Postharvest Biology and Technology 138 (2018) 64–73

Contents lists available at ScienceDirect

Postharvest Biology and Technology

journal homepage: www.elsevier.com/locate/postharvbio

Characterization of an innovative device controlling gaseous exchange in T

packages for food products
⁎
Giuseppe Altieri , Francesco Genovese, Attilio Matera, Antonella Tauriello,
Giovanni Carlo Di Renzo
SAFE − Scuola di Scienze Agrarie, Forestali, Alimentari ed Ambientali, Università degli Studi della Basilicata, Viale dell'Ateneo Lucano, 10, 85100 Potenza, Italy

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

Keywords: The work describes both the characterization carried out and the experimental results from an innovative device
Modified atmosphere packaging allowing the gas exchange across a sealed package. The device is especially suitable for fresh fruit and vegetable
Micro-device preservation during cold storage, as it allows for the management of the headspace atmosphere inside sealed
Gas exchange packages containing food products presenting metabolic activity.
Post-harvest storage
Characterization results confirm the device's suitability for managing the bidirectional gas flow. For each
tested device, the hydro-dynamical parameter and the diffusion parameter have been measured to define its
steady-state behaviour. The steady-state gas concentrations reached in the package show the aptness of the
devices for use with specific products; these equilibrium gas concentrations depend on the device type and on the
mass of product contained in the package.
The results of simulated storage demonstrate that some products (pomegranate arils and table grapes) require
preconditioning of the package free volume to reach the optimal gas concentrations in a short time. Other
products (black truffle) with very high metabolic activity could reach the optimal condition without package
preconditioning depending on the free volume.

1. Introduction minimal effect on the nutritional value or organoleptic quality of the

Packaging methods and surrounding atmosphere composition are
crucial for maintaining the organoleptic and nutritional characteristics
of a product over time and to extend its shelf-life (Kalia and Parshad,
2015). Market demand for organic and pesticide/sulphite-free products
creates increasing interest in treatments based on modified atmosphere
shock treatment and/or packaging using low levels of oxygen and other
gases having an effect on the metabolic activity of vegetable tissue or
pathogens. Consequently, the food market requires continuous product
innovation in packaging, thus creating a growing demand for new
technological solutions (Dainelli et al., 2008).
Until now, the approach has been the development of polymeric
films to achieve improvements in the organoleptic and microbial
quality of the fruits and vegetables treated, as it is well known that, for
proper preservation of foodstuffs, it is necessary to maintain the correct
gas composition within the package, and that this objective can be
achieved using the correct packaging materials.
Modified atmosphere packaging (MAP) technology gives the ad-
vantage of packaging perishable products after harvest allowing for
extension of their shelf-life, thereby restraining distribution costs with
product (Mastromatteo et al., 2010).
MAP technology basically consists of the substitution of the sur-
rounding atmosphere of a packaged product with a modified atmo-
sphere. The gases mainly used in MAP are CO2, N2 and/or O2
(Rutherford et al., 2007). In addition, MAP, in combination with low-
temperature storage, is an effective tool to prolong shelf-life of food
products because, generally, low O2 or high CO2 concentrations de-
crease produce respiration, postharvest pathogens and the rate of de-
terioration (Kader et al., 1989). Therefore, product deterioration pro-
cesses are slowed down, respiration decreases and shelf-life is
prolonged.
However, to prevent anaerobic conditions as a result of oxygen
consumption, it is necessary to use permeable materials so that oxygen
can freely flow from external air into the package. The choice of
packaging material is related to the required barrier effect against gases
and water vapour, depending on the desired residual amount of each
gas to be exchanged over time with the outside environment at steady-
state (Agar et al., 1999). Therefore, the packaging material is very
important, and it represents a strategic choice as it affects the final
characteristics of the product and a significant part of production costs

⁎
Corresponding author.
E-mail address: giuseppe.altieri@unibas.it (G. Altieri).

https://doi.org/10.1016/j.postharvbio.2017.12.012
Received 21 August 2017; Received in revised form 20 December 2017; Accepted 22 December 2017
0925-5214/ © 2017 Elsevier B.V. All rights reserved.
G. Altieri et al. Postharvest Biology and Technology 138 (2018) 64–73

