 -2.421454 -2.688558 -2.640009 -2.666309
b 6.877598 6.864049 6.872123 6.064842 6.049799 6.058570
c 105 3.333160 - - 4.663870 - -
d 102 -3.004796 - - -3.315373 - -
e 7.126061 - - 6.290342 - -
f 108 -4.610000 -3.610000 - -7.700000 -5.977264 -
g 105 7.787260 6.462290 - 10.46253 8.618384 -
h 102 -4.403784 -3.842411 - -4.760146 -4.133162 -
i 8.554439 7.803983 - 7.479953 6.808940 -
k 106 - - 4.983710 - - 6.922362
n - - 1.304615 - - 1.296889
correlation RMSE 0.01094 0.01971 0.06481 0.00874 0.01640 0.06726
parameters r2 0.99997 0.99990 0.99888 0.99997 0.99991 0.99840
χ2 0.0001201 0.0003899 0.0042085 0.0000768 0.0002704 0.0045352
apple 45
C 50
C
linear+empirical linear+empirical linear+Page linear+empirical linear+empirical linear+Page
model model model model model model
(two FDRP*) (one FDRP*) (two FDRP*) (one FDRP*)
model constants a 102 -2.996639 2.945945 -2.978738 -3.508272 -3,349337 -3.399622
b 5.758742 5.744749 5.754283 5.337561 5.304671 5.317321
c 105 5.487220 - - 6.669760 - -
d 102 -3.583882 - - -3.966049 - -
e 5.921480 - - 5.434810 - -
f 108 -9.690000 -9.150000 - -18.02000 -17.19502 -
g 105 11.75382 11.23324 - 17.48122 16.82145 -
h 102 -4.759035 -4.601809 - -5.635577 -5.471018 -
i 6.604801 6.458611 - 6.211207 6.083763 -
k 106 - - 8.881999 - - 9.418408
n - - 1.294241 - - 1.319492
correlation RMSE 0.01173 0.01401 0.05582 0.01363 0.01562 0.06331
parameters r2 0.99994 0.99992 0.99874 0.99991 0.99988 0.99806
χ2 0.0001385 0.0001973 0.0031245 0.0001870 0.0002457 0.0040212
banana 35
C 40
C
linear+empirical linear+Page linear+empirical linear+Page
model model model model
(one FDRP*) (one FDRP*)
model constants f 109 -1.370608 - -2.646832 -
g 106 5.027515 - 7.884100 -
h 103 -6.261850 - -7.904981 -
i 2.875060 - 2.892177 -
k 105 - 9.475404 - 10.55466
n - 0.921827 - 0.931220
correlation RMSE 0.03822 0.04330 0.03675 0.05230
parameters r2 0.99724 0.99646 0.99741 0.99476
χ2 0.0014638 0.0018764 0.0013536 0.0027380
banana 45
C 50
C
linear+empirical linear+Page linear+empirical linear+Page
model model model model
(one FDRP*) (one FDRP*)
model constants f 109 -5.967737 - -8.528760 -
g 106 14.33405 - 18.07846 -
h 103 -11.60724 - -12.61829 -
i 3.429281 - 3.155837 -
k 105 - 9.846221 -
16.73897
n - 0.957166 - 0.918859
correlation RMSE 0.01797 0.06263 0.02326 0.08700
parameters r2 0.99956 0.99465 0.99905 0.98673
χ2 0.0003237 0.0039284 0.0005424 0.0075795
*
Falling Drying Rate Period.
7
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

Figure 3. Deviations between experimental data and predictions obtained with the tested models (linear þ empirical model) with one FDRP and two FDRP, and
(linear þ Page model) for Bravo de Esmolfe apples dried at 40
C.

Figure 4. Histogram for deviations between experimental data and predictions obtained with the tested models (linear þ empiric model) with one FDRP and two
FDRP, and (linear þ Page model) for Bravo de Esmolfe apples dried at 40
C.

