Permeable Gas Cavity at Elevated Pressure

Enhances Modified Atmosphere Packaging of

Fresh Produce
Ziynet Boz and Bruce A. Welt

Abstract: Modified atmosphere packaging (MAP) of fresh produce involves exploiting package properties to satisfy
respiration activity of produce. While effective, package material properties are not infinitely adjustable to match needs of
all products. Additional ways of providing favorable in-package gaseous environments are needed. This work explores the
use of permeable inserts filled with gas at elevated pressures as a means to achieve in-package gaseous atmospheres that
may not be possible by the package alone. Mathematical models were developed to predict transient package atmospheres
for packages containing respiring produce and pressurized permeable inserts. The model was validated for semirigid
tray packages containing grape tomatoes and Granny Smith apples. With inserts initially pressurized with oxygen at
approximately 200 to 300 kPa (about 30 to 45 psi), about 2 weeks additional shelf life was observed relative to controls
for both tomatoes and apples in test packages. Additionally, simulations provide design guidance for pressurized inserts
for the case of very high respiration rate produce such as spinach.

Keywords: apple, insert, modified atmosphere packaging, spinach, tomato

Food Engineering, Materials

Science, & Nanotechnology
Practical Application: This work promises to expand application of modified atmosphere packaging (MAP). Currently,
applications are limited by gas transfer material properties of existing packaging films. However, packaging offers other
important functions that may not be well served by materials that satisfy critical gas permeation requirements. This work
demonstrates an approach that disconnects packaging material specifications from MAP design.

Introduction Also, since retaining CO2 is challenging, desirable antimicrobial

Modified atmosphere packaging (MAP) has been widely used
as an effective way to preserve food. Global revenue for MAP of
foods is expected to reach $6.4 billion by 2020 (Shahani Dewan,
2015). MAP design is often challenging due to limited availabil-
ity of materials that match the variety of application requirements.
This problem is particularly acute when designing MAP for respir-
ing produce. MAP equilibrium is achieved by matching source
(that is, film permeation) and sink (that is, produce respiration).
Ideally, equilibrium is reached rapidly, within a desirable range
of favorable atmospheres, by maintaining a balance between gas
transmission through packaging and gas consumption/production
via produce respiration (Zagory, 1995). Availability of packaging
materials that meet specific requirements of produce is essential
(Al-Ati & Hotchkiss, 2002; Paul & Clarke, 2002; Sousa-Gallagher
& Mahajan, 2013). Gas transfer rates of existing commercial films
span at least three orders of magnitude, whereas respiration rates of
fresh produce differ by only about one order of magnitude (Al-Ati
& Hotchkiss, 2002), which limits the number of appropriate pack-
aging films to relatively few. Further, relative permeation rates of
common films to carbon dioxide (CO2 ) are about 3 to 6 times that
of oxygen (O2 ) (Guisheng, Incarnato, Di Maio, & Acierno, 1995),
causing loss of respiration produced CO2 from packaging more
rapidly than O2 can be gained. This results in package deflation.
JFDS-2017-1835 Submitted 11/8/2017, Accepted 3/3/2018. Authors are with packages (Sivertsvik, Rosnes, & Bergslien, 2002). Despite recent
Packaging Engineering Program, Agricultural and Biological Engineering Dept., Univ.
of Florida, 1741 Museum Road, Gainesville, FL 32611, USA. Direct inquiries to advances, minimal availability and high cost of selective barrier
author Welt (E-mail: bwelt@ufl.edu).
benefits due to acidification are often not achievable.
Temperature dependency of respiration rate may also cause
nonoptimal and even deleterious anaerobic conditions under abu-
sive thermal conditions if not compensated for with proportional
changes in gas permeability. For example, activation energy for
respiration rate of broccoli varies between 43 and 105 kJ/mol
(Fonseca, Oliveira, & Brecht, 2002), whereas O2 permeability of
common polyethylene and polypropylene films range from about
35 to 43 kJ/mol (Mangaraj, Goswami, & Mahajan, 2009), sug-
gesting respiration rate is much more sensitive to temperature
changes than oxygen permeability of packaging. When thermal
abuse occurs, MAP may actually lead to reduced shelf life when
the package atmospheres are solely dependent on film permeability
(LeBlanc, Stark, MacNeil, Goguen, & Beaulieu, 1996; Simpson,
Almonacid, Acevedo, & Cortés, 2003). Packaging films are not
only expected to provide produce-specific barrier properties, but
also functional properties such as heat sealability, performance at
low temperatures, water vapor barrier , mechanical strength, pin-
hole resistance, machinability, and printability (Lange, 2000; Sand-
hya, 2010). Films often fail to provide desired oxygen transmission
rates (OTR), gas selectivity, and desired mechanical properties
simultaneously. Microperforated films offer effective solutions for
MAP products. However microperforated films are not considered
as hermetic packaging, and have disadvantages of possible dehy-
dration and microbial contamination under wet handling of the
films limit applications of MAP.

C 2018 Institute of Food Technologists

 R

doi: 10.1111/1750-3841.14139 Vol. 83, Iss. 6, 2018 r Journal of Food Science 1639
Further reproduction without permission is prohibited
 Permeable gas cavity at elevated . . .

This work presents an alternative approach to MAP that dis-
connects the package, in whole or part, from the gas delivery
role in MAP. Specifically, a pressurizable gas permeable reservoir
was charged with specific gases in order to satisfy gas delivery
requirements of MAP packaging of fresh produce. The principle
is to throttle gas delivery using varying partial pressures of gases
contained within a permeable, pressurizable package insert. Dy-
namic release of gas to package headspace can potentially mitigate
a variety of problems in MAP applications such as package defla-
tion, elevated CO2 egress for low barrier films, and thermal abuse
compensation using super-atmospheric oxygen. Further, a passive
permeation device filled with atmospheric gases offers clean label,
“all-natural” options for preserving quality.
Therefore, the objective of this study was to develop and
demonstrate use of a prototype permeable, pressurizable package
insert for respiring produce MAP. Specific objectives include pre-
diction and validation of respiration rates and transient headspace
gas concentrations in fresh produce packages, and to demonstrate
package-insert scenarios based on the validated model.
Theoretical Considerations
Gas flux from a pressurized insert within a package can as-
Food Engineering, Materials
Science, & Nanotechnology
Based on the principle of increased partial pressure differ-
ences, an accelerated, high-pressure/high-performance OTR test
method (that is, ASTM F3136-15, 2015) was demonstrated by
Welt (2015). In this context, a permeable package component,
appropriately pressurized with a specific gas mixture, can serve
as a gas reservoir within a package. Pressurized permeation reser-
voirs can be adjusted to deliver application-specific gas mixtures
as an additional option to reducing reliance on packaging film
permeation and fixed atmospheric gas partial pressures.
Materials and Methods
Experiments were conducted in temperature-controlled envi-
ronmental chambers in three steps : (i) measurements of respira-
tion rates of the fresh produce and gas transmission rates (GTR)
of package and pressurized insert materials, (ii) gas permeation
test from pressurized inserts into otherwise empty and sealed sam-
ple packages, and (iii) dynamic gas changes in a fresh produce
packaged with and without pressurized inserts. Respiration rates
were predicted by fitting respiration data to the Michaelis–Menten
(MM) enzyme kinetics equation. Parameters from enzyme kinet-
ics and permeation measurements were defined as model input
parameters for experimental validation. For simulations, package

