.......

 

‘6‘. .s

4:. “3;; .LEfiARY .-
37 O 3 Michigan State
Unwersuty

This is to certify that the
thesis entitled

EFFECT OF PACKAGE CONFIGURATION ON BARRIER
PROPERTIES AND SENSORY PERCEPTION OF FLAVOR

presented by
PANKAJ KUMAR

has been accepted towards fulfillment
of the requirements for the

 

 

 

Master of degree in Packaging Science
Science
Majorfl’rofés‘Ec'rTs Signature

I1 Illllostl-

 

Date

MSU is an affinnative-action, equal-opportunity employer

4 l1-I-¢-.--..~

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

{ DATE DUE DATE DUE DATE DUE

EFFECT OF PACKAGE CONFIGURATION ON BARRIER PROPERTIES AND
SENSORY PERCEPTION OF FLAVOR
By

Pankaj Kumar

A THESIS
Submitted to:
Michigan State University
in partial fulfillment of the requirements
for the degree of
Master of Science

School of Packaging

2007

 

ABSTRACT
EFFECT OF PACKAGE CONFIGURATION ON BARRIER PROPERTIES AND
SENSORY PERCEPTION OF FLAVOR
By

Pankaj Kumar
The effect of processing on barrier properties of thermoformed containers was
studied. Two containers with the same surface area, one a semi-spherical shape and the
other a polyhedron shape with six comers, were designed and thermoformed with
polypropylene resin. Crystallim'ty of the polymer in various parts of the containers was
measured. The transmission rates of these containers to oxygen (OTR) and water
(WVTR) were determined. These properties were studied in tandem with change in
sensory properties of a simulated drink. A beverage system with benzaldehyde as flavor
compound was created and stored at 23 °C and 50% RH for 3, 7, 14, 21 and 28 days in
the two types of container. The drink was served to a trained panel to assess the flavor
strength and was analyzed with HPLC to quantify residual benzaldehyde. WVTR and
OTR of the polyhedron containers were 21% and 60%, respectively more than semi-
spherical containers. Sensory studies and l-IPLC results showed stronger benzaldehyde
character in semispherical containers after 21 days onwards. This study confirms that
there is a relationship between package configuration and barrier properties of containers

causing difference in the sensory flavor perception.

 

I I " ;‘. J: .

' p [6' ' AI ' J

.‘1 I ‘g t . 'fl ‘t ‘1
.U I~.‘-’O‘-H .mmm Nth»; (”I WWW

ll II- (hil’tullt tt 1px.» 'Iu rvcn'v OI W I have ”ww
Ralarl .Mm‘ He tau mm 1.,» sed \s w' u gm“ [1* mane ”M
darkness. II; ha: been Jul 1:.l\ L v :~. mend 2106 mm
from u gunk I‘IS- cznm- m u‘tuw-xl ,.'( mums m C. M M .
Rubmu and Z'-. Emu-s Hum, b'uLlI luv: hen-a M h m '

  
  
  
  
  
  
  
  
  
  
  
  
  
 
  
  

gmw up to clam: ms chnc 1.; 5.1.: .IL-r nr I‘xlnginfl Science. In”,
here. both at them tendered (u mc MU: mor'vctx I M M M
Janice ”NW. In what «Canary I tell LIL}: Ming In myo‘”
was Ihmwmtx ui' mt}: . may. l mu“ .iltr N mack Mfr! IIIII.
One person M» was not In my mm In W “I,
.‘h a“ undo do W to” male I! W
Indirquestion was Dr Bruce Harte. i look Allis M“'
m simplicity and In: him. I uyto keep my tea w. m In“

with him. Ithhkilmmlcwmmblmlmdflm. 2 ‘ 1 “I'M ‘ ‘
lwould like to acknowhdgc Dr Sara M‘! ~
me march forum; on appropriate mm lulu, ..
W. M and a. ”JD!“ link M‘MM'
MIWWIIIMMWQ~QH
”Myra-banned: my stay Ind M at!” -‘
I, _u-m “.m I weak! mo H mm
5; .. “Wthdm «awn-nun

     
 

‘.' \‘rni" .77 Isl-”K

ACKNOWLEDGEMENTS

It is difficult to express the extent of gratitude I have for my major adviser, Dr.
Rafael Auras. He has truly proved what a guiding light means when one is stuck in
darkness. He has been an advisor, a friend and beyond everything what one would expect
from a guide. His cause in context of mentoring me was greatly helped by Dr. Maria
Rubino and Dr. Janice Harte. Both have been instrumental in helping me mature and
grow up to claim the degree of Master of Packaging Science. In absence of my mother
here, both of them tendered to me like mothers. I would especially like to mention Dr.
Janice Harte, in whose company I felt like being in company of my mother, who actually
was thousands of miles away. I would like to thank her for that.

One person who was not in my committee but was always available whenever I
had .a question was Dr. Bruce Harte. I look at his personality and I always try to emulate
his simplicity and like him, I try to keep my feet grounded. Though I took only one class
with him, I think I learnt lessons worth a lifetime from him.

I would like to acknowledge Dr. Sara Risch’s contribution as well, as she guided
me towards forming an appropriate drink product. I owe many thanks to Dr. Paul Singh,
Dr. Twede and Dr. Selke, for little chitchats and occasional advises on many matters and
their availability whenever I needed their attention. I must say that they in addition to all
other faculty members made my stay and work at SOP more enjoyable. They made me
feel welcome and important. I would also thank Robert Hurwitz who helped me many
times and went beyond the call of duty. If it would not have been for him, I would have

made countless, and abyss mistakes. Besides, he mentored me in life’s many virtues.

 

I would like to thank CFPPR for funding this project. Tolas Healthcare, BASF
chemicals and Colloides Naturales became special companies for me for the way they
tendered help to an unknown student with all that they could do. My trained panel, whose
participants I can not disclose here was something that I felt very lucky about. They were
dot on target and sacrificed many times while going out of the way to be able to make it
to my study. Consumer panel also deserves to be thanked. I am thankful to School of
Packaging for supporting me financially and giving me an Opportunity to hone my
teaching skills. I am thankful to all the students who survived through my imperialistic
teaching style. Special thanks to Richard Wolthius, his efficiency came in very handy
during the last part of my thesis work. I must say he is a master of the art of mechanical
techniques. Linda, Colleen, Jane, Kristin and especially April were very helpful during
my stay here. I can not forget the contribution they made by taking my problems off my
head and taking care of my business many times.

It is needless to say the extent of support and love I received from my family
during all these years was something that only some exceptional family members can
present. My father has been an extreme force and support during all my years of struggle
while sustaining through all the pain of being away from me. My mother has been no less
in handling the distance from me while trying to hide all the pain of having a son, getting
to see whose face is a luxury for her now. Not to forget my nephews and nieces, sweet
voices of whose have been life-force for me. My brothers and sister-in-laws have been
very supportive and I am well-short of words to express my gratitude to them.

I would need another ten pages to express my gratitude towards friends and

graduate students in and outside of SOP. I would like to thank all of them for the kind

 

support they tendered. Special thanks to all Indian and Thai friends who made it
worthwhile here. Dr. Eva Alrnenar has been instrumental and a great support. Dr.
Joongmin Shin and Dr. Kate Tanprasert were no less when it came to supporting me and
my work. Victor and Ying can not go unnamed in this acknowledgement. Special thanks
to Maria-Paz, Praveen Rawal, Mahesh Khurana, Nitin Gulati, Bhupinder Kumar, Dr.
Napawan and Oh for help with sensory panel and everything else they did for me.
Rajeev, Sunil Bhutani and Gaurav Kale were also there whenever I needed them. Xin,
Neung, Abhishek, Turk and Ploy undoubtedly deserve a special thank for all the debates
and discussions I had with them. Hayati, LeRonda, Josh, Audrey and Isinay also deserve
a special mention. Outside SOP, I am greatly indebted to Mitzi Ma. I feel very lucky to
have her as a friend. Her calming presence and abilities to alleviate all pain made me feel
alive just when I was about to be dead with frustration. Besides, she also helped me
during many stages in the study. Ritu Saini, Aileen Tanojo and Dharmendra Mishra were
outstanding and so was Shiva. Dave Lockhart has to be mentioned specially for enduring
all those calls that I never returned and when I was always caught guilty for being
missing from messenger. Diana deserves a special mention and so do Kristi and her life
reviving jams. Lindsay and Ki were there whenever I looked up for friends.

In the end, I would like to thank God, who gave me the strength to learn and
spread the knowledge. He bestowed me with Karma, and the skills to pursue the same.
His presence is not possible not to acknowledge.