Nomenclature L-MMAD Long type device

Mi Molar mass of gas “i” (kg mol−1)
Indices in vector and subscripts Mm Molar mass of gas mixture (kg mol−1)
MAP Modified atmosphere packaging
1 is O2 mc Model coefficient of CO2 respiration rate eq. (3)
2 is CO2 mo Model coefficient of O2 respiration rate eq. (2)
3 is N2 MMAD Micro-machined adhesive device
μ Gas viscosity (Pa s) N Vector of Ni
ΔC Vector of ΔCi Ni Moles of gas “i” in package (mol)
ΔCi Ci,out-Ci, differential molar concentration of gas “i” across NP Overall moles of gas in the package (mol)
the device (mol m−3) NF Vector of NFi
ΔPP Pout-PP, differential pressure across the device (Pa) NFi Moles of gas “i” in the package (mol) due to diffusion and
ΔPi Pi,out-Pi, differential pressure of gas “i” across the device hydro-dynamical mechanism
ΔPP0 Pout-PP0: differential pressure across the device (Pa) at nc Model coefficient of CO2 respiration rate eq. (3)
time t = 0 no Model coefficient of O2 respiration rate eq. (2)
Σνi “diffusion volume” of gas “i” (Fuller et al., 1966) Pi Partial pressure of gas “i” in the package (Pa)
Σνm “diffusion volume” of the gas mixture (Fuller et al., 1966) PP Pressure in the package (Pa)
ac Model coefficient of CO2 respiration rate eq. (3) PP0 Pressure in the package (Pa) at time t = 0
ao Model coefficient of O2 respiration rate eq. (2) Pi,0 Partial pressure of gas “i” in the package (Pa) at time t = 0
A Device active cross-sectional area (m2) Pi,out Partial pressure of gas “i” external to the package (Pa)
Ci Ni/V, molar concentration of gas “i” in the package Pi,x =Pi if ΔPP < 0, =Pi,out if ΔPP ≥ 0
(mol m−3) Pout Pressure external to the package (Pa)
Ci,out Molar concentration of gas “i” external to the package RRO2 Respiration rate expressed as consumption of O2
(mol m−3) (mol kg−1 s−1)
Cout Molar concentration of gas mixture external to the RRCO2 Respiration rate expressed as production of CO2
package (mol m−3) (mol kg−1 s−1)
Cx Vector of Ci,x RR Vector of respiration rates RRi
Ci,x =Ci if ΔPP < 0, =Ci,out if ΔPP ≥ 0 (mol m−3) RRi Component related to gas “i” of vector of respiration rates
CP NP/V, molar concentration of gas mixture in the package (mol kg−1 s−1)
(mol m−3) RQ Respiratory quotient (RRCO2/RRO2)
d Device active cross-sectional area diameter (m) PO2 Partial pressure of O2 (Pa)
Dij Diffusivity coefficient of gas “i” into gas “j” (m2 s−1) PCO2 Partial pressure of CO2 (Pa)
Dim Diffusivity coefficient of gas “i” into gas mixture (m2 s−1) R Gas constant (8.314 J mol−1 K−1)
dc Model coefficient of CO2 respiration rate eq. (3) S-MMAD Short type device
do Model coefficient of O2 respiration rate eq. (2) STP Standard temperature and pressure (0 °C and 105 Pa)
Fim Diffusion parameter of the device with regard to gas “i” T Constant temperature (K)
(m3 s−1) V Constant container volume (m3)
G Mass matrix of the differential equations set Vo Measuring apparatus internal volume (m3)
H Hydro-dynamic parameter of the device (m3 Pa−1 s−1) zc Model coefficient of CO2 respiration rate eq. (3)
HPVR Headspace to product volume ratio zo Model coefficient of O2 respiration rate eq. (2)
J Vector of Ji w Mass of product in the container (kg)
Ji Molar flow of gas “i” (mol m−2 s−1) Xi Ni/NP = Pi/PP, molar fraction of gas “i” in the package
JFi Molar flow of gas “i” (mol m−2 s−1) due to diffusion and Xi,out Pi,out/Pout, molar fraction of gas “i” external to the package
hydro-dynamical mechanism Xi,x =Xi if ΔPP<0, = Xi,out if ΔPP ≥ 0
JF Vector of JFi Xi,out Pi,out/Pout, molar fraction of gas “i” external to the package
kc Model coefficient of CO2 respiration rate eq. (3) XO2 Molar fraction of O2
ko Model coefficient of O2 respiration rate eq. (2) XCO2 Molar fraction of CO2
L Device active length (m)

Fig. 1. Working principle of the MMAD and its application to a container.

65
G. Altieri et al. Postharvest Biology and Technology 138 (2018) 64–73

(Rojas-Grau et al., 2009). 2002); therefore, RRO2 takes into account the O2 consumption rate, and
Furthermore, packaging manufacturers have created a multitude of RRCO2 takes into account the CO2 generation rate.
plastic films made with different materials and layers, different surface Even though the most suitable form of respiration rates is described
treatments or micro-holes in order to adapt the packaging material to using a Michaelis-Menten kinetic equation with a term that takes into
stored food due to the importance of packaging material being selec- account respiratory inhibition due to CO2 (Paul and Clarke, 2002;
tively permeable to gases (Restuccia et al., 2010). Geysen et al., 2005; Ersan et al., 2010; González-Buesa et al., 2009);
However, in most cases, these solutions have several drawbacks in nonetheless, equations providing a good empirical fit were used, in
terms of efficacy, cost-effectiveness and adaptability to existing order to easily use experimental data arising from previous works found
packaging machines (e.g., the requirement to be thermoweldable in in literature.
“flow pack” machinery or in welding machines). As fitting, a generic function composed of the ratio of two power-
Currently, manufacturers are searching for solutions to overcome law expressions was selected taking into account a possible competitive
the previously mentioned drawbacks, in particular, the requirement for inhibition due to CO2.
specific films to be adapted each time to the various product type and
desired storage duration. ao +ko(XO2 ) no
RRO2 =
The purpose of the present paper is to show the characterization 1 + mo(XO2 )do + zo XCO2 (2)
results of a device (BLOW®, Di Renzo et al., 2015, 2016) that makes it
possible the passive management of gas mixture inside a food packa- ac+kc(XO2 ) nc
ging. The device was characterized by the fluid-dynamic point of view RRCO2 =
1 + mc(XO2 )dc + zc XCO2 (3)
also using a simulation approach to test its suitability with some fruits
and vegetables. Separate trials are currently going on real living pro- The MMAD device was modelled as a cylindrical hole of active
duces as rocket leaves, pomegranate, pomegranate arils and table length “L” with a cross-sectional area “A” and diameter “d”.
grapes. The flow of gas “i” through the MMAD is produced by two super-
imposing mechanisms (Del-Valle et al., 2003): the first one being or-
2. Theoretical background dinary diffusion flow (as Knudsen flow is considered negligible), and
the second being hydro-dynamical flow due to differential pressure
The innovation is a micro-machined adhesive device (MMAD) di- across the MMAD, as described by Poiseuille's law in laminar flow
rectly applied to the package; it is equipped with a micro-structure conditions. Therefore, the overall molecular flow (JF) is composed of
inserted in the thickness of its main body allowing bidirectional gas diffusion flow and hydro-dynamical flow (Schneider, 1978; Del-Valle
flow across the container. et al., 2003; Webb, 2006; González-Buesa et al., 2009), as expressed by
The MMAD (Fig. 1) comprises a main body applicable to the con- Eq. (4).
tainer and having a first opening, a second opening and a capillary duct The equations for each considered gas “i” are:
made in the thickness of the main body and extending between the first
and the second opening. The first opening is intended to be interposed
⎛ ⎞ D π d4
between the inside of the container and the capillary duct and the JF i = Xi ⎜∑ JF k⎟ + im ΔCi + ⎜⎛ ΔPP ⎟⎞ Ci,x
L ⎝ 128 μLA ⎠ (4)
second opening is intended to be interposed between the capillary duct ⎝ k ⎠
and the outside of the container. The capillary duct extends along a
segmented and/or curved line (e.g. spiral) and its length is greater than The term ( π d4
128 μ L A
ΔPP )C i,x being the molar flow of gas “i” due to
the maximum linear dimension of the main body. hydro-dynamical volumetric flow (Poiseuille's law) caused by the dif-
In order to study and characterize the MMAD using MAP, the gas ferential pressure across the MMAD. The molar flow depends on ΔPP
flow was considered isothermal and package volume (V) constant; across the MMAD. If ΔPP < 0 then Ci,x = Ci as the gas “i” flows from
moreover, a simplified approach was used, as defined by González- the inside of the package to the outside atmosphere, if ΔPP > = 0 then
Buesa et al. (2009), Paul and Clarke (2002), and Del-Valle et al. (2003). Ci,x = Ci,out as the gas “i” flows from the outside atmosphere into the
The simplification regards the approximation of the isothermal package.
gaseous phase with a binary-like mixture comprising the gas “i” and the Because dNi/dt = Ji *A, Eq. (4) lead to Eq. (5) expressed in terms of
rest of the mixture itself, defining the diffusivity (Dim) of gas “i” into the moles of gases in the package (Ni and its first derivative) where, in
gas mixture (Zanderighi, 2001; Paul and Clarke, 2002; Del-Valle et al., addition, it has been considered the product respiration rate to take
2003). Indeed, rigorous application of Stefan-Maxwell equations to a account of generation and depletion mechanism of CO2 and O2.
multi-component gas mixture (Lam, 2006; Cussler, 2009) brings un-
necessary model complexity, as the binary diffusivities of gases in- A
⎧ Fim = L Dim
volved are very similar (Zanderighi, 2001; Paul and Clarke, 2002; Del- ⎪
π d4 1
Valle et al., 2003). ⎪ H = 128 L μ
⎪
The binary diffusivity Dij of gas “i” into gas “j” can be calculated
using the equation proposed by Fuller et al. (1966). In addition, this ⎪ dt k ( )
⎨ dNFi = Xi ∑ dNFk + Fim ΔCi + H ΔPP Ci,x
dt
equation can be used to estimate Dim for a binary-like gas mixture, ⎪ dNi dNF
⎪ = dt i + w RRi (5)
using the average molecular mass and diffusion volumes (Σνi and Σνm) ⎩ dt