are indicated, corresponding to the third-degree polynomial function and banana dehydration are higher for both tested models, as illustrated in
the exponential function used to describe the FDRP. Figure 5.
The parameters used to evaluate the model's statistical fitness in the The drying rate curves obtained from the best fitting functions
case of banana are presented in Table 3 and it can be concluded that the referred before, for both fruits, are shown in Figure 6 for the different hot
best fit was obtained when the polynomial function was used to represent airstream temperatures used in the experiments.
the FDRP. Therefore, this option was considered the most appropriated to During the apple drying, the moisture evaporation rate is constant for
fit drying kinetics data for the banana. However, it should be noted that higher moisture contents in the fruit and is observed a CDRP for all the
although a good statistical performance was obtained, when comparing used temperatures (Figure 6a), followed by a period where the drying
with the apples results the correlation parameters point out for a poor rate decreases continuously (FDRP) until the apple equilibrium moisture
correlation of data. In fact, deviations between predictions and data for content is reached. In opposition, Madeira banana does not exhibit an

Figure 5. Deviations between experimental data and predictions obtained with the tested models (linear þ empirical model and linear þ Page model) for Madeira
Island bananas dried at 40
C.

8
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

initial CDRP in all the conditions used. A unique FDRP was observed until modelled as a Fickian diffusion problem through micro-channels. The free
the end of the drying process (Figure 6b). The difference in drying be- moisture (10–20% of water is in an intercellular environment) probably
haviors of the two fruits probably can be indorsed to the dissimilarity in migrates to the surface by capillarity at the early drying stage, where
cellular structure. evaporates to the airstream passing over the upper surface of the thin-layer
The structure of apple tissue is highly porous and amorphous. Khan et al. of apple in the tray. Until this evaporation rate is compensated by the flow
(2016b) referred that about 80–90% of the water within the apple structure rate of (free) water coming by capillarity from the solid interior, heat
is inside the cells. Therefore, it is expected that during drying the moisture transfer in the airstream governs the drying process and a CDRP is observed.
migration within apple tissue occurs essentially from the intracellular As the driving force for heat transfer in the hot airstream passing tangen-
environment, which is a slow process unless the cell membranes rupture tially to the solid surface is higher when the temperature of the airstream
occurs. This cellular level mechanism of water transport is frequently increases, greater water evaporation rates from apple surface exposed to

Figure 6. Drying rate curves obtained with the air-drying stream (1.6 m s1) at different temperatures for: (a) Bravo de Esmolfe apple samples and (b) Madeira
banana samples.

