USING HANSEN SOLUBILITY PARAMETERS (HSPS) TO DEVELOP
ANTIOXIDANT-PACKAGING FILM TO ACHIEVE CONTROLLED
RELEASE
By
Jing Gao

A THESIS
Submitted to

Michigan State University

in partial fulfillment of the requirements
for the degree of
Packaging-Master of Science
2014

ABSTRACT
USING HANSEN SOLUBILITY PARAMETERS (HSPs) TO DEVELOP
ANTIOXIDANT-PACKAGING FILM TO ACHIEVE CONTROLLED RELEASE
By
Jing Gao
The Hansen Solubility Parameters (HSPs) methodology was used to estimate the affinity
between poly (butylene adipate-co-terephthalate) (PBAT) and 25 commonly used antioxidants.
The HSPs of PBAT and the antioxidants were calculated by the Hoftyzer and Van Krevelen
group contributions method, and the HSPs of PBAT were also experimentally determined. The
HSPs of PBAT were δd =18.97 ± 0.30, δp =4.83 ± 0.75 and δh =9.10 ± 0.30. Of the selected
25 antioxidants, two antioxidants-α-tocopherol (α-TOC) and propyl gallate (PG), were
incorporated into the PBAT matrix according to their likelihood to migrate from PBAT.
Migration test at 10, 20, and 30

from the produced antioxidant films (α-TOC-PBAT (PBA) and

PG-PBAT (PBP)) into 95% ethanol were conducted to verify the initial migration estimation.
The Fick’s diffusion equations were used to estimate the diffusion coefficients (D) and the mass
of antioxidants release at equilibrium (M∞). The partition coefficients (

) were determined

from the M∞. D values obtained at all three temperatures for the two films were of the same order
of magnitude (10-9 cm2/s) indicating a very fast release of α-TOC and PG from PBAT. Due to the
fast release of the antioxidants from PBAT, the HSP method was not able to properly predict the
likelihood of α-TOC and PG migration from the PBAT films. The activation energy (Ea) was
determined by the Arrhenius equation, as 48.5kJ/mol for PBA film and 54.4kJ/mol for PBP film.
The optical properties of PBA and PBP films are also reported in this study.

ACKNOWLEDGMENTS

First of all, I would like to give my precious appreciation to my major advisor, Dr. Rafael Auras
for accepting me in to his research group and giving me the opportunity to work on this
interesting project. As my advisor, he gave me many guidance and valuable suggestions on my
experiment and thesis writing. I appreciate his patient and encouragement, making me brave to
finish my research. I appreciate his tolerant heart on the mistakes I made in my research. I also
appreciate his educational and professional support for many books and articles he offered me.
Thank you very much for Dr. Auras. I would not finish this project without your mentoring.
I would like to thank my committee members Dr. Maria Rubino and Dr. Kirk Dolan for
their kindly guidance and support throughout my projects. I am very thankful to Aaron Walworth
and Ajay Kathuria for training and helping me in all the machines that I have used for my
research purpose in school of packaging at MSU. I would like to thank Dr. Daniel Jones and Dr.
Jiangyin Bao from the department of Biochemistry & Molecular Biology for their assistance
with my Mass Spectrometry analysis. I would also like to thank all the professors and advisors
from School of Packaging, thank them for spreading the useful knowledge and organizing the
interesting seminars. I really learned a lot during my master study.
I would like to convert my gratitude to all RAA research group members: Hayati, Kang,
Ajay, Turk, Sukkeewan, Fabiola, Edgar, Saleh, Soohyung, Javiera, Marcelo, Rodolfo, Ning,
Siyuan for their kindness and timely assistance and sharing the invaluable expertise with me. A
special thanks to Hayati for sharing her knowledge and being a great guidance in migration
modeling analysis. Without her help, I could not easily analyze my migration data. I would also
like to thank Yining Xia for our discussion about my project and Saleh for his help in the

iii

statistical analysis of the results. Not to forget, Rijosh John Cheruvathur, thank you very much
for your support on my film process.
I could not thank enough to all my friends for their encouragement and kindness. It is a
magic that we met each other. And it will be a great memory in my life that we share the
happiness and sorrow together.
Last but not least, I would like to express my endless thanks to my dear parents, my
brothers and my sisters. Thank you for your concerns and trust. I am extremely grateful for your
support during my entire master’s year. You are always in my heart.

Jing Gao
July, 2014

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................................................... vii
LIST OF FIGURES .................................................................................................................................. ix
KEY TO SYMBOLS AND ABBREVIATIONS ................................................................................. xi
CHAPTER 1 INTRODUCTION & MOTIVATION ............................................................................ 1
1.1
Introduction ...................................................................................................................... 1
1.2
Motivation ........................................................................................................................ 3
1.3
Goal and objectives .......................................................................................................... 4
REFERENCES ............................................................................................................................ 5
CHAPTER 2 LITERATURE REVIEW .................................................................................................. 8
2.1
A brief introduction of food packaging ............................................................................ 8
2.2
Antioxidant packaging ................................................................................................... 10
2.2.1
Antioxidants ............................................................................................................ 10
2.2.2
Plastics added with antioxidants ............................................................................. 12
2.2.3
Biodegradable Polymers ......................................................................................... 15
2.3
Interaction between package and food ........................................................................... 16
2.3.1
Solubility parameter ................................................................................................ 17
2.3.2
Hansen solubility parameters .................................................................................. 18
2.3.3
Methods used to measure HSP ............................................................................... 20
2.3.4
The experimental method ....................................................................................... 20
2.3.5
Group contributions calculation .............................................................................. 22
2.4
Migration process ........................................................................................................... 23
2.4.1
Effect of temperature on diffusion .......................................................................... 24
2.4.2
Diffusion models for migration process ................................................................. 24
2.5
Methods for measuring migration .................................................................................. 26
2.5.1
Migration cell .......................................................................................................... 26
2.5.2
Food simulant selection .......................................................................................... 27
2.5.3
Temperature and time selection for migration testing ............................................ 28
2.6
Liquid chromatography for characterization and analysis ............................................. 28
REFERENCES .......................................................................................................................... 30
CHAPTER 3 MATERIALS AND METHODS ................................................................................... 35
3.1
Materials ......................................................................................................................... 35
3.2
Methods .......................................................................................................................... 35
3.2.1
Hansen solubility parameters (HSP) ....................................................................... 35
3.2.1.1 Experimental method for determining HSP ........................................................ 35
3.2.1.2 Hoftyzer and Van-Krevelen theoretical estimation method of HSP ................... 37
3.2.2
Production of PBAT- α-tocophorol (PBA) and PBAT-Propyl gallate (PBP) Film 38
3.2.3
Extraction and quantification of selected antioxidants for processed film ............. 41
v

3.2.4
Diffusion of antioxidants from processed film into food simulant. ........................ 43
3.2.4.1 Preliminary migration study ................................................................................ 44
3.2.4.2 Designed of the final migration test .................................................................... 44
3.2.5
Antioxidants release models ................................................................................... 45
3.2.6
Activation energy for the diffusion (Ea) of selected antioxidants into food simulant
……………………………………………………………………………………..47
3.2.7
Optical properties .................................................................................................... 47
3.2.8
Statistic analysis ...................................................................................................... 48
REFERENCES .......................................................................................................................... 49
CHAPTER 4 RESULTS AND DISCUSSIONS .................................................................................. 51
4.1
Hansen Solubility parameters determined by the experimental method ........................ 51
4.2
Hoftyzer and Van-Krevelen theoretical calculation method .......................................... 54
4.3
Analysis of HSP parameters ........................................................................................... 57
4.4
Quantification of α-TOC and PG after film processing ................................................. 61
4.5
Migration of α-TOC and PG into food simulant (95% ethanol) .................................... 62
4.5.1
Preliminary migration study ................................................................................... 62
4.5.2
Final migration study .............................................................................................. 64
4.6
Activation energy (Ea) of the diffusion of selected antioxidants into 95% ethanol ...... 71
4.7
Optical property.............................................................................................................. 73
REFERENCES .......................................................................................................................... 75
CHAPTER 5 CONCLUSIONS .............................................................................................................. 78
5.1
Future work .................................................................................................................... 80

vi

LIST OF TABLES

Table 2.1 Structure of antioxidants ............................................................................................... 11
Table 2.2 Food simulants recommended by the FDA…………………………………………...27
Table 3.1 Solvents used for determining the Hansen’s solubility parameters .............................. 36
Table 3.2 Scoring scale for determining the Hansen solubility parameters ................................. 37
Table 3.3 Weight of master batch and PBAT resin used for producing the PBA, PBP and PBC
film ................................................................................................................................................ 41
Table 3.4 MS parameters for the analysis of PG .......................................................................... 43
Table 3.5 Types of boundary conditions and associated number designation, adapted from
(http://exact.unl.edu/exact/home/number.php) ............................................................................. 45
Table 4.1 Scoring of PBAT in various solvents after two days of storage ................................... 51
Table 4.2 Dispersion, polar, hydrogen parameters and interaction radius for PBAT ................... 54
Table 4.3 Solubility parameter components according to the group contribution method of
Hoftyzer-Van Krevele. .................................................................................................................. 55
Table 4.4 Addition of the group contributions for BHT ............................................................... 56
Table 4.5 Summary of HSPs for the polymers, solvents and antioxidants studied ...................... 57
Table 4.6 HSP for PBAT from two methods ................................................................................ 58
Table 4.7 Information summary for selected antioxidants ........................................................... 60
Table 4.8 Concentrations of antioxidants in the film after film processed ................................... 61
Table 4.9 Timing table for PBA final migration test at 10

....................................................... 64

Table 4.10
and values of α-TOC and PG for the diffusion from PBAT films into 95%
ethanol at 10 , 20 , and 30 * .................................................................................................. 65
Table 4.11 D and RMSE of α-TOC and PG for the diffusion from PBAT films into 95% ethanol
at three different temperature * ..................................................................................................... 65

vii

Table 4.12
and RMSE of α-TOC and PG for the diffusion from PBAT films into 95%
ethanol at three different temperature * ......................................................................................... 66
Table 4.13 CIE L*a*b values and total color change (

viii

) for PBC, PBA, PBP * ....................... 73

LIST OF FIGURES

Figure 2.1 Chemical structure of poly (butylene adipate-co-terephthalate) (PBAT) ................... 16
Figure 2.2 HSP sphere with dots representing the selected solvents or materials, modified from
Jing Ma’ et al.[40]......................................................................................................................... 22
Figure 2.3 Migration cell .............................................................................................................. 27
Figure 3.1 Schematic process for the production of antioxidants film ......................................... 39
Figure 3.2 Master batch extrusion process: Left picture shows the feeder machine for PBAT
resin and AOx. Right picture shows the water bath used to cool extruded filament .................... 40
Figure 3.3 PBC, PBA and PBP blow film extrusion process: Left figure presents hopper, screw
and blow film sections; Right figure presents reception of processed film. ................................. 41
Figure 3.4 Migration test cell ........................................................................................................ 43
Figure 4.1 Dissolution of PBAT in dichloromethane, chloroform, THF, tetrachloroethane and
phenol (a- left to right: 1-dichloromethane, 7- chloroform, and 9-THF, b-PBAT in 3tetrachloroethane and 14-phenol after 15 min of sonication, c-PBAT in 3-tetrachloroethane and
14-phenol after two days of storage) (For the better visualization of this dissolution phenomenon,
the reader is suggested to view the electronic version of this document) ..................................... 52
Figure 4.2 Dissolution of PBAT in various solvents after 15 min sonication (from left to right,
named number 1-17: 1-dichloro methane, 2-toluene, 3-tetra chloro ethane, 4-methyl ethyl ketone,
5-cyclohexane, 6-diethylene glycol, 7-chloroform, 8-acetone, 9-THF, 10-acetonitrile, 11dimethyl sulfoxide, 12-ethyl acetate, 13-methanol, 14-phenol, 15-DMF, 16-hexane, 17-ethanol )
....................................................................................................................................................... 53
Figure 4.3 Dissolution of PBAT in various solvents after two days storage (from left to right,
named number 1-17: 1-dichloro methane, 2-toluene, 3-tetra chloro ethane, 4-methyl ethyl ketone,
5-cyclohexane, 6-diethylene glycol, 7-chloroform, 8-acetone, 9-THF, 10-acetonitrile, 11dimethyl sulfoxide, 12-ethyl acetate, 13-methanol, 14-phenol, 15-DMF, 16-hexane, 17-ethanol)
....................................................................................................................................................... 53
Figure 4.4 BHT chemical structures ............................................................................................. 54
Figure 4.5 3D plot of the HSP parameters for the 25 antioxidants, 2 biodegradable polymers and
water and ethanol. The spheres show the radius of interaction of the solvents and the polymers.
....................................................................................................................................................... 59
Figure 4.6 3D plot of selected HSP parameters ............................................................................ 60
ix

Figure 4.7 Scaled sensitivity coefficient plot for PBA migration trial at 30, 20, and 10 and PBP
migration trial at 20 : (a) PBA migration trial at 10 , (b) PBA migration trial at 20 , (c) PBA
migration trial at 30 , (d) PBP migration trial at 20 ................................................................ 63
Figure 4.8 The activation energy for diffusion of α-TOC (y= -2515.8649+0.1886; R2=0.9450) (in
the left side of Figure) and PG (y= -2859.4411x+1.4215; R2=0.9211) (in the right side of Figure)
from processed film into 95% ethanol. The slope of each line was equal to –Ea/2.03R. ............. 72
Figure 4.9 The appearance of PBC (left), PBA (middle), PBP (right) ......................................... 73
Figure 4.10 UV-visible light transmission spectra of PBC, PBA and PBP films. The film
thicknesses of PBC, PBA and PBP were 63.5 ± 5.1, 68.6 ± 2.5 and 76.2 ± 5.1
, respectively.
....................................................................................................................................................... 74

x

KEY TO SYMBOLS AND ABBREVIATIONS

AOx

Antioxidants

APF

Antioxidant packaging film

HSPs

Hansen solubility parameters

MAP

Modified atmosphere packaging

PG

Propyl gallate

OG

Octyl gallate

BHA

Butylhydroxyanisole

TBHQ

Tert-Butyl-hydroquinone

-TOC

-tocopherols

AA

Ascorbic acid

IR 1076

Irganox 1076

BHT

Butylated hydroxytoluene

PBAT

Poly (butylene adipate-co-terephthalate)