of the gas mixture (Fuller et al., 1966; Del-Valle et al., 2003).

Eq. (5) shows that the terms H and Fim are characteristic parameters
0.0101325 1 1 T 7/4 of the MMAD: they represent a geometric parameter and a physical
Dim = + 2
PP Mi Mm parameter.
⎡
⎣
3 ∑ νi + 3 ∑ νm ⎤
⎦ (1) After defining the vector of respiration rates RR,
Respiration of packaged produce modifies the O2 and CO2 content,
this is taken into account considering how the product respiration rates ⎡− RRO2 ⎤
RR = ⎢ RRCO2 ⎥
are strictly related to O2 and CO2 content and to storage temperature.
⎢
⎣ 0 ⎥ ⎦
The most suitable form of this dependency is described using a
Michaelis-Menten kinetic equation with a term that takes account of
and the following matrices,
respiratory inhibition due to CO2 (Fonseca et al., 2002; Paul and Clarke,

66
G. Altieri et al. Postharvest Biology and Technology 138 (2018) 64–73

⎡1 0 0⎤ having solution:
I= ⎢ 0 1 0 ⎥ ,
⎣0 0 1⎦ Pi (t) − Pi,out F
= exp ⎛− im t⎞
Pi,o − Pi,out ⎝ V ⎠ (7)
⎡ F1m 0 0 ⎤
F= ⎢ 0 F2m 0 ⎥, When PP > Pout the test has been carried out with an experimental
⎢ 0 0 F3m ⎥
⎣ ⎦ apparatus that measured the ΔPP decrease versus time using the same
⎡ X 1 X 1 X1 ⎤ gas concentrations across the device, the flow due to diffusion vanishes
X= ⎢ X2 X2 X2 ⎥, and, in this case, Eq. (6) simplifies to:
⎢ X3 X3 X3 ⎥
⎣ ⎦ d Ni
= H ΔPP Ci
⎡1 − X1 − X1 − X1 ⎤ dt
G= I− X= ⎢ − X2 1 − X2 − X2 ⎥
⎢ − X3 − X3 1 − X3 ⎥ Summing over all gases:
⎣ ⎦
d NP
then, Eq. (5) can be transformed in matrix form considering all the = H ΔPP CP
dt
components, leading to Eq. (6) which represents the final equation in
matrix form to be solved. Hence, slightly transforming:
dN d ΔPP H
G = F Δ C+ H ΔPP C x + w G RR = ΔPP (ΔPP − Pout)
dt (6) dt V
The differential equations set (6) is stiff and cannot be solved by The solution is:
popular ordinary algorithms (e.g., Runge-Kutta); indeed, the mass ΔPP0
matrix G is singular because its determinant is 1-X1-X2-X3 and the sum ΔPP (t) PP0 ΔPP0 H Pout
= ≅ exp ⎛− t⎞
of all Xi equals 1 by definition. In addition, G is strongly linked to
variables to be found. Therefore, Eq. (6) requires specific algorithms to
Pout ΔPP0
PP0
+ exp ( H Pout
V )
t PP0 ⎝ V ⎠