sist in creating a desired headspace atmosphere. Gas exchange
between pressurized insert and package and package and environ-
ment can be described mathematically. Gas species were assumed
to be transferred through isotropic permeable films governed by
Fickian diffusion in Eq. (1) (Crank, 1975):
headspace, volume and total pressure of the inserts, GTR of mate-
rials, density, weight, and respiration parameters of fresh produce
were defined as inputs. Once the model was validated from experi-
ments, the model was used to predict desirable insert specifications
for particular scenarios and desired performance.

dC Materials
F = −D (1) Grape tomatoes and Granny Smith apples were supplied from
dx
local supermarkets in Gainesville, Florida. Storage temperatures
where F, D, C and x are gas flux (mol m−2 s−1 ), diffusion coef- of 21 °C were selected for testing (Dadzie & Dadzie, 1992; Yang
ficient (m2 s−1 ), concentration of gas species (mol m−3 ) and the & Chinnan, 1988). Samples were stored at testing temperatures
thickness of material (m), respectively. under dark conditions for 24 hr prior to use. Extruded low-
For convenience, gas concentration gradients are expressed in density polyethylene (LDPE) tubing with the wall thickness and
terms of partial pressures adjacent to film surfaces. Diffusion co- outer diameter of 0.6604 ± 0.025 mm (0.026 ± 0.001 inch)
efficients are modified via Henry’s Law, which describes the rela- and 26.60 ± 0.0254 mm (1.05 ± 0.001 inch), respectively, was
tionship between partial pressure of gas and its solubility, resulting supplied from a plastic extrusion manufacturer (GSH Industries,
in a permeation coefficient, P (Geankoplis, 1993). Thus, gas per- Cleveland, Ohio, U.S.A.). Glass containers (Mason jars, 491 ± 6.4
meation through a plastic film is given as and 934.8 ± 4.1 mL capacity) were used to perform the oxy-
gen permeation and respiration rate experiments via noninvasive
Qi Pi A( p i,Ext − p i,In ) fluorescence-quenching sensors (OxyDot R
, Oxysense, New Cas-
= (2)
t x tle, Del., U.S.A.). Sample jars were sealed with lids and gas perme-
able films (OTR 8700 ± 274 cm3 (STP)/m2 /day) for gas perme-
where Pi , pi,Ext , pi,In , are the permeability and partial pressures of ation experiments with Granny Smith apples and grape tomatoes,
gas i at the high and low concentration sides of the film, respec- respectively.
tively.
Equation (2) indicates that the driving force for gas transfer isExperimental procedures
governed by differences in partial pressure across film. Thus, the Gas permeation experiments. Oxygen concentration
greater the partial pressure differences, the greater the magnitude changes in packages with and without fresh produce and with
of gas flux. The relationship among partial pressure, molar fraction,
and without pressurized inserts were measured noninvasively. A
and total pressures are represented by the Dalton’s Law of partial diagram of the package, insert, and fresh produce combinations
pressures (Eq. 3 and 4): used in model development and validation studies is provided in
Figure 1.
 n
P = pi (3) Inserts were mechanically sealed using 1-inch copper crimp
i =1
rings (PexUniverse.com, Brooklyn, NY, U.S.A.). A pressure regu-
lator was connected to the LDPE insert tubing through a custom
mounting fixture fabricated from schedule 1 inch 40 PVC pipe,
p i = P yi (4) which permitted sealing of pressurized tubing while pressurized.
Insert tubing slipped over the PVC pipe fixture and was secured
where P and are yi total pressure and molar fraction of gas species with a hose clamp. The oxygen supply regulator was set to a mini-
i, respectively. mum gauge pressure of 241.32 kPa (35 psig) and connected to

1640 Journal of Food Science r Vol. 83, Iss. 6, 2018

Permeable gas cavity at elevated . . .
Figure 1–Experimental and modeling approach in the current study; (A) respiration of fresh produce in a closed system, (B) only insert permeation in
an empty package, and (C) fresh produce, insert, and package permeation.
the PVC pipe fixture. Although oxygen flushed unsealed tubes,
two crimp rings sealed both ends of an insert using a custom-
made, spring-loaded press tool mounted in a 12-ton hydraulic shop
press (Central Machinery, Harbor Freight Tools, CA, U.S.A.). A
wireless pressure sensor was placed in each insert before sealing.
After setting mechanical seals, protruding tubing was sealed us-
ing an ultrasonic tube sealer as a secondary seal. Resulting insert
lengths and diameters were measured by a 6 ± 0.001 in composite
digital caliper (PittsburghR
, USA). Figure 2 illustrates the insert
pressurization and sealing setup (A) and the pictures of the sealing
system components used in this study (B to D).
Fluorescence oxygen sensors were placed on interior walls of
test jars using double sided tape and calibrated via standard gases
of 100% nitrogen (0% O2 ) and air (20.9% O2 ) using a nonde-
structive oxygen measurement system (Oxysense R
Inc., Dallas,
TX, U.S.A.). Fresh produce and pressurized inserts were placed in
jars, which were initially flushed by air–N2 mixture using a mul-
ticomponent gas mixer (Series 2000; Environics R
, CN, U.S.A.)
to achieve 10 ± 1% initial O2 to observe transient gas differ-
ences in shorter time via equilibrium MAP. Mouth of the jars in
the experiments with grape tomatoes were covered with high O2
transmitting to allow certain amount of gas transmission, and se-
cured by vacuum silicone grease. Final CO2 concentrations were
measured in jars and inserts at the end of the permeation tests by
a multiple gas headspace analyzer equipped with a nondispersive
infrared sensor for CO2 determination (Model 900131; Bridge
analyzers, OH, U.S.A.). For comparison, an additional set of per-
meation tests was conducted in jars without inserts, containing
only fresh produce. Each condition was run in triplicate.
Pressure, volume, and density measurements. Total pres-
sure changes of the inserts were monitored by wireless pressure
sensors (EK2 Universal, China). Response times and displayed
pressures were calibrated by placing sensors in a sealed PVC pipe
connected to a gas pressure regulator. Supplied pressures were
compared to pressures measured by sensors. Sensor dimensions
were measured and calculated from circular truncated cone ge-
ometry as 4.84 ± 0.1 cm3 . Free volumes of the jars and densities
of fresh produce were measured gravimetrically by water displace-
Food Engineering, Materials
Science, & Nanotechnology
ment at 20 °C. Apparent density of grape tomatoes was determined
by filling the cylindrical containers of 500 mL capacity with pro-
duce, whereas density of the individual apples was measured by
weighing submerged in water. Calculated densities were assumed
to be uniform within individual batches and were used to calculate
volume occupied by the samples in jars, of which volumes were
measured by water filling method. Thus, the difference between
the capacity of the jars and the volume of produce was used as the
free headspace in the model. The weights were measured with a
digital laboratory scale (Adventurer ARD110, Ohaus Corp., N.J.,
U.S.A.) with 4100 ± 0.05 g maximum capacity.
OTR of packages and inserts
OTR of the LDPE insert tubing material (thickness of 26 ±
0.1 mm) was measured using dynamic accumulation according
ASTM Intl. F3136-15 (2015), at 10 and 21 °C, by OxySense R
Gen III- 310 (Oxysense Inc.). The dynamic accumulation method
was described and compared to steady state method by Abdellatief
and Welt (2013). Briefly, the permeation chamber had two cells,
an oxygen rich chamber and an oxygen accumulation chamber,
separated by the test specimen. The accumulation chamber was
initially flushed with nitrogen. The oxygen rich side was open to
ambient air and assumed to stay constant with 20.9% O2 through-
out the test. Free volume and transmission area of the dynamic
accumulation chamber were 8.3 and 16.6 cm2 , respectively.
OTR of the packaging film was also measured using ASTM
F3136-15 at 21 °C. CO2 and N2 transmission rates (CO2 TR and
N2 TR) were obtained from literature and set to 5 and 0.4 times
OTR values, respectively, for simulations (Siracusa, 2012).
Produce respiration
Respiration rates were measured using a closed (static) method.
The method is based on determining the rate at which produce
modifies the atmosphere within a sealed container. Derivatives of
empirical functions are used to provide respiration rates over a
range of oxygen and carbon dioxide conditions. Time-dependent
O2 depletion and CO2 evolution data are then fitted to a re-
gression function before applying more generalized MM kinetics,
Vol. 83, Iss. 6, 2018 r Journal of Food Science 1641
Food Engineering, Materials Permeable gas cavity at elevated . . .
Science, & Nanotechnology