9
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

drying agent are obtained for the experiments conducted at higher tem- drying of untreated samples carried out at temperature conditions
peratures (Figure 6a). As can be seen in Table 2, the value of Rc increases ranging from 40 to 70
C, were controlled by the internal resistance of
from 8.10 to 14.47 g m2 min1 when the apple was dried at 35 and 50
C, moisture transfer during the entire process. Doymaz (2010) when
respectively. When the water in the intercellular environment becomes drying thin-layer banana (Cavendish variety) slices in a hot air dryer at
scarce, the drying process starts to be governed by the migration of moisture temperatures varying from 50 to 80
C also reported the non-existence
located at intracellular spaces, and the drying rate decreases as the moisture of a CDRP. Baini and Langrish (2007) pointed out that it is not clear
content in the apple drops and an FDRP initiates. the existence of an unhindered drying period for bananas.
As in the early stage the moisture removed from Bravo de Esmolfe The moisture distribution in the food tissue changes if the cell
apples is unbound moisture, theoretically, the drying process is not membranes are damaged or suffer a rupture during the drying process.
influenced by the solid. The external conditions of the airstream (tem- Halder et al. (2011) dried potato and eggplant samples in an oven at
perature, humidity, and flow rate) control the water evaporation rate. For different temperatures and regularly measured the intercellular water
the sake of confirmation, an experiment with similar airstream condi- content by bio-electrical impedance analysis. The authors observed that
tions but with the tray filled with water instead of apple samples was at ca. 52
C the intercellular water content sharply decreased indicating
performed. In such an experiment, for an airstream at 40.0  0.5
C and cell membranes rupture at one single stage. However, Khan et al. (2017)
1.64  0.02 m s1 passing over the tray filled with water, the free water found a contradictory behavior and referred to the cell membrane
evaporation rate, per unit of the exposed area, was 10.59 g m2 min1. collapsing at different stages within potato tissue when drying occurs at
This value compares very well with the Rc value obtained in similar air- 60
C. The authors used an NMR method to identify the type of water at a
drying conditions (10.09 g m2 min1, see Table 2), being only 5% less, cellular level. Khan et al. (2018) extended the previous work, concerning
confirming that unbound moisture is removed from the apples at the the water transport at a cellular level in low porous food tissues, as is the
CDRP. case of potato tissue, to food products highly porous as apples. They used
Like the majority of the cells from mature plant tissues, in banana Granny Smith apple samples in drying experiments carried out in a
tissue, the cells are characterized to have a large central vacuole within cabinet dryer at different temperatures. They observed one period where
the cytoplasm besides the presence of several organelles and the cell wall. the amount of intracellular water remains almost constant until the apple
Starch granules also exist in banana cells in a large number at the early surface rose to 52–53
C, followed by a period with a rapid decrease.
ripening stages. When a banana presents a green peel, starch granules Once more, the authors argue that cell membrane rupture occurs at
occupy a considerable fraction of the volume cell, representing about different stages. According to the studies referred, it is expected that
20% of its weight (Lii et al., 1982). Banana ripening is accompanied by damage/rupture of cell membranes do not occur, at least significantly, as
chemical composition modifications with starch being rapidly converted the highest airstream temperature used in the experiments is lower than
to soluble sugars. the cell rupture conditions identified.
Proton relaxation times measurements have been used to charac- Observing Figure 6a) and Table 2, it can be seen that the solid
terize water distribution within plant tissues and food at a cellular critical moisture content (Xc) slightly decreases with increasing drying
level (Snaar and Van As, 1992; Hills and Remigereau, 1997). Raffo and temperature (only visible for apples, since experiments with bananas
coauthors (2005) used transverse relaxation times of water protons did not present a CDRP as already referred). This probably is due to
and self-diffusion coefficients NMR measurements to monitor the the mass transfer intensification within the fruit layer in the tray
ripening of bananas. Banana (Musa sp.) at a premature ripening stage towards the surface when the temperature is higher that easily
was used in the work of Raffo et al. (2005). From the results obtained maintains a continuous water film on the solid surface. As a result,
with samples of the same banana during seven days, the water present the FDRP starts at slightly lower (critical) solid moisture content
in the different sub-cellular environments was evaluated. The shortest values for higher temperatures. The Xc value decreases gradually from
relaxation time was attributed to the water in the cell walls and within 4.800 to 4.134 kgwater/kgdry solid when the air drying temperature in
starch granules and represented the smallest contribution to the the experiments performed with the Bravo de Esmolfe apples changed
overall banana moisture content. The intermediate relaxation time was from 35 to 50
C.
associated mainly with the water resident in the cytoplasm and the
highest time relaxation component was related to the water resident in 3.3. Normalization of drying rate curves
the vacuole. The vacuolar water represented during all the banana
ripening period more than 50% of the banana moisture content. The Drying rates for a specific product depend on complex phenomena as
authors concluded that no significant changes occurred in the water capillarity, diffusion, and heat transfer. Thus, the reduction to one
distribution inside the cells during banana ripening but no reference is normalized drying curve presenting a good performance in capturing the
made to the water in the intercellular environment. However, banana influence of the temperature of the drying agent on drying rates is a very
is a low porous food and similar characteristics in water distribution at useful method. If the method can properly explain this dependency
a cellular level to identical tissues are expected. Madiouli et al. (2011) probably exists a proportionality between the influence of drying agent
determined the apparent volume of a banana during convective drying temperature on the drying rate period governed by moisture diffusion
(26.7
C) using a non-intrusive stereovision technique, which allows within the solid (the FDRP) and the influence of drying agent tempera-
the 3D shape measurements from the images recorded. The banana ture on the drying rate period governed by the external conditions (the
shrinkage coupled with the information of banana's weight loss CDRP) (Suherman and Tsotsas, 2007).
allowed the porosity calculation. The authors referred to an initial Figure 7 presents the normalized drying rate curves for both fruits. As
value of 5% for banana porosity, which remains almost constant until referred before, there is no CDRP in the banana experiment and, as by
its moisture content decrease ca. 70%. Potato, also a low porous food, Baini and Langrish (2007) and Jannot et al. (2004), Xc was replaced by Xi
has mostly intracellular water at low temperatures (when cell mem- and Rc by the highest value of R to obtain the normalized curve for
branes are not damaged), and only ca. 2% of water is in the intercel- Madeira banana.
lular spaces (capillaries) (Halder et al., 2011). Probably similar Observing Figure 7, a unique curve appears to describe data obtained
characteristics can be attributed to the banana. Following this idea, the from drying experiments for the studied temperature range. This result
absence of a CDRP, as observed in Figure 6b, is expectable since the indicates that the shape of drying curves for Bravo de Esmolfe apples and
capillary water contribution to banana moisture content is not signif- Madeira bananas does not differ significantly for the airstream temper-
icant. In the literature, several studies corroborate that trend. Demirel atures used in the experiments. However, for higher temperatures a
and Turhan (2003) studied the air-drying behavior of two varieties of difference starts to be perceptible in the transition region between the
banana (Dwarf Cavendish and Gros Michel) and concluded that the two FDRP.