be solved, being differential-algebraic equations (DAE) of the first ΔPP0 H Pout

as ≪ exp ⎛ t⎞
order. A specific algorithm to solve first order DAE that has been suc- PP0 ⎝ V ⎠ (8)
cessfully used to solve this specific problem can be found in Matlab's
ordinarily differential equations (ODE) suite (i.e., ode15 s function)
3. Materials and methods
(Shampine and Reichelt, 1997).
The steady-state equilibrium gas concentrations are obtained from
Two versions of MMAD (short S-MMAD and long L-MMAD, having
Eq. (6) when its derivative term vanishes; steady-state gas concentra-
different length of capillary duct) have been tested. MMAD fluid-dy-
tions are cleanly obtained by setting the ODE final time to a very high
namic behaviour has been characterized measuring its hydro-dynamic
value (e.g., 1000 d), and these final gas concentrations are independent
parameter H: trials have been carried out using pure air to eliminate the
of the package free volume and starting gas concentrations.
effect of the diffusion term, starting with a pressure differential of ap-
To measure MMAD characteristic parameters H and Fim, we con-
proximately 400 Pa decreasing to 10 Pa. The data have been collected
sider RR = 0 and two different experimental conditions, PP = Pout and
using a DAQCard acquisition board (AI-16X-E50, 16 ADC channels, 16
PP > Pout.
bits resolution, National Instrument S.r.l., Italy) and a differential
When ΔPP = 0, the hydro-dynamical flow vanishes and the flow is
pressure sensor (Gems Sensors 5266500LBHTI, ± 500 Pa, RS
only due to diffusion. Moreover, as PP = constant, NP is also constant,
Components S.r.l., Italy). The apparatus consists of two stainless steel
dN
( )
and the term ∑ dtk is zero; thus Eq. (6) simplifies to:
k
plates (Fig. 2) sealing the MMAD applied on an impermeable film of
dN high-density polyethylene. Gas is subsequently charged from one end
= F ΔC
dt (gas loading) into the internal apparatus volume (Vo, to be measured),
and the discharge begins through the MMAD after closing the inlet
Hence leading to three decoupled equations for each type of con-
valve. The apparatus has been enclosed into a thermostatic chamber
sidered gas “i”:
settled at 20 °C.
d ΔPi F Subsequently, diffusion parameters Fim have been measured: two
= − im ΔPi
dt V containers (internal volume of 1050 mL) have been sealed applying the

Fig. 2. Apparatus used to measure the hydro-dynamical parameter H of the two types of MMAD. The determination of apparatus internal volume, Vo, was carried out by adding two
known volumes.

67
G. Altieri et al. Postharvest Biology and Technology 138 (2018) 64–73

Table 1
Recommended storage parameters for tested produce items.

Produce Days Storage Temperature (°C) O2 (%) CO2 (%) Reference

Black truffle 30 4 7–10 10–15 Rivera et al. (2010)

Pomegranate arils 7–15 4 2–5 5–10 López-Rubira et al. (2005)
Cherry, sweet 14–21 0 3–10 10–15 Kader (1997), Cantwell (2001)
Strawberry 7–10 0 5–10 15–20 Kader (1997); Cantwell (2001)
Rocket leaves 15 4 3–7 5–10 Pereira et al. (2017)
Table grapes 60 0 2–5 1–10 Kader (1997), Mencarelli et al. (2005), Crisosto et al. (2002), Cantwell (2001)

Table 2
RRO2 and RRCO2 for selected products as obtained from listed references.

Produce Gas (range %) Respiration Rate (m3 kg−1 s−1) 10−9 STP Temp. (°C) Adj. R2 Reference

a (oorc) k (oorc) n (oorc) m (oorc) d (oorc) z (oorc)

Black truffle O2 (5–21) 2.338 24.85 1.078 0 0 0 4 1.00 González-Buesa et al. (2009)
CO2 (0–20) 2.708 25.99 0.9994 0 0 0 1.00

Pomegranate arils (Hicaz) O2 (2–21) 0 22.66 1 26.32 1 0.8681 4 0.87 Ersan et al. (2010)
CO2 (0–20) 0 16.48 1 26.32 1 0.8681 0.87

Cherry, sweet (Bing) O2 (1–19) −0.01349 7.098 0.5866 13.65 2.294 0 0 1.00 Wang and Long (2014)
CO2 (0–18) 0.1711 46.41 1.255 28.59 1 0 1.00

Strawberry (Elsanta) O2 (1–21) 2.015 0 0 0.009 −1 0 0 1.00 Geysen et al. (2005)

CO2 (0–20) 1.332 0.006981 −1 0.01304 −1 0 1.00

Rocket Leaves (Diplotaxis tenuifolia) O2 (2–18) 1.661 0 0 0.01023 −1 0 4 1.00 Pereira et al. (2017)
CO2 (0–16) 1.867 0 0 0.01023 −1 0 1.00

Table grapes O2 (1–21) 0 6.402 1 0 0 6.992 5 1.00 Del Nobile et al. (2008)
CO2 (0–13) 0 4.706 1 0 0 6.992 1.00

Table 3 Table 5
Results of the trials carried out to measure the hydro-dynamical parameter H of the two Parameters H and Fim for each MMAD and for each gas referenced to 5 °C, 4 °C and 0 °C.
types of MMAD. The data were fitted using Eq. (8). Matlab's Curve Fitting Toolbox was
used to fit the data to the model “a/(b + exp(t/τ))". The standard deviation of the mean is T (°C) Fim (O2) Fim (CO2) Fim (N2) H (m3 Pa−1 s−1)
shown in parenthesis. (m3 s−1) 10−9 (m3 s−1) 10−9 (m3 s−1) 10−9 10−11

Device type Repetitions Time constant τ (s) H (m3 Pa−1 s−1) 10−11 S-MMAD
20 1.51 2.21 1.11 2.65
S-MMAD 9 10.49 (0.19) 2.65 (0.096) 5 1.38 2.02 1.01 2.76
L-MMAD 8 20.56 (0.53) 1.35 (0.069) 4 1.37 2.00 1.01 2.77
0 1.33 1.95 0.981 2.80