Figure 2–Components used to seal pressurized insert (A) illustration of system, (B) custom crimping tool, (C) hydraulic press, and (D) completed and
pressurized inserts containing wireless pressure sensors.

demonstrated in the subsequent sections on mathematical model- based on the effects of dissolved O2 and CO2 in produce on RR
ing and by Hagger and Yam (1992). This closed system approach is the most widely utilized model due to simplicity and compati-
offers advantages of simple experimental setup, suitability for the bility (Fonseca et al., 2002). A method proposed by Hagger, Lee,
low Respiration Rate (RR) products, and ability to determine and Yam (1992), based on fitting transient O2 and CO2 changes
the RR under several O2 and CO2 combinations (Fonseca et al., from closed system respiration experiments into the regression
2002). equation was utilized here. Coefficients in Eq. (5) and (6) were
In order to reflect normal aerobic respiration kinetics, exper- estimated from experimental time and gas concentration data by
iments were stopped when CO2 concentration reached 16%, minimizing sum of squared errors in nonlinear solver option in
which has been shown to be upper tolerable limit for produce Microsoft Excel R
solver:
(Hagger & Yam, 1992). Jars (3 ± 0.1 L) were fitted with rub-
ber septa for sampling. A gas chromatograph (Varian CP-3800, t
O2 = 21 − (5)
Varian, CA, U.S.A.) with thermal conductivity detector was cal- (A1 t + B1 )C1
ibrated using standard gas. Temperatures at injector, column, and
detector, and carrier gas flow rate were 220, 50, and 130 °C, and
20 mL/min, respectively. Gas sample of 5 mL was collected with t
CO2 = (6)
gas-tight syringes. Sampling was done from three separate jars for (A2 t + B2 )C2
statistical significance.
Derivatives of Eq. (5) and (6) provide rate of change of gas con-
Mathematical model development centrations (Eq. 7 to 8). Calculated derivatives at each time are then
Respiration parameter estimation. Various models have normalized with weight, pressure, temperature, and headspace vol-
been proposed to estimate O2 and CO2 concentration-dependent ume to estimate respiration rates (Eq. 9 and 10):
respiration rate in fresh produce (Beaudry, Cameron, Shirazi, &
Dostal-Lange, 1992; Cameron, Boylan-Pett, & Lee, 1989; Lee, d O2
Hagger, & Yam, 1991). However, single enzyme MM kinetics, = A1 C1 t (A1 t + B1 )(−1−C1 ) − (A1 t + B1 )−C1 (7)
dt

1642 Journal of Food Science r Vol. 83, Iss. 6, 2018

Permeable gas cavity at elevated . . .

crosoft Excel
R
solver can be found in the literature (Brown, 2001):
12

10 Vm a x,O2 [O2 ]
T
RRO = (11)
O2 Concentration (%)

2
Km + [O2 ]
8

6 Vmax,O2 [O2 ]
T
RRO2 =   (12)
Model Experimental [O2 ] + Km O2 1 + Km [CO2 ]
c CO2
4

2
Vmax,O2 [O2 ]
T
RRO2 =   (13)
0 Km O2 + [O2 ] 1 + Km [CO2 ]
uCO2
0.0 5.0 10.0 15.0
time (days)
Vmax,O2 [O2 ]
Figure 3–Measured and predicted O2 concentrations in a non-permeable RRTO2 =    (14)
rigid package initially flushed with 100% N2 containing an insert initially Km O2 + [O2 ] 1 + [CO2 ]
Km nCO2
flushed with O2 at 331 kPa (48 psia) 22.3 ± 0.1 °C.
where Vmax , O2 , KmO2 , [O2 ], [CO2 ], Km with subscripts c, u,
20 n are maximum O2 consumption rate (ml kg−1 hh−1 ), O2 con-
18
578.6 kPa centration (%), CO2 concentration (%), Michaelis constant for

Food Engineering, Materials

Science, & Nanotechnology
496.5 kPa competitive, uncompetitive and noncompetitive inhibition, re-
16
O2 Concentration (%)

413.4 kPa spectively.