10
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

Figure 7. Normalized drying rate curves obtained from the experiments performed with different temperatures of the air-drying stream for: (a) Bravo de Esmolfe apple
samples and (b) Madeira banana samples.

4. Conclusions have an effect directly proportional on the resulting drying rate

magnitude.
Using a simple mathematical approach, different adjusting functions As far as the authors know, it was not available in the literature similar
were essayed to represent the moisture content evolution with time during information for both studied fruits of the Portuguese cultivars. Thus, this
drying of Bravo de Esmolfe apple, from Beiras province of Portugal, and paper contributes to reducing the lack of information regarding the ki-
Cavendish banana, from Madeira Island. The best fitting was obtained with netics of the drying process of these two fruits and gives relevant infor-
a linear function followed by two polynomial functions for the case of mation to support the industrialization of healthy snacks production that
Bravo de Esmolfe apples, indicating the existence of a constant drying rate have been showing an increasing acceptance for commercial purposes.
period and the presence of two falling drying rate periods. As expected, the
constant drying rate value is higher when the apples were dried at higher Declarations
temperatures, changing from 8.103 to 14.474 g m2 s1, and the critical
moisture content, at the instant the constant drying rate period changes to a Author contribution statement
falling drying rate period, slightly decreases from 4.800 to 4.134 kgwater/
kgdry solid, when the drying temperature increases from 35 to 50
C. M. N. Coelho Pinheiro, Luis M. M. N. Castro: Conceived and designed
For the case of Madeira banana a unique polynomial function was the the experiments; Analyzed and interpreted the data; Contributed re-
best fitting model, showing that the constant drying rate period was not agents, materials, analysis tools or data; Wrote the paper.
perceived. A falling drying rate period during all drying process exists. A R. O. Madaleno: Performed the experiments; Analyzed and inter-
possible reason for the different dehydration behavior, supported in the preted the data.
different structural nature of the fruits based on published information,
was presented and discussed. However, an effort to get a deep compre- Funding statement
hension of the phenomena involved must pursuit.
A unique curve was obtained when the drying rate curves at different This research did not receive any specific grant from funding agencies
temperatures were normalized, indicating that the airstream temperature in the public, commercial, or not-for-profit sectors.