L-MMAD
two types of MMAD. Each container, in turn, has been flushed (2 min at 20 1.38 2.06 1.06 1.35
50 L/h) with three types of commercial gaseous mixtures (5% CO2–5% 5 1.26 1.88 0.967 1.41
O2, 10% CO2–10% O2 and 10% CO2–5% O2). Subsequently, the con- 4 1.25 1.87 0.961 1.41
0 1.22 1.82 0.937 1.43
tainer has been enclosed into a thermostatic chamber settled at 20 °C.
Three trials for each of the two MMAD versions have been carried out
with a duration of approximately 7 d. A CheckMate3 instrument literature and fitted to Eqs. (2) and (3). Table 2 shows the coefficients of
(Dansensor Italia S.r.l., Italy) has been used to measure the gas con- respiration rates in Eqs. (2) and (3) used in the simulation for the se-
centration (O2% and CO2%) hourly inside the container. lected produce items.
The average Lab atmospheric pressure was 922.0 hPa.
To test the suitability of the MMADs used in MAP, some produce
items have been selected for use in simulations (black truffle, pome- 4. Results and discussion
granate arils, cherry, strawberry, rocket leaves and table grape). From
references listed in Table 1, the recommended storage temperature and 4.1. Determination of the hydro-dynamical parameter H
commonly advised gas concentrations used in MAP have been collected.
The respiration rates of selected products have been found in Preliminarily, the internal apparatus volume, Vo, was measured by
adding two known volumes to it. A discharge trial series has been

Table 4
Results of the trials carried out to measure the time constant τ and the diffusion parameter Fim for each gas (O2 and CO2) of the two types of MMAD. The data were fitted using Eq. (7).
Matlab's Curve Fitting Toolbox was used to fit the data to the model “exp(-t/τ))". The standard deviation of the mean is shown in parenthesis.

Device type Repetitions O2 CO2 N2

Time constant τ (h) Fim (m3 s−1) 10−9 Time constant τ (h) Fim (m3 s−1) 10−9 Time constant τ (h) Fim (m3 s−1) 10−9

S-MMAD 3 193.5 (4.8) 1.51 (0.036) 132.0 (3.3) 20.56 (0.054) 263.9 (11.9) 1.11 (0.053)
L-MMAD 3 211.6 (3.7) 1.38 (0.024) 141.6 (1.5) 2.06 (0.022) 274.8 (3.0) 1.06 (0.012)

68
G. Altieri et al.
Fig. 3. Steady-state simulation results of S-MMAD applied to black truffle (a) and pomegranate arils (b), both stored at 4 °C with varying product mass in a sealed package. The dotted
lines represent the recommended interval for O2 and CO2 gas concentration.
carried out using a calibrated hole (diameter 0.25 mm) and pure air to
further guarantee the vanishing of terms due to diffusion. Using Eq. (8)
as a model and Matlab's Curve Fitting Toolbox permitted the time
constant (τ = V/H/Pout) of Eq. (8) to be measured for each of two
added volumes (V1 = 25.5 mL and V2 = 76.9 mL). The ratio of mea-
sured time constants is related to volumes because τ1/τ2 = (V1 + Vo)/
(V2 + Vo). Fig. 2 shows the apparatus used for the trials carried out.
The measured internal volume of the apparatus was Vo = 28.2 mL with
a standard deviation of the mean of 0.9 mL.
Subsequently, the two versions of the MMAD were fixed on a
polyethylene disk and then positioned between the two stainless steel
plates of the measuring apparatus. Again, pure air was used to eliminate
the effect of diffusion terms, starting with a differential pressure of
approximately 400 Pa. Eq. (8) was used to fit the experimental data
obtaining the time constants related to each of the two types of MMAD.
Table 3 shows the results of the trials carried out to measure the
hydro-dynamical parameter H of the two types of MMAD.
4.2. Determination of the diffusion parameter Fim
The diffusion parameter Fim for each gas (O2 and CO2) and for each
MMAD was measured by measuring hourly the molar fraction of gases
inside a sealed container positioned in a temperature controlled
69
Postharvest Biology and Technology 138 (2018) 64–73
chamber settled at 20 °C.
Eq. (7) was used to fit the experimental data estimating the time
constant (V/Fim). N2 diffusion parameters were determined by fitting
the data coming out from the simulation.
Table 4 shows the results of the trials carried out to measure the
diffusion parameter Fim for each gas (O2 and CO2) and for each MMAD.
4.3. Simulation of MMAD behaviour at steady-state
Some simulations of the MMAD were carried out using Eq. (6). The
simulation temperature was hypothesized to be 5 °C, 4 °C or 0 °C. To this
aim, the MMAD parameters have been referred to these temperatures
using the temperature dependency as established by Eq. (1) for Fim. For
H the temperature dependency of viscosity in Sutherland's formula was
used (Sutherland, 1893). These new values, used in the simulation, are
listed in Table 5. The probable temperature effect on the shrinkage of
the MMAD material has not been considered.
Matlab's ODE suite has been used to solve Eq. (6) in matrix form
with added product interference using Eqs. (2) and (3).
The steady-state equilibrium gas concentrations reached in the
package headspace during MAP storage show the suitability of the
MMAD for use with specific products; these equilibrium concentrations
depend on the MMAD and on the mass of product in the package.
G. Altieri et al.
Fig. 4. Steady-state simulation results of S-MMAD applied to cherry (a) and strawberry (b), both stored at 0 °C with varying product mass in a sealed package. The dotted lines represent
the recommended interval for O2 and CO2 gas concentration.
Fig. 3 (a) presents the simulation results of the S-MMAD applied to
black truffle. The atmosphere in the package is within the re-
commended range for MAP storage (Table 1) when the product mass
varies from 35 to 49 g. When using the L-MMAD (data not shown), the
product mass variation guaranteeing the recommended MAP atmo-
sphere is from 32 to 45 g. Therefore, a standard commercial gas mixture
(e.g., 10% O2 and 10% CO2) could be used to safely start the MAP
headspace for this specific product as long as the product mass is within
the allowed range.
Fig. 3 (b) shows the simulation results of the S-MMAD used for MAP
storage at 4 °C of pomegranate arils. The atmosphere in the package is
within the recommended range for MAP storage (Table 1) when the
product mass varies from 440 to 671 g. Using the L-MMAD (data not
shown), the product mass variation guaranteeing the recommended
MAP atmosphere ranges from 403 to 667 g. Therefore, a standard
commercial gas mixture (e.g., 5% O2 and 5% CO2) could be used to
safely start the MAP headspace for this specific product as long as the
product mass is within the allowed range.
Fig. 4 presents the simulation results of the S-MMAD applied to
cherry and strawberry. The data show the unsuitability of MMAD use
for cherry storage in MAP. The same negative results are obtained when
considering strawberry.
Fig. 5 (a) shows the simulation results of S-MMAD used for MAP
70
Postharvest Biology and Technology 138 (2018) 64–73
storage at 4 °C of rocket leaves (Diplotaxis tenuifolia). The data show the
unsuitability of MMAD use for rocket leaves in MAP. However, some
preliminary trials confirm the results of the suitability of MMAD use in
MAP for rocket leaves (Eruca Sativa) (Matera et al., 2017). In these
trials, the MMAD significantly improved vitamin C content irrespective
to MAP used, significantly improved appearance score with MAP and
significantly improved odour score irrespective to MAP. In addition, the
suitability of MMAD use in MAP for table grapes, pomegranates and
pomegranate arils is also emerging (unpublished results).
Fig. 5 (b) shows the simulation results of S-MMAD used for MAP
storage at 5 °C of table grapes. The atmosphere in the package is within
the recommended range for MAP storage (Table 1) when the product
mass varies from 1059 to 2060 g. Using the L-MMAD (data not shown),
the product mass variation guaranteeing the recommended MAP at-
mosphere is from 965 to 2115 g. Therefore, a standard commercial gas
mixture (e.g., 5% O2 and 5% CO2) could be used to safely start the MAP
headspace for this specific product as long as the product mass is within
the allowed range.
4.4. Simulation of MMAD transient behaviour
To further examine the products suitable for use in conjunction with
MMADs (i.e., black truffles, pomegranate arils and table grapes), it has
G. Altieri et al.
Fig. 5. Steady-state simulation results of S-MMAD applied to rocket leaves (Diplotaxis tenuifolia) (a) stored at 4 °C and table grapes (b) stored at 5 °C, with varying product mass in a sealed
package. The dotted lines represent the recommended interval for O2 and CO2 gas concentration.
been determined the package free volume that could create, using
product respiration starting from the atmospheric environment, the
recommended gas concentration inside the package within 12 h of
storage without having to preliminarily fill the package free volume
with a modified atmosphere.
To this aim, the headspace to product volume ratio (HPVR) has been
defined as the ratio between the free volume and the volume occupied
by the product itself. The product volume has been calculated using a
true density of 1000 kg m−3 for black truffle, 1089 kg m−3 for pome-
granate arils (bulk density of 679 kg m−3, Riyahi et al., 2011) and
1000 kg m−3 for table grapes (bulk density of 638 kg m−3). HPVR has
been set from 0.05 to 1.45 with increments of 0.20. The product mass
has been set to 45 g for black truffle, 500 g for pomegranate arils and
1400 g for table grapes, the storage temperature being the same as
previously considered. In all cases an S-MMAD has been used to control
the MAP.
HPVR is not dependent on product mass, the ratio between free
volume and volume occupied by the product being an alternate form to
take account of porosity. Pomegranate arils can reach a minimum
HPVR of 0.60, whereas table grapes show a minimum HPVR of 1.57.
Fig. 6 shows the simulation results related to MAP storage at 4 °C of
45 g of black truffle. The steady-state gas concentration values reached
are 7.7% for O2 and 11.7% for CO2 (Table 6), inside the recommended
71
Postharvest Biology and Technology 138 (2018) 64–73
MAP storage values for this product. At low values of HPVR, corre-
sponding to a small free volume, O2 is rapidly consumed and CO2 is
quickly produced by the product respiration and steady-state con-
centrations levels are speedily reached. At high values of HPVR, cor-
responding to a big free volume, the steady-state concentrations levels
are reached slowly. The maximum allowable HPVR to meet the re-
commended gas concentration for MAP within 12 h of cold storage is
1.42; beyond this limit value, the free volume is too large, and the re-
commended range of gas concentration for storage is entered at a time
after 12 h.
Table 6 shows the simulation results related to MAP storage of 500 g
of pomegranate arils stored at 4 °C and 1400 g of table grapes stored at
5 °C.
Pomegranate arils reach a steady-state gas concentration of 4.2% for
O2 and 9.3% for CO2, and these values are inside the recommended
storage values for this product. The maximum allowable HPVR to meet
the recommended gas concentration for MAP within 12 h of cold sto-
rage is 0.09; beyond this limit value, the free volume is too large, and
the recommended range of gas concentration for storage is entered at a
time after 12 h. As 0.60 > 0.09, the MAP storage of pomegranate arils
cannot be started at atmospheric gas concentrations.
Table grapes reach a steady-state gas concentration of 4.2% for O2
and 9.9% for CO2, which is inside the recommended values for MAP
G. Altieri et al. Postharvest Biology and Technology 138 (2018) 64–73

Fig. 6. Transient simulation results of S-MMAD applied to black truffle stored at 4 °C with varying headspace, starting from gas concentrations equal to the atmospheric environment. The
upper graphs represent the evolution versus time of the gas concentration inside the package varying HPVR from 0.05 to 1.45 with steps of 0.20. The maximum HPVR to meet the
recommended gas concentrations for MAP storage within 12 h is determined. The dotted lines in the upper graphs represent the recommended interval for O2 and CO2 gas concentration.

Table 6 parameter have been measured in order to characterize the device's

Transient simulation results of S-MMAD applied to various types of produce with varying steady-state behaviour. The steady-state equilibrium gas concentrations
HPVR, starting from gas concentrations equal to the atmospheric environment. The reached in the package free volume during MAP storage show the fit-
maximum allowed HPVR to meet the recommended gas concentrations for MAP storage
within 12 h is determined.
ness of the MMAD for use with specific products; these equilibrium
concentrations depend on the MMAD type and on the mass of product
T (°C) Produce Mass (g) Steady- Steady- Maximum contained in the package.
state CO2 state O2 allowed HPVR The results obtained with the simulated storage conditions demon-
(%) (%)
strate that for some products (pomegranate arils and table grapes)
4 Black truffle 45 11.7 7.7 1.42 preconditioning of the package free volume is required to reach the
4 Pomegranate arils 500 9.3 4.2 0.09 optimal MAP in a short time. Other products (black truffle) with very
(Hicaz) high metabolic activity could reach the optimal condition without
5 Table grapes 1400 9.9 4.2 0.07
preconditioning if the maximum allowed headspace to product volume
ratio is not outreached.
storage of this product. The maximum allowable HPVR to meet the On the basis of these results, MMAD represents a suitable device to
recommended gas concentration for MAP within 12 h of cold storage is be used in MAP applications, especially for fruit and vegetables, despite
0.07; beyond this limit value, the free volume is too large, and the re- not all produce items being appropriate for use in conjunction with the
commended range of gas concentration for storage is entered at a time MMAD models tested during the experimental trials. Each product re-
after 12 h. Once again, as 1.57 is beyond 0.07, even in this case, un- quires an MMAD breathing capacity adequate to its metabolism.
fortunately the MAP storage of table grapes cannot be started at at- Moreover, the opportunity to manage, during long-term storage, the
mospheric gas concentrations. internal atmosphere composition gives MMADs great potential for fu-
ture application on all fresh vegetable products and, in particular, on
organic vegetable products, further improving the shelf-life and redu-
5. Conclusions cing the product loss and waste due to physiological and pathological
deterioration at all levels of the distribution/consumption chain, with
The MMAD characterization results confirm the device's suitability particular reference to the domestic one.
for management of the bidirectional gas flow across a container. For
each tested MMAD, the hydro-dynamical parameter and the diffusion

72
G. Altieri et al.
Acknowledgments
“This research was carried out in the framework of the project
‘Smart Basilicata’ which was approved by the Italian Ministry of
Education, University and Research (Notice MIUR n.84/Ric 2012, PON
2007–2013 of 2 March 2012) and was funded with the Cohesion Fund
2007–2013 of the Basilicata Regional authority."
References
Agar, I.T., Massantini, R., Hess-Pierce, B., Kader, A.A., 1999. Postharvest CO2 and
ethylene production and quality maintenance of fresh-cut kiwifruit slices. J. Food Sci.
64, 433–440.
Cantwell, M., 2001. Properties and Recommended Conditions for Long-Term Storage of
Fresh Fruits and Vegetables. UC-Davis Postharvest Technology(http://ucanr.edu/
sites/Postharvest_Technology_Center_/files/230191.pdf. Accessed March 2017).
Crisosto, C.H., Garner, D., Crisosto, G., 2002. Carbon dioxide-enriched atmospheres
during cold storage limit losses from Botrytis but accelerate rachis browning of
‘Redglobe’ table grapes. Postharvest Biol. Technol. 26, 181–189.
Cussler, E.L., 2009. Diffusion: Mass Transfer in Fluid Systems, 3rd edition. Cambridge
University Press.
Dainelli, D., Gontard, N., Spyropoulos, D., Zondervan-van den Beuken, E., Tobback, P.,
2008. Active and intelligent food packaging: legal aspects and safety concerns. Trends
in Food Science & Technology, vol. 19. Elsevier, pp. 103–112.
Del Nobile, M.A., Sinigaglia, M., Conte, A., Speranza, B., Scrocco, C., Brescia, I.,
Bevilacqua, A., Laverse, J., La Notte, E., Antonacci, D., 2008. Influence of postharvest
treatments and film permeability on quality decay kinetics of minimally processed
grapes. Postharvest Biol. Technol. 47, 389–396. http://dx.doi.org/10.1016/j.
postharvbio.2007.07.004.
Del-Valle, V., Almenar, E., Lagarón, J.M., Catalá, R., Gavara, R., 2003. Modelling per-
meation through porous polymeric films for modified atmosphere packaging. Food
Addit. Contam. 20 (2), 170–179. http://dx.doi.org/10.1080/
0265203021000042869.
Di Renzo, G.C., Altieri, G., Genovese, F., 2015. Dispositivo per controllare scambi gassosi
fra l’interno e l’esterno di un contenitore per prodotti alimentari solidi o liquidi.
Patent: 102015000018707 (in italian).
Di Renzo, G.C., Altieri, G., Genovese, F., 2016. Device for controlling gaseous exchanges
between the inside and outside of a container for solid or liquid food products. Patent:
PCT/IB2016/0506600.
Ersan, S., Gunes, G., Zor, A.O., 2010. Respiration rate of pomegranate arils as affected by
O2 and CO2, and design of modified atmosphere packaging. Acta Hortic. 876,
189–196. http://dx.doi.org/10.17660/ActaHortic.2010.876.24.
Fonseca, S.C., Oliveira, F.A.R., Brecht, J.K., 2002. Modelling respiration rate of fresh
fruits and vegetables for modified atmosphere packages: a review. J. Food Eng. 52
(2), 99–119. http://dx.doi.org/10.1016/S0260-8774(01)00106-6.
Fuller, E.N., Schettler, P.D., Giddings, J.C., 1966. A new method for prediction of binary
gas-phase diffusion coefficients. Indus. Eng. Chem. 58 (5), 18–27. http://dx.doi.org/
10.1021/ie50677a007.
Geysen, S., Verlinden, B.E., Conesa, A., Nicolaï, B.M., 2005. Modelling Respiration of
Strawberry (cv. ‘Elsanta’) as a Function of Temperature, Carbon Dioxide, Low and
Superatmospheric Oxygen Concentrations. Information and Technology for
Sustainable Fruit and Vegetable Production, FRUTIC 05, 12–16 September 2005,
Montpellier, France. http://symposcience.lyon.cemagref.fr/exl-doc/colloque/ART-
00001606.pdf.
González-Buesa, J., Ferrer-Mairal, A., Oria, R., Salvador, M.L., 2009. A mathematical
model for packaging with micro-perforated films of fresh-cut fruits and vegetables. J.
Food Eng. 95, 158–165.
73
Postharvest Biology and Technology 138 (2018) 64–73
Kader, A.A., Zagory, D., Kerbel, E.L., 1989. Modified atmosphere packaging of fruits and
vegetables. CRC Crit. Rev. Food Sci. Nutr. 28, 1–30.
Kader, A.A., 1997. A summary of CA requirements and recommendations for fruits other
than apples and pears. In CA'97 Controlled Atmosphere Research Conference, July
13–18, 1997, University of California, Davis. Proceedings Volume 3: Fruits other than
apples and pears. Postharvest Horticulture Series no. 17, ed. A.A. Kader.
Kalia, A., Parshad, V.R., 2015. Novel trends to revolutionize preservation and packaging
of fruits/fruit products: microbiological and nanotechnological perspectives. Crit.
Rev. Food Sci. Nutr. 55, 159–182.
López-Rubira, V., Conesa, A., Allende, A., Artés, F., 2005. Shelf life and overall quality of
minimally processed pomegranate arils modified atmosphere packaged and treated
with UV-C. Postharvest Biol. Technol. 37, 174–185.
Lam, S.H., 2006. Multi-component diffusion revisited. Phys. Fluids 18 (073101), 1–8.
http://dx.doi.org/10.1063/1.2221312.
Mastromatteo, M., Conte, A., Del Nobile, M.A., 2010. Combined use of modified atmo-
sphere packaging and natural compounds for food preservation. Food Eng. Rev. 2,
28–38.
Matera, A., Genovese, F., Altieri, G., Tauriello, A., Di Renzo, G.C., 2017. An innovative
smart device to control Modified Atmosphere Packaging (MAP) of fruit and vege-
tables. CIOSTA 2017, Research and Innovation for the Sustainable and Safe
Management of Agricultural and Forestry Systems, 13–15 June 2017, Palermo
−Italy. Chemical Engineering Transactions, 58, 193–198. 10.3303/CET1758033.
Mencarelli, F., Bellincontro, A., Di Renzo, G.C., 2005. GRAPE: post-harvest operations. In:
Mejia, D. (Ed.), FAO, INPhO – Post-harvest Compendium. FAO, 1-42. http://www.
fao.org/3/a-av004e.pdf.
Paul, D.R., Clarke, R., 2002. Modeling of modified atmosphere packaging based on de-
signs with a membrane and perforations. J. Membr. Sci. 208, 269–283.
Pereira, M.J., Amaro, A.L., Pintado, M., Poças, M.F., 2017. Modeling the effect of oxygen
pressure and temperature on respiration rate of ready-to-eat rocket leaves. A prob-
abilistic study of the Michaelis-Menten model. Postharvest Biol. Technol. 131, 1–9.
http://dx.doi.org/10.1016/j.postharvbio.2017.04.006.
Restuccia, D., Spizzirri, G., Parisi, O.I., Cirillo, G., Curcio, M., Iemma, F., Puoci, F., Vinci,
G., Picci, N., 2010. New EU regulation aspects and global market of active and in-
telligent packaging for food industry applications. Food Control, vol. 21. Elsevier, pp.
1425–1435.
Rivera, C.S., Blanco, D., Salvador, M.L., Venturini, M.E., 2010. Shelf-Life extension of
fresh tuber aestivum and tuber melanosporum truffles by modified atmosphere
packaging with microperforated films. J. Food Sci. 75 (4), 225–233. http://dx.doi.
org/10.1111/j.1750-3841.2010.01602.x.
Riyahi, R., Rafiee, S., Dalvand, M.J., Keyhani, A., 2011. Some physical characteristics of
pomegranate: seeds and arils. J. Agric. Technol. 7, 1523–1537.
Rojas-Grau, M.A., Oms-Oliu, G., Soliva-Fortuny, R., Martin-Belloso, O., 2009. The use of
packaging techniques to maintain freshness in fresh-cut fruits and vegetables: a re-
view. Int. J. Food Sci. Technol. 44, 875–889.
Rutherford, T.J., Marshall, D.L., Andrews, L.S., Coggins, P.C., Schilling, M.W., Gerard, P.,
2007. Combined effect of packaging atmosphere and storage temperature on growth
of Listeria monocytogenes on ready-to-eat shrimp. Food Microbiol. 24, 703–710.
Schneider, P., 1978. Multicomponent isothermal diffusion and forced flow of gases in
capillaries. Chem. Eng. Sci. 33, 1311–1319.
Shampine, L.F., Reichelt, M.W., 1997. The MATLAB ODE suite. SIAM J. Sci. Comput. 18,
1–22.
Sutherland, W., 1893. LII. The viscosity of gases and molecular force. Philos. Mag. Ser. 5
36 (223), 507–531. http://dx.doi.org/10.1080/14786449308620508.
Wang, Y., Long, L.E., 2014. Respiration and quality responses of sweet cherry to different
atmospheres during cold storage and shipping. Postharvest Biol. Technol. 92, 62–69.
http://dx.doi.org/10.1016/j.postharvbio.2014.01.003.
Webb, S.W., 2006. Gas transport mechanisms. In: Clifford, K.H., Webb, S.W. (Eds.), Gas
Transport in Porous Media. Pub. Springer, The Netherlands, pp. 5–26 (chapter 2).
Zanderighi, L., 2001. How to design perforated polymeric films for modified atmosphere
packs (MAP). Packag. Technol. Sci. 14, 253–266. http://dx.doi.org/10.1002/pts.557.