14 330.3 kPa Contribution of RR on total molar changes of O2 and CO2
12 248.2 kPa at each time step calculation was expressed in molar change
10 101.3 kPa
of rates by selecting the best fitting model parameters among
8 Eq. (11) to (14) and calculating the contribution of respiration
6 activity to molar gas changes in the package (Eq. 15 and 16):
4
d n RRO2 RRO2 WP
2 = (15)
0 dt RTabs
0.0 5.0 10.0 15.0 20.0 25.0
time (days)
d n RRCO2 RRCO2 WP
= (16)
Figure 4–Simulated transient package O2 concentration changes based on
dt RTabs
varying flushed pressures.
The ratio of RR in terms of CO2 production (RRCO2 ) to O2
consumption (RRO2 ) is expressed as respiratory quotient (RQ)
d CO2
= −A2 C2 t (A2 t + B2 )(−1−C2 ) + (A2 t + B2 )−C2 (8) (Eq. 17), which varies typically between 0.8 and 1.5 depending
dt on the type of substrate used in respiration (Beaudry et al., 1992)
and often assumed 1 for a range of produce. In this study, exper-
imental RQ values were calculated for each produce to predict
  CO2 production from O2 consumption in the model:
d O2 MWO2 P V
RRO2 =− (9)
dt 100 RWT RRCO2
RQ = (17)
RRO2
 
Transient gas changes in the package-produce-insert
d CO2 MWCO2 P V
RRCO2 = (10) system. Gas transfer through polymeric films in MAP of fresh
dt 100 RWT
produce are modeled by combining the contributions of respira-
tion rate (that is, Eq. 15 and 16) and gas exchange through the
where MWO2 , P, V, R, W and T are the molecular weight of packaging film by Eq. (18), which is commonly used for MAP
O2 , total pressure, volume, Universal gas constant, weight and modeling for fresh produce, was solved for all three gas species:
temperature, respectively.
Change of respiration rates by O2 and CO2 concentrations were d n i,Pkg OTR A( p Ext − p Pkg ) d n RRi
predicted by assessing inhibitory nature of CO2 on respiration rate = ∓ (18)
dt RT dt
in MM equations with: (i) no CO2 inhibition, (ii) competitive, (iii)
uncompetitive, and (iv) noncompetitive inhibitions via Eq. (11) where n, A, pExt , pPkg are the moles of gas i, transfer area, partial
to (14), respectively. MM parameters of Vmax,O2 , Km , Km c CO2 , pressure of gas i in the higher and lower concentration side of the
Km uc CO2 , and Km nCO2 were estimated through nonlinear mul- package, respectively.
tiple regression in Microsoft Excel R
solver by minimizing sum In the packages containing both pressurized insert and fresh
of the residuals. Details of regression and equation fitting via Mi- produce, another term to represent permeation from insert was

Vol. 83, Iss. 6, 2018 r Journal of Food Science 1643

 Permeable gas cavity at elevated . . .
included to obtain Eq. (19), which was also used to simulate Table 1–Equation (5) regression parameters.
packages containing only pressurized insert: Regression parameters
d n i,Insert GTRi,Insert AInsert ( p i,Insert − p i,Pkg )
= (19) B1
dt RTabs
where GTRi , Insert , pi, Insert are the gas transfer rate of the insert
and partial pressure of gas i in the insert, respectively.
Combining and rearranging 15, 16, 18, and 19 for O2 , CO2 ,
and N2 , Eq. (20) is obtained. The sign of RR terms is negative and
positive for O2 consumption and CO2 production, respectively:
d n i,Pkg GTRi,Pkg A( p Ext − p Pkg ) d n RRi d n i,Insert
= ∓ + (20)
dt RT dt dt
For modeling of the packages without insert, Eq. (18) was used.
Runge–Kutta fourth order with classical constant time step was
used to solve Eq. (18) to (20) simultaneously to predict transient
changes of O2 , CO2 , and N2 in package and pressurized insert.
The code was executed in Microsoft Excel R
for Visual Basic
Applications (VBA). All the permeation and respiration rate ex-
periments were conducted in rigid packages. For rigid packages,
Food Engineering, Materials
Science, & Nanotechnology
the total volume stays constant while package pressure is varied.
Pressure changes associated with molar gas transfer were included
in the model based on the package type (that is, rigid or flexible).
Oxygen permeation experiments including fresh produce and in-
serts were conducted in rigid jars. In this case, total volume remains
constant while total pressure changes. Volume and lateral surface
area of the insert was calculated by cylindrical geometry equations.
Also, dimensional measurements showed a negligible increase in
the volume of inserts, due to stretching when pressurized. For
this reason, the volume of the insert was assumed constant, while
pressure changes were included in the model. Pressure and volume
changes of the package and inserts were calculated by the ideal gas
law equation (Equation 21):
P Vheadspace = n Tot RTabs (21)
where Vheadspace , nTot and Tabs are the headspace volume, total
moles of gas and absolute temperature, respectively.
The goodness of fit of MM parameters was determined by
calculating coefficient of determination (R2 ). Mathematical model
evaluation was done by calculating root mean square error (RMSE)
values (Eq. 22) between experimental and model results:
n
1  by permeation area and material gas transmission properties (that
RMSE =  (xi,exp − xi,pred )2 , (22)
n i =1
where n, xi,exp , and xi,pred are the number of measurements,
value of the experimental and model results at point i, respectively.
Experimental uncertainties and standard errors were included in
the plots as error bars for independently repeated experiments.
Results and Discussion
OTRs of the films
OTR of the insert tubing material at 5, 10, and 21 °C were
124, 212, and 300 cm3 (STP)/m2 /day/atm, respectively. OTR of
packaging film at 21 °C was 8700 cm3 (STP)/m2 /day/atm. As
expected, OTR increased with increasing temperature. Reduced
OTR at lower temperatures is a limiting factor for fresh produce
1644 Journal of Food Science r Vol. 83, Iss. 6, 2018
Grape tomatoes Granny Smith apples
A1 2.47 × 10–3 2.75 × 10–4
1.5631 1.1287
C1 3.226 14.1829
RMSE 0.6878 0.9439
R2 0.9127 0.9490
A2 1.5419 1.50 × 10–4
B2 69.244 1.090
C2 0.3424 20.688
RMSE 3.002 1.003
R2 0.9872 0.9669
stored at lower temperatures with respect to O2 ingress, which
may be mitigated with pressurized permeable insert.
Package-insert system. When fresh produce was not in-
cluded in nonpermeable glass jars, the only gas modification in
the package occurred via permeation from the insert initially filled
with 100%. Figure 3 shows experimental and simulation results in
a 933 cm3 package initially flushed with N2 , and an insert ini-
tially partially flushed with 100% O2 at 22.28 ± 0.11 °C for over
15 days. Total absolute pressure, volume, area, and initial O2 and
N2 concentrations of the inserts were 330.95 kPa (48 psia), 72.62
cm3 , 106.14 cm2 , 72%, and 28%, respectively. Simulations show
good agreement with experimental data with R2 and RMSE val-
ues of 0.9834% and 1.95%, respectively. After 15 days, O2 con-
centration reached approximately 10%. This system can be com-
pared to a nonrespiring product packaging containing chemical
gas (that is, CO2 or O2 ) emitters. Achieved gas concentrations
are a function of gas solubility in product and package as well as
any consumption/production via biochemical reactions, and the
capacity of the emitter.
The rate of gas accumulation and range of concentrations de-
livered by pressurized inserts depends upon similar factors. For
example, increasing initial partial gas pressures increases perme-
ation rates and effective permeation duration. Using the same
model as shown in Figure 3, with all parameters constant, but
increasing initial O2 pressures from 101.33 to 579.16 kPa (14.7
to 84 psia) is shown in Figure 4. Increasing oxygen pressure leads
to higher maximum achievable O2 concentrations with in pack-
ages for extended periods. Simply flushing inserts with oxygen
at atmospheric pressure of 101.325 kPa (14.7 psia) also provides
gas egress, but at a reduced rate and capacity. The rate of gas ex-
change through the nonpressurized insert could also be elevated
is, GTR) of the insert.
Stern, Fang, and Frisch (1972) studied the effect of pressure
on permeability coefficients in Fluoroform-polyethylene and 1,1-
Difluoroethylene-polyethylene systems at pressures differentials
across films of 0 to 4053 kPa (0 to 40 atm) and temperatures
from –18 to 70 °C. Results showed that permeability coefficients
increased slightly and linearly over this large pressure differen-
tial range. While magnitude of permeability coefficients decreased
with temperature, sensitivity to pressure differentials increased, as
indicated by increasing slopes, as temperature decreased. However,
changes in permeability coefficients were modest with less than
approximately two orders of magnitude over large temperature
ranges from 0 to 50 °C. In this work, pressure differentials were
relatively small, ranging from 0 to 400 kPa (0 to 4 atm). Tem-
peratures were bounded by what is optimal for retail display and
storage of produce studied. Depending on the cultivar, optimum
Permeable gas cavity at elevated . . .