11
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

Data availability statement Jannot, Y., Talla, A., Nganhou, J., Puiggali, J.R., 2004. Modeling of banana convective
drying by the drying characteristic curve (DCC) method. Dry. Technol. 22 (8),
1949–1968.
Data included in article/supplementary material/referenced in Kaleta, A., G
ornicki, K., Winiczenko, R., Chojnacka, A., 2013. Evaluation of drying models
article. of apple (var. Ligol) dried in a fluidized bed dryer. Energy Convers. Manag. 67,
179–185.
Kemp, I.C., Fyhr, B.C., Laurent, S., Roques, M.A., Groenewold, C.E., Tsotsas, E.,
Declaration of interests statement Sereno, A.A., Bonazzi, C.B., Bimbenet, J.J., Kind, M., 2001. Methods for processing
experimental drying kinetics data. Dry. Technol. 19 (1), 15–34.
Khan, M.I.H., Wellard, R.M., Pham, N.D., Karim, M.A., 2016a. Investigation of cellular
The authors declare no conflict of interest. level of water in plant-based food material. In: The 20th International Drying
Symposium (IDS 2016) Gifu, JAPAN, D-6-1, pp. 7–10.
Khan, M.I.H., Wellard, R.M., Nagy, S.A., Joardder, M.U.H., Karim, M.A., 2016b.
Additional information Investigation of bound and free water in plant-based food material using NMR T2
relaxometry. Innovat. Food Sci. Emerg. Technol. 38A, 252–261.
No additional information is available for this paper. Khan, M.I.H., Karim, M.A., 2017. Cellular water distribution, transport, and its
investigation methods for plant-based food material. Food Res. Int. 99 (1), 1–14.
Khan, M.I.H., Wellard, R.M., Nagy, S.A., Joardder, M.U.H., Karim, M.A., 2017.
References Experimental investigation of bound and free water transport process during drying
of hygroscopic food material. Int. J. Therm. Sci. 117, 266–273.
Aghilinategh, N., Rafiee, S., Gholikhani, A., Hosseinpur, S., Omid, M., Mohtasebi, S.S., Khan, M.I.H., Nagy, S.A., Karim, M.A., 2018. Transport of cellular water during drying: an
Maleki, N., 2015. A comparative study of dried apple using hot air, intermittent and understanding of cell rupturing mechanism in apple tissue. Food Res. Int. 105,
continuous microwave: evaluation of kinetic parameters and physicochemical quality 772–781.
attributes. Food Sci. Nutr. 3 (6), 519–526. Lee, J.H., Kim, H.J., 2009. Vacuum drying kinetics of Asian white radish (Raphanus
ANSI/ASAE, 2014. Thin-layer drying of agricultural crops. In: Revised Approved ASABE sativus L.) slices. LWT - Food Sci. Technol. (Lebensmittel-Wissenschaft -Technol.) 80,
Standards. American Society of Agricultural and Biological Engineers. S448.2). 200–207.
Available from: https://elibrary.asabe.org/azdez.asp?JID¼2&AID¼45096 Lii, C.-Y., Chang, S.-M., Young, Y.-L., 1982. Investigation of the physical and chemical
&CID¼s2000&T¼2&redir¼none&redirType¼standards.asp. properties of banana starches. J. Food Sci. 47 (5), 1493–1497.
Baini, R., Langrish, T.A.G., 2007. Choosing an appropriate drying model for intermittent Madaleno, R.O., Castro, L.M., Coelho Pinheiro, M.N., 2017. Drying kinetics of granulated
and continuous drying of bananas. J. Food Eng. 79 (1), 330–343. cork: effect of air drying stream conditions and granule size. Biomass Bioenergy 107,
Barroca, M.J., Guin
e, R.P.F., 2013. Variation of physical properties of banana along 8–9.
drying for cvs. In: Musa nana and Musa cavendishii. VII Congresso Ib
erico de Madaleno, R.O., da Silva, P.M., Castro, L.M., Coelho Pinheiro, M.N., 2018. Potatoes
Agroingeniería y Ciencias Hortícolas, Madrid, pp. 26–29. Agosto, Ref nº C0158. drying process – a vehicle to put into practice mathematical competencies in
Castro, L.M.M.N., Pinheiro, M.N.C., 2016. A simple data processing approach for drying engineering students. In: The 19th SEFI Mathematics Working Group Seminar on
kinetics experiments. Chem. Eng. Commun. 203 (2), 258–269. Mathematics in Engineering Education, 39–44.
Corr^ea, J.L.G., Braga, A.M.P., Hochheim, M., Silva, M.A., 2012. The influence of ethanol Madiouli, J., Sghaier, J., Orteu, J.J., Robert, L., Lecomte, D., Sammouda, H., 2011. Non-
on the convective drying of unripe, ripe, and overripe bananas. Dry. Technol. 30 (8), contact measurement of the shrinkage and calculation of porosity during the drying
817–826. of banana. Dry. Technol. 29 (12), 1358–1364.
Cruz, A.C., Guin
e, R.P.F., Gonçalves, J.C., 2015. Drying kinetics and product quality for May, B.K., Perr
e, P., 2002. The importance of considering exchange surface area
convective drying of apples (cvs. Golden delicious and Granny Smith). Int. J. Fruit reduction to exhibit a constant drying flux period in foodstuffs. J. Food Eng. 54 (4),
Sci. 15 (1), 54–78. 271–282.
Demirel, D., Turhan, M., 2003. Air-drying behavior of dwarf cavendish and Gros Michel Mujumdar, A.S., 2006. Handbook of Industrial Drying (Advances in Drying Science and
banana slices. J. Food Eng. 59 (1), 1–11. Technology). In: Dekker, M. (Ed.), third ed. New York.
Doran, P.M., 2013. Bioprocess Engineering Principles, second ed. Academic Press, Nguyen, M.H., Price, W.E., 2007. Air-drying of banana: influence of experimental
London; San Diego. parameters, slab thickness, banana maturity and harvesting season. J. Food Eng. 79
Doymaz, I., 2010. Evaluation of mathematical models for prediction of thin-layer drying (1), 200–207.
of banana slices. Int. J. Food Prop. 13 (3), 486–497. Nowak, D., Lewicki, P.P., 2005. Quality of infrared dried apple slices. Dry. Technol. 23