PBCM

PBAT without antioxidants control master batch

PBAM

-TOC of PBAT master batch

PBPM

PG of PBAT master batch

PBC

PBAT control film

PBA
PBP

-TOC of PBAT film
PG of PBAT film
Diffusion coefficient
Permeant concentration
Diffusion time

xi

Glass transition temperature
Activation energy
Gas constant
Absolute temperature
Migrant concentration in the food or food simulant
Migration level at time t
Initial migrant concentration
Polymer film thickness
Volume of the polymer
Volume of the food or food simulant
Partition coefficient of the migrant
Migration level at equilibrium
Area of the polymer in contact with the food simulant
Root mean-square error
Solubility parameter distance
Dispersion parameter
Polar parameter
Hydrogen bonding parameter

xii

CHAPTER 1 INTRODUCTION & MOTIVATION
1.1

Introduction
In the food industry, packaging plays a vital role to protect the products from the outer

environment and to extend their shelf life. There are also other functions performed by a
package like containment of the products and providing detailed information about the package
about the products [1]. Still, for food products quality assurance is one of the most important
functions of the package. For oxygen-sensitive foods like milk and other dairy products
containing protein and lipid, which are subjected to oxidation and will generate oxidative
rancidity when they are exposure to oxygen, then packages play an important role to delay and/or
avoid oxygen exposure. Oxidative rancidity is a major cause of food quality deterioration,
leading to the formation of undesirable off-flavors as well as unhealthful compounds [2, 3]. Thus,
oxidized food products have decreasing acceptability by consumers. Several methods are
developed to retard oxidation based on directly protecting the food by adding antioxidant and/or
using the packaging system to delay oxygen permeation and/or releasing antioxidants.
Antioxidants can inhibit the oxidation of foods by scavenging free radicals, chelating prooxidative metals and quenching singlet oxygen [4, 5]. They can be added to the food and/or the
packaging system. If added to the food products, they are highly effective; but their type and
amount are highly regulated by the United States Food Drug Administration (FDA) and other
international agencies to assure human’s health [6]. The other method, and the subject of this
thesis, is introducing the antioxidants on the packaging system for slow release. This method
provides the advantage of controlling the amount of antioxidant releasing to the product in
function of time, but it also provides a number of challenges such as degradation temperature,
release kinetic control, and antioxidant stability.
1

Additionally methods used to delay oxygen intake are the use of oxygen scavenger in
form of sachets that contains a compounds to scavenge the oxygen through oxidation reactions.
These sachets are placed into the package together with the products. Some disadvantages of the
sachets are the need for an additional packaging operation to add the sachet to the package, and
the limit oxygen scavenger capacity of the sachet.
Antioxidant active packaging made as compounded polymeric membranes based on a
polymer matrix and the addition of one or more antioxidants to control the release of the
antioxidants in contact with the food has gaining attention as a feasible option to control food
products’ oxidation. The use of antioxidant polymeric membranes benefits both the packaging
and the food products themselves because the addition of the antioxidant in the polymeric
packaging may help to stabilize the polymer during processing and also favors to inhibit the
products’ oxidation [7-9]
Of the produced antioxidant polymeric films most of them are based on the use of
commercial polymers such as polyolefins and/or polyester. Due to environmental concerns, a
current trend is to develop APF made from biodegradable polymers and natural antioxidants.
Polymers such as poly (lactic acid) -PLA, soy-protein, chitin, and starch-based polymers are
being used to reduce the dependence on petroleum based resources. The most used natural
antioxidants are a-tocopherol, ascorbic acid, quercetin among others. Most of these APF are
produced based on the trial and error or base on previous experience. However, researcher have
demonstrated that it is difficult to achieve a suitable release rate for different food systems and
degradation reactions. The release rates of the antioxidants added to the films are either too slow
or too fast. The interactions between the polymer and antioxidants are a key point to estimate
their affinity and the rate and total amount of antioxidants released to the food product from the

2

APF. Several methods can be used to determine the interaction of chemical compounds with
polymers such as a Flory-Huggins approach, quantitative structure property relationship (QSPR)
[10, 11], and the Hansen solubility parameters [12]. A fast method to determine the relative
affinity between a polymer and an additives is the solubility parameter [13]. The term solubility
parameter was first used by Hildebrand and Scott. They correlated solubility with cohesive
properties of the solvents [12]. The earlier solubility parameter work has a shortcoming, which is
limited to regular solutions and does not consider the association between molecules like polar
and hydrogen-bonding interactions. Hansen (1967) proposed an extension of the solubility
parameter approach to polar and hydrogen- bonding systems. Thus, the Hansen solubility
parameters (
polar part and

are defined as

representing nonpolar part,

representing

representing hydrogen bonding part. The basic principle of HSP is “like

dissolves like”. That means the materials with similar HSP have high affinity for each other.
Then, the likeness between a polymer and chemical could be estimated by using their HSP [14].
Therefore, the estimation of compatibility between a polymer and chemicals especially
antioxidants can help to save time, efforts and money, which are always expended on a new
designed packaging system.
1.2

Motivation
As discussed above, more and more attention is given on developing APF based on

biodegradable polymers and natural antioxidants. However, most APF are designed based on
trials and errors, and the results showed that appropriate release rate for the new designed APF
are hard to achieve. To help address this problem, we are proposed to use the HSP theory to
estimate the interaction between natural antioxidant and biodegradable polymers. If the
compatibility and the release rate between natural antioxidants and biodegradable polymers and
3

the surrounding environment can be estimated properly, time, resources and funding can be
optimally used.
1.3

Goal and objectives
The overall goal of this thesis is to find a method to estimate the release rate of

antioxidant from polymers. In this study, the HSP theory is used to describe the affinity between
biodegradable polymer and antioxidants. The following tasks are listed to achieve the goal:
1 Determine the HSP for the selected biodegradable polymers and antioxidants by using the
Hoy and Van-krevelen method
2 Determine the HSP for the selected biodegradable polymers by using an experimental
technique
3 Identify the APF system based on HSP theory to achieve an optimal release of the
antioxidant
4 Conduct migration studies on the final processed APF;
5 Analyze and compare the experimental result with the theoretical HSP estimations

4

REFERENCES

5

REFERENCES

1.

Selke, S.E., J.D. Culter, and R.J. Hernandez, Plastics packaging: Properties, processing,
applications, and regulations. 2004: Hanser Munich.

2.

Decker, E., R. Elias, and D.J. McClements, Oxidation in Foods and Beverages and
Antioxidant Applications: Understanding mechanisms of oxidation and antioxidant
activity. Vol. 1. 2010: Elsevier.

3.

Marsh, K. and B. Bugusu, Food packaging—roles, materials, and environmental issues.
Journal of food science, 2007. 72(3): p. R39-R55.

4.

Choe, E. and D.B. Min, Mechanisms of antioxidants in the oxidation of foods.
Comprehensive Reviews in Food Science and Food Safety, 2009. 8(4): p. 345-358.

5.

Wanasundara, P. and F. Shahidi, Antioxidants: Science, technology, and applications.
Bailey's Industrial Oil and Fat Products, 2005.

6.

Ashby, R., Food packaging migration and legislation. 1997.

7.

Miltz, J., et al., Loss of antioxidants from high-density polyethylene. Food and Packaging
Interactions, ed. by Hotchkiss JH. American Chemical Society, 1988: p. 83-93.

8.

Sharma, G., C. Madhura, and S. Arya, Interaction of plastic films with foods. II, Effect of
polyethylene and polypropylene films on the stability of vegetable oils. Journal of food
science and technology, 1990. 27(6): p. 328-331.

9.

Van Aardt, M., et al. Controlled release of antioxidants from polymeric films to control
lipid oxidation in milk. in International Animal Agriculture and Food Science Conference,
Indianapolis. 2001.

10.

Tehrany, E.A., F. Fournier, and S. Desobry, Simple method to calculate partition
coefficient of migrant in food simulant/polymer system. Journal of food engineering, 2006.
77(1): p. 135-139.

11.

Luan, F., et al., QSPR study of permeability coefficients through low-density polyethylene
based on radial basis function neural networks and the heuristic method. Computational
materials science, 2006. 37(4): p. 454-461.

12.

Hansen, C.M., Hansen solubility parameters: a user's handbook. 2007: CRC.

13.

Auras, R., B. Harte, and S. Selke, Sorption of ethyl acetate and d‐limonene in poly
(lactide) polymers. Journal of the Science of Food and Agriculture, 2005. 86(4): p. 648656.
6

14.

Hansen, C.M., Polymer additives and solubility parameters. Progress in organic coatings,
2004. 51(2): p. 109-112.

7

CHAPTER 2 LITERATURE REVIEW
Introduction
This chapter provides a brief introduction of food packaging, including the function of packaging,
and the methods used to protect food in food packaging. Then, a section about the main
antioxidants (AOx) used in packaging and food is presented. After that the main antioxidant used
in plastics and their Hansen solubility parameters (HSP) used to estimate the affinity between
polymer and additives are discussed. Finally, the method to test for migration from plastics and
the main instruments used to conduct this determination are described.
2.1

A brief introduction of food packaging

Packaging is a very important process to maintain the quality of products during storage,
transportation and end-use. The basic functions of packaging are containment, protection,
information and convenience [1]. Among them, the main function of food packaging is to
achieve preservation and the safe delivery of food products. So, food packaging can contribute to
extend the shelf life and maintain the quality and safety of food products. Many packaging
materials are used for food packaging, such as metal, glass, paper and plastics. Compare to other
materials, plastics have a number of advantages, such as, low cost, light weight, easy to shape,
and tailored barrier properties [2]. Food packaging functions have evolved from simple delivers
and preservation methods to packaging systems that enhance convenience, safety as well as
reducing the environmental impact of the entire product package system [3]. In the last decades,
food technologists have focused on how to increase food safety and security [4]. To achieve that,
food packaging must actively protect the food product by enhancing their shelf life, which can be
fulfilled by adding active or selective barrier properties to the packaging structure. In the food
packaging industry, several functional packaging systems including active packaging, modified
8

atmosphere packaging (MAP) and edible films and coating have been developed to maintain and
enhance the quality and shelf-life of food products [4].
Many applications of active packaging have been investigated and created to enhance the
safety and security of food products [5]. Active packaging is designed to extend the shelf life of
food products, prevents the deterioration of quality and inhibits microbial growth, thus ensuring
food safety and quality. Examples of active packaging include oxygen scavenger, moisture
absorbers, flavor releasing/absorbing systems, and antioxidants releasing systems.
For oxygen sensitive foods, reducing their exposure to oxygen in the packaging system is
very important to keep the shelf life of the product. One method currently uses to delay oxygen
uptake or retard food oxidation and rancidity is active packaging. There are different active
packaging methods. The inclusion of oxygen scavengers in the form of sachets within the cavity
or interior of a package is extensively observed in the market, due to its lower cost and
adaptability to different system. However, disadvantage of sachets are the additional packaging
operations needed to put a sachet into a package, which increases packaging operations and costs,
the need for a particular atmospheric conditions to achieve a targeted scavenging rate, and the
chances of being accidentally ingested by minors [6]. Also, scavenging sachets are difficult to be
used for liquid food, since direct contact of the liquid food with the sachets may cause leakage of
the sachet contents. Other means of controlling the oxygen exposure by the food product is the
incorporation of scavenging material (i.e., antioxidants) in the packaging itself and control its
release. This method can provide a more uniform scavenging effect throughout the package and
its shelf life. Because incorporating the oxygen scavenger into the packaging structure is an
efficient way to scavenge oxygen, which will permeate from the outside of the package. So, the
lowest possible oxygen in the package system can be achieved [7]. Packaging structures

9

containing AOxs is an alternative method to reduce the oxidation of food. For a good AOx active
packaging to work, the appropriate packaging material and AOxs should be selected. There are
certain concerns raised about active AOx packaging, such as whether the added materials will
affect processing, and the final properties of the active packaging.
2.2

Antioxidant packaging

2.2.1 Antioxidants
Antioxidants are widely used food additives to improve food stability and extend the shelf-life of
oxygen-sensitive foods. AOxs, both synthetic and natural, are very important chemical
compounds to prevent oxidative degradation and consequently extend the shelf-life of foods. The
structures for some commonly used antioxidants are shown in Table 2.1. Synthetic antioxidants
such as propyl gallate (PG), octyl gallate (OG), butylhydroxyanisole (BHA), BHT and tertbutyl-hydroquinone (TBHQ) are commercially used following the regulations of the European
Union Directives and Regulations, the FDA in USA, and the Food Standards of Australia and
New Zealand [8]. The maximum allowances of these AOxs are regulated by these organizations.
Among the synthetic AOxs, PG and BHA are powerful antioxidants widely used to stabilize food
from oxidation. Among the natural AOxs,

-tocopherols (Vitamin E), and ascorbic acid

(Vitamin C) are commonly used. Catechin, epicatechin, and resveratrol among others AOx are
newly being considered.

-tocopherol is a lipid-soluble antioxidant and can be found in many

foods like vegetable oils [9]. It is one of most effective radical-chain breaker in unsaturated fatty
food. -tocopherol has be recognized as a safe food additive by The Code of Federal Regulations
[10]. The incorporation of -tocopherol with low- and high- density polyethylene (LDPE and
HDPE) has shown to inhibit oxidation of food oxygen sensitive product [11, 12]. Catechin and
epicatechin belong to the flavonoid group of AOx and can be found in green tea or in fruits like
10

grapes [13]. Resveratrol is a phenolic compound of the stibene family, with similar structure to
catechin and epicatechin. Resveratrol is commonly found in wines and in various parts of grapes
including the skin and the seeds [14]. Radical-scavenging capacity of propyl gallate, ascorbic
acid and alpha-tocopherol has been reported lower than resveratrol [15]. Due to consumer
demands and market trends, natural AOxs are promising AOxs to be incorporated in polymer
films to exert antioxidants effects.
Table 2.1 Structure of antioxidants
Antioxidants

IUPAC Nomenclature

Structure

BHT

2,6-ditert-butyl-4-methylphenol

BHA

2-tert-butyl-4-methoxyphenol; 3tert-butyl-4-methoxyphenol

TBHQ

2-tert-butylbenzene-1,4-diol

Propyl gallate

propyl 3,4,5 -trihydroxybenzoate

Octyl gallate

Octyl 3,4,5-trihydroxybenzoate

Dodecyl
gallate

dodecyl 3,4,5-trihydroxybenzoate

11

Table 2.1 (cont’d)
Antioxidants

IUPAC Nomenclature

Ascorbic acid

(5R)-[(1S)-1,2-dihydroxyethyl]-3,4dihydroxyfuran-2(5H)-one

Irganox 1076

octadecyl-3-(3,5-ditert-butyl-4hydroxyphenyl) propanoate

Catechin

(2R,3S)-2-(3,4-dihydroxyphenyl)-3,4dihydro-2H-chromene-3,5,7-triol

Epicatechin
gallate

[(2S-3S)-2-(3,4-dihydroxyphenyl)5,7-dihydroxy-3,4-dihydro-2Hchromen-3-yl]3,4,5trihydroxybenzoate

Alphatocopherol

5,7,8-trimethyltocol

Resveratrol

5-[(E)-2-(4hydroxyphenyl)ethenyl]benzene-1,3diol

Structure

2.2.2 Plastics added with antioxidants
Plastics were introduced as main materials at the beginning of 20th century. Properties such as
softness, transparency and low cost, have made plastics essential materials in many industries.
Most commercial plastics are petroleum-based synthetic polymers, such as low density LDPE,
12

polystyrene (PS) and poly (ethylene terephthalate) - PET. Since they are resistant against
microbial attack, it cannot be degraded when they are discarded into the environment [16]. With
the increasing concerns on environment issues, polymers derived from bio-based resources,
which are recyclable and biodegradable have been developed with the intention to reduce the
environment pollution caused by plastics. Furthermore, food package are usually contaminated
with food, so they cannot practically be recycled. Bio-based and compostable plastics made from
renewable resources such as poly (lactic acid) (PLA), poly (hydroxyl-alkanoates) (PHAs), poly
(3-hydroxybutyrate) (PHBs) and thermoplastics starch have been developed to try to deal with
these issues. These polymers are biodegradable and compostable. Biodegradability depends on
the polymer structure, and it is not related to the origin of the material [17]. So, biodegraded
polymer can be produced from non-renewable and renewable sources, and it can be defined as
biodegradable if they can be mineralized by a natural process.
However, there are still drawbacks with these newly developed polymers. For example,
PLA is brittle and have low barrier properties, limiting its application to some extent. As for
starch-based materials, they are often very sensitive to moisture conditions, so they mechanical
properties are dramatically reduced in humid environment. Another disadvantage of bio-based
materials is that they still relatively more expensive than their counterpart petroleum-based
polymers.
Polymers unlike glass interact with small molecules, which can be sorb, diffused and
desorbed across their structure. AOx polymer uses this principle to release AOx from the
polymers and to protect oxygen sensitive food products. AOx packaging is one type of active
packaging, in which the antioxidants are incorporated into the plastic films to reduce oxidation of
the packed food. The incorporated of AOx can serve dual functions: (1) avoid polymer

13

degradation during processing, and (2) reduce food oxidation. Extensive research has been done
into these systems, mostly from polymers made from non-renewable resources. Research done
on oatmeal [18] and vegetable oil [19] have demonstrated that impregnated AOx within plastic
films-AOx packaging films (APF)- can increase the storage stability of food since they are
released from the polymer. In oatmeal cereals packaged in HDPE film with 0.022% (w/w) BHT
and 0.32% (w/w) BHT, HDPE films with lower BHT showed higher oxidation after six weeks
storage at 39℃. Marcato et al. investigated the migration of AOx additives pentaerythrityl
tetrakis(3,5-di-tert-butyl-4-hydroxyphenyl) propionate (Irganox 1010) and tris(2,4-di-tertbutylphenyl) phosphite (Irgafos 168) from polyolefin into oily simulants [20]. The studied
polyolefin were isotactic polypropylene homopolymer (PP), ethylene-co-propylene random
copolymer (RACO), ethylene-propylene heterophasic copolymer and ethylene-propylene
amorphous copolymer blend (EP) and HDPE. The result showed the amount of Irganox 1010
transfer from these plastic materials at 25℃ were different, decreasing in the order EP > RACO >
PP > HDPE. The same polyolefin ranking was obtained for Irgafos 168. This information can be
used for the selection of polyolefin as raw materials for the production of pharmaceutical and
cosmetic containers. However, BHT, Irganox 1010 and Irgafos 168 are synthetic AOx, the
addition amount of synthetic antioxidant to food packaging is specifically regulated by the FDA
due to human health concerns. There is a growing interest in using natural AOxs in food
packaging applications. Therefore, research has been studied combination of natural AOx and
bio-based material or biodegradable materials. For example, release of -tocopherol from PLA
films delayed the induction of the oxidation of soybean oil when stored with 2.5% (w/w) tocopherol PLA film at 20 and 30℃, comparing with that of oil in contact with PLA without
APF [10]. A number of natural AOxs have been incorporated into PLA [10, 13, 21, 22]; however,
14

few research has been conducted into other polymers like poly (butylene adipate-co-terephthalate)
(PBAT), which is also compostable. PBAT is currently made from non-renewable resources, but
in the future it is expected to be produced from renewable resources.
2.2.3 Biodegradable Polymers
Biodegradable polymers can be identified as: (1) polymers directly extracted from
biomass like proteins, lipids, etc.; (2) polymers synthesized by a classical polymerization
procedure such as aliphatic-aromatic copolymers, aliphatic polyesters, polylactide aliphatic
copolymer (CPLA), using renewable bio-based monomers such as poly (lactic acid) and oilbased monomers like polycaprolactones; and (3) polymers produced by microorganisms and
bacteria like polyhydroxylalkanoates [23]. Aliphatic polyesters such as poly(hydroxyl alkanoates)
–PHA and synthetic polycaprolactone (PCL), have good biodegradable properties. PHA is
produced in microorganism cells and PCL is produced from non-renewable resources. The lack
of thermal and mechanical properties of PHAs limits their wide application. Aliphatic-aromatic
copolyesters are a more promising alternative as conventional plastics. These plastics present
good material properties and biodegradability. These copolyesters are produced from the random
polymerization of the diester oligomers of adipic acid/butanediol, and terephthalic
acid/butanedoil [24][25].
Poly (butylene adipate-co-terephthalate) (PBAT) is an aliphatic-aromatic copolyester that
is biodegradable and can be degraded within a few weeks with the aid of naturally occurring
enzymes [26]. The structure of PBAT consists of butylene adipate (BA) unit and butylene
terephthalate (BT) unit, as showed in figure 2.1. The biodegradability and the material properties
are relative to the composition of BA and BT. It has been reported that the optimal compromise
between biodegradability and physical properties of PBAT copolyesters can be obtained by a

15

range from about 35 to 55 mol% of terephthalic acid (TA) (with regard to the total amounts of
acid components) [27]. So, adjusting the composition of the aromatic fraction can fulfill the
desired application of PBAT copolymers. The great advantage of PBAT copolyesters is that can
be synthesized from cheap and widely available bulk chemicals by traditional technical process.
Research has shown that thermal and mechanical properties, and the rate of biodegradation all
depend on ratio of BA/BT. A range of glass transition temperature (Tg) for PBAT copolymers
can be estimated from -61 to -23 °C. Melting temperature (Tm) from 80 to 140 °C can be
obtained by altering copolymer composition [27]. PBAT is a flexible and tough polymer
designed mainly for film extrusion and extrusion coating [28]. Since PBAT has great
biodegradability and high toughness, it becomes an excellent candidate for tough biodegradable
polymers such as poly(lactic acid) (PLA) and poly(hydroxybutyrate-co-hydroxyvalerate)
(PHBV)[29].

Figure 2.1 Chemical structure of poly (butylene adipate-co-terephthalate) (PBAT)
2.3

Interaction between package and food

For oxygen sensitive food, active packaging not only should absorb the excess oxygen but it
should also avoid the ingress of oxygen to the package and the release of the oxygen scavenger
from the package to protect the food product. In the case of plastic in contact with food, there is
always an interaction between the polymer matrix and the food in which small molecules diffuse
from one part of the system to another. This interaction is part of the mass transfer occurring
within the packaging system. Migration is a mass transfer process, in which low molecular mass

16

substances initially present in package are released into the contained food, and it is the result of
diffusion, dissolution and an equilibrium processes [30]. APF used this mechanism to protect
oxygen sensitive food by incorporating AOx in the plastic materials intended to migrate to the
food [31]. The AOx transfer through a polymer depends on the interaction between the AOx and
the packaging materials as well as the solubility and affinity of the AOx to the food product [32].
If there is a high affinity between the polymer and the AOx, it is hard for the AOx to migrate
from the polymer to protect the product. When the likeness between the polymer and the AOx is
weak, the AOx migrates to the food too easily, which is also not good for a long extension of the
product’s shelf life. So, a tailored release is needed to get the right protection. To predict the
mass transfer process, it is necessary to estimate the compatibility and affinity between the
polymer and the AOx. There is not very much research conducted and reported to estimate the
produce compatibility and affinity between the polymer and AOx to design APF.
2.3.1 Solubility parameter
Solubility parameters, in particular the Hansen solubility parameters (HSP), have been
shown useful to correlate polymer solution phenomena and chemical resistance since the HSP
are easy to calculate or experimentally determined [32]. The initial concepts of the solubility
parameter was developed by Scatchard and subsequently greatly extended by Hildebrand [33].
The solubility parameters have been used to select solvents for coatings materials for many years.
A series of improvements and other application in solubility parameter has been made over the
years, such as to predict compatibility of polymers, chemical resistance, and permeation rates,
and characterize the surfaces of pigments or fibers [34].

17

The total solubility parameter was first introduced by Hildebrand and Scott [34]. The
Hildebrand solubility parameter

is defined as the square root of the cohesive energy density

(CED):
2.1
2.2
As the equation 2.1 expresses,
for the pure solvent,

is calculated by dividing the energy of vaporization

by the molar volume, V, of the involved liquid, and making the square

root of this number. The energy of vaporization for the liquid is calculated by equation 2.2,
where

is the heat of vaporization for the liquid, R is the universal gas constant, and T is the

absolute temperature. The common used unit for

in the USA is (cal/cm3)1/2. The SI unit for it

is MPa1/2, which is 2.0455 times larger than that in the former unit. The basic principle to using
the solubility parameter is “like dissolves like”. Liquids with similar solubility parameters will be
miscible. Also, polymers will dissolve in solvents whose solubility parameters are similar with
them. Therefore, the affinity of a polymer and solvent can be predicted by using solubility
parameters.
2.3.2 Hansen solubility parameters
A shortcoming of the earlier solubility parameter work is that the approach was limited to
regular solutions [35], since other interaction could exist between molecules such as polar and
hydrogen-bonding interactions. To attempt to improve on this, several researchers have been
working on it. A recolection of the research and advances can be found in Barton’s extensive
book [36]. The work made by Gardon and Teas is also a good source [37]. The approach of
Blanks and Prausnitz divided the solubility parameter into “nonpolar” and “polar” [38]; the
approach of Burrell divided solvents into hydrogen bonding classes. Blanks and Prausnitz’s
18

approach greatly influenced Hansen’s earlier work [34]. Hansen proposed an extension of the
single Hildebrand solubility parameter to three components solubility parameters (nonpolar,
polar and hydrogen-bonding parameters), now call the Hansen solubility parameters (HSP). This
is based on the fact that all types of physical bonds are broken when they are evaporated
including those three kinds of bonds. As mentioned before, materials with similar Hildebrand
solubility parameters should have high affinity for each other. However, in some cases, this is
not applicable. For example, the Hildebrand solubility parameter of ethylene carbonate and
methanol are identical, 29 MPa1/2, but their solvencies are different [39]. This can be illustrated
by HSP of these two solvent. For ethylene carbonate,
5.1 MPa1/2 respectively, and for methanol,

,

and

,

and

are 18 MPa1/2, 21.7 MPa1/2,

are 15.1 MPa1/2, 12.3 MPa1/2, 22.3

MPa1/2 respectively. It is easily noticed that the HSP of these two solvents are not the same. This
phenomenon can never been predicted by the Hildbrand solubility parameters. There are three
types of interaction included in HSP concept.
Nonpolar interaction (

) — The most general one is nonpolar interaction, which is

derived from atomic forces and also named dispersion interactions in the literature [34]. All
molecules contain this type of attractive forces since molecules are built up from atoms.
Polar interaction (

) — Polar interaction are inherently molecular interactions and found

in most molecules to one extent or another [34]. It is caused by permanent dipole. The molecular
dipoles occur when there is an unequal sharing of electrons between atoms in a molecule. The
dipole moment is the primary parameter which is used to calculate polar interactions.
Hydrogen bonding interaction (

) — The basis of the hydrogen bonding interaction is

the molecule attraction because of hydrogen bond. This can be called electron exchange

19

parameter. In HSP approach, the hydrogen bonding parameters are defined to more or less
collect the energies from interactions not included in other two parameters [34].
2.3.3 Methods used to measure HSP
According to the HSP concept, the total cohesion energy E must be the sum of three individual
energies, as expressed in equation 2.3. Also the relationship of Hildebrand solubility parameters
with HSP is found as equation 2.4.
2.3
2.4
Thus, the HSP is considered as a more reliable method to estimate the affinity of a polymer and a
solvent.
Theoretical and experimental methods can be used to estimate the HSP of a material.
When the experimental data for the test material is not available, the group contribution can be
used as a first estimate of the solubility behavior of a material. It is recommended to always
confirm the HSP for polymers determined by group contributions with experimentally
determined HSP [34].
2.3.4 The experimental method
The experimental method to determine HSP for a polymer is to evaluate whether or not it
dissolves in selected solvents. Experimental data is based on the observation of the interaction
between studied materials and well-known solvents. A lot of different phenomena will be
observed including full solution, degree of swelling by visual observation, clarity, surface attack,
etc. The solvents that have similar or close HSP to the studied material will dissolve the material;
on the other hand, if any visual change is observed, there is assumed that little interaction
between the solvent and studied material exist. These different interactions are used to divide the
20

solvents into two groups, one which is considered “good” and the other which is considered “bad”
[34]. “Good” solvents have strong interaction with the studied material, indicating that they have
closer HSP to materials studied than “bad” solvents. Various ways can be used to generate and
treat experimental data, like using a score scale to rank the different interactions happening with
the selected solvents. This data can then be processed to determine the HSP parameters. The
three HSP parameters and the radius value, Ro, of the sphere of interaction for materials can be
obtained from these calculations. Mainly, a sphere of interaction is determined for the solvents
and the material. Ro is the radius of interaction for the studied material, which gives the
maximum difference allowed to define a good interaction between the solvents and materials or
other two materials. To evaluate whether or not a material belongs to a sphere of high affinity of
another material, the distance Ra between these two materials can be calculated by the equation
2.5.

,

and

refer to the dispersion, polar and hydrogen parameters respectively. The

subscripts present the different materials.
2.5
2.6
The ratios between Ra and Ro is called the Relative Energy Difference (RED), see the
equation 2.6. This parameter can be referred to find whether a material is within the sphere of
affinity of other studied material or solvents. Good interactions of materials always show a RED
number less than 1.0. The RED number 1 means the material is right on the soluble/insoluble
border [39]. Fig 2.1 shows a three-dimensional graph of equations 2.5 and 2.6. It is easy to
visualize the solubility properties, and the distance between materials. The HSP coordinates of a
material are located in the center of sphere with radius Ro. Ra is the distance between the studied
material and specified solvent or material.
21

Figure 2.2 HSP sphere with dots representing the selected solvents or materials, modified from
Jing Ma’ et al.[40]
2.3.5 Group contributions calculation
There are a limited number of solvents and material, which HSP values have been
already determined. Measuring the HSP through experimental method is time consuming.
Therefore, methods to predict the HSP are highly valuable. One of the useful prediction method
is based on the molecular structure of a material. However, the available experimental data
indicates that it is impossible to use a calculation methods from the chemical structure for an
accurate prediction of solubility parameter components [41]. However, such prediction method
can provide a rough estimation of HSP, which is useful when there is no available data. There
are two highly published methods, Hoftyzer and Van Krevelen (1976) and Hoy (1985). Both
methods have the same basic assumption, as expressed in equations 2.3 and 2.4. Van Krevelen
and Hoy both proposed a set of group contribution values of cohesive energy E, and the molar
attraction constant F which is related with the cohesive energy E [41]. Van Krevelen (1965)
derived a set of atomic contributions to calculate F, which derive in an indirect way to the values
of E for polymers. The structure of studied materials can be broken down into several different
groups, which have their own F and E values referred from the group contribution values. Values
for F can be obtained by the published tables depending on the used method [41]. Based on F

22

and E values of the group contribution method, the HSP of a material could be estimated by the
equations by either the Hoftyzer-Van Krevelen or Hoy methods. The usages of both methods are
easily implemented.
2.4

Migration process

Due to the increasing concerns about health matters among consumers, the importance of
migration of substances from food packaging to food has attracted increasingly attention.
Migration can be considered as the desorption process of chemical compounds from the
packaging material into the product. The interaction of packaging materials with the food can
strongly affect the migration process. The released components called migrants such as
monomers and additives are common migrants in packaging materials. Migrants can be sensitive
to human health, so they are regulated by the EU (European Union) and the FDA (Food and
Drugs Administration). Legally, polymers for packaging are regulated through global or specific
migration levels [42]. Global migration measures the total amount of compounds migrated to a
food simulant under the designed temperature and time conditions. However specific migration
is to measure the migration of a specific chemical component. Due to the toxicity of a specific
component, it is important to conduct specific migration.
The migration process in packaging systems specially in polymers is influenced by the
diffusion and partition coefficients [43]. The diffusion coefficient (D) provides information on
the migration rate, and it is affected by the molecular structure, weight, and affinity of the
migrant to the food simulant and/or food, and the affinity of the polymer to the migrant [44]. The
partition coefficient (Kp/s) provides information on the quantity transferred from a polymer to a
food at equilibrium, and it is affected by the solubility and affinity between the polymer phase
and food product phase. The diffusion process in migration is mostly described by Fick’s laws of
23

diffusion [45]. The Fick first law of diffusion at steady state for one dimension of a migrant from
a polymer is expressed by equation 2.7. For unsteady state in one dimension, the Fick second law
is used, as expressed in equation 2.8.
2.7

2.8
where F is the transfer rate of the migrant per unit area,
substance in the polymer (cm2/s),

is the diffusion coefficient of the

is the migrant concentration in the polymer (g migrants/g

polymer), X is the diffusion distance (cm), and t is the elapsed time (s).
2.4.1 Effect of temperature on diffusion
During the diffusion process, the effect of temperature is an important factor, which affects the
mobility of migrants in polymeric materials and influences the diffusion process. When
temperature is higher than the Tg of the polymer, the polymer is at the rubbery state, which
enhances the diffusion process. When the temperature is lower than Tg of the polymer, the
polymer is at glass state. Due to the restriction of the polymer chains, it requires more time for
the migrants to be transported through the polymer matrix and to reach an equilibrium state.
2.4.2 Diffusion models for migration process
Migration processes are described by the diffusion kinetics of migrant in the film and it is
expressed by the diffusion coefficient (D). According to the Fick’s second law, different solution
of a system can be obtained depending on the boundary conditions of the system that we are
solving for as expressed by equations 2.9 and 2.12 for one-dimension diffusion[44]. A the initial
time (t=0), there is no migration happening between the package and the food, so the migrant
concentration in the food is zero. When the migration begins to occur in the film and the food

24

system, the amount of migrant, which is migrated from the film, will be equal to the amount
existing in the food during the migration process. When the diffusion is in one-dimension and in
a limited volume of film in finite volume of solution, equation 2.9 is used.
∑
where

2.9

is the concentration of the antioxidant diffused at time t divided by the

concentration of the antioxidant diffused at equilibrium; l is the thickness of the film and the
and α as expressed by equation 2.10:

are the non-zero positive roots of tan

2.10
where

and

are the volume of the simulant and the polymer.

is the partition coefficient

of antioxidant between the polymer and the simulant, which can be assumed constant at a lower
concentration and calculated from the ratio of the concentration of the antioxidant in the film
(

, µg/cm3) and the simulant (

, µg/cm3) at equilibrium according to following equation:
2.11

When the diffusion happens in one-dimension and in a limited volume of film in infinite
volume of solution, equation 2.12 is applied,
∑

2.12

Equation 2.12 is a simplified solution, which is used for the case that the amount of simulant can
be considered infinite (i.e.,

since

and /or

25

).

2.5

Methods for measuring migration

Migration testing is mostly carried out according to specific regulation. In the US, the FDA is the
governing body providing regulations and methods for migration testing. The FDA also provides
the guidance and recommendations for industry to perform migration studies. Migration testing
is controlled by appropriate laboratory conditions. The main intention in a migration experiment
is always to simulate the migration process as happened in a real case. The purpose of rules and
legislations related to migration is to assure the safety of the materials used in contact with food
and to guarantee consumer’s safety.
2.5.1 Migration cell
The migration cells to run migration experiments are normally made of a glass container
or unreactive plastic in general of amber color to avoid photo degradation. According to the FDA
guidance, for a general migration study a specimen of known surface area and a known volume
of food simulant are used in the migration cell [46]. The two-sided migration cell is
recommended as shown in Figure 2.3. Although this specific cell may not be universally
applicable, the modified design with two essential features are recommended by FDA: (1) inert
spacers (such as glass beads) are used to separate polymer films or sheets to allow the food
simulant flows freely around each film or sheet; (2) a gas-tight and liquid-tight seals are
maintained where the headspace is minimized. For some application, such as lamination
structure, two-sided migration cell would be replaced by other cell designs as explained by the
FDA.

26

Figure 2.3 Migration cell
2.5.2 Food simulant selection
Migration testing is usually performed using food simulant to mimic real food and to
facilitate detection of the migrant chemical compound. The food simulants recommended by the
FDA are listed in table 2.2.
Table 2.2 Food simulants recommended by the FDA
Food-type

Recommended Simulant

Aqueous & Acidic foods

10% ethanol

Low- and High-alcoholic foods
Fatty foods

10% or 50% ethanol
Food oil (e.g corn oil), HB307,
Miglyol 812, or others

The information about migration cells provided by the FDA serves as recommendation.
Actually, there are many other testing methods provided by organizations such as the European
Commission, MERCOSUR and Japan legislations. For unsaturated food oil (like olive oil, corn
oil), since they are susceptible to oxidation, especially at high temperature, it is difficult to test
them. So, HB307 and Miglyol 812 could be acceptable as a fatty-food simulant since they are
composed of saturated carbons. The FDA provides a list of various polymers and their
27

recommended fatty-food stimulants to effectively help to simulate food oil [46]. The volume of
simulant to specimen surface area ratio should reflect properly the value in actual food packaging,
so a ratio between 155 and 0.31mL/cm2 is generally recommended.
2.5.3 Temperature and time selection for migration testing
Migration protocols recommended by the FDA use short-term accelerated testing to simulate
migration in real situation. For room temperature applications, a test with temperature 40

(104

) for 10 days is recommended. This test method can approximately reflect the equivalent levels
of migration obtained from extended storage (6 -12 months) at 20

(68

refrigerated food application, the test temperature recommended is 20

). For frozen or
(68

). Other

temperature and exposure time could be recommended depending on different situations. For
each migration experiment, the FDA recommends that at least for four time intervals the test
solutions should be analyzed. For example, for the 10 days (240 hours) exposure, the
recommended sampling times are 2, 24, 96 and 240 hours. Also the analysis of a blank or control
is recommended to use.
2.6

Liquid chromatography for characterization and analysis

Liquid chromatography (LC), especially high-performance liquid chromatography (HPLC) is
commonly applied to identify and quantify specific migrants in migration studies. Sample
preparation is the first step for an HPLC analysis. Solvent extraction is a commonly used
technique for antioxidant extraction from polymers. The chemical compounds from a polymeric
film are extracted in a solvent at specific temperature and for a certain time. The extracts can
then be injected into an HPLC system directly for the quantification analysis. The separation of
the studied antioxidant from other chemical compounds in the extract is carried out in the
column which is installed into the HPLC equipment and designed to identify these compounds.
28

Different type of columns are used as stationary phase and different mix of solvents are used as
carried or mobile phases [47]. The elution of the mobile phase in the HPLC system can be either
isocratic or gradient depending on the composition of the sample and the detector used in the
HPLC system. Various detectors are used for an HPLC, such as ultra violet (UV) detector,
fluorescence (FL) detection and mass spectrometry (MS) detection. UV absorption detector can
analyze the chemical compounds based on their varying degrees of absorption in the UV regions.
Compounds with aromatic rings, C=C double bonds or some other functional groups, they all
have a specific positive absorption of UV light at certain wavelength. These wavelengths are
easily measured by UV-HPLC. As for fluorescence detector in HPLC, molecules are excited
with intense UV emission lines from a mercury lamp or wavelengths isolated by filters from a
xenon lamp [47]. This permits an optimal sensitivity for a wide range of analytes. Even though
fluorescence detector can provide several orders of magnitude greater sensitivity for those
compounds which do display fluorescence emission compared with UV absorption detector,
there is only limited subset of analytes that can display fluorescence emission. MS detection is
carried out by using electrospray ionization (ESI) and atmospheric pressure chemical ionization
(APCI), and it provides more reliable identification compared to UV and fluorescence detectors.

29

REFERENCES

30

REFERENCES

1.

Barlow, S., The role of the Scientific Committee for Food in evaluating plastics for
packaging. Food Additives & Contaminants, 1994. 11(2): p. 249-259.

2.

Coles, R., D. McDowell, and M.J. Kirwan, Food packaging technology. 2003: Blackwell
Oxford.

3.

Stilwell, E.J., Packaging for the environment: a partnership for progress. 1991.

4.

Han, J.H., New technologies in food packaging: overview. Innovations in food packaging,
2005. 1.

5.

Rooney, M.L., Overview of active food packaging. 1995: Springer.

6.

elangovan, D., Charaterization of metal organic framework and polymer composites and
the method of their preparation, in School of package. 2010, Michigan state univeristy:
east lansing.

7.

Ching, T.Y., Oxygen scavenging ribbons and articles employing the same. 1998, Google
Patents.

8.

André, C., et al., Analytical strategies to evaluate antioxidants in food: a review. Trends
in Food Science & Technology, 2010. 21(5): p. 229-246.

9.

Min, D. and J. Boff, Chemistry and reaction of singlet oxygen in foods. Comprehensive
Reviews in Food Science and Food Safety, 2002. 1(2): p. 58-72.

10.

Manzanarez-López, F., et al., Release of α-Tocopherol from Poly (lactic acid) films, and
its effect on the oxidative stability of soybean oil. Journal of Food Engineering, 2011.
104(4): p. 508-517.

11.

Mallégol, J., D. Carlsson, and L. Deschenes, Antioxidant effectiveness of vitamin E in
HDPE and tetradecane at 32 C. Polymer Degradation and Stability, 2001. 73(2): p. 269280.

12.

Al‐Malaika, S., H. Ashley, and S. Issenhuth, The antioxidant role of α‐tocopherol in
polymers. I. The nature of transformation products of α‐tocopherol formed during melt
processing of LDPE. Journal of Polymer Science Part A: Polymer Chemistry, 1994.
32(16): p. 3099-3113.

13.

igue -Franco, F., et al., Antioxidant Activity and Diffusion of Catechin and Epicatechin
from Antioxidant Active Films Made of Poly (l-lactic acid). Journal of agricultural and
food chemistry, 2012. 60(26): p. 6515-6523.
31

14.

Soleas, G., et al., A derivatized gas chromatographic-mass spectrometric method for the
analysis of both isomers of resveratrol in juice and wine. American journal of enology
and viticulture, 1995. 46(3): p. 346-352.

15.

Soares, D.G., A.C. Andreazza, and M. Salvador, Sequestering ability of butylated
hydroxytoluene, propyl gallate, resveratrol, and vitamins C and E against ABTS, DPPH,
and hydroxyl free radicals in chemical and biological systems. Journal of agricultural and
food chemistry, 2003. 51(4): p. 1077-1080.

16.

Madhavan Nampoothiri, K., N.R. Nair, and R.P. John, An overview of the recent
developments in polylactide (PLA) research. Bioresource Technology, 2010. 101(22): p.
8493-8501.

17.

Witt, U., et al., Biodegradable polymeric materials—not the origin but the chemical
structure determines biodegradability. Angewandte Chemie International Edition, 1999.
38(10): p. 1438-1442.

18.

Miltz, J., et al., Loss of antioxidants from high-density polyethylene. Food and Packaging
Interactions, ed. by Hotchkiss JH. American Chemical Society, 1988: p. 83-93.

19.

Sharma, G., C. Madhura, and S. Arya, Interaction of plastic films with foods. II, Effect of
polyethylene and polypropylene films on the stability of vegetable oils. Journal of food
science and technology, 1990. 27(6): p. 328-331.

20.

Marcato, B., et al., Migration of antioxidant additives from various polyolefinic plastics
into oleaginous vehicles. International journal of pharmaceutics, 2003. 257(1): p. 217-225.

21.

Ortiz-Vazquez, H., et al., Release of butylated hydroxytoluene (BHT) from Poly (lactic
acid) films. Polymer Testing, 2011. 30(5): p. 463-471.

22.

Soto‐Valdez, H., R. Auras, and E. Peralta, Fabrication of poly (lactic acid) films with
resveratrol and the diffusion of resveratrol into ethanol. Journal of Applied Polymer
Science, 2011. 121(2): p. 970-978.

23.

Siracusa, V., et al., Biodegradable polymers for food packaging: a review. Trends in
Food Science & Technology, 2008. 19(12): p. 634-643.

24.

Javadi, A., et al., Processing and characterization of solid and microcellular
PHBV/PBAT blend and its RWF/nanoclay composites. Composites Part A: Applied
Science and Manufacturing, 2010. 41(8): p. 982-990.

25.

Herrera, R., et al., Characterization and degradation behavior of poly (butylene adipate‐
co‐terephthalate) s. Journal of Polymer Science Part A: Polymer Chemistry, 2002. 40(23):
p. 4141-4157.

32

26.

Witt, U., et al., Biodegradation of aliphatic–aromatic copolyesters: evaluation of the
final biodegradability and ecotoxicological impact of degradation intermediates.
Chemosphere, 2001. 44(2): p. 289-299.

27.

Witt, U., R.-J. Müller, and W.-D. Deckwer, Biodegradation behavior and material
properties of aliphatic/aromatic polyesters of commercial importance. Journal of
environmental polymer degradation, 1997. 5(2): p. 81-89.

28.

Gu, S.-Y., et al., Melt rheology of polylactide/poly (butylene adipate-< i> co</i>terephthalate) blends. Carbohydrate Polymers, 2008. 74(1): p. 79-85.

29.

Javadi, A., et al., Processing and characterization of microcellular PHBV/PBAT blends.
Polymer Engineering & Science, 2010. 50(7): p. 1440-1448.

30.

rubino, M., Permeability ans shelf life, S.o. package, Editor. 2012: east lansing.

31.

Contini, C., et al., Development of active packaging containing natural antioxidants.
Procedia Food Science, 2011. 1: p. 224-228.

32.

Auras, R., B. Harte, and S. Selke, Sorption of ethyl acetate and d‐limonene in poly
(lactide) polymers. Journal of the Science of Food and Agriculture, 2005. 86(4): p. 648656.

33.

Hildebrand, J. and R. Scott, The Solubility of Nonelectrolytes, Reinhold Publ. Corp, New
York, 1950.

34.

Hansen, C.M., Hansen solubility parameters: a user's handbook. 2007: CRC.

35.

Hildebrand, J.H. and R.L. Scott, Regular solutions. 1962: Prentice-Hall Englewood Cliffs,
NJ.

36.

Barton, A., Handbook of Solubility Parameters and other Cohesion ParametersCRC.
Baton Rouge, FL, 1983: p. 82.

37.

Myers, R.R. and J.S. Long, Characterization of coatings: physical techniques. Vol. 2.
1976: Dekker.

38.

Blanks, R.F. and J. Prausnitz, Thermodynamics of polymer solubility in polar and
nonpolar systems. Industrial & Engineering Chemistry Fundamentals, 1964. 3(1): p. 1-8.

39.

Auras, R.A., et al., Poly (lactic acid): synthesis, structures, properties, processing, and
applications. Vol. 10. 2011: Wiley.

40.

Ma, J. and R.M. Larsen, A COMPARATIVE STUDY ON DISPERSION AND
INTERFACIAL PROPERTIES OF SWNTS/POLYMER COMPOSITES USING HANSEN
SOLUBILITY PARAMETERS. ACS applied materials & interfaces, 2013.
33

41.

Van Krevelen, D.W. and K. Te Nijenhuis, Properties of polymers: their correlation with
chemical structure; their numerical estimation and prediction from additive group
contributions. 2009: Elsevier Science.

42.

Garde, J.A., R. Catalá, and R. Gavara, Global and specific migration of antioxidants from
polypropylene films into food simulants. Journal of Food Protection®, 1998. 61(8): p.
1000-1006.

43.

Koszinowski, J. and O. Piringer, Food/package compatibility and migration. Journal of
Plastic Film and Sheeting, 1987. 3(2): p. 96-111.

44.

Crank, J., The mathematics of diffusion. 1979: Oxford university press.

45.

Hernandez, R. and J. Giacin, Factors affecting permeation, sorption and migration
processes in package-product systems. Food storage stability, 1998: p. 269-330.

46.

FDA, U., CFSAN. Guidance for Industry-Preparation of Premarket Notifications for
Food Contact Substances: Chemistry Recommendations.

47.

Robinson, J.W., E.M.S. Frame, and G.M. Frame II, Undergraduate instrumental analysis.
2004: CRC Press.

34

CHAPTER 3 MATERIALS AND METHODS
3.1

Materials

Alpha-tocopherol (α-TOC) (95% HPLC), propyl gallate (PG) (≥98% HPLC), butylated
hydroxytoluene (BHT) (≥99% HPLC), ethanol (200 proof HPLC), and methanol (≥99.9% HPLC)
were all purchased from Sigma-Aldrich, St. Louis, MO, USA. Water (HPLC grade) was obtained
from J.T.Baker, Center Valley, PA, USA. Poly (butylene adipate-co-terephthalate) (PBAT) resin
with a density of 1.25-1.27 g/cc (Ecoflex F Blend C1200) was obtained from BASF chemical
company (TX, USA). Migration cell of 40 mL amber vials with slide valve caps with PTFEsilicon septa was procured from Sigma-Aldrich, St. Louis, MO, USA. Stainless steel wire, glass
beads, and magnetic stirrer were obtained from VWR, Radnor, PA, USA.
3.2
3.2.1

Methods
Hansen solubility parameters (HSP)

In order to determine the HSP an experimental method and theoretical group contribution
methods were used.
3.2.1.1 Experimental method for determining HSP
Table 3.1 shows the HSP parameters of the 17 solvents used to determine the HSP of PBAT.
Equal amounts (2 g) of PBAT were dissolved in 10 ml of solvent for each vial. Vials with PBAT
were sonicated for 15 min before visual observation.

35

Table 3.1 Solvents used for determining the Hansen’s solubility parameters
Solvent
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

Solvent
Dichloromethane
Toluene
Tetrachloroethane
Methyl Ethyl Ketone
Cyclohexane
Diethylene Glycol
Chloroform
Acetone
THF
Acetonitrile
Dimethyl Sulfoxide
Ethyl Acetate
Methanol
Phenol
DMF
Hexane
Ethanol

(mJ/m2)
18.2
18
18.8
16
16.8
16.6
17.8
15.5
16.8
15.3
18.4
15.8
15.1
18
17.4
14.9
15.8

( mJ/m2)
6.3
1.4
5.1
9
0
12
3.1
10.4
5.7
18
16.4
5.3
12.3
5.9
16.7
0
8.8

(mJ/m2)
6.1
2
5.3
5.1
0.2
20.7
5.7
7
8
6.1
10.2
7.2
22.3
14.9
11.3
0
19.4

The basic principle of HSP is that “like dissolves like”, which means materials with
similar HSP have high affinity for each other. In the three dimensional Hansen space, three
parameters can be treated as coordinates to help estimate the affinity between materials. Based
on the basic principle of HSP, the nearer two materials positioned in Hansen space, the more
likely they are to dissolve into each other. The interaction radius (Ro) is introduced to determine
if the parameters of materials (the antioxidant, polymer and food simulant in this study) are
within in range of solubility. This value is the radius of sphere in the Hansen space and the center
of this sphere is determined by HSP. The distance between different HSP in the Hansen space
was calculated by equation 3.1:
3.1

36

The subscripts 1 and 2 in the equation represent the dispersive, polar and hydrogen bond
energy of the two studied materials, respectively.
After calculation of the distance between chemical and/or materials, the relative energy
difference (RED) is also recommended to use, which is easy for the visual estimation. RED is
given by following equation:
RED=

3.2

The general guideline:
RED

: the molecules are alike and will dissolve each other

RED

: the system is right on the soluble/insoluble border

RED

the system will not dissolve

In order to obtain the sphere of solubility, the mix of the 17 solvents used to solve PBAT
were labeled in an arbitrary scale between 1 and 6 based on the degree of solubility of PBAT in
the solvent as shown in Table 3.2.
Table 3.2 Scoring scale for determining the Hansen solubility parameters
Score
1
2
3
4
5
6

comments
Easy to be dissolved
Take a bit of efforts to be dissolved completely
Take a bit of efforts to be dissolved most
Swell a lot, and gel
Swell a little, settle down at bottom
No change, just settle down

3.2.1.2 Hoftyzer and Van-Krevelen theoretical estimation method of HSP
The Hoftyzer and Van-Krevelen method can be used to predict the three solubility parameter
components through the group contribution approach. Equations 3.3 to 3.5 were used to
determine them:
∑

3.3
37

√∑

3.4
∑

√

3.5

where,
is the group contributions to dispersion component of molar attraction constant, J
3/2

1/2

.cm

/mol
is the group contributions to polar component of molar attraction constant, J 1/2.cm 3/2/mol
is the group contributions to hydrogen bonding energy component, J/mol

V is molar volume of calculated material, cm3/mol
Analysis of HSP parameters —The experiment method and the Hoftyzer and VanKrevelen method were conducted to get the HSP values of PBAT. All the HSP values for the
antioxidants, polymers and solvents were obtained through the Hoftyzer and Van-Krevelen
method. Selected compounds were also obtained from literature. Base on the basic principle of
HSP- “like dissolves like”, the distance between the two different materials can estimate the
affinity between them. In order to clearly display the distance between them, a 3D plot of three
HSP parameters was created. These three parameters can be treated as coordinates of the radius
of a sphere in a 3D space. Plots were created by using MATLAB R2011b (MathWorks, Natick,
MA, USA).
3.2.2 Production of PBAT- α-tocophorol (PBA) and PBAT-Propyl gallate (PBP) Film
Figure 3.1 shows a sketch of the entire production process of the antioxidant films. In order to
obtain 2% w/w PBA films and 2% w/w PBP films and PBAT control (PBC) film, 20% α-TOC of
PBAT master batch (PBAM), 20% PG of PBAT master batch (PBPM) and PBAT without
antioxidants as a control (PBCM) were first prepared by ZSK 30 Twin-screw extruder (Werner
and Pfleiderer, NJ, USA). α-TOC (200 g) was mixed directly with 800 g PBAT resin then
38

introduced into the hopper. To get 20% PBPM, PBAT resin from a feeder machine Schenck
AccuRate (Whitewater, WI, USA) was introduced at 90 g/min and mixed with PG in the hopper
which was came from another feeder machine Schenck AccuRate (Whitewater, WI, USA) at
22.2 g/min of feed rate (Figure 3.2). Extrusion temperatures for the master batch from zone 1
through zone 6 of the extruder were 110-125-145-150-160-155 , respectively. A screw speed of
123 rpm was used. The extruded filament was immediately cooled through a water bath, Then, a
pelletizer from Scheer Bay Co. (MI, USA) was used to ground the filaments.

Figure 3.1 Schematic process for the production of antioxidants film

39

Figure 3.2 Master batch extrusion process: Left picture shows the feeder machine for PBAT
resin and AOx. Right picture shows the water bath used to cool extruded filament
After obtaining the PBAM, PBPM and PBAC, a specific amount (presented in table 3.3)
of pure PBAT resin was mixed with 20% PBAM, 20% PBPM and PBCM and were separately
extruded in a Killion KLB 100 blown film extruder (Davis-Standard LLC, Pawcatuck, CT),
respectively (Figure 3.3). The temperature profile of all film extrusion process was 177-182-182177-177-162-154

for barrel zone 1, 2, 3, clamp ring, adapter, die 1 and 2, respectively [1]. A

screw speed of 10 rpm and take up speed of 0.018m/s were used. During the production of the
master batch and films, the heads and tails of each batch were discarded to obtain a
homogeneous distribution of the AOx. The overall thickness of the produced PBC, PBA and
PBP films for the treatments were 63.5 ± 5.1
76.2 ± 5.1

(2.5±0.2 mil), 68.6 ± 2.5

(2.7±0.1 mil) and

(3.0±0.2 mil), respectively, which were measured by micrometer model 549M,

Testing machine Inc. (N.Y., USA).

40

Table 3.3 Weight of master batch and PBAT resin used for producing the PBA, PBP and PBC
film
2% PBA
2% PBP
2% PBC

20% AOx –PBAT master bacth (g)
80
80
80

PBAT resin (g)
720
720
720

Figure 3.3 PBC, PBA and PBP blow film extrusion process: Left figure presents hopper, screw
and blow film sections; Right figure presents reception of processed film.
3.2.3 Extraction and quantification of selected antioxidants for processed film
α-TOC and PG were extracted from the processed film (PBA and PBP) by stirring pieces of
polymers (0.1 g) with 20 mL of methanol at 40

for 24 h in the dark. To protect the

antioxidants from degradation during the extraction period, 100 µg/mL BHT was added to the
solutions. Several extraction times were performed on the same material to ensure complete
extraction of antioxidants. Three replicates were carried out. The amount of α-Tocopherol from
the extruded films was quantified by using a high performance liquid chromatography (HPLC)
(Waters model 2695, Milford, MA, USA) equipped with an auto sampler and a dual absorbance
detector (Waters 2487) at 295 nm wavelength. Aliquots (10 µL) of the extraction were injected
41

into a C18 Nova-Pak column (3.9*150 mm; Waters) protected with a C18 guard column with run
time of 12 min per injection. An isocratic elution of methanol - water (98:02) at 0.8 mL/min and
25

- was used. Calibration curves from 0.2 to 10 µg/mL of α-TOC in methanol were carried

out for qualification,

was 0.99. Retention time for α-TOC was around 5.5 min and the limit of

quantification (LOQ) was 0.1µg/mL.
The amount of PG from the extruded films was quantified by using a 3200
QTRAP® liquid chromatography-tandem mass spectrometry (LC/MS/MS) System (AB Sciex,
Framingham, MA, USA) and a Shimadzu LC-20AD HPLC system and SIL HTc auto-sampler
(Shimadzu Scientific Instruments, Addison, IL, USA). The instrument was operated by the
Analyst® software. Aliquots (10 µL) of the extraction were injected into an Acenti® Express
C18 column (2.1×50 mm; 2.7 µm particles) (Sigma-Aldrich, St. Louis, MO, USA) protected
with a C18 guard column with run time of 7 min per injection. The temperature of the column
was maintained at 40

via CTO-10A vp column oven. A gradient of 0.15% formic acid/ water

(mobile phase A) and methanol (mobile phase B) at flow rate of 0.4 mL/min was set as follows:
95% A and 5% B was initial set and maintained for 0.5 min, then the ratio of 5% A and 95% B
was reached in the following 2 min and kept this ratio from 2.5 to 5 min. Then a return to the
initial settings (95% A and 5% B) was applied at 5.01 min and maintained at this proportion
from 5.01 to 7 min. The samples were analyzed by electrospray ionization in negative mode
(ESI-) and multiple reactions monitoring (MRM) (detail parameters presented in Table 3.4) by
using a hybrid triple quadrupole linear ion trap (“Q-Trap”). on source settings for Gas 1
(nebulizer gas) and Gas 2(heater gas) were 10 and 10, respectively. Source temperature was
550 . Ionspray voltage was set to -4500 V, curtain gas was set to 20 psi, and collision CAD gas
was set to High mode. The results were analyzed by Analyst® software.

42

Table 3.4 MS parameters for the analysis of PG
Q1
(amu)
211.1

PG

Q3
(amu)
169

Dwell
(s)
0.1

DP volt.
(V)
-60

Collision volt.
(V)
-25

Calibration curves from 0.1 to 10 µg/mL of PG in methanol were carried out for
quantification,

was 0.96. Retention time for PG was 4.9 min and the limit of quantification

(LOQ) was 0.01µg/mL.
3.2.4 Diffusion of antioxidants from processed film into food simulant.
The amount of antioxidants diffused from the process film into 95% ethanol was determined by
using a migration cell according to ASTM D4754-11. The migration cell consisted of a 40 mL
amber glass vial with a screw cap and silicon septum. Six round disc (2 cm in diameter) cut from
the film were inserted in a stainless steel wire and placed in the amber vial containing 35 mL of
95% ethanol. Glass beads were used to separate each disc in order to expose both sides to liquid,
which was displaced in Figure 3.4. A magnetic stirrer was used inside each vial to provide a
constant agitation. The liquid volume/area ratio was 0.9 mL/cm2 complying with ASTM D475411, which establishes a ratio between 155 and 0.31 mL/cm2.

Figure 3.4 Migration test cell

43

3.2.4.1 Preliminary migration study
Preliminary migration studies were performed for processed PBA and PBP films into 95%
ethanol. Two trials of PBA migration test at 40
and 30

and one trial of PBA migration test at 10, 20

respectively were done to get preliminary migration data to design the full migration

experiment. Additionally, one trial of PBP migration test at 20
temperature chamber was used and controlled within ±0.5

as conducted. Constant

. A designed sample collection

schedule was applied for each trial until equilibrium was reached. At each sample collection, 50
µL solution from the migration cell was diluted with 950 µL 95%ethanol. Then, the dilution
solution was analyzed by HPLC for quantification. Antioxidants qualifications were performed
as described in the extraction section. To establish the equilibrium stage for each temperature,
the concentration (mg/L) versus time (s) of each vial at each temperature was plotted until the
slope of the curve reached zero. Using the preliminary data, the scaled sensitivity coefficient
(SSC) was obtained and an optimal experimental design (OED) was conducted to determine the
sampling time interval for the final migration test.
3.2.4.2 Designed of the final migration test
According to the information obtained in the preliminary migration study, the final migration
test for PBA and PBP were conducted at 10

, 20

and 30

Four cell replicates

were used per each temperature. Samples were taken out periodically from each cell during 4
days for 10 , 2 days for 20

and 2 days for 30 . At each sampling time, 50 µL solution from

migration cell was dilute with 950 µL 95%ethanol. Then the dilution solution was injected to
HPLC for quantification. Antioxidants quantifications were performed as the same as described
in extraction section.

44

3.2.5 Antioxidants release models
Migration processes for test conducted in migration cells are mostly described by
standard Fick’s law equations. Several types of boundary conditions are used to categorize
solutions of the heat and mass (diffusion) equation. Five types of boundary conditions are
defined at physical boundaries and a “zeroth” type defines the case with no physical boundaries.
Table 3.5 shows the types of boundary conditions from numbers zero through five.
Table 3.5 Types of boundary conditions and associated number designation, adapted from
(http://exact.unl.edu/exact/home/number.php)
Name of boundary condition
No Physical boundary

Equation

Number

T is bounded

0

|

Dirichlet

1

Neumann

2

Robin

3

|

Thin, high-conductivity film

[

]

Thin, high-conductivity film

4

5

with addition of convection

Type 1 is called the Dirichlet condition or referred to as a fixed boundary condition. It
corresponds to the situation for which the surface is maintained at a fixed temperature in heat
transfer. Type 2 is called the Neumann condition, which describes the existence of a fixed or
constant heat flux at the surface. Type 3 is called the Robin condition, sometimes called
convective boundary conditions. It corresponds to the existence of convection heating (cooling)
45

at surface. Type 4 represents a thin, high-conductivity film at the body surface. Type 5 represents
a thin, high-conductivity film at the body surface with the addition of convection heat losses
from the surface. Type 0 is used to identify the case with no physical boundary exists.
In this study, the migration process occurred between well-mixed food simulant and
packaging material with well-distributed AOx. The mass transfer resistance on the side of food
can be negligible. Before migration process happen, the concentration of AOx in the packaging
material equals to the initial concentration and the concentration of AOx in food simulant is 0.
When migration occurs, the amount of AOx migrated from packaging material is the same
amount existing in food simulant. Then, boundary conditions type 3 or 4 mentioned above can be
used to solve the studied migration. According to Fick’s second law equation for D, two
analytical solutions were applicable for this experiment conducted in one-dimension, and they
are expressed by equation 3.6 and equation 3.9 [2, 3]. When the diffusion is in one-dimension
and in a limited volume of film is in contact with a finite volume of solution, equation 3.6 is used.
∑
where

3.6

is the concentration of the antioxidant diffused at time t divided by the

concentration of the antioxidant diffused at equilibrium; l is the thickness of the film and the
are the non-zero positive roots of tan

and α is expressed in 3.7:
3.7

where

and

are the molar volume of the simulant and the polymer.

is the partition

coefficient of antioxidant between PBAT and the simulant, which can be calculated from the
ratio of the concentration of the antioxidant in the PBAT film (
equilibrium according to following equation:
46

) and the simulant (

) at

3.8
When the amount of simulant can be considered infinite (i.e.,
/or

since

and

), equation 3.9 is used
∑

3.9

Diffusion coefficients (D, cm2/s) were calculated for each antioxidant at different
experimental temperatures. The value of D was determined by minimizing the sum of the
squares errors (SSE) of the measured and estimated values. To determine the fit of the
experimental data to equation 3.6 and/or 3.9, a nonlinear regression (nlinfit) function in
MATLAB R2011b ( MathWorks, Natick, MA, USA) was applied to the data [4].
3.2.6 Activation energy for the diffusion (Ea) of selected antioxidants into food simulant
To determine the effect of temperature on the diffusion of α-Tocopherol and propyl gallate from
PBAT films into 95% ethanol, the activation energy (

was calculated using the Arrhenius

equation:
3.10
where D is the diffusion coefficient,

is the pre-exponential factor of diffusion,

is the

activation energy of diffusion, R is the ideal gas constant (8.3145 J/Kmol), and T is the
temperature in K.

was obtained from the slope of a plot of the reciprocal of temperature (1/T)

vs the logarithm of D (

= - slope×2.303R) [5].

3.2.7 Optical properties
The transmission to UV-visible light through the films was measured with a PerkinElmer Lambda 25, UV-visible spectrophotometer (Perkin-Elmer Instruments, Beaconsfield, UK)
between 190 and 800 nm at a rate of 480 nm/min, a slit width of 1.0 nm and an interval of 1.0
47

nm. An integrating reflectance spectroscopy accessory (model RSA-E-20, Labsphere®, North
Sulton, NH) was also employed.
Films color was determined by a LabScan XE (HunterLab, Reston, VA, USA). The
Hunter L*, a*, b* values were obtained by analysis software Easymatch QC version 3.8. Three
replicates for each type of film were measured. Total difference in color

was calculated as

expressed in equation 3.12:
√

3.12

where

,

3.2.8 Statistic analysis
One-way analysis of variance (ANOVA) was used to analyze the data. Comparisons of
experimental data were conducted by using Tukey‘s HSD test with 95% level of confidence
( =0.05). Statistical software SAS (Version 9.4, SAS Institute Inc., Cary, NC, US) was used

48

REFERENCES

49

REFERENCES

1.

Chidambarakumar, M., et al., Biodegradable polymeric nanocomposite compositions
particularly for packaging. 2009, Google Patents.

2.

Crank, J., The mathematics of diffusion. 1979: Oxford university press.

3.

Manzanarez-López, F., et al., Release of α-Tocopherol from Poly (lactic acid) films, and
its effect on the oxidative stability of soybean oil. Journal of Food Engineering, 2011.
104(4): p. 508-517.

4.

Dhoot, G., et al., Determination of eugenol diffusion through LLDPE using FTIR-ATR
flow cell and HPLC techniques. Polymer, 2009. 50(6): p. 1470-1482.

5.

Limm, W. and H.C. Hollifield, Modelling of additive diffusion in polyolefins. Food
Additives & Contaminants, 1996. 13(8): p. 949-967.

50

CHAPTER 4 RESULTS AND DISCUSSIONS
4.1

Hansen Solubility parameters determined by the experimental method
The Hansen solubility parameters (HSPs) of PBAT were determined by an experimental

method as described in section 3.2. The scores for the 17 solvents are listed in table 4.1.
Table 4.1 Scoring of PBAT in various solvents after two days of storage
Solvent Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

IUPAC Name (CAS Number)
Dichloromethane (75-09-2)
Methylbenzene (108-88-3)
1,1,2,2-Tetrachlorethane (79-34-5)
Butan-2-one (78-93-3)
Cyclohexane (110-82-7)
(2-hydroxyethoxy) ethan-2-ol(11146-6)
Chloroform (67-66-3)
Propanone (67-64-1)
Oxacyclopentane (105-99-9)
Acetonitrile (75-05-8)
Dimethyl Sulfoxide (67-68-5)
Ethyl Acetate (141-78-6)
Methanol (67-56-1)
Hydroxybenzene (108-95-2)
N,N-dimethylmethanamide (68-122)
Hexane (110-54-3)
Ethanol (64-17-5)

Solvent
Dichloromethane
Toluene
Tetrachloroethane
Methyl Ethyl Ketone
Cyclohexane
Diethylene Glycol

Score
1
3
2
3
6
6

Chloroform
Acetone
THF
Acetonitrile
Dimethyl Sulfoxide
Ethyl Acetate
Methanol
Phenol
DMF

1
4
1
4
6
5
5
2
3

Hexane
Ethanol

6
6

Figure 4.1 shows that PBAT was completely dissolved in dichloro methane, chloroform
and THF after 15 min of sonication, but not in tetrachloro ethane and phenol solvents, which
took two days of storage for PBAT to dissolve. According to these observations, a score of 1 was
given to dichloro methane, chloroform and THF, and a score 2 was given to the tetra
chloroethane and phenol solvents.

51

Figure 4.1 Dissolution of PBAT in dichloromethane, chloroform, THF, tetrachloroethane and
phenol (a- left to right: 1-dichloromethane, 7- chloroform, and 9-THF, b-PBAT in 3tetrachloroethane and 14-phenol after 15 min of sonication, c-PBAT in 3-tetrachloroethane and
14-phenol after two days of storage) (For the better visualization of this dissolution phenomenon,
the reader is suggested to view the electronic version of this document)
The dissolution behavior of PBAT pellets in other solvents such as toluene, methyl ethyl
ketone, DMF, acetone, acetonitrile, ethyl acetate and methanol did not vary very much after 15
min of sonication and 2 days of storage. PBAT was not dissolved completely in these solvents,
instead it settled down to the bottom, or swelled with different levels when settled. Also, it gelled
at some point. PBAT was swelled most in toluene, methyl ethyl ketone and DMF, and form gell.
Then, a score of 3 was given for these three solvents. Swelling of PBAT in acetone, acetonitrile,
ethyl acetate and methanol was less than that in the former solvents. Meanwhile PBAT in
cyclohexane, diethylene glycol, dimethyl sulfoxide, hexane and ethanol did not change much
comparing to the previous mentioned solvent. The scores were graded based on the interaction
between PBAT and the solvents. Figure 4.2 and 4.3 indicate how PBAT was dispersed in various
solvents after 15 min of sonication and two days of storage, respectively. PBAT pellet was used
in the experimental method to measure HSPs instead of PBAT film, since HSPs of pellets and
films are assumed the same since the geometry and chemical structures are not changing. So,
based on the group contribution calculation methods, their HSPs should be the same.

52

Figure 4.2 Dissolution of PBAT in various solvents after 15 min sonication (from left to right,
named number 1-17: 1-dichloro methane, 2-toluene, 3-tetra chloro ethane, 4-methyl ethyl ketone,
5-cyclohexane, 6-diethylene glycol, 7-chloroform, 8-acetone, 9-THF, 10-acetonitrile, 11dimethyl sulfoxide, 12-ethyl acetate, 13-methanol, 14-phenol, 15-DMF, 16-hexane, 17-ethanol )

Figure 4.3 Dissolution of PBAT in various solvents after two days storage (from left to right,
named number 1-17: 1-dichloro methane, 2-toluene, 3-tetra chloro ethane, 4-methyl ethyl ketone,
5-cyclohexane, 6-diethylene glycol, 7-chloroform, 8-acetone, 9-THF, 10-acetonitrile, 11dimethyl sulfoxide, 12-ethyl acetate, 13-methanol, 14-phenol, 15-DMF, 16-hexane, 17-ethanol)
The HSPiP® software (Horsholm, Denmark) was used to estimate the HSPs values of
PBAT based on the data from Table 4.1 obtained by dissolving PBAT in the 17 mentioned
solvents. Table 3.1 provided in chapter 3 lists the unique dispersion, polar and hydrogen forces

53

values for each solvent. These

data were input into the HSPiP® software to obtain the

dispersion, polar, hydrogen forces and interaction radius of PBAT as shown in Table 4.2. The
HSPiP® software used the scores define for each solvent to fit a sphere of PBAT in the HSP
space, which allows good solvents to be located inside the sphere and bad solvents outside [1, 2].
The

calculation

indicates

that

the

HSP

of

PBAT

are

18.97,

9.83

and

9.10

and the radius of dissolution is 3.9, which means that a wide rage
of solvents located in or on this sphere will dissolve, interact or at least swell PBAT. The
solvents close to the center of the sphere are more likely to totally dissolve the polymer.
Table 4.2 Dispersion, polar, hydrogen parameters and interaction radius for PBAT
Material

Dispersion
(MPa 1/2)

PBAT

18.97±0.3

4.2

Polar
(MPa 1/2)
4.83±0.75

Hydrogen Force
(MPa 1/2)

Interation Radius
(

9.10±0.3

3.9

Hoftyzer and Van-Krevelen theoretical calculation method
The HSP parameters of two biodegradable polymers and twenty-five antioxidants were

predicted according to the Hoftyzer and Van-Krevelen methods. Butylated hydroxytoluene (BHT)
is next used as an example to explain how to use the Hoftyzer and Van-Krevelen methods was
used to calculate the HSP. Figure 4.4 shows the chemical structure for BHT.

Figure 4.4 BHT chemical structures
The BHT chemical structure can be broken down in four parts, which are seven –CH3,
one –OH, two –C and one benzene ring. Based on Table 4.3 [3], the group contributions values
of

can be found for each structural group. The sum of

,

and

can be

calculated based on the types and number of structural groups present on BHT. Table 4.4
54

summarizes the values of

for BHT. Table 4.4 shows the detailed calculation for

generating the HSPs of BHT. The molar volume for BHT is 244.3 cm3/mol [4].
Table 4.3 Solubility parameter components according to the group contribution method of
Hoftyzer-Van Krevele.
Structural group

ring

420

0

Ehi (J/mol)
0

270

0

0

80

0

0

-70

0

0

400
200
70

0
0
0

0
0
0

1620

0

0

1430

110

0

1270

110

0

(220)
450
(550)
430
210
100
470
290
530
390
530
280
160
20

550
1100
500
400
800
770
420
490
210
800

400
2500
20000
3000
4500
2000
10000
7000
8400
3100
5000

500
440
740
190

1070
1890
-

1500
13000
-

Fdi (

Fpi (

55

Table 4.3 (cont’d)
Structural group
One plane of
symmetry
Two plane of
symmery
More plane of
symmery

-

0.5×

Ehi (J/mol)
-

-

0.25×

-

-

0

0×

Fdi (

Fpi (

Table 4.4 Addition of the group contributions for BHT
The group
contributions

Numbers

-CH3
-OH
-CBenzene
Total

7
1
2
1

(J 1/2.cm3/2/mol)

(J 1/2.cm3/2/mol)

2940
210
-140
1270
4280

0
250000
0
12100
262100

(J /mol)
0
20000
0*2
0
20000

According to equations (3.3) to (3.5) in chapter 3 and table 4.4:
∑

J 1/2/cm 3/2 =17.52 MPa 1/2

√∑

∑

√

√

√

J 1/2/cm 3/2 =2.10 MPa 1/2
J 1/2/cm 3/2 =0.58 MPa 1/2

The three solubility parameter components (

for BHT are 17.52 MPa

1/2

, 2.10

MPa 1/2 and 0.58 MPa ½, respectively.
Table 4.5 lists the HSP values for 25 antioxidants, two biodegradable polymers and the
solvents studied.

56

Table 4.5 Summary of HSPs for the polymers, solvents and antioxidants studied
Category
Polymer

Material
PLA
PBAT
Solvent
Ethanol
Water
Antioxidants BHT
Ascorbic acid
Catechin
Quercetin
Octyl gallate
Dodecyl gallate
TBHQ
Ralox 35
Irganox 1135
Irganox 1076
Nordihydroguaiaretic acid
Epicatechin gallate
Epigallocatechin gallate
Epigallocatehin
Propyl gallate
2,3,5-trihydroxybutyrophenone
Butylated hydroxyanisole
4-Hydroxymethyl-2,6-di-tertbutylphenol
α-tocopherol
Resveratrol
Rutin
Astaxathin
Lycopene
β-carotene
Melatonin
4.3

(MPa 1/2)
18.9
18.5
15.8
15.5
17.5
10.2
13.3
23.0
18.2
17.8
51.5
17.1
18.5
18.6
18.63
15.1
14.3
13.7
19.2
19.6
17.3
15.7

(MPa 1/2)
0.73
4.9
8.8
16
2.1
7.5
4.5
15.8
6.9
5.4
10.5
3.0
2.6
1.9
0.13
4.3
3.6
4.7
7.3
7.2
2.8
2.9

(MPa 1/2)
10.3
7.8
19.4
42.3
0.58
22.0
19.7
25.0
1.0
0.79
1.9
9.3
8.1
6.9
1.07
8.5
20.4
21.1
19.6
19.6
11.2
9.1

15.01
21.26
19.3
16.3
15.6
15.47
18.6

1.41
9.03
16.1
3.3
0.0
0.0
0.2

7.12
10.91
25.4
8.9
0.0
0.0
5.8

Analysis of HSP parameters

The HSP values of PBAT were obtained by the experimental method and the Hoftyzer and VanKrevelen calculation procedure. Table 4.6 presents the data from these two methods. The values
for the three components of the HSP were found to be very similar to each other according to
these two different methods of calculation. In this study, the data from the experimental method

57

was then used. Large variation in

is normal due to the unpredictable nature of the chance of

forming hydrogen bonds.
Table 4.6 HSP for PBAT from two methods
PBAT
Experiment method
Calculation method

(MPa 1/2)
18.47±0.3
18.55

(MPa 1/2)
4.83±0.75
4.90

(MPa 1/2)
9.10±0.3
7.82

Base on the basic principle of HSP- “like dissolves like”, the distance between two
different materials can be estimated by the affinity between them. The greater the distance
between the two materials, the less alike they are. Figure 4.5 shows a 3D plot of all HSPs of the
antioxidants, polymers and solvents and their relative positions in the Hansen space.
According to the information provided by the 3D plot, there are some antioxidants
located close to the sphere of interaction of ethanol, while others are presenting closer to the
sphere of PBAT and PLA. In this study, PBAT was selected as the main studied polymeric
material. So, a zoom out of the 3D plot is displayed in Figure 4.6. Only the sphere of water,
ethanol, PBAT and the antioxidants around or located in the sphere of ethanol and/or PBAT are
plotted. Four antioxidants, named in the graph propyl gallate (PG) and ascorbic acid (AA)
around the sphere of ethanol, and Irganox 1076 (IR1076) and α-tocopherol (α-TOC) around the
sphere of PBAT are individually marked. Based on the HSP methodology and the graph, PG and
AA have higher affinity to ethanol comparing to R 1076 and α-TOC. R 1076 and α-TOC have
higher affinity to PBAT than PG and AA. So, one antioxidant close to PBAT (i.e., inside the
interaction regions of polymer) - α-TOC and one antioxidant close to ethanol (i.e., inside the
interaction regions of the solvent) -PG were selected and later on incorporated into PBAT.
Migration test in 95% ethanol were conducted for these antioxidants to verify the initial
estimation that PG incorporated in PBAT will have higher affinity towards ethanol than α-TOC.

58

Figure 4.5 3D plot of the HSP parameters for the 25 antioxidants, 2 biodegradable polymers and
water and ethanol. The spheres show the radius of interaction of the solvents and the polymers.
To further justify the selection of one antioxidant among PG, AA, IR 1076 and α-TOC,
their information is listed in Table 4.7. Between PG and AA, which are close to the sphere of
ethanol, the antioxidant capacity of PG is greater than that of AA according to DPPH
measurement of the antioxidant capacities [5]. The reference used the IC

50

to indicate the

antioxidant capacity. The IC50 values for AA and PG were 11.8 µM and 4.4 µM in methanol and
11.5 µM and 4.7 µM in buffered methanol as reaction medium, respectively. While in the case of
IR 1076 and α-TOC, which is close to sphere of interaction of PBAT, the antioxidants capacity of
α-TOC is greater than that of R 1076 [6]. Samples containing α-TOC have higher level of

59

retention of parent antioxidant than that for R 1076 through HPLC analysis of dichloromethane
extracts of polymer samples. So, PG and α-TOC were selected and procured for this study.

Figure 4.6 3D plot of selected HSP parameters
Table 4.7 Information summary for selected antioxidants
Molar
Density(g/cm3)
Melting
3
volume(cm /mol)
point
PG1
169.7
1.25
214
150
2
AA
176.1
1.65
220 [7]
190-192
3
IR 1076
453.38
1.01
230 [8]
50-55
α-TOC4
528
0.95
241
2.5-3.5
Note: * = thermal decomposition point;
1 the information was provided by Sigma Aldrich where PG bought and referred to Wikipedia.
2 the information was referred to Wikipedia.
3 the information was referred to Wikipedia.
4 the information was provided by Sigma Aldrich where PG bought and referred to Wikipedia.
Antioxidants

60

4.4

Quantification of α-TOC and PG after film processing
The final concentrations of α-TOC and PG in their master batch added with 20% wt/wt of

the antioxidants were 17.3 ± 0.4% and 17.8 ± 0.0%, respectively. The loss of α-TOC and PG
during the processing of PBAM and PBPM were 13.5 ± 0.4% and 11.0 ± 0.0% respectively. The
final concentrations of α-TOC and PG in the produced cast film were 1.5 ± 0.0 % and 1.8 ±
0.05 %, respectively. The loss of antioxidant during processing was expected and attributed to
several aspects. Since α-TOC is a stickier liquid, some α-TOC in a prepared mixture of α-TOC
with PBAT was obviously left in the container during the master batch process. While PG is
obtained as powder, so it is better handled when introduced to the extruder. During the master
batch production and film processing, the material was poured into the feeder of the master batch
extruder at 110

and at 177

in the feeder of blow film extruder. For both processes, the

material took 5 to 10 min to come out from the feeder to the die, so a large friction was applied
between the barrel and the screw. Additionally, in the stage of pelletization and film production
process, it is normal cause that the AOxs present in matrix to degrade [9, 10]. Adherence of
antioxidants to the screw and the barrel could also contribute to further loss of the antioxidants.
Manzanare et al. also reported similar loss of α-TOC for production of poly (lactic acid) (PLLA)
compounded with α-TOC [11].
Table 4.8 Concentrations of antioxidants in the film after film processed
Films
PBP
PBA

Antioxidant (w %)
Before extrusion
After extrusion
1.78
1.8±0.05
1.73
1.5±0.01

61

4.5

Migration of α-TOC and PG into food simulant (95% ethanol)

4.5.1 Preliminary migration study
Initial migration studies were conducted to determine the sampling interval needed to
conduct the migration test. Sampling interval time (every 30 min in the first 3 h) was used for
one trial of PBA migration test at 40 . The equilibrium (steady state) of migration of α-TOC
was reached at 30 min of the study indicating a very fast diffusion of α-TOC from PBAT. There
was not obvious change of peak area of the analyzed solution detected by HPLC from the first
sampling time until 24 h. An adjusted sample collection schedule every 5 min for 35 min at the
beginning of the migration test was applied. The data for this adjusted test confirmed that
migration of PBA in 95% ethanol reached equilibrium after 30 min. PBA migration into 95%
ethanol at 40

was too fast to determine the unsteady state. One trial of PBA migration at 30, 20,

and 10 , respectively and one trial of PBP at 20

were done at designed sampling time. The

data from all these trials were analyzed by determining the scaled sensitivity coefficient (SSC),
by MATLAB R2011b (MathWorks, Natick, MA, USA) provided by Samsudin et al [12]. SSC
are needed to determine whether the migration parameters can be estimated with relative
smallest error. To scale of the sensitivity coefficient, it is obtained by multiplying the sensitivity
coefficient with its respective parameters (Equation 4.1). SSC plots was used to determine the
correlation among the main parameters studied in this research [13]. The red line in Figure 4.7
stands for

and the blue line stands for the D. In Figure 4.7 a to d can easily be seen that the

behavior of the red line and blue line are different, and that these two parameters are not
correlated and can be determined simultaneously for the same experiment. The SSC plot also
provides important timing information to design the full migration experiment. Taking PBA
migration trial at 10

as an example, the blue line is increasing to reach a maximum peak from

62

time 0 to 1.4 h. So during this period more sampling points should be collected to guarantee a
proper estimation of D. After this time period, less frequent sampling points can be taken. Based
on the plot, a timing table was created for conducting the final migration test. Table 4.9 shows an
example for PBA final migration test at 10 . Then the final migration test for PBA and PBP
were conducted according to the timing table created based on SSC.
4.1

Figure 4.7 Scaled sensitivity coefficient plot for PBA migration trial at 30, 20, and 10 and
PBP migration trial at 20 : (a) PBA migration trial at 10 , (b) PBA migration trial at 20 , (c)
PBA migration trial at 30 , (d) PBP migration trial at 20

63

Table 4.9 Timing table for PBA final migration test at 10
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

Time (h)
0.25

1

2
3
4
6
8
12
27
41

Vial 1
Time (min)
15
30
45
60
75
90
105
120
180
240
360
480
720
1620
2460

Vial 2
Time (min)
17.5
32.5
47.5
62.5
77.5
92.5
107.5
122.5
182.5
242.5
362.5
482.5
722.5
1622.5
2462.5

Vial 3
Time (min)
20
35
50
65
80
95
110
125
185
245
365
485
725
1625
2465

Vial 4
Time (min)
22.5
37.5
52.5
67.5
82.5
97.5
112.5
127.5
187.5
247.5
367.5
487.5
727.5
1627.5
2467.5

4.5.2 Final migration study
In the experimental migration of α-TOC and PG from PBAT films to 95% ethanol at 10,
20, and 30 , the diffusion of both antioxidants showed a typical Fick’s behavior. Equilibrium at
10, 20, and 30

was obtained for both PBA film and PBP film at 1.75 h, 1 h and 0.75 h,

respectively. The experiments were stopped at 2 d for 10 , 1 d for 20

and 1 d for 30 , as

designed according to the preliminary study. At the end of the experiment, the total release of αTOC was 99.63, 99.63 and 99.62% at 10, 20, and 30 , respectively while the release of PG was
99.26, 99.46 and 99.21% at 10 , 20, and 30 . At each temperature, almost a total migration of
α-TOC and PG was observed (˃99%) at a fast rate.

at the three temperatures for PBA and

PBP were determined and shown in table 4.10. The

value was calculated from the ratio of

the concentration of the antioxidant left in the film (

) and the concentration of antioxidants

in the simulant (

) at equilibrium.

˃1 indicates a higher concentration of migrant in the

polymer compared to that in the solvent at steady state [14]. In this study,
64

for α-TOC at the

three temperatures were all less than 1 and not significantly different was found among them.
Values of
The low

for PG were slightly above 1, there was no significant difference between them.
values at each temperature for both α-TOC and PG agreed with the release of more

than 99% α-TOC and PG into 95% ethanol. Also the low

values indicated that α-TOC and

PG migration systems can be considered and solved with the Fickian’s equation for an infinite
simulant volume [15] [16], as expressed in equation 3.9 in Chapter 3.
Figure 4.9 and Figure 4.10 display the diffusion curves according to the Fick’s second
law at 10 , 20 , and 30

in 95% ethanol and their respective residuals plots. The residual plot

is used to check whether the assumptions made in the regression analysis are accurate. Based on
the Figures 4.9 and 4.10, the residuals plots do not exhibit characteristic signatures and/or
patterns, which means that the used regression model is accurate enough.
Table 4.10
Temperature
( )
10
20
30

and

values of α-TOC and PG for the diffusion from PBAT films into 95%
ethanol at 10 , 20 , and 30 *

α-TOC
0.66±0.02 a
0.73±0.12 a
0.73±0.07 a

PG
1.18±0.11 b
1.03±0.47 b
1.12±0.12 b

α-TOC
411.10±14.48
371.30±54.33
388.45±56.85

PG
225.02±20.62
212.63±59.64
229.82±26.94

*

The data are the averages of four samples ± the standard deviation. Different letters within the
same columns indicate statistically significant different values (p<0.05)
Table 4.11 D and RMSE of α-TOC and PG for the diffusion from PBAT films into 95% ethanol
at three different temperature *
Temperature
( )
10
20
30

D ×10-9 (cm2/s)
α-TOC
2.05±0.37 a, A
3.25±0.18 a, B
7.94±0.71 a, C

PG
1.86±0.13 a, A
4.42±0.61 b, B
8.59±0.12 a, C

*

RMSE ×10-6 *
(g antiox. /g 95%ethanol)
α-TOC
PG
3.54
6.59
5.01
0.11
9.36
0.14

The data are the averages of four samples ± the standard deviation. Different lower case letters
within the same row indicates statistically significant different values (p<0.05). Different upper
case letters within the same column indicates statistically significant different values (p<0.05).
RMSE: root mean square error.
65

Table 4.12
Temperature
( )
10
20
30

and RMSE of α-TOC and PG for the diffusion from PBAT films into 95%
ethanol at three different temperature *
×10-4 (g antiox. /g 95%ethanol)
α-TOC
0.71±0.03 a
1.08±0.04 a
1.02±0.06 a

PG
0.86±0.02 b
1.02±0.08 a
0.96±0.07 b

RMSE ×10-6 *
(g antiox. /g 95%ethanol)
α-TOC
PG
3.54
6.59
5.01
0.11
9.36
0.14

*The data are the averages of four samples ± the standard deviation. Different letters within the
same row indicate statistically significant different values (p<0.05). RMSE: root mean square
error.

Figure 4.9 Diffusion of α-TOC into 95% ethanol at 10 , 20 , and 30 according to the
second Fick’s law and their respective residuals plot. The y-axis shows the mass of antioxidants
diffused at time t in h.
66

Figure 4.9 (cont’d)

67

Figure 4.10 Diffusion of PG into 95% ethanol at 10 , 20 , and 30 according to the second
Fick’s law and their respective residuals plot. The y-axis shows the mass of antioxidants diffused
at time t in h.

68

Figure 4.10 (cont’d)

Tables 4.11 and 4.12 show the D and
equation 3.9, since

≈ 1 and

values and RMSE values obtained by using

for both diffusion of α-TOC and PG. At 10 , D for α-

TOC and PG were 2.05×10-9 and 1.86×10-9 cm2/s, respectively. At 20 , D for α-TOC and PG
were 3.25×10-9 and 4.42×10-9 cm2/s, respectively. Meanwhile, D values at 30
order of magnitude than at 10 and 20

has the same

, giving values of 7.94×10-9 cm2/s for α-TOC and

8.59×10-9 cm2/s for PG. D values for α-TOC at the three different temperatures were significantly
difference from each other and values increase with an increas in temperature. Although D
values for α-TOC were different, they were in the same order of magnitude (10-9 cm2/s).
Indicating that temperature made a difference on the diffusion process of antioxidants but its
69

influence was low in this study due to the fast release at all temperatures. The same trend also
appeared to D values for PG. When comparing the D values for α-TOC and PG at each
temperature, only at 20℃, they were significantly difference, being the D of PG larger than the
D of α-TOC. In the HSP three-dimension graph, the location of PG is closer to the sphere of
ethanol, which should indicate that PG has higher affinity to ethanol than α-TOC. However, at 10
and 30 , D for α-TOC and PG were not significantly different (p-value>0.05). So, indicating that
the precision of the HSP may not be sufficient to differentiate the affinity between these
antioxidants and ethanol.
Reported D values for α-TOC from PLA to ethanol at 23, 33 and 43

were 3.16×10-11,

5.29×10-11 and 3.8×10-10 cm2/s, respectively[11]. Heirlings et al [17] reported D values for αTOC from polymers like LDPE and ethylene vinyl acetate (EVA) in 95% ethanol at 7

as

2.64×10-11 and 4.23×10-11cm2/s, respectively. Since there is a little research done on the
migration of PG in polymer systems, it is hard to compare the D values of PG in other polymer
systems. Result from this study shows that D values for α-TOC and PG at each temperature were
of the same magnitude (10-9 cm2/s), which is faster than that in PLA, LDPE and EVA. Figure 4.5
showed that the location of PBAT in the HSP space is closer to ethanol compared with PLA,
which means ethanol has a good solubility to PBAT. So, if we compare ethanol as solvent,
ethanol should be better solvent for PBAT than PLA and has more interaction penetrating into
the PBAT chains and promote the diffusion of the antioxidants. To the best of the author’s
knowledge, extensive research of interaction of solvents with PBAT is not found in the peerreview literature. Based on equation 3.1, the distance
the distance

between α-TOC and PBAT is 8.0 and

between PG and PBAT is 10.8. The distance of these two antioxidants to PBAT

is not quite different. Even though based on Figure 4.6 PG is closer to ethanol, the similar

70

distance to PBAT and the aggressive ethanol solvent could contribute to the similar diffusion
process of α-TOC and PG obtained in this study. Even at 10

the diffusion of α-TOC and PG

were very fast and there was not a significant difference between them. So, it is understandable
that the diffusion difference between α-TOC and PG at 30

is not significantly different since

the values are so high to start. Further studies are needed to understand if the estimation of the
HSP can be useful to predict chemical compounds, polymers and solvents compatibilities, and to
determine if proper estimation can be achieved for different molecules with different functional
groups and basic chemical structures.
4.6

Activation energy (Ea) of the diffusion of selected antioxidants into 95% ethanol
The Arrhenius equation was used to determine the Ea of α-TOC and PG into 95% ethanol.

The Ea can be defined as the energy required for an additive molecule to move through the
polymer matrix [18]. Figure 4.11 shows the Arrhenius plots of α-TOC and PG from PBA and
PBP films, respectively. The plot of Log D versus 1/T gave a linear line with an Ea of 48.5 ± 8.8
kJ/mol for PBA, 54.2 ± 6.9 kJ/mol for PBP. The Ea values of α-TOC and PG are not statistically
significant

difference

71

(p˃0.05).

-7.8

-7.4

-tocopherol
-tocopherol regression line
95% Confidence Band
95% Prediction Band

-8.0

PG
PG regression line
95% Confidence Band
95% Prediction Band

-7.6
-7.8
-8.0

-8.2
Log(D)

Log(D)

-8.2
-8.4

-8.4
-8.6

-8.6

-8.8
-9.0

-8.8

-9.2
-9.0
0.00325

0.00330

0.00335

0.00340

0.00345

0.00350

-9.4
0.00325

0.00355

1/T (K)

0.00330

0.00335

0.00340

0.00345

0.00350

0.00355

1/T (K)

Figure 4.8 The activation energy for diffusion of α-TOC (y= -2515.8649+0.1886; R2=0.9450)
(in the left side of Figure) and PG (y= -2859.4411x+1.4215; R2=0.9211) (in the right side of
Figure) from processed film into 95% ethanol. The slope of each line was equal to –Ea/2.03R.
The Ea values reported for α-TOC diffusion from PLA to ethanol was 96.2 kJ/mol [11]
and for α-TOC diffusion from a multilayer active packaging (made of HDPE, eEVA and a layer
of LDPE containing the antioxidant) to whole milk powder was 11.3 kJ/mol [19], so it can be
observed that α-TOC in different polymer systems have large differences. This difference could
be mainly attributed to the nature of PBAT being a flexible polymer and mostly amorphous
which allows a fast release of α-TOC from its chains. No diffusion of PG in polymer has been
reported, so the comparison of PG in different polymer cannot be made. In this study, the relative
low Ea for α-TOC and PG can be interpreted as the low energy required for these two
antioxidants to diffuse from the processed films into 95% ethanol at temperatures from 10 to
30 .

72

4.7

Optical property

PBC film presents a milky color. PBA film displays a beige color with varying intensities, while
PBP film shows an obvious orange color (Figure 4.12). Table 4.14 shows the CIE L*a*b values
for these three different films.

Figure 4.9 The appearance of PBC (left), PBA (middle), PBP (right)
Table 4.13 CIE L*a*b values and total color change (
Film
PBC
PBA
PBP

L*
93.57±0.06 a
92.31±0.02 b
87.28±0.06 c

a*
-1.19±0.03 b
-1.07±0.03 b
1.62±0.08 a

) for PBC, PBA, PBP *

b*
0.61±0.04 c
5.37±0.25 b
21.92±0.30 a

1

4.93
22.40

*

PBC was used as a reference. Results are means ± the standard deviation of three replicates.
Different letters within the same columns indicate statistically significant different values
(p<0.05)
1

√

.
According to table 4.14, the addition of α-TOC and PG did affect the lightness (L*)

(p<0.05) and yellowness (b*) (p<0.05) of the films. A minor difference of L* were found
between PBC and PBP, while PBP has the lowest L* compared to PBC and PBA. PBP has the
highest b* value compared to other two films. The greenness (a*) of PBC and PBA were not
significantly different, while the greenness (a*) of PBP was higher than PBC and PBA. The total
difference in color

for PBA was 4.93 and for PBP was 22.40. PBP film has a bigger color

change, which can be seen by its orange color (Figure 4.12). PBA also shows color change, it is
73

barely perceptible to the eye in the single film layer but the color was perceptible in the film rolls
to the naked eye.
Figure 4.13 shows the UV-visible light transmission spectra of PBC, PBA and PBP films.
The three films exhibited nearly no light transmission from 300 to 200nm. There is a slight
difference between PBC and PBA from 300 to 500nm, and the difference is more obvious
between PBC and PBP from 300 to 500 nm, which is due to the color changes in the film and the
absorption of the functional group C=O and –OH [20].

Figure 4.10 UV-visible light transmission spectra of PBC, PBA and PBP films. The film
thicknesses of PBC, PBA and PBP were 63.5 ± 5.1, 68.6 ± 2.5 and 76.2 ± 5.1
, respectively.

74

REFERENCES

75

REFERENCES

1.

Hansen, C.M., Hansen solubility parameters: a user's handbook. 2007: CRC.

2.

Auras, R.A., et al., Poly (lactic acid): synthesis, structures, properties, processing, and
applications. Vol. 10. 2011: Wiley.

3.

Van Krevelen, D. and K. Nijenhuis, Properties of polymers. Their correlation with
chemical structure: their numerical estimation and prediction from additive groups
contributions. Elsevier, New York, 1990.

4.

Ortiz-Vazquez, H., et al., Release of butylated hydroxytoluene (BHT) from Poly (lactic
acid) films. Polymer Testing, 2011. 30(5): p. 463-471.

5.

Sharma, O.P. and T.K. Bhat, DPPH antioxidant assay revisited. Food Chemistry, 2009.
113(4): p. 1202-1205.

6.

Al‐Malaika, S., H. Ashley, and S. Issenhuth, The antioxidant role of α‐tocopherol in
polymers. I. The nature of transformation products of α‐tocopherol formed during melt
processing of LDPE. Journal of Polymer Science Part A: Polymer Chemistry, 1994.
32(16): p. 3099-3113.

7.

Reda, S.Y., Evaluation of antioxidants stability by thermal analysis and its protective
effect in heated edible vegetable oil. Ciência e Tecnologia de Alimentos, 2011. 31(2): p.
475-480.

8.

chemicals, C.s., Ciba® IRGANOX® 1076. Ciba® product introduaction, 2004.

9.

Soto-Cantú, C., et al., Release of butylated hydroxytoluene from an active film packaging
to asadero cheese and its effect on oxidation and odor stability. Journal of dairy science,
2008. 91(1): p. 11-19.

10.

Graciano-Verdugo, A.Z., et al., Migration of α-tocopherol from LDPE films to corn oil
and its effect on the oxidative stability. Food research international, 2010. 43(4): p. 10731078.

11.

Manzanarez-López, F., et al., Release of α-Tocopherol from Poly (lactic acid) films, and
its effect on the oxidative stability of soybean oil. Journal of Food Engineering, 2011.
104(4): p. 508-517.

12.

Samsudin, H., et al. Application of Parameter Estimation to Predict Migration of
Antioxidant Films. in 6th Shelf Life International Meeting. 2014. New Brunswick, NJ.

76

13.

Dolan, K.D. and D.K. Mishra, Parameter Estimation in Food Science. Annual review of
food science and technology, 2013. 4: p. 401-422.

14.

Tehrany*, E. and S. Desobry, Partition coefficients in food/packaging systems: a review.
Food additives and contaminants, 2004. 21(12): p. 1186-1202.

15.

Limm, W. and H.C. Hollifield, Modelling of additive diffusion in polyolefins. Food
Additives & Contaminants, 1996. 13(8): p. 949-967.

16.

Crank, J., The mathematics of diffusion. 1979: Oxford university press.

17.

Heirlings, L., et al., Influence of polymer matrix and adsorption onto silica materials on
the migration of α-tocopherol into 95% ethanol from active packaging. Food additives
and contaminants, 2004. 21(11): p. 1125-1136.

18.

igue -Franco, F., et al., Antioxidant Activity and Diffusion of Catechin and Epicatechin
from Antioxidant Active Films Made of Poly (l-lactic acid). Journal of agricultural and
food chemistry, 2012. 60(26): p. 6515-6523.

19.

Granda-Restrepo, D.M., et al., Migration of α-tocopherol from an active multilayer film
into whole milk powder. Food research international, 2009. 42(10): p. 1396-1402.

20.

Hong, X., et al., Visible-light-activated nanoparticle photocatalyst of iodine-doped
titanium dioxide. Chemistry of Materials, 2005. 17(6): p. 1548-1552.

77

CHAPTER 5 CONCLUSIONS
This study explored the use of the Hansen Solubility Parameters (HSP) methodology to estimate
the affinity between polymer and antioxidants (AOxs) in AOx packaging films (APF), in order to
understand the control release of AOx in different packaging systems. Poly (butylene adipate-coterephthalate) (PBAT) was targeted as studied biodegradable polymer, and twenty-five
commonly used AOxs were considered as candidates. The HSPs of PBAT were measured by an
experimental method and the Hoftyzer and Van-Krevelen theoretical calculation method. The
values from both methods were found similar. The data from the experimental method was used
to conduct this work, which were 18.97, 4.83 and 9.10 for the

,respectively. The

HSPs of selected AOxs were also calculated based on the Hoftyzer and Van-Krevelen method.
After determining and plotting the HSPs of PBAT ethanol (a food simulant) and selected AOxs
in a 3D plot, propyl gallate (PG) and α-tocopherol (α-TOC) were chosen to incorporate in PBAT.
According to the basic principle of HSP, the greater distance between the HSP of the two
materials, the less alike they are. So, PG is closer located in the region of interaction of ethanol
and α-TOC is closer located in the region of interaction of PBAT. So, migration test in 95%
ethanol were conducted for these antioxidants-PBAT films to verify if the HSP estimation that αTOC incorporated in PBAT will have less affinity towards ethanol than PG.
n order to conduct the migration test, two PBAT films added with α-TOC and PG were
produced (2% PG-PBAT (PBP) and 2% α-TOC-PBAT (PBA).) The final concentration of
antioxidants in the films were 1.8% (w/w) for PBP and 1.5% (w/w) for PBA due to the adherence
of antioxidants to the screw and the barrel. Preliminary migration study was performed to
determine the optimal experimental design for the migration test for these two films. Based on
the scaled sensitivity coefficient (SSC) analysis during the preliminary migration study, the final

78

migration tests were conducted at 10, 20 and 30

for the produced films. The diffusion of both

antioxidants into 95% ethanol followed Fick’s law for diffusion. The equilibrium at 10 , 20
and 30

for both PBP and PBA films was obtained at 1.75 h, 1h and 0.75 h, respectively. At the

end of the migration study, more than 99% of α-TOC and PG was migrated from PBA film into
95% ethanol. Parameters related to migration process such as diffusion coefficients (D) and
partition coefficients (

) were determined by solving the Fick’s diffusion equations.

α-TOC at three temperatures were all less than 1, while
slightly above 1. The low

for

for PG at all temperature were

for both α-TOC and PG agreed with the release of more than 99%

AOxs into 95% ethanol. D values for α-TOC and PG increase with the increasing temperature,
which means that temperature made a difference on the diffusion process. However, D values for
both PBA and PBP films at all three temperatures were in the same order of magnitude (10-9
cm2/s) due to the fast release at all temperatures. When comparing the D values between α-TOC
and PG at each temperature, only at 20

there was a significantly different between them. D of

PG was larger than D of α-TOC. The difference was not found at 10

and 30 , indicating that

the HSP may not estimate the affinity between the antioxidants and ethanol due to the fast
release of these compounds from this polymer. The activation energy (Ea) of α-TOC and PG
were calculated by Arrhenius equation, resulting as 48.5 kJ/mol for the release of α-TOC from
PBAT and 54.2 kJ/mol for the release of PG from PBAT.
There were obvious changes on the optical properties of PBA and PBP compared with the
PBAT control (PBC) film. PBA films display a beige color while PBP display a heavy orange
color. The transmission spectra of UV visible light of PBC, PBA and PBP films also presented
difference on the light transmission from 300 to 500 nm, due to the color changes in the film and
the functional groups of the AOxs.

79

5.1

Future work
According to the outcomes obtained from this study, the future work may concentrate on

the followings:
1) To determine if the HSP can be used to estimate the affinity between PBAT and AOxs
that are further located in the Hansen solubility parameter space.
2) To use other solvents to investigate the release of these antioxidants such as water and oil.
3) To determine the limitation of the HSP to estimate the affinity between polymer and
AOxs.
4) To further understand the use of prediction methods to determine the affinity between
polymers and different chemical compounds.

80