30 0.3
CO2 O2 A B

O2 and CO2 concentration

O2 Predicted CO2 Predicted
Gas concentrations (%)
25 0.2

rate of change (1/h)

20 0.1 dO2/dt
dCO2/dt
15 0.0
0 50 100 150 200
10 -0.1

5 -0.2

0 -0.3
0 50 100 150 200
time (h)
time (h)

25 0.20
C D

O2 and CO2 concentration

CO2 O2
O2Predicted CO2 Predicted 0.15
Gas concentrations (%)

rate of change (1/h)

20
0.10 dO2/dt
0.05 dCO2/dt
15
0.00
10 -0.05 0 50 100 150 200

Food Engineering, Materials

Science, & Nanotechnology
-0.10
5
-0.15
0 -0.20
0 50 100 150 200
time (h) time (h)

Figure 5–Demonstration of the O2 and CO2 concentration and rate of concentration change obtained experimentally and from the regression model
for grape tomatoes (A and B) and Granny Smith apples (C and D).

Table 2–Michaelis–Menten (MM)parameters for predicting RR based on predicted O2 and CO2 concentrations for grape tomatoes
and Granny Smith apples at 21 °C.

Parameter Crop type No inhibition Competitive inhibition Uncompetitive inhibition Non-competitive inhibition
Vmax GT 97.307 92.250 62.692 29.871
GSA 40.550 40.545 40.550 40.549
Km GT 103.40 97.218 57.899 16.567
GSA 60.905 60.896 60.905 60.904
Kmu GT N/A N/A 21.888 N/A
GSA N/A N/A 5.17 × 106 N/A
Kmc GT N/A 8.3 × 107 N/A N/A
GSA N/A 5.46 × 105 N/A N/A
Kmn GT N/A N/A N/A 35.934
GSA N/A N/A N/A 5.16 × 106
RMSE GT 0.1132 0.1349 0.0458 0.0534
GSA 0.1381 0.1381 0.1381 0.1381
R2 GT 0.9979 0.9977 0.9995 0.9993
GSA 0.9944 0.9944 0.9944 0.9944
GT, grape tomatoes; GSA, Granny Smith apples; N/A, not applicable

temperatures were suggested to be approximately 10 to 21 and 0
to 3 °C for tomatoes and apples, respectively (Maul et al., 2000;
Univ. of California Davis, Produce Fact Sheets). However, retail
display temperatures usually are at room temperature for these
crops. Therefore, permeability coefficients would not be expected
to vary significantly from values measured using dynamic accu-
mulation, which is done with no pressure differential. If certain
combinations of materials and permeant result in greater sensitiv-
ity to pressure differentials such that common gas transfer methods
(for example, ASTM F3136 or ASTM D3985) are less reliable
in predicting gas transfer rates under application condition, then
other methods, such as parameter estimation could be used to ob-
tain better parameter estimates of apparent permeation coefficients
or standardized gas transfer rates.
Fresh produce respiration parameters. Respiration regres-
sion parameters (Eq. 7 and 8), used to predict gas concentrations
with respect to time, are shown in Table 1. For both tomatoes and
apples, simulations show good agreement with experimental data.
Figure 5 shows O2 and CO2 concentration and rate of change
of concentrations with respect to time. Rate of change in CO2
concentration was not affected greatly by concentration over time,
indicating little inhibition of CO2 on respiration rate.

Vol. 83, Iss. 6, 2018 r Journal of Food Science 1645

 Permeable gas cavity at elevated . . .

12
A B
10

RRO2 (cm3/kg/h)
8

6
MM No Inhibition
NormalizedRR
4

0
0.0 5.0 10.0 15.0 20.0 25.0
O2 (%)

Figure 6–Prediction of respiration rate based on changing O2 concentrations via non-CO2 inhibition Michaelis-Menten (MM) kinetics and normalized
respiration rate from experiments by regression model for grape tomatoes (A) and Granny Smith apples (B).

Table 2 shows MM parameters obtained from nonlinear mul-

tiple regressions performed on Eq. (11) to (14). Parameters Kmu 14
A
Food Engineering, Materials
Science, & Nanotechnology

and Kmc generally represent influence of CO2 on respiration rate. With Insert (Experimental)
12 Without Insert (Experimental)
Units of these parameters are expressed in mL /100 mL CO2 ;
O2 concenttation (%) With Insert (Model)
therefore, such high values are unrealistic to be included in the 10 Without Insert (Model)
respiration rate calculations. In the literature, this is a pronounced
phenomenon associated with the effect of CO2 on respiration 8
rate calculations. For example, Peppelenbos and Leven (1996) as-
6
sessed four types of inhibition models on available data of ap-
ple, tomato, and broccoli, and conducted additional experiments 4
on Golden Delicious and Elstar apples, asparagus, broccoli, mung
bean sprouts, and cut chicory. They observed no influence of CO2 2
on O2 consumption on Golden Delicious and Elstar apples. How-
ever, CO2 appeared to affect respiration of asparagus, mung bean 0
sprouts, and chicory. They suggested that influence of CO2 could 0 5 10 15 20
be related to data collection periods, (that is, the longer the storage, time (days)
the higher the effect of CO2 ). Other data support the notion that 12
carbon dioxide has little impact on oxygen consumption (Beaudry, B With Insert (Experimental)
1993; Talasila, Chau, & Brecht, 1992). Beaudry (1993) surmised 10 Without Insert (Experimental)
that lack of evidence for carbon dioxide inhibition of respiration With Insert (Model)
O2 concenttation (%)