DREM - Direç~ ao Regional de Estatística da Madeira, 2020. Estatísticas da Agricultura e (4), 831–846.
Pescada - Regi~ ao Aut
onoma da Madeira (Agriculture and Hake Statistics - Oliveira, L., Freire, C.S.R., Silvestre, A.J.D., Cordeiro, N., Torres, I.C., Evtuguin, D., 2005.
Autonomous Region of Madeira). ISBN978-989-8755-58-2 (in Portuguese). Acessed Steryl glucosides from banana plant Musa acuminata Colla var cavendish. Ind. Crop.
from. https://estatistica.madeira.gov.pt/download-now/economica/agricultura-flo Prod. 22 (3), 187–192.
resta-e-pesca/prod-veg-prd-animal-pesca-pt/prod-vegetal-publicacoes-pt/send/74-p Oliveira, L., Freire, C.S.R., Silvestre, A.J.D., Cordeiro, N., Torres, I.C., Evtuguin, D., 2006.
roducao-vegetal-publicacoes/12357-agricultura-e-pesca-2019233.html. Lipophilic extractives from different morphological parts of banana plant “Dwarf
Fadhel, M.I., Abdo, R.A., Yousif, B.F., Zaharim, A., Sopian, K., 2011. Thin-layer drying Cavendish”. Ind. Crop. Prod. 23 (2), 201–211.
characteristics of banana slices in a force convection indirect solar drying. In: Recent Omolola, A.O., Jideani, A.I.O., Kapila, P.F., 2014. Modeling microwave drying kinetics
Researches in Energy and Environment - 6th IASME/WSEAS International and moisture diffusivity of mabonde banana variety. Int. J. Agric. Biol. Eng. 7 (6),
Conference on Energy and Environment, EE’11. 107–113.
FAOStat - Food and Agriculture Organization of the United Nations, 2020. http://faostat Onwude, D.I., Hashim, N., Janius, R.B., Nawi, N.M., Abdan, K., 2016. Modeling the thin-
.fao.org/. layer drying of fruits and vegetables: a review. Compr. Rev. Food Sci. Food Saf. 15,
Feliciano, Rodrigo P., Antunes, C., Ramos, A., Serra, Ana T., Figueira, M.E., 599–618.
Duarte, Catarina M.M., Carvalho, A., Bronze, Maria R., 2010. Characterization of Pinheiro, M.N. Coelho, Madaleno, R.O., Castro, L.M.M.N., Santos, L.M.T., 2015.
traditional and exotic apple varieties from Portugal. Part 1 – nutritional, Comparative study between infrared and hot air drying of courgette – kinetics curves
phytochemical and sensory evaluation. J. Funct.Foods 2 (1), 35–45. ISSN 1756-4646. and effective moisture diffusivity. In: ICEUBI2015 - International Conference on
Feng, H., Tang, J., Dixon-Warren, S.J., 2000. Determination of moisture diffusivity of red Engineering.
delicious apple tissues by thermogravimetric analysis. Dry. Technol. 18 (6), Pires, T.C.S., In^es Dias, M., Barros, L., Jos
e Alves, M., Oliveira, M.B.P.P., Santos-
1183–1199. Buelga, C., Ferreira, I.C.F., 2018. Antioxidant and antimicrobial properties of dried
Golestani, R., Raisi, A., Aroujalian, A., 2013. Mathematical modeling on air drying of Portuguese apple variety (Malus domestica Borkh. cv Bravo de Esmolfe). Food Chem.
apples considering shrinkage and variable diffusion coefficient. Dry. Technol. 31 (1), 240, 701–706.
40–51. Pires, T.C.S., In^es, M. Dias, Calhelha, Ricardo, C., Alves, Maria Jos
e, Santos-Buelga,
Guerreiro, Adriana C., Cust
odia, M.L.Gago, Faleiro, Maria L., Maria, G.C. Miguel, Celestino, Ferreira, Isabel, C.F.R., Barros, Lillian, 2021. Development of new bilberry
Antunes, Maria D.C., 2017. The effect of edible coatings on the nutritional quality of (Vaccinium myrtillus L.) based snacks: nutritional, chemical and bioactive features.
‘Bravo de Esmolfe’ fresh-cut apple through shelf-life. LWT 75, 210–219. Food Chem. 334, 127511.
Guin
e, R.P.F., Barroca, M.J., 2014. Evaluation of the browning kinetics for bananas and Pontes, M., Pereira, J., C^amara, J.S., 2012. Dynamic headspace solid-phase
pears submitted to convective drying. Curr. Biochem. Eng. 1 (2), 165–172. microextraction combined with one-dimensional gas chromatography–mass
Guin
e, R.P.F., Barroca, M.J., Gonçalves, F.J., Alves, M., Oliveira, S., Mendes, M., 2015. spectrometry as a powerful tool to differentiate banana cultivars based on their
Artificial neural network modelling of the antioxidant activity and phenolic volatile metabolite profile. Food Chem. 134 (4), 2509–2520.
compounds of bananas submitted to different drying treatments. Food Chem. 168 (1), Raffo, A., Gianferri, R., Barbieri, R., Brosio, E., 2005. Ripening of banana fruit monitored
454–459. by water relaxation and diffusion 1H-NMR measurements. Food Chem. 89 (1),
Guin
e, R.P.F., Cruz, A.C., Mendes, M., 2014. Convective drying of apples: kinetic study, 149–158.
evaluation of mass transfer properties and data analysis using artificial neural Reis, Sofia F.A.R., Rocha, Sílvia M., Barros, Ant
onio S., Delgadillo, Ivonne,
networks. Int. J. Food Eng. 10 (2), 281–299. Coimbra, Manuel A., 2009. Establishment of the volatile profile of ‘Bravo de Esmolfe’
Halder, A., Datta, A.K., Spanswick, R.M., 2011. Water transport in cellular tissues during apple variety and identification of varietal markers. Food Chem. 113 (2), 513–521.
thermal processing. AIChE J. 57 (9), 2574–2588. Roberts, J.S., Kidd, D.R., Padilla-Zakour, O., 2008. Drying kinetics of grape seeds. J. Food
Hills, B.P., Remigereau, B., 1997. NMR studies of changes in subcellular water Eng. 89 (4), 460–465.
compartmentation in parenchyma apple tissue during drying and freezing. Int. J. Rocha, A.M.C.N., Barreiro, M.G., Morais, A.M.M.B., 2004. Modified atmosphere package
Food Sci. Technol. 32 (1), 51–61. for apple ‘Bravo de Esmolfe. Food Control 15 (1), 61–64.