Pankaj Kumar

vi

Table of Contents

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

List of Tables x
List of Figures xi
]_ Intrndnr‘tinn 1
1.1 Background 1
12 Hypothesis
2. Literature Review 6
2_1 Intrndnrtinn 6
2.2 Mass Transfer PL 8
2.2.1 Diffusion 8
2.2.2 Snrptinn 9
2.2.3 Permeation 10
2.2.4 Migration 12
2.3 Permeability Measurement 12
2.3.1 Isostatic terhniqne 12
2.3.2 Quasi-isostatic technique 13
2.3.3 Gravimetric Technique 15
2.3.4 Miscellaneous Techniques 15
2.4 Parameters governing permeability 16
2.4.1 Polymer 18
2.4.1.1 Crystallinity 18
2.4.1.2 Distribution of Crystallites 18
2.4.1.3 Orientation 18
2.4.1.4 Free Volume 19
2.4.1.5 Cohesion & Adhesion Forces 20
2.4.1.6 Additives 20
2.4.2 Permeant 21
2.4.2.1 Molecular Size & Weight 21
2.4.2.2 Vapor Pressure 21
2.4.3 Affinity 22
2.4.3.1 Chemical Similarity 22
2.4.3.2 Polar Similarity 22
2.4.4 Environment 23
2.4.4.1 Partial Pressure 23
2.4.4.2 Temperature 23
2.4.4.3 CO-permeants 24
2.4.4.4 Relative Humidity 25
2.5 Flavors 25
2.5.1 Flavor vs Off-flavors 25
2.5.2 Flavor Constituents 26
2.5.3 Flavor Scalping 26
2.6 Flavor compounds of some major fruits and their thresholds 27

 

vii

3. Materials and Methnds
3.1 I ‘ J tinn
3.2 Simulated Drink
3.2.1 Drink Design
3.2.2 Consumer Panel Benzaldehyde threshold
3.3 Molds

 

3.3.1 Designs

332 Prndnr‘tinn
3.4 Polypropylene Sheet Extntsinn
3.5 Container Thermoforming

 

3.6 le Configllrnfinn
3.7 Shelf Life Studies
3.7.] Phase I: Pre-run Testing
3.7.1.1 Filling
3.7.1.2 Sealing
3.7.1.3 Package Storage

 

3.7.1.4 Quantification of Benzaldehyde

 

3.7.1.5 Sensory Testing with Consumer Panel

3.7.2 Phase II: Final Testing

 

 

3.7.2.1 Selection, Threshold and Training of Panelists ....................................

3.7.2.2 Filling
3.7.2.3 Sealing
3.7.2.4 Package Storage

 

 

3.7.2.5 Quantification of Benzaldehyde
3.7.2.6 Sensory Testing with Trained Panel

 

3.8 Container Properties
3.8.1 Thickness Profile

 

3.8.2 Water Vapor Transmission Rate
3.8.3 Oxygen Transmission Rate

 

3.8.4 Crystallinity

 

 

3.9 Statistical Techniques

4_ Results

4.1 Container Prnperties
4.1.1 Thickness

 

4.1.2 Water Vapor Transmission Rate (WVTR)

4.1.3 Oxygen Transmission Rate (OTR)

 

 

 

4.1.4 Crystallinity
4.2 Shelf Life Studies
4.2.1 Sensory Results

 

4.2.1.1 Consumer Panel

 

4.2.1.2 Trained Panel

 

4.2.2 Benzaldehyde Concentration Determination

5. F- ‘
5.1 Outcomes from the study

 

 

5.2 Recommendations for Future work

 

viii

31
31
35
35
36
38
38

41
41
42
42
42
43
43
43
45

47
47
47
48
48
48
48
48
49
50
5 l

52
52
52
54
55
56
58
58
S8
59

65

A 1‘

Appendix A Approval from Institutional Review Board (IRB) .................................
Appendix B Process Flows
8.] Drink composition
32 Making - ‘ ' s
Appendix C Stability testing of drink
Appendix D Sensory promotional flyer
Appendix E Consent Forms
E.l Consumer Panel Consent form
E.2 Trained Panel Consent Form
Appendix F Sensory Q "
F.1 Questionnaire for measuring sensory threshold of benzaldehyde ...............
F.2 Questionnaire for consumer panel sensory testing ......................................
F.3 Trained panel rating qurst' '
Appendix G Panelist Response for Benzaldehyde threshold study ............................
Appendix H Codes for creating designs in RhinocerosO 3 0
H1 Semispherical Mold
H.2 Polyhedron Mold
Appendix I Deskproto Milling Parameters
Appendix J Calibration curve for benzaldehyde quantification
Appendix K Panelist Performance Charts

 

 

 

 

 

 

 

 

ht
[—Q

List of Tables

Table 1 Major flavor compounds of some fruits and their threshold values in water ...... 27
Table 2 Concentrations of various constituents of the drink 35
Table 3 Solvent flow gradient to determine benzaldehyde- ‘ " 45

Table 6.1 Number of correct responses for different benzaldehyde concentrations ........ 88

 

List of Figures

Figure 1 Wheel of Permeation

 

Figure 2 Process Flow chart for overall prnject
Figure 3 Semispherical Container

Figure 4 Polyhedron Container

 

Figure 5 Semispherical Mold

Figure 6 Polyhedron Mnld

 

Figure 7 Lid configuration

Figure 8 Water Vapor Transmission Rate test setup

 

Figure 9 Oxygen Transmission Rate test setup

 

Figure 10 Thickness distribution of semispherical container

Figure 11 Thickness distribution of polyhedron container

 

Figure 12 Water Vapor TransmisSion Rate (gms/ pkg—day) of Containers ......................
Figure 13 Oxygen Transmission Rate (cc/ pkg-day) of Containers .................................

Figure 14 Crystallinity in different regions of ‘ '

 

Figure 15 Benzaldehyde concentration rating
Figure 16 Residual Benzaldehyde concentration of product in two containers ................
Figure 17 Benzaldehyde panel rating and concentration measured by HPLC .................
Figure B-i Process Flow for drink mmpnsitinn

Figure B-ii Process Flow for making containers

 

Figure J-i Benzaldehyde Calibration Curve for HPI I‘

 

Figure K-i Response in training week 1

Figure K-ii Response in training week 4

 

xi

34

38

38

39

42

49

50

52

53

54

55

57

61

62

73

74

93

94

95

Chapter 1

Introduction

1.1 Background

Polymers are fast growing into as preferred choice for packaging food and
beverage products due to their competitive advantages such as customizable barrier and
physical properties, and appearance and light weight. The ever increasing marketing
requirements for a longer shelf life make it ever more challenging for packaging engineer
to meet the expectations. In this regard, metal and glass have been packaging engineer’s
best friends for a long time, but now polymers and polymer based packaging systems
have taken the center stage in packaging of food and pharmaceuticals. However,
polymers, unlike metal and glass, are semi—crystalline materials and their properties such
as morphology and barrier are subject to change during their processing stage and usable
life. Since barrier properties dictate the shelf life in most food packaging applications,
changes on them will affect the product performance. Extensive and rigorous research
has been conducted to determine these properties and to assess different polymer barrier
properties [1-7] and novel materials are being developed wherever deemed needed [8—
10].

Polymers are vastly used now-a—days to package beverage products. The beverage
industry has always been an intensely competitive market. New and enhanced products
keep on appearing on a regular basis, and the focus has constantly been shifting on
younger consumer generation. The products are not only carrying nutritional values these
days but are more palatable and lasting. A lot of innovation goes in to improving flavor

profile and stability of products making complex flavor systems take over the main stage.

Such innovative and complex products need excellent protection to preserve the delicate
flavor balance. This delicate flavor balance is intended to be preserved for the whole
shelf life. One of the major causes Of spoilage is mass transfer that includes exchange of
moisture, oxygen and flavor compounds between the inside and external environment of
a package. There are different approaches used alone and in combination with each other
to extend the acceptability and shelf life of packaged food products. Incorporating food
preservatives, improving package performance and controlling the storage environment
are major approaches. The incorporation of preservatives like Butylated hydroxyanisole
(anti-oxidant) is limited because of regulation on concentration of many of preservative
ingredients. Besides, some of them like ascorbic acid (anti-oxidant and anti-microbial
agent) have a flavor of their own, and thus they have to be used in sensible concentrations
so as not to influence the desired flavor balance too much. In addition, these preservatives
are not intended for reducing loss of flavor compounds and moisture. Sustaining a
controlled environment in storage and in transit involves huge cost factor. Thus, it
becomes imperative for the container to work beyond the function of containment and
provide mandatory protection to the food product.

As suggested earlier, polymeric materials are vulnerable to changes because of
processing. Change in orientation and crystallinity are prime examples of such changes
occurring in the polymer matrix [11-14]. TO the best of the author’s knowledge very
limited studies look into the barrier changes occurring during conversion from one format
to another. For instance, what happens when the polymer is processed from resin to sheet
form, or what changes occur when resin is formed in to bottles or when the sheet is

thermoformed into containers. There have not been enough studies concerning how these

properties change as a result of number and severity of processing stages. Some times,
the polymers are rendered inappropriate because of inadequate processing conditions
bringing in undesirable changes in the polymer matrix. This hinders the applicability of a
polymer in a particular scenario and the packaging engineer ends up selecting a higher
cost material or higher thickness of material which are not cost favorable.

Another issue overlooked is differences in container shapes. When same material
is processed in two or more different container shapes or configurations, the ultimate
barrier properties achieved of the containers might be different or might not be different.
In this regard, there has not been any significant research. Possibility of barrier properties
being affected differentially can not be ignored in the light of differential stresses being
created because of geometry of the containers. This can cause morphological changes
including but not exclusively crystallinity, distribution of crystallites, orientation etc.
Looking into this might be of paramount importance, especially for flavor intensive
products where the flavor concentrations used are minute.

As discussed above, barrier of containers might be affected by the processing
based upon the container shape or configuration. If there is such change in the barrier
properties, it would be of any importance to the food industry only if it is significant
enough. In other words, if processing into different shapes is affecting the barrier of a
polymer, it should be large enough to be detected by human sensory organs. For example,
if the permeation rates of a flavor compound out of two different containers are different
enough to be detected by the instrumental techniques but are not perceived differentially
by human taste buds, then this difference is not of practical significance for food or other

flavor-centric industries as its effects over marketability of the products would be null or

 

very limited. In such a case, enhancing packaging would result in increase in cost without
any change in consumer preference, ultimately resulting in losing profitability. In case of
pharmaceutical or medical industry, where the functionality of compound is more
important than the sensory perception of the product, enhancing packaging might still be
required for the compounds that are included for their medicinal qualities, but they also
carry flavor value. Benzaldehyde is one such highly aromatic compound that is
commonly used in cough syrups [15]. On the other hand, if the flavor loss is different to a
great deal so that consumers are able to perceive it, this would greatly affect the
preference of consumers as well as shelf life of the product. The concentration of flavors
used in various food beverages is very low to start with, and any significant loss can
seriously damage the delicate balance. In such a case, it becomes imperative to move
towards better design principles to ensure appropriate protection of permeating
components. Thus, it becomes mandatory to study the possibility and extent of changes in

barrier properties of the polymer as a result of processing it in to different shapes.

 

1.2 Hypothesis

The main goal of this work is to assess the extent of impact on the barrier
properties of containers affected by configuration or shape. The null and alternate
hypothesis of this work can be expressed as follows:
Ha
Two polymeric containers made with same material and same surface area but in
different shapes have the same barrier properties and the product stored will have the

same sensory flavor perception.

Ha:
Two polymeric containers made with same material and same surface area but in
different shapes have different banier properties and the product stored will have

different sensory flavor perceptiim.

Chapter 2
Literature Review

2.1 Introduction

Food has two important aspects: nutrition and flavor. While nutrition makes the
food essential, flavor makes it palatable. All the five human senses are involved in
judging the flavor of foods. Aroma, taste, texture, visual appearance and even the sound
are essential parts of the flavor [16] . With the introduction of more volumes and varieties
of food products in packaged form, the job of packaging engineer is getting more
challenging by the day. If the package is not able to protect the product for its aroma
(volatile permeation), taste (lipid and microbial deterioration), texture (moisture uptake/
removal), visual appearance (non-enzymatic/ enzymatic browning etc.) or sound
(crispiness, affected mainly by moisture content), the onus is on the packaging scientist.
If a product loses the nutritional content and that goes below the label claim, the
regulating agencies impose a penalty on the food company, but if a product loses its
flavor or palatability, the consumer rejects the product and this turns out to be a much
bigger penalty for the food company. Thus, it becomes essential for the packaging
engineer to look at the finer aspects of flavor/ packaging interactions. The crucial factor
that compounds the problem is that flavors are just 10'14 to 10'8 % of the food
composition [17]. More over, hundreds of compounds are responsible for flavor of a
particular product. Even though, the character impact compounds are just one or two but
loss of any of those flavor compounds alters the flavor profile, and the product may not

be able to win consumer’s preference. In this context, one of the most challenging areas

is that of fruit flavors. Not only the slow growing market for fruits and fruit juice/ flavor
based products is very competitive, but also with the introduction of innovative products
like cherry, vanilla and grape flavored carbonated drinks, the challenge to packaging
industry has become more incisive.

There are many facets to flavor/ packaging interactions: sorption, permeation and
migration. All play specific roles in altering the flavor profile of a product making it
difficult to keep it acceptable for the intended storage period (called shelf life). The
sorption of aromatic compounds by the polymeric packaging materials can cause
unbalanced flavor profile [18]. After the sorption of the flavoring compound, the polymer
behaves as a channel for the flavor molecule to diffuse into, and finally liberate itself into
the environment by being desorbed on the other end of the polymeric package.
Desorption of undesirable flavors from polymers can also take place if the materials]
packages are being reused as was observed with PET refillable bottles [19]. Thus, these
flavor compounds become part of the packaging material in the virgin cycle while they
migrate to the product in the second and later cycles even though they were not intended
or expected as additives or migrants. Such an event involves participation of the entire
spectrum of mass transfer phenomenon. Research is going on in all the three aspects of
flavor/ package interactions, and this review tries to bring together the fundamentals as
well as recent developments in the field of mass transfer, factors affecting mass transfer,

techniques to measure it and available data for important fruit flavors.

2.2 Mass Transfer Phenomenon

2.2.1 Diffusion

Diffusion is the process of movement of substance within itself or another
substance [20]. Crank [21] defined it as the process by which matter is transported from
one part of a system to another as a result of random molecular motions. In context of
diffusion through polymer, it involves movement of flavor compounds, moisture vapors,
gases and additives across the polymer matrix. This phenomenon is mainly governed by
the Fick’s first law [22] at steady state:

F = —D(6c/ 6x) (1)

where F is the flux of mass,

D is the Diffusion Coefficient,

5c/5x is the partial of concentration gradient and distance

D, the average effective diffusion coefficient can be defined as:

1 C2 D
D = —— j —"’c- dC (2)
C2 - C1 C1 1— a)
where C2 and C] are concentrations of the diffusant on the two sides of the polymer wall

Dloc is the binary mutual diffusion coefficient of the diffusant in the polymer
to is the mass fraction of the diffusant in the polymer

Fick’s law identifies concentration gradient to be the only driving force and such
diffusion is called Fickian diffusion. It has been observed that sometimes, besides
concentration gradient, other factors like gradients in stress or temperature might also
drive diffusion. Also, concentration itself besides concentration gradient affects diffusion.
Such a type of diffusion is called Non-fickian diffusion and can not be explained with

Fick’s laws of diffusion. In such cases, concentration dependence of diffusion is

explained by two mechanisms: Flory-Huggins [23] and clustering [24]; wherein the first
mechanism accounts for increase in diffusion coefficient, the other explains decrease in
diffusion coefficient.

Diffusion plays an important role in determining the permeation rate of a
permeant through a barrier material. This is a phenomenon that is active during both
steady state as well as unsteady state. In simpler words, large diffusion coefficient means
easy movement of permeant through the polymer wall and thus increasing the

permeability coefficient values.

2.2.2 Sorption

Sorption is the uptake of product components, such as moisture, flavor, aroma or
colorant compounds by the polymeric packaging material [11]. These components are
called sorbates. This phenomenon is also referred as negative migration since the
polymer behaves as the receptor of the sorbate instead of being source of the
contaminant. Sorption is primarily governed by similarity between the polymer and the
sorbate, chemical or polar. This is in fact related to solubility of the sorbate into the
polymer. Mixing of a compound such as oxygen or an aroma vapor (a gas) with a
polymer (solid) (or a liquid for that matter) at molecular level is called a solution [11]. At
low concentration levels, it behaves like an ideal solution and follows Henry’s law:

6.- = S - p. (3)

where c, is concentration of the solute (sorbate)

S is Henry’s proportionality coefficient called Solubility coefficient

p, is the equilibrium vapor pressure of the solute (sorbate)

The solubility coefficient defines the rate at which a sorbate would get absorbed
by the polymer or in other words what would be the extent of sorption. But Henry’s law
holds only when there is no interaction between the polymer and the sorbate. Henry’s law
works very well with Oz and N2, at essentially all pressures of practical interest but many
organic vapors including many flavor compounds and C02 tend to deviate from Henry’s
law at sufficiently high pressures [l l]. Langmuir— Henry’s law equation is useful to
describe high pressure permeant sorption in polymers like that of C02 in Polyethylene
Terephthalate (PET) bottles. Other equations like Flory-Huggins equation [23] are more
appropriate to use when there is an interaction between polymer and permeant. It defines
the solubility relationship by including an interaction parameter correlating the vapor
activity of the solute in the gas phase with its volume fraction in the polymer.

Whereas the solubility coefficient defines the rate of sorption at which the sorbate
would be absorbed into the polymer, the partition coefficient defines the overall sorption
that would occur. Partition coefficient is the ratio of concentration of the sorbate at

equilibrium in two phases, here product and polymer phases.

[C 1;
Partition c0efi7cient= eq

— (4)
[Ceq ]II

where numerator and denominator denote the equilibrium concentrations of the sorbate in

the two phases. Both solubility and partition coefficient are temperature dependent.

2.2.3 Permeation
“Permeation is the movement of gases, vapors, or liquids (called permeants)
across a homogenous packaging material, and excludes the travel of materials through

perforations, cracks, or other defects” [11]. Thus, this iszthe transport phenomenon of a

10

permeant molecule from one side of the polymeric wall to the other side, from higher
concentration] pressure side to the lower concentration/ pressure side. Permeation is
related to permeability or permeability coefficient, which is a characteristic of a barrier
material for a specific permeant at a specific temperature. Permeability thus is
independent of thickness of the barrier, partial pressure difference, area of exposure and
exposure/ permeation time. All these variables are normalized in the following equation

for permeability coefficient:

_ Q'l

P-t-A-Ap (5)

where P is Permeability coefficient at steady state

Q is total quantity of permeant that permeated through the polymer wall

1 is the thickness of the polymer wall

t is the total time of exposure! permeation

Ap is the partial pressure gradient between the two sides of the polymer wall.

Thus, permeability coefficient is a function of polymer, permeant and
temperature. Steady state permeability coefficient can also be expressed as a function of
Diffusion (D) and Solubility coefficients (s).

P=D-S (6)

The above equation holds at low concentrations of the permeant and when there is
no interaction between permeant and polymeric material. Materials change behavior
when environmental temperature changes, and in some cases materials change behavior

when environmental relative humidity (RH) changes. Thus, permeability Of a material for

a particular permeant, changes as the temperature changes and sometimes is also affected

by the change in environmental RH.

2.2.4 Migration

Migration is the phenomenon of polymer additives and components moving in to
the product matrix. Examples of compounds that can migrate include residual monomers,
solvents, catalysts, additives and even residual flavor compounds from the earlier
polymer cycle if the polymer is being reused. Migration affects not only the organoleptic
qualities but also the toxicological attributes of the food product. This can cause serious
health problems as well as damage to the flavor profile. The flavor change caused by

such migration of polymer components is deemed as off—flavor.

2.3 _ Permeability Measurement

Permeability can be measured using many different techniques. Isostatic and
quasi-isostatic, gravimettic techniques are the most popular techniques. These techniques

are more or less applicable depending on scenario. These techniques are briefed below.

2.3.1 Isostatic technique

Isostatic technique involves maintaining the partial pressure gradient of permeant
on the two sides Of the packaging material constant. The high pressure side is kept
constant while the low pressure side is maintained at zero in most cases. This is achieved
either by sweeping a carrier gas or a trapping agent. Diffusion and permeability
coefficients can be measured with the following equations, once the steady state flow has

been established.

 

 

= 7.211/2 (7)
Fss-l
P = (8)
Ap

where l is thickness of the barrier permeability

F3, is the steady state flux

Ap is the partial pressure difference

t1 ,2 is the time to achieve half of the steady state mass uptake

Solubility coefficient can be calculated once values for P and D are available
using the Fick’s law. It is a common practice to use this technique to measure the
permeation of generally non-interacting permeants like moisture vapor, 02, CO2 and N2
[25-27]. There have been attempts to measure organic vapor permeability too using this
technique [7, 28] but it becomesessential to conduct a consistency test to make sure that
the results are not adversely affected by the interaction between the permeant and the
packaging material. Permeability equipments like Mocon Permatran, Oxtran, Illinois
instruments utilize this technique. ASTM E96, F1249 and D3985 are also based on this

technique.

2.3.2 Quasi-isostatic technique

Quasi-isostatic technique is applied in scenarios where it is not feasible to
continuously keep the low pressure side at zero and detect the permeant simultaneously.
This is a pressure—variable method. The diffusion coefficient is calculated by lag time (0),
which is the time required for the particular permeant— polymer system to reach steady

state marked by constant rate of mass transfer.

13 ‘

12
‘E

_ g 1
LL]...

where 1 is the thickness of the barrier material

D (9)

9 is lag time

(q/t)ss is the slope of the accumulated quantity of permeant plotted against time at

steady state

A is area of barrier material

Ap is the partial pressure gradient

The quasi-isostatic method involves change in pressure gradient (usually below
5%) but still assumes the changes to be effectively small enough to assume changes in Ap
to be zero. The apparatus for the quasi-isostatic technique involves use of a permeability
cell with two parts separated by the barrier material. The low pressure side has a
sampling port used to sample the headspace that is analyzed to quantify the vapor of
interest. This method is heavily utilized to measure the permeability of organic vapors
individually as well as in combination to each other [29-31]. A gas chromatograrn is
typically utilized for separation and quantification with a Flame Ionization Detector
(FID), Thermal Conductivity Detector (TCD) or Mass Spectrophotometer (MS) detector.
This gives an opportunity to measure permeability of two or more organic vapors in

tandem, bringing it closer to real field conditions.

 

2.3.3 Gravimetric Technique

Gravimetric technique is considered as a type of isostatic technique, but it has
grown as a very significant technique [25, 32, 33] and deserves a separate mention. With
this technique, the permeability coefficient is calculated indirectly by measuring
solubility and diffusion coefficients. The polymer is exposed to the vapors of interest, and
the weight change is observed at very short time intervals. The solubility coefficient is
calculated from the total sorption by the polymer once the steady state is achieved. The
diffusion coefficient is calculated with the data available for non—steady state. For non-

interacting permeants, in general, following gravimetric diffusion coefficient is calculated

 

by the following equations:
2
= l (11)
7.2!1/2
W —W-
s = ————f ' (12)
W.- -Ap

where Wf is the weight of the polymer after attaining steady state

W, is the dry weight of the polymer
The product of D and S provides the permeability coefficient. Exciting developments
have taken place with this technique with novel kinds of detectors and microbalances

utilizing it for water and organic vapor permeation measurements.

2.3.4 Miscellaneous Techniques
Most of the novel techniques utilize isostatic or quasi-isostatic dynamics in one

form or another to measure permeation rates. The development is usually limited to the

 

detector and the techniques to derive information from the varied kinds of detectors. One
of the exciting advances is the utilization of Quartz Crystal Microbalance (QCM) to
measure the permeability of extremely thin films [34-39]. The mass uptake by the
polymer film is measured by change in resonance frequency of the quartz crystal which
has been coated with the polymer film. Once the mass uptake has been calculated, the
same equations as for the gravimetric technique are utilized to calculate the mass transfer
coefficients. Another technique of importance is Fourier Transform Infrared (FI'IR)
spectroscopic measurement. With this method, the variable being measured is the
particular IR wavelength that is characteristic of the permeant compound. This has
proved to be an important tool when the need is to study permeation in a dynamic
environment in tandem with some other effects [40-44]. Small Angle Neutron Scattering
(SANS) is an upcoming technique being applied in some studies [45, 46]. A big
advantage with this technique is that it not only helps understand the diffusion process,
but also lets to look at the morphological changes including glass transition temperature
and d-spacing between the polymeric chains, which is defined as the perpendicular

distance between two planes inside a crystal.

2.4 Parameters governing permeability

There are many factors that affect the permeability of packaging materials
towards moisture, gases and organic compounds. Permeation occurs mainly in two
stages, unsteady state and steady state. During the unsteady state, the rate of
concentration change of permeant across the matrix of the polymer is variable and
gradually increasing while it becomes constant in the steady state. Diffusion and sorption

both play significant role in the unsteady state while diffusion takes over and sorption

 

reaches dynamic equilibrium during the steady state. Both sorption and diffusion are
affected by myriad factors that can mainly be classified originating from 4 sources:
Polymer, permeant, affinity between polymer and permeant, Environmental conditions.
Because of such multiple factors, every permeation phenomenon becomes unique and
attention has to be paid to each individual element in order to understand it. Figure 1
depicts the process of permeation and the factors affecting it. This “wheel of permeation”

provides a snapshot of the factors involved during permeation.

Molecule Size
& Weight

 

Figure 1 Wheel of Permeation

17

 

2.4.1 Polymer

2.4.1.1 Crystallinity

Crystalline materials are known to be impervious to gases, moisture and organic
compounds. Metal and glass are perfect example of that. Polymers being semi-crystalline
are affected by the extent of crystallinity present in the polymer. It has been shown that
crystallinity has a bearing in the effectiveness of a polymer as a barrier [4]. In the
crystalline part of a material, the mobility of a permeant is very limited, and it is
considered that all mass transport through semi-crystalline materials is through the
amorphous regions [11]. It has been also suggested that the mass transfer rate is highest
through the interfacial regions between crystalline and amorphous regions owing to the

enhanced stresses in those areas [1 1].

2.4.1.2 Distribution of Crystalline

Distribution of crystallites affects the permeability in a unique way. An
appropriate distribution of crystallites presents a tortuous path of diffusion to the
permeant. Permeant thus require higher activation energy to diffuse through such a
polymer. Same reasoning is applied to create novel barrier material by utilizing nano-
materials based on various kinds of crystalline materials. In such composite materials
also, getting an appropriate dispersion is very important to achieve optimal properties.
Besides the crystallites present in the polymer chain restrict the movement of amorphous

parts and thus increasing the activation energy and barrier against permeation [47].

2.4.1.3 Orientation
Random and disorderly polymeric chains do not provide equivalent barrier as an

oriented polymer. The orientation of polymer brings the chains closer to each other as

18

 

well as aligning them. Thus, density increases and the molecular mobility in oriented
regions decreases. This decreases the permeability by limiting diffusion of the permeant
through film. This is a generally accepted argument, but there have been some anomalies
reported to that as well. One such reported behavior was that of Vinyl Dichloride
copolymer that showed increased permeability to oxygen when biaxially oriented as
compared to unoriented extrusion cast film [48]. The researchers attributed the increase

to microvoid development in the polymer matrix caused by orientation.

2.4.1.4 Free Volume

Free volume represents the volume available in the polymer matrix that is not
occupied by the polymer chains or additives. These are the gaps created in between the
entangled polymer chains and thus are available for the permeant to occupy and provide a
channel for diffusion. Thus, availability of more free volume means both more diffusion
and solubility. This mainly relates to the density of the polymer [49]. Less density would
mean more free volume and thus more permeability. This further can be related to
molecular weight and its distribution. As the molecular weight distribution increases, the
sorption increases [11, 50]. This probably can be explained on the increased irregularities
in the polymer matrix providing more channels and voids at the surface for the permeant
to invade. In this context LLDPE which has narrow molecular weight distribution is
known to be a better barrier compared to LDPE which has relatively broader molecular

weight distribution. The density of a polymer also works against sorption [49] and ease of

diffusion.

 

2.4.1.5 Cohesion & Adhesion Forces

Secondary forces between the polymeric chains play a critical role in defining the
barrier of a polymer. For example, polyvinyl chloride (PVC) being completely
amorphous would be expected to be a poor barrier. On the contrary, unplasticized PVC is
an excellent barrier to gases, organic vapors and moisture. The partial charge on chlorine
and hydrogen atoms that allows making H-bonds and thus providing secondary forces
between polymer chains, restrict the chain mobility. That raises the glass transition
temperature and the activation energy, and limits the diffusion and permeation. Similarly
introduction of metal ions in polymer chains like that of LDPE raises its barrier properties

as evident from the increased barrier of ionomers like surlyn.

2.4.1.6 Additives

Additives usually behave as plasticizers increasing the chain mobility and thus
permeability. Additives can decrease the free volume available but this effect is expected
to be very small because of the small quantities used. This effect is also nullified by the
very nature of the additives themselves. Many additives have inorganic parts that create
stresses at interface and cause more free volume in these regions thus aiding further in the
permeation process. Those additives behave like impurities breaking continuity of
otherwise continuous phase. There are also some additives that might act as bridges
between the polymeric chains and in such a case they limit the permeability by enhancing

the intermolecular forces and reducing the mobility of polymer chains [51].

20

2.4.2 Permeant

2.4.2.1 Molecular Size & Weight

Smaller and lighter molecules can diffuse easily and fast, simply because they
require less amount of energy to move [1 1]. Besides, they can reside in smaller spaces as
compared to bigger molecules, and can travel through narrow channels and defects
available in the polymer matrix. This is the reason that H2 and He are the most permeable

permeants through all non—impervious materials.

2.4.2.2 Vapor Pressure

Vapor pressure of the permeant makes it more or less mobile at room temperature.
This relates to molecular weight and boiling point of the permeant. The molecules of
gases are very mobile because at room temperature they are in gaseous phase and thus
exhibit high diffusion. Other viscous and high molecular weight permeants need
relatively higher amount of energy not necessary available abundantly at room
temperature. The permeants exhibit relatively lower diffusion capability. The vapor
pressure of a particular permeant would also change depending upon what kind of system
it is trapped in. For example, different foods containing same flavor will have different
partition coefficient with air for the same flavor compound. This would cause different
vapor pressure of the permeant on top of the different food matrices. High vapor pressure
also causes deviation from Henry’s law of solubility, where Langmuir- Henry’s law

equation holds more relevance [11].

21

 

2.4.3 Affinity

2.4.3.1 Chemical Similarity

Chemical similarity between the polymer and the permeant increases the
solubility of permeant in the polymer. For this reason, the long chain permeants have a
greater solubility in polymers since polymers themselves are long chain organic
compounds. The effect of carbon chain length is observed especially in case of
aldehydes. Aldehydes in general show relatively low affinity towards polyethylene. But
as the carbon chain length increases, the sorption also increases [49]. On the other hand
such long chain permeants would have decreased diffusivity for the requirement of higher
activation energy to mobilize the larger molecules. Similarly, same functional groups
enhance the solubility as well. Such chemical similarity reduces the surface tension
between the two phases and may also cause relaxation of polymer chain. If the affinity is
very high, causing clustering, diffusion coefficient will go down. In all scenarios,
chemical similarity will increase the solubility coefficient and depending upon the
dominance of diffusion or solubility, the permeability coefficient may increase or

decrease.

2.4.3.2 Polar Similarity

Polar similarity affects the surface tension between the polymer and the permeant
and thus the solubility of permeant and the polymer. This is the reason that PET is
relatively a better barrier against organic vapors while poly ethylene and other
polyolefins make for better barrier against moisture vapor. PET with its partial polar
nature stops the organic vapors to permeate through it while polyolefins being completely

apolar in nature stop a polar molecule like moisture to pass through. Similarly straight

22

chain hydrocarbons like d-limonene are the fastest to get absorbed and permeate when in
contact with LDPE while compounds like hexanol and linalool are relatively slow and
less in extent given the presence of hydroxyl group. In this context, the affinity towards
getting absorbed in an apolar polymer like poly ethylene has follows this sequence:
hydrocarbons> ethyl esters> aldehydes> alcohols [49].

Enhanced solubility of long carbon chain permeants can be also explained on the
relative contribution of the polar part which is decreasing as the carbon chain length
increases in relation to the permeants that have non-polar functional groups [52]. The
same phenomenon was observed with esters. For the polymers that are prevalently apolar
in nature, such permeant would absorb and diffuse faster while in polymers that are

relatively polar, shorter carbon chain would absorb and diffuse faster.

2.4.4 Environment

2.4.4.1 Partial Pressure
The partial pressure gradient defines the driving force for the permeation process.
For permeants like 02, it is relatively a universal gradient. For organic permeants, it

would vary based on the product under consideration, or the specific environment.

2.4.4.2 Temperature

Effects of temperature on permeability are very well recorded. Generally,
permeability increases with increasing temperature. As the temperature increases,
polymer chains become more mobile as they get closer to the glass transition temperature
(in case of glassy polymers) or further higher than it (in case of rubbery polymers).

Besides, because of the increase temperature more and more permeant molecules now

23

 

have the required energy to diffuse in and permeate through. The effects of temperature

are best defined by an Arrhenius-type relationship:

P = Poe—Ep/RT (13)

Temperature has two pronged effect on the permeability. One, it provides more
energy to the permeant molecule to go through the polymer wall, and two, as the
temperature goes up, the chains in the amorphous region of polymer become more
mobile, thus allowing the permeant to pass through. In other words, the barrier of
polymer goes down while permeant molecules have more energy to cross the barrier.
But, there are a few anomalies. Polystyrene and polylactide have negative activation
energy [53]. It is hypothesized that some kind of cross—linking activity takes place in
polystyrene increasing its barrier at higher temperature. Polylactides exhibit regular
barrier trend against oxygen and C02, but with moisture the trend reverses [53].
Temperature affects the partition coefficient also, thus affecting solubility. With
increasing temperature the solubility usually decreases, but there are exceptions to it. The
partition coefficients (polymer/ gas) of ethyl acetate, n-hexanal, and d-limonene at vapor
activity of 0.2 were found to decrease as the temperature increased from 25 °C to 40 °C
for barrier materials such as foil, metallized- PET (M-PET), SARANTM and ethyl vinyl
Alcohol (EVOH) [54]. In the same study it was found that partition coefficient (polymer/

gas) of a—terpineol with all the materials and SARAN with ethyl acetate increased.

2.4.4.3 Co-permeants
If there is a significant interaction between a permeant and the polymer, it is quite
possible that the polymer would swell by absorbing the molecule. This would mean

increased voids and spaces between the chains paving the way for other permeants to

24

 

pass through. This phenomenon has been noted with d-limonene and ethyl acetate where

d—limonene helps ethyl acetate escape from polyolefin matrix [55].

2.4.4.4 Relative Humidity

Relative humidity affects only those polymers that are hydrophilic meaning they
have some functional group in the chain that can create hydrogen bonds. In most
circumstances, as the relative humidity goes up the polymer’s barrier towards other
permeants goes down. EVOH is a classic example of such a polymer. Water molecules
get themselves attached through H-bonding with the hydroxyl functional group and act as
a plasticizer for the polymer making the chains further apart and increasing the
permeability [50]. Thus, the permeability in this case is a funcu'on of not just temperature
but relative humidity as well. In some cases, there have been claims of increasing barrier
with increasing RH. Some trademarked amorphous type of nylons (SelarTM from
DuPont®) are claimed to exhibit such a property. It can be explained in terms of creating

pseudo-crosslinks with limited H-bondings between the polymeric chains [56].
2.5 Flavors

2.5.1 Flavor vs Off-flavors

Flavor is often defined as a sensation realized by the human brain when a food or
beverage is placed into the oral cavity, and this is dependent upon the reactions of the
taste and olfactory receptors to the chemical stimuli of the particular compounds present
in the foods [57]. In oversimplified terms, flavor is the product of a physiological reaction
that, defies objective measurement [58]. But this definition does not define the distinction

between flavors and off-flavors. More effective definition will be “flavor substances are

25

 

strong-smelling organic compounds with characteristic and usually pleasant odors” [59].
Thus, we can distinguish flavors and off-flavors as both belonging to the same chemical
species category of compounds producing sensations through taste and olfactory
receptors but flavors being pleasant and “in-place” while off-flavors being either
unpleasant or “out of place” e.g. milk picking up onion smell because of being stored

along with onions in refrigerator.

2.5.2 Flavor Constituents

Flavor is composed of three parts namely character impact items, contributory
items and differential items. Character impact items are those which when smelled and/
or tasted are reminiscent of the named flavors and it provides most of flavor’s
organoleptic effect. Contributory items are those that when smelled and/ or tasted help to
create, enhance or potentiate the named flavor. These bring the character impact items
closer to the flavor. Differential items are additives and combination of additives that
have no or little reminiscent of the named flavor and are added just to provide

individuality and difference (this concerns specifically to synthetic flavorings).

2.5.3 Flavor Scalping

Flavor Scalping is defined as the loss of flavor compounds (mainly of character
impact and contributory items) due to either absorption by the package or permeation
through the package after the product has been packed. This causes flavor imbalance and
thus loss of consumer preference. The sorption of flavor compounds by the packaging
materials and into the seals might severely affect the polymer as well as overall package

performance.

26

 

2.6 Flavor compounds of some major fruits and their thresholds

Presented below is a list of major flavor compounds of some fruits and their odor

and flavor thresholds whichever is applicable. The list has been compiled by acquiring

threshold values from mainly two great sources [60, 61] which are online database

extracted from many resources with references provided. The flavor compounds found in

different fruits have been from many different sources [57, 59, 61-74]

Table 1 Major flavor compounds of some fruits and their threshold values in water

 

 

Important Ma'or Fruits Odor threshold Flavor threshold
Constituents J (Ppb) (PPb)
2,5-Dimethyl-4—
methoxy-2H-furan—3- Strawberry 10
one
2-heptanone Raspberry 140-3000 1000
2-methylbutan-l-ol Apple 4000
2—methylpropan-1-ol Apple 7000
2-methylpropyl acetate Apple 800
Raspberry
2-phenylethanol Apple 750
3-methylbutan-1-ol Apple 250-300 170
3~methylbutanoate Peach 60-76
Benzaldehyde Apple 350-3500 1500
Cherry
Peach

 

27

 

 

 

Important Ma'or Fruits Odor threshold Flavor threshold
Constituents J (ppb) (ppb)
Benzyl alcohol Peach 10000 5500
Raspberry
Ethanol Apple 100000 52000
Ethyl 2— Apple .006-. 1 0.1
methylbutanoate
Ethyl Butanoate Strawberry 1 450
Apple
Raspberry
Ethyl Butyrate Grapes 1 450
Ethyl Hexanoate Strawberry 3
Peach
Raspberry
Eugenol Cherry 6-30
Geraniol Cherry 4075
Raspberry
H-(4—hydroxyphenyl- Raspberry 100
butan-Z-one) (raspberry
ketone)
hex-2-en-1-ol Apple
hex—2-enal Apple 17
hex-3-en-l-ol Apple 70
hex-3-enal Apple 0.25
Hexan-l-ol Cherry 6

 

 

 

Important Ma “or Fruits Odor threshold Flavor threshold
Constituents J (ppb) (ppb)
hexanal Apple 4.5-5 16-76
4.5-5
Cherry
Linalool Strawberry 6
Cherry
Blue Berry
Peach
Raspberry
methyl anthranilate Grapes 3
nonanol Peach 50
Nona-trans-2,Cis-6- Cherry 0.1
dienal
Octana] Cherry 0.7 5-45
pentan-Z-one Apple 70000
pentyl acetate Apple 5-5000 3000-6600
Phenylacetaldehyde Cherry 4
p-Menth-l-en-9-ol Cherry
propyl butanoate Apple 1 8- 124
trans-hex-Z-en- 1 -ol Blue Berry
trans-hex-Z-enal Cherry 17
Strawberry
Blue Berry
trans-Non-Z-enal Cherry 008-0. 1 6

 

29

 

 

Important Major Fruits Odor threshold Flavor threshold
Constituents (ppb) (ppb)
a—ionone Raspberry 0.4
a-terpineol Peach 330-350
B-damascenone Grapes 0.002

Raspberry
B-ionone Cherry 0.007

Raspberry
B-Phenylethanol Grapes 750
y-decalactone Raspberry l l 88
y-dodecalactone Peach 7
y-hexalactone Peach 1600
y-undecalactone Peach
8—decalactone Peach 100 90-160
8-dodecalactone Peach 1000
8-undecalactone Peach 150

 

30

Chapter 3
Materials and Methods

3.1 Introduction

The major stepping stones of this work were making containers, making the drink,
developing analytical techniques and ultimately conducting the testing after filling and
storage of the product in the containers.

As part of the preliminary work, the packaging material and flavor compound
were selected. Polypropylene (PP) was chosen as the packaging material. PP is a very
common polymer used in food and pharmaceutical industry. PP is also known to have
very limited ability in protecting against flavor loss as it is a polyolefin, being of non-
polar nature like most of the flavor compounds. This gives the possibility to demonstrate .
the hypothesis in reasonable time frame and providing efficient tools for effective cost
management with the use of a relatively cheaply available polymer.

Benzaldehyde was chosen as the flavor compound. Benzaldehyde is a character
impact compound for cherry and almonds and is also known as bitter almond flavor. It is
also found in many fruits of commercial importance including apples, peaches, plums and
many types of berries. Benzaldehyde has both aromatic and taste qualities, and thus
involves the use of all flavor sensing organs. Benzaldehyde also has medicinal effects and
is also used in pharmaceutical industry [15, 75].

Two different package configurations were needed that would be different to each
other in configuration and also practically feasible. A round semi-spherical shape and a
polyhedron shape with six comers were used for this purpose. This allows comparison

between two configurations, one with no comers and one with multiple comers. The

31

designs were created so as to allow for equal surface area. This did not include the
lidding part which was ensured to be a total barrier achieved by the use of foil based
materials. Therrnoformability of containers was also a consideration. Very complicated
and intricate shapes were not feasible to produce with simple machines available in-
house.

Simulated cherry drink was created by starting with a simple composition of a
fruity flavored drink based on an acid—sugar system. Gum acacia was added to the drink
to make benzaldehyde (natural oil) soluble in the aqueous system and vitamin E was
added to serve the purpose of anti-oxidant and fortifier. Appropriate levels of flavoring
compound were found using threshold study for benzaldehyde concentration in the base
system.

Design files for the molds were created for both the two shapes identified for the
containers. For milling the molds, many parameters were defined when converting from a
3-dimensional design file to a machine understandable command file. These parameters
were Optimized by trial and error.

Polypropylene sheets were cast extruded. The optimal extrusion conditions were
found to get appropriate thickness with least variability achievable with both thickness
(measured after regular periods with a micrometer) and crystallinity (measured in terms
of visual clarity of the plastic sheets).

The molds were used to thermoform the containers from the polypropylene
sheets. Thermoforming conditions were optimized to provide maximum productivity with
both types of containers. After the drink composition was finalized and the containers

were manufactured, the product was manufactured and packed in clean conditions. High

32

Pressure Liquid Chromatography (HPLC) technique was developed for quantification of
benzaldehyde in the drink. Lid configuration and sealing conditions were identified.

The product was filled and sealed in the containers for the Phase I pre-run and
stored under room environment at 23 °C and 50% RH. Sensory and HPLC evaluation
was conducted after 2 and 7 days’ storage. For the sensory evaluation, consumer panel
was used and subjected to triangle test. The sensory studies were approved by
Institutional Review Board (IRB) and the relevant documents are attached in appendix A.

Based on the experience of this preliminary run, many changes were made to the
overall process. Individuals were screened for their ability to rate the strength of
benzaldehyde in the drink. Selected individuals were trained for 5 weeks and so
developed trained panel was used for final testing. For the final stage of testing, drink
was prepared under nitrogenous environment. Product filling and sealing was carried out
under nitrogenous environment also. Filled containers were stored in a nitrogen filled
chamber for 3, 7, 14, 21, 28 days.

Testing at 28 days was replicated. Trained panel rated the concentration of
benzaldehyde in the product, and HPLC was used to quantify the concentration.
Permeability of the two containers for water, oxygen and benzaldehyde was also
measured. Crystallinity in the walls and comers of the containers was also measured.

Figure 2 provides schematics of the materials and methods section.

33

80.3%. 2803 :8 52.0 30E 388m m 8%.".

....u:m,
-111. M grew-Fact; N
o_a>_ac< 30.3305 M

 

111- 1 ..1.

.r m.m>.m:< bomcow .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9:30... ..
.11-. b_:___£mbo .r. 1.1a:1_n1____1maBn1.-11 1.-...
_. . : .0 Egan-mam: -1. 1:1 .. 9.8 mm _
fl _ .71 w .1...-N.,.:.k..m-.o.mm1mbm1
_ _
3.2.32“. _ 333m _ . . -, 1111-...
o EoEoSmmo _ M _ . _ .ocmn. .0
.1 .50 w 1111- 1&- T _ 352.30 . m 2... =25 . w-1om_m_m_1+-mcm @898
1 1 - fl _ r u _
E..>>> ..o .c-mE193-mm-o5 .l jam—.11.
.- 95ch u. 32E -_ 11- 1 11.1
. m.m>_m:.< bogemz
.11 111 _ ., . 1-...-_.-..1...«e.... 1.
mcoeecoo . M .. -11..
m:__mom1mmo=m:_E.2oo t. . 9.055200 3 _
u H m man s new m to. 1: $8
. ...... 1, .- . 1 _1m..a o. a m meme-@- 1.
m 52139500 c... Q 1.. . (3.61.... 0.5.. 3:1...131:
v- mcflmem. 05E £890 -
.1-1 - 11 1 - -111-xc_.u .9...me
-111 :o.m..3.;x.w -.om..:..m -11-- _ U W m. . ---...-..-.11
1-11 11 11 . m _ . _ . _ A cozmcEcoEc _
. -1 1 1 . . 3820.. . " v.02 .. . v.55 M “ e_o.:m£m.-mn>1:oe_m~com
. :o..oo..om .eEzoL _ H _ m n - .-1
. 1- .1.1411. 11-- 5.11. .. >111: fil-1mmzdoo-cmo ...1
. v.56 Em_3E_m

34

3.2 Simulated Drink

3.2.1 Drink Design

The process flow for drink design is attached in appendix B. In any natural drink
system, there are few essential components and flavorings other than the identifying
flavor. It usually constitutes of a sour-sweet system with anti—oxidants. The drink system
that we used as a simulant had sugar as the sweetener and citric acid as the souring agent.
a-Tocopherol acetate or Vitamin E served dual purpose of anti-oxidant and fortifying
agent. Benzaldehyde being natural oil required an emulsifying agent to dissolve in water
therefore gum acacia was used for the purpose. Food grade benzaldehyde (98.5% purity)
and citric acid were procured from Sigma-Aldrich, St. Louis, MO. “Spartan” brand
household granular sugar was provided by Spartan Stores Distribution, LLC, Grand
Rapids, MI. Gum Acacia (grade: IRX 4765) used as emulsifying agent was provided by
Colloides Naturels, Bridgewater, NJ. Vitamin B (grade: 500 BG) used as anti-oxidant
was provided by BASF, Florham Park, NJ. The required concentration of benzaldehyde
was defined after threshold testing (discussed later). It was decided to use 45.75 ppm as
targeted concentration of benzaldehyde in this drink system. Vitamin E and gum acacia
were used in accordance with the recommended usage levels by the manufacturer. Table
2 shows the concentrations of the various components used:

Table 2 Concentrations of various constituents of the drink

 

Component Sugar Citric Acid Vitamin E Gum Acacia Water

Concentration (g/L) 84.55 1.06 2 0.66 Balance

 

35

This composition was tested for stability for benzaldehyde. The procedure and
results of the stability test are presented in appendix C. The final drink was prepared by
adding gum in water to give 6.6% gum-water solution. Benzaldehyde was added to the
concentrated gum-water solution. The solution was stirred at ambient temperature until
all benzaldehyde was dissolved and individual droplets were not visible. Citric acid,
sugar and vitamin E were added and solution was stirred for some more time. The
solution was then brought to the correct concentration by diluting it. The surroundings
were sanitized by spraying 0.05% Sodium Hypochlorite solution. For the preliminary
testing, the product was prepared under room environment. For the second phase of
testing, the preparation was carried out in nitrogenous environment inside a small

chamber to avoid possibility of oxidation of benzaldehyde.

3.2.2 Consumer Panel Benzaldehyde threshold

Threshold of benzaldehyde perception in this simulated drink system was found
in accordance with ASTM E 679—04 [76] using a 5-series 3-AFC test. The test was run on
SIMS® 2000 software provided by Sensory Computer Systems, Morristown, NJ in the
Sensory Laboratory, Department of Food Science, MSU. For each of the 5 concentration
steps, the consumers were served with approximately 15 mL of three samples in 60 ml
identical cups coded with unique 3-digit codes. Out of the three samples, two were blank
drink samples containing no benzaldehyde while the third sample had specific
benzaldehyde concentration. Samples were provided in random sequence. Concentrations
used were 4.9, 14.7, 44.1, 132.3 and 396.9 ppm. The threshold study used 30 consumer
panelists that mainly were students, faculty and staff members of Michigan State

University. The promotional flyer is attached in appendix D and consent forms are

36

attached in appendix E. Panelists were asked to pick the odd sample out. The
questionnaire presented to the panelists is presented in the appendix F.l. To calculate
group threshold, first individual thresholds were calculated. Individual thresholds were
calculated to be the geometric mean of the last concentration step at which the respondent
could not identify the correct sample and the next higher concentration level. If a panelist
identified the correct sample at all concentration steps, then it was assumed that he would
have answered incorrectly at one step lower than the smallest concentration level. Group

threshold was calculated as geometric mean of individual thresholds.

Individual Threshold(lT) = (Cn me,l )05 (14)

Group Threshold (GT) = (IT] x 1T2 x...x [TN )(“N’ (15)

where C,I is the highest concentration step at which the panelist gave an incorrect answer

CM. is the next concentration step to Cn

N is the number of panelists

With this experimental design, it was found that the consumer threshold of
benzaldehyde in this particular design was 15.25 ppm. The table of responses is presented
in the appendix G. Based on the threshold value, as suggested in the earlier section, 45.75
ppm was used. Primary reasons were that this concentration lets enough allowance so that
consumer could taste the flavor at various time stages, and it is also close to 44.1 ppm,

the concentration in the threshold test at which people liked the product the most.

37

3.3 Molds

3.3.1 Designs

The mold designs were created using Rhinoceros® 3.0 provided by McNeel
Corporation, Seattle, WA. The semispherical and polyhedron containers had an overflow
volume of 100.2 mL and 81.3 mL respectively. The containers had equal surface area of
82.8 cmz. The semi-spherical and polyhedron containers are shown in the figures 3 & 4

below:

 

Figure 3 Semispherical Container

 

 

 

 

 

/ Draft Angle =45 °
19i/

Figure 4 Polyhedron Container

*All dimensions in centimeters

38

The command codes for designing the molds are listed in the appendix H. The
original file format of the mold designs corresponded to 3-dimensional design (.3dm)
which was converted to mesh or stereo lithographic file format (.stl). This file format was
used by the Deskproto® 4.0 provided by Delft Spline Systems, Utrecht, Netherland. In
Deskproto‘D 4.0, milling parameters like mill tool, step height, free movement height were
defined. These parameters are listed in the appendix 1. Parameters were optimized for
milling time estimate and fineness achieved. More milling steps would bring more
fineness meaning less sanding later on while coarse milling required more sanding.
Deskproto® 4.0 converted the design to a machine understandable file format (.ncm).
This file format is essentially a command text file understood by the mill software
(CPmill®) used by the Minitech minimill/2 desktop CNC machine. Both the mill machine
and the operating software were provided by Minitech Machinery Corporation, Norcross,

GA. Figures 5 & 6 show the semi-spherical and polyhedron molds.

 

Figure 5 Semispherical Mold

39

 

Figure 6 Polyhedron Mold

*All dimensions in centimeters

3.3.2 Production

Renwood procured from Huntsman Advanced Materials, East Lansing, MI was
cut in blocks of 12.7 x 12.7 x 5.08 cm. These blocks were used to make each mold.
Milling operation of the molds took approximately 6-7 hours for each mold. The molds
were then sanded using sandpapers of different fineness (#80 to #120). On the bottom
side of the molds, channels were created for the suction of air. Vacuum holes that
connected to the channels on the bottom side were drilled on the surface of the molds

with #80 drill bit (diameter: 0.3429 m) using air-drill.

3.4 Polypropylene Sheet Extrusion

Polypropylene resin (grade: PP4612 E2) was provided by ExxonMobil Chemical
company, Baytown, TX. Polypropylene sheets with a thickness of 12 i 5 mil were cast
extruded using Killion KLB 100 cast film extruder manufactured by Davis-Standard,
LLC, Pawcatuck, CT. The temperature of the three zones of the extruder was 400 °F (204

°C). The adapter and die temperatures were 410 °F (210 °C) and 390 °F (199 °C),

40

respectively. The melt temperature was 360 °F (182 °C). The screw speed was 75 rpm.
The temperature and speed of the chill roll were maintained at 75° F (24 °C) and 85 rpm
respectively. The rubber roll pressure was 30 psi (0.21 MPa), and the rewinder speed was

40 rpm.

3.5 Container Thermoforming

The containers were thermoformed on a Hydro-trim 1620 thermoforrner
manufactured by Hydrotrim Corp, Valley Cottage, NY. Prior to thermoforming, the
molds were sprayed with Ultra4 anti-stick food grade mold release agent manufactured
by Price-Driscoll Corporation, Waterford, CT. The polypropylene sheets were heated by
keeping upper and lower heating platens at 420 °F (216 °C) so as to allow for optimal
sagging. The average heating time was 45 seconds. The forming time was kept at 25
seconds, and the suction pressure was 40 psi (0.28 MPa). Because the starting
polypropylene sheets had a variable thickness profile, the formed containers were
visually inspected for appropriate formation, and then were accepted depending on the
thickness profile and surface smoothness. Excessively thin containers and containers with
very rough texture at sealing rims were rejected. The excess material was trimmed off the

good containers and the containers were then stored at 23 °C and 50% RH.

3.6 Lid Configuration

The lidding material was composed of two layers of foil based materials. As
shown in figure 7, one of them was a scalable foil that was supplied by Tolas Healthcare,
Feasterville, PA. This sealable foil had universal sealing material coated on 1.1 mil

Aluminum foil making it scalable with polypropylene and also providing for peelability.

41

The food side layer was a 3 mil adhesive backed foil provided by Uline, Waukegan, IL. It
was introduced in the lidding structure in order to prevent contact of sealant material with
the beverage. For semi-spherical containers, the inner material was 8 x 8 cm2 while the
sealing material was 12.5 x 10 cm2. For polyhedron containers, the inner adhesive backed

material was 8 x 11.25 can2 while the sealing material was 14 x 10 cm2.

'IIIIIIIIIIIIIIIIIV [II/[seathe F0"

.\\\\\ Sealant
wm Adheslve

— Adheslve backed Foll

Figure 7 Lid configuration

3.7 Shelf Life Studies

3. 7.1 Phase I: Pre-run Testing

3.7.1.1 Filling

The filling of the containers was carried out in a sanitized environment. The
sanitizer was produced by diluting Clorox (6% Sodium Hypochlorite) manufactured by
The Clorox Company, Oakland, CA so as to get at least 100 ppm active chlorine
evolution. This was achieved by adding at least 2 mL of Clorox in 1 L of clean water.
Sanitizer was sprayed all around the working area. The product was filled in such a way
to have minimal headspace without interfering with the sealing area. Average fill volume
achieved in semi-spherical and polyhedron containers was 87.7 i 2.2 mL and 71.1 i 3.3

mL respectively.

42

3.7.1.2 Sealing

The sealing of the containers was carried out in sanitized environment. For this
phase of the testing, it was carried out in room environment with air in the headspace.
The sealing of the containers was carried out in-house using a hydraulic pressure blister
heat sealer. The conditions applied were 275 ”F (135 °C) for 10 seconds at 100 psi (0.69
MPa). These conditions were found to be optimal in confirmation with ASTM D 3078
[77] in accordance to which a vacuum of 15 mm Hg (1999.8 Pa) was applied in dye leak
detector equipment for 2 minutes.

The containers filled with the product were placed in a fitting cavity. Cushioning
tape was applied at the edges of the cavity, so as to provide padding for good sealing. The
lid was placed on top of the container and sealed. The containers and the lids were
sanitized just before filling and sealing using Sodium Hypochlorite sanitizing solution. It
was made sure that the sanitizing solution dried before filling. After sealing the
containers and letting the seal cool down, the containers were checked visually for

possible leakages by inverting the container and shaking the product inside the container.

3.7.1.3 Package Storage

After sealing, the containers were placed on corrugated partitions prepared for the
purpose and placed under controlled conditions. For the first phase of testing, the
containers were kept under room environment at 23 °C and 50% RH. Packages were

stored in these conditions for 2 days and 7 days.

3.7.1.4 Quantification of Benzaldehyde

High Performance Liquid Chromatographic (HPLC) technique was used to

quantify benzaldehyde in the drink system. Waters 2695 HPLC system with dual channel

43

UV detector (model 2487) Nova-pak® C18 column (35 mm x 150 mm x 0.4 pm)
manufactured by Waters Corporation, Milford, Massachusetts was used. HPLC grade
water and acetonitrile were used as the solvents in ratio of 65: 35. Table 3 lists the flow

gradient used:

Table 3 Solvent flow gradient to determine benzaldehyde concentration

 

 

Start time End time Flow rate and condition

0 min 0.5 min 0 ml] min- 0.1 mL/ min, gradient
0.5 min 2.5 min 0.1 mL/ min, isocratic

2.5 min 3 min 0.1 mL/min— 0.25 mL/ min, gradient
3 min 7 min 0.25 mL/min, isocratic

7 min 9 min 0.25 mIJ min- 1 mUmin, gradient
9 min 12 min 1 mI/min, isocratic

12 min 15 min 1 mL/ min- 0 mI/min, gradient

 

The injection volume was 5 uL and the injection run time was 15 mins.
Benzaldehyde eluted at around 9.5 mins. Each vial was used for 3 injections. At the end
of each storage stage, samples were collected from 2 or more containers for analysis.
Calibration curve was prepared using the following concentrations of benzaldehyde in

methanol: 10.45, 41.8, 83.6, 167.2 ppm. The calibration curve is shown in the appendix J.

3.7.1.5 Sensory Testing with Consumer Panel

First stage of testing was carried out with consumer panel. This was used as a
precursor to the final sensory testing and to make appropriate corrections. The study
involved 40 panelists. This sensory test was conducted after 2 days and 7 days of storage
of the product as a precursor for more intensive tests to follow. Testing was carried out in
a standard triangle test format with no replication because of intense flavor of the
product. The test was run on SIMS® 2000 software provided by Sensory Computer
Systems, Morristown, NJ. The samples were randomized and presented in a balanced

incomplete block (BIB) design to the panelists. Each panelist was presented with three

45

samples out of which two were from one type of container and one was from other type
of container. Consumers were asked to pick the odd sample out. They had to guess, if
they could not perceive any difference. Questionnaire for sensory testing is attached in

the appendix F2.

3. 7.2 Phase II: Final Testing

3.7.2.1 Selection, Threshold and Training of Panelists

Based on the experience from the previous stage of testing, some corrections were
made in the overall methodology. It was decided that a trained panel needs to be
developed to be absolutely sure that the panelists are actually judging the benzaldehyde
strength in the product and nothing else. The panelists were selected to comply with
ASTM E 460-04 [78] which states that panelists should be screened in order to test for
sensory effects of packaging on the product. The panelists were selected for further
training. Rating test was used to select 15 panelists out of 37 panelists who participated in
the screening test. Panelists were presented with 2 reference samples and 5 unidentified
samples with different concentration of benzaldehyde, and they were asked to rate the
samples on a scale of 0—7. Reference samples represented 0 and 7 on that scale. At the
start of training of the trained panel, threshold study was conducted in confirmation with
ASTM E679-04 [76] to find out the threshold of trained panel towards benzaldehyde in
that particular drink system. The threshold of the panel was found to be 6.43 ppm. The
panelists were trained for 5 weeks with varying concentrations of benzaldehyde and
varying concentrations of citric acid and sugar that can affect perception of
benzaldehyde. Only after satisfactory improvement in the performance of the panel as

measured by the consistency and the proximity of perception in the group was observed,

46

actual testing was started. The panelist performance curves showing performance

enhancement are attached in appendix K.

3.7.2.2 Filling
The filling was carried out in the same way as for the pre-run testing in sanitized

environment.

3.7.2.3 Sealing

The