might be related to conditions where tolerable CO2 limits are not Without Insert (Model)
8
exceeded.
In this study, O2 and CO2 concentrations were measured by the 6
closed system method and respiration rates were expressed in terms
of O2 changes. Dynamic decrease in O2 concentrations can be seen 4
as the proxy of all the inhibitive reactions occurring simultaneously
in tissues, leading to the calculation of “apparent” effects. Due to 2
good correlation and simplicity, the simplest kinetics equation
with no inhibition of CO2, was used (Eq. 11). Figure 6 shows the 0
RRO2 calculated by MM with no-CO2 inhibition with respect to 0 5 10 15
measured O2 concentrations. time (days)
Respiration rates for apples measured in this study were be-
tween 2 and 10 cm3 /kg/ hr, which agrees well with Kader et al. Figure 7–Measured O2 concentration changes from permeation experi-
(1989) who reported 7.4 cm3 /kg/hr at 20 °C. Published data for ments and simulated predictions in grape tomatoes (A) and Granny Smith
respiration rate of mature pink grape tomatoes were not found, apples (B)
however, RRO2 of the mature green tomatoes have been reported
to be 18 cm3 /kg/hr under ambient air conditions at 20 °C, which reservoir/sink and can provide oxygen required to support aerobic
compares well to the results of this study of 16.65 cm3 /kg/hr. respiration. When glass jars were sealed with high gas transmitting
Combined insert-fresh produce-package system. When film in place of the lid, as in the case of grape tomatoes, the pack-
pressurized inserts and fresh produce were included in nonperme- age also contributed to dynamic gas concentration changes. Ide-
able rigid packages, atmosphere modification occurs through fresh ally, optimal gas concentrations can be maintained for prolonged
produce respiration and insert permeation. The insert acts as a gas periods with proper specification and design of the pressurized

1646 Journal of Food Science r Vol. 83, Iss. 6, 2018

Permeable gas cavity at elevated . . .

0.973 6.0 × 10–1

insert (that is, volume, area, material GTRs, internal gas partial

40.55
GSA
0.82

140
pressures, and so on) and package relative to produce respiration

21

1
requirements. Figure 7 shows experimental and simulation results
for grape tomato and apple packages initially containing 10% O2 .

103.39
1.022

84.48
GT
The mathematical model predicted cases for both packages, with

1.2
21
and without insert. Input parameters used to generate predicted
curves in Figure 7 are provided in Table 3.

Produce parameters

Km (molar fraction)

Vmax (cm3 /kg/hr)

Optimal gas concentrations for storage of tomato (O2 :CO2

Temperature (°C)
(g/cm3 )
ratios) vary between 2.5% and 5%:5% at 12 °C (Salunkhe, Jad-

Weight (g)
hav, & Yu, 1974). Off flavor development of tomatoes was delayed

RQ
by combinations of 5% O2 and 5% to 15% CO2 at 21 °C (Yang

Density
& Chinnan, 1988). Storage temperature of 21 °C (ambient) is
recommended for tomatoes. Shelf life of tomatoes can be ex-
tended with 2% to 4% O2 (Cameron et al., 1989). Using 2% to
4% O2 as a minimum limit for ripen grape tomatoes, this study

288.35:115.34:1441.8
showed that a flexible package made from gas barrier film (OTR =
1 cm3 /m2 /day/atm) containing an insert initially charged with

374.90

65.81
97.22
GSA

1:0:0
100% O2 at about 304 kPa, can provide about 12 to 14 addi-

4.84
Table 3–Model input parameters for packages containing grape tomatoes and Granny Smith apples, with and without inserts.
tional days before oxygen levels fall below the aerobic tolerance
limit. Carbon dioxide concentrations were measured at the end of
each test due to the measurement’s destructive nature to packaged
samples. Final CO2 concentration predictions of 8.9% and 5.3%

Food Engineering, Materials

Science, & Nanotechnology
(at 19 days) with and without insert, respectively, indicate close

288.35:115.34:1441.8
agreement with measured concentrations of 8.32 ± 0.9% and
6.31 ± 1.04%. Errors indicate the standard error based on mea-

344.51

54.86
81.39

1:0:0
surements from three different sample jars.

4.84
GT
Apple packages without inserts have O2 concentrations be-
low 2% after the fifth day. However, packages with inserts were
still above 2% at the end of the experiments (days 15), suggest-
ing more than 10 days of extension were possible with apples
packaged in glass jars. Optimum oxygen concentrations for con-

Wireless sensor volume

Initial absolute pressure

OTR:N2 TR:CO2 TR
Insert molar fractions
Insert volume (cm3 )
Insert parameters

(cm3 /m2 /d/atm)

trolled atmosphere storage of Granny Smith apples have been re-

(O2 :N2 :CO2 )

Insert area (cm2 )
ported to be 1.5% to 2% and 1.5% (Kupferman, 2001). Utiliz-
(cm3 )
(kPa)

ing the insert when selecting a barrier material can be beneficial

GT, grape tomatoes; GSA, Granny Smith apples; WI, with inserts; WOI, without inserts; N/A, not applicable
to prevent gas concentrations reaching intolerable limits of the
produce.
Packaging materials with higher barrier properties may be de-
sirable for reasons other than barrier. Typically, packaging films
suitable for horticultural MAP are limited to thin polyolefins. Such 0.11:0.89:0 (WOI)
films can be difficult to print, offer inferior optical properties, and
0.11:0.89:0 (WI);

may tear easily on packaging machinery. Thicker films and/or

101.325
50000

Rigid

934.8

different materials may be more suitable for branding, printing,

GSA

N/A

N/A
1

quality perception, and so on. The opportunity to delink package

gas permeation properties from product respiration support might
offer opportunities to expand applications for premium quality
MAP fresh produce.
Experimental and simulation data compared well according to
0.12:0.88:0 (WOI)
8663.6:3465.4:43318

the calculated RMSE values of 1.196%, 0.479%, 0.48%, and 0.54%

0.10:0.89:0 (WI);

in grape tomato packages with and without insert and in apple

101.325
50000

Rigid
23.46

packages with and without insert, respectively. A RMSE value of

GT

491
1

2% or lower is commonly used as a criteria for a good compatibility

of MAP of fresh produce models (Charles, Sanchez, & Gontard,
2003; Yang & Chinnan, 1988). After 15 days of storage, exudates
along with white and gray mold were observed in grape tomatoes
packaged without insert. No visible mold formation was observed
Package molar fractions

in packages with insert. Mold formation may be promoted as

OTR:N2 TR:CO2 TR
Package parameters
Simulation time (min)

Total package volume

Initial pressure (kPa)

(cm3 /m2 /d/atm)

a consequence of stresses due to anoxic conditions in packages

Package area (cm2 )

(O2 :N2 :CO2 )

Time step (min)

without inserts. Alternatively, inserts might also support mainte-

nance of mold inhibiting concentrations of CO2 due to main-
tenance of aerobic respiration. Figure 8 shows images of grape
Rigidity

(cm3 )

tomatoes obtained from packages without (A and C) and with

Vol. 83, Iss. 6, 2018 r Journal of Food Science 1647

 Permeable gas cavity at elevated . . .

Figure 8–Images showing the visual

quality and mold formation in grape
tomatoes used in two different sets of
permeation tests in packages without
(A and C) and with inserts (B and D) at
21 °C.
Food Engineering, Materials
Science, & Nanotechnology

Table 4–Model input parameters for case scenario of baby spinach packages, with and without pressurized inserts at 5 °C.

Package parameters Insert parameters Produce parameters

Simulation time (min) 50000 Initial absolute pressure (kPa) 405.3 Density (g/cm3 ) 0.6038
Time step (min) 1 Insert volume (cm3 ) 300 Volume (cm3 ) 447.17
Initial pressure (kPa) 101.325 Insert area (cm2 ) 350 Weight (g) 230
Package area (cm2 ) 1186 Insert molar fractions (O2 :N2 :CO2 ) 1:0:0 Temperature (°C) 5
Rigidity Flexible OTR:N2 TR:CO2 TR (cm3 /m2 /d/atm) 500:500:2500 RQ 1.2
Total package volume (cm3 ) 2000 Km (molar fraction) 6.03 × 10–3
Package molar fractions (O2 :N2 :CO2 ) 0.08:0.92:0 Vmax (cm3 /kg/hr) 13.882
OTR:N2 TR:CO2 TR (cm3 /m2 /d/atm) 1400:560:7000

14 Figure 9–Dynamic O2 and CO2

O2 and CO2 concenttations (%)

O2 With Insert O2 Without Insert concentration changes in spinach

CO2 With Insert CO2 Without Insert MAP with (dashed line) and without
12
(continuous line) inserts.
10
Opmum O2
8 Opmum CO2
6

2
Minimum tolerable O2
0
0 5 10 15 20
time (days)

(B and D) inserts at two separate tests of permeation exceeding tions of high respiration rate produce at low storage temperature,
15 days. which reduces gas permeation rates. Baby spinach was packaged
Model scenario: Spinach package with insert. A chal- in barrier multilayer film (OTR = 1 cm3 /m2 /day/atm) with and
lenge study was performed to test inserts under extreme condi- without inserts and stored at 5 °C. Unfortunately, at the time of

1648 Journal of Food Science r Vol. 83, Iss. 6, 2018

Permeable gas cavity at elevated . . .
this study our ability to produce inserts was limited to only one
material and method, which limited our ability fabricate inserts
according to specifications. Using materials and methods available,
it was not possible to extend the time to reach anoxia more than
approximately 1 day. One option is to increase flushing pressure
and volume/area of the insert. Model input parameters designed
for this scenario are given in Table 4.
Respiration parameters for spinach were obtained as described
earlier. Input parameters of volume, initial total pressure, and
GTRs of the insert were compared to previously validated model
for spinach. Figure 9 shows predicted O2 and CO2 concentrations
obtained from simulations. In addition to the minimum tolerable
O2 concentrations of 0.4% to 0.8%, a window of optimum MAP
concentrations for shelf life extension has been identified 7% to
10% and 5% to 10% for O2 and CO2 concentrations, respectively
(Saltveit, 1997). In accordance with these values, simulations sug-
gest that a spinach MAP specified with insert parameters given in
Table 4 would provide approximately 12 additional days at optimal
concentrations and more than 20 days extension before packages
reach deleterious anaerobic conditions.
Conclusions
In this study, a mathematical model was developed to simulate
transient O2 , CO2 , and N2 changes in a novel MAP consisting of
a permeable pressurized insert and fresh produce. The model was
assessed by comparing with experimental results. Compatibility
of the model was dependent upon the prediction of respiration
rate and film permeation parameters. Results of this work can
be used to improve and expand opportunities for MAP. Modified
atmospheres can also be beneficial for high value products such as
meat, fish, seafood, poultry, and so on.
References
Abdellatief, A., & Welt, B. A. (2013). Comparison of new dynamic accumulation method for
measuring oxygen transmission rate of packaging against the steady-state method described
by ASTM D3985: Dynamic accumulation for OTR measurement. Packaging Technology &
Science, 26, 281–288. https://doi.org/10.1002/pts.1974.
Al-Ati, T., & Hotchkiss, J. H. (2002). Application of packaging and modified atmosphere to fresh-
cut fruits and vegetables. In L. Olusola (Ed.), Fresh-cut fruits and vegetables: Science, technology,
and market. Boca Raton, FL: CRC Press.
ASTM International. (2015). ASTM F3136-15 Standard test method for oxygen gas transmission rate
through plastic film and sheeting using a dynamic accumulation method. ASTM International Report
No.: ASTM F3136-15.
Beaudry, R. M. (1993). Effect of carbon dioxide partial pressure on blueberry fruit res-
piration and respiratory quotient. Postharvest Biology and Technology, 3, 249–258. https://
doi.org/10.1016/0925-5214(93)90060-G.
Beaudry, R. M., Cameron, A. C., Shirazi, A., & Dostal-Lange, D. L. (1992). Modified-
atmosphere packaging of blueberry fruit: Effect of temperature on package O2 and CO2 .
Journal of the American Society for Horticultural Science, 117, 436–441.
Brown, A. M. (2001). A step-by-step guide to non-linear regression analysis of experimental
data using a Microsoft Excel spreadsheet. Computer Methods and Programs in Biomedicine, 65,
191–200.
Cameron, A. C., Boylan-Pett, W., & Lee, J. (1989). Design of modified atmosphere packaging
systems: Modeling oxygen concentrations within sealed packages of tomato fruits. Journal of
Food Science, 54, 1413–1421.
Charles, F., Sanchez, J., & Gontard, N. (2003). Active modified atmosphere packaging of fresh
fruits and vegetables: Modeling with tomatoes and oxygen absorber. Journal of Food Science,
68, 1736–1742. https://doi.org/10.1111/j.1365-2621.2003.tb12321.x
Crank, J. (1975). The mathematics of diffusion. Oxford: Clarendon Press.
Dadzie, B. K., & Dadzie, B. K. (1992). Gas exchange characteristics and quality of ap-
ples: A thesis presented in partial fulfilment of the requirements for the degree of Doc-
tor of Philosophy in Plant Science. Massey University, New Zealand. Retrieved from
https://mro.massey.ac.nz/handle/10179/2309
Fonseca, S. C., Oliveira, F. A., & Brecht, J. K. (2002). Modelling respiration rate of fresh fruits
and vegetables for modified atmosphere packages: A review. Journal of Food Engineering, 52,
99–119.
Geankoplis, C. J. (1993). Transport processes and unit operations (3rd Sub ed.). Engelwood Cliffs,
NJ: Prentice Hall PTR.
Guisheng, F., Incarnato, L., Di Maio, L., & Acierno, D. (1995). Discussion about the use
of relative values of permeabilities between two gases for high molecular weight polymers.
Polymer, 36, 4345–4346. https://doi.org/10.1016/0032-3861(95)92232-4.
Hagger, Pe, Lee, Ds, & Yam, K. (1992). Application of an enzyme kinetics based on respiration
model to closed system experiments for fresh produce. Journal of Food Process Engineering, 15,
143–157. https://doi.org/10.1111/j.1745-4530.1992.tb00148.x
Hagger, P. E., & Yam, K. L. (1992). Application of an enzyme kinetics based respiration model
to closed system experiments for fresh produce. Journal of Food Process Engineering, 15, 143–
157.
Kader, A. A., Devon, Z., Eduardo, L. K., & Chien, Y. W. (1989). Modified atmosphere packaging
of fruits and vegetables. Critical Reviews in Food Science & Nutrition, 28 (1), 1–30.
Kupferman, E. (2001). Controlled atmosphere storage of apples and pears. VIII Inter-
national Controlled Atmosphere Research Conference, 600, 729–735. Retrieved from
https://www.actahort.org/books/600/600_111.htm
Lange, D. L. (2000). New film technologies for horticultural products. HortTechnology, 10, 487–
490.
LeBlanc, D. I., Stark, R., MacNeil, B., Goguen, B., & Beaulieu, C. (1996). Perishable food
temperatures in retail stores. New developments in refrigeration for food safety and quality. Re-
frigeration Science and Technology Proceedings of the Meeting of Commission C2, with
Commissions B2, D1, and D2-3; 1996 Oct 2-4. Paris: International Institute of Refrigeration,
42–51.
Lee, D. S., Hagger, P. E., & Yam, K. L. (1991). Model for fresh produce respiration in mod-
ified atmosphere based on principles of enzyme kinetics. Journal of Food Science, 56, 1580–
1584.
Mangaraj, S., Goswami, T. K., & Mahajan, P. V. (2009). Applications of plastic films for modified
atmosphere packaging of fruits and vegetables: A review. Food Engineering Reviews, 1, 133–158.
https://doi.org/10.1007/s12393-009-9007-3.
Maul, F., Sargent, Sa, Sims, Ca, Baldwin, Ea, Balaban, Mo, & Huber, Dj (2000). Tomato flavor
and aroma quality as affected by storage temperature. Journal of Food Science, 65, 1228–1237.
https://doi.org/10.1111/j.1365-2621.2000.tb10270.x
Food Engineering, Materials
Science, & Nanotechnology
Paul, D. R., & Clarke, R. (2002). Modeling of modified atmosphere packaging based on designs
with a membrane and perforations. Journal of Membrane Science, 208, 269–283.
Peppelenbos, H. W., & van’t Leven, J. (1996). Evaluation of four types of inhibition for modelling
the influence of carbon dioxide on oxygen consumption of fruits and vegetables. Postharvest
Biology and Technology, 7, 27–40. https://doi.org/10.1016/0925-5214(96)80995-1
Saltveit, Jr., M. E. (1997). A summary of CA and MA requirements and recommenda-
tions for harvested vegetables. Proceeding of the 7th International Controlled Atmosphere Re-
search Conference. Vol. 4,Davis, CA: University of California, 98–117. Retrieved from
https://scholar.google.com/scholar?cluster=11991562286401651609&hl=en&oi=scholarr
Salunkhe, D. K., Jadhav, S. J., & Yu, M. H. (1974). Quality and nutritional composition of
tomato fruit as influenced by certain biochemical and physiological changes. Plant Foods for
Human Nutrition (Formerly Qualitas Plantarum), 24, 85–113.
Sandhya (2010). Modified atmosphere packaging of fresh produce: Current status and fu-
ture needs. LWT - Food Science and Technology, 43, 381–392. https://doi.org/10.1016/
j.lwt.2009.05.018.
Shahani Dewan, S. (2015). Active, controlled, and intelligent packaging for foods and beverages. BCC
Research Report No.: FOD038D.
Simpson, R., Almonacid, S., Acevedo, C., & Cortés, C. (2003). Mathematical model
to predict effect of temperature abuse in map systems applied to Pacific Hake (Mer-
luccius australis). Journal of Food Process Engineering, 26, 413–434. https://doi.org/10.
1111/j.1745-4530.2003.tb00610.x
Siracusa, V. (2012). Food packaging permeability behaviour: A report, food packaging per-
meability behaviour: A report. International Journal of Polymer Science, 2012, e302029.
https://doi.org/10.1155/2012/302029.
Sivertsvik, M., Rosnes, J. T., & Bergslien, H. (2002). Modified atmosphere packaging. In Minimal
processing Technologies in the Food Industry (pp. 61–86). Boston, New York, Washington, DC:
CRC Press.
Sousa-Gallagher, M. J., & Mahajan, P. V. (2013). Integrative mathematical modelling for MAP
design of fresh-produce: Theoretical analysis and experimental validation. Food Control, 29,
444–450. https://doi.org/10.1016/j.foodcont.2012.05.072.
Stern, S. A., Fang, S-M., & Frisch, H. L. (1972). Effect of pressure on gas permeability coeffi-
cients. A new application of “free volume” theory. Journal of Polymer Science Part B: Polymer
Physics, 10, 201–219. https://doi.org/10.1002/pol.1972.160100202
Talasila, P. C., Chau, K. V., & Brecht, J. K. (1992). Effects of gas concentrations and temperature
on O2 consumption of strawberries. Transactions of the ASAE, 35, 221–224.
University of California Davis, Produce Fact Sheets. Produce Fact Sheets. Retrieved from
https://postharvest.ucdavis.edu/Commodity_Resources/Fact_Sheets
Welt, B. (2015). Elevated test pressure significantly reduces dynamic accumulation oxygen trans-
mission rate (ASTM F3136) measurement time for barrier packaging films. Journal of Applied
Packaging Research, 7, 4.
Yang, C. C., & Chinnan, M. S. (1988). Modeling the effect of O2 and CO2 on respiration and
quality of stored tomatoes. Transactions of the ASAE, 31, 920–0925.
Zagory, D. (1995). Principles and practice of modified atmosphere packaging of horticultural
commodities. In J. M. Farber & K. Dodds (Eds.), Principles of modified-atmospjere and sous vide
product packaging (pp. 175–204). Lancaster: Tecnomic Publishing Co.Inc.
Vol. 83, Iss. 6, 2018 r Journal of Food Science 1649