12
M.N.Coelho Pinheiro et al. Heliyon 8 (2022) e09341

Snaar, J.E.M., Van As, H., 1992. Probing water compartments and membrane permeability Vilela, C., Santos, S.A.O., Villaverde, J.J., Oliveira, L., Nunes, A., Cordeiro, N.,
in plant cells by 1H NMR relaxation measurements. Biophys. J. 63, 1654–1658. Silvestre, A.J.D., 2014. Lipophilic phytochemicals from banana fruits of several Musa
Suherman, M., Tsotsas, E., 2007. On the applicability of normalization for drying kinetics. species. Food Chem. 162, 247–252.
Dry. Technol. 26 (1), 90–96. Villaverde, J.J., Oliveira, L., Vilela, C., Domingues, R.M., Freitas, N., Cordeiro, N.,
Vega-G
alvez, A., Andr
es, A., Gonzalez, E., Notte-Cuello, E., Chacana, M., Lemus- Silvestre, A.J.D., 2013. High valuable compounds from the unripe peel of several
Mondaca, R., 2009. Mathematical modelling on the drying process of yellow squat Musa species cultivated in Madeira Island (Portugal). Ind. Crop. Prod. 42, 507–512.
lobster (Cervimunida jhoni) fishery waste for animal feed. Anim. Feed Sci. Technol. Wang, G.Y., Singh, R.P., 1978. Single Layer Drying Equation for Rough rice. Paper -
151 (3-4), 268–279. American Society of Agricultural Engineers, pp. 78–3001.
Veli
c, D., Planini
c, M., Tomas, S., Bili
c, M., 2004. Influence of airflow velocity on kinetics
of convection apple drying. J. Food Eng. 64 (1), 97–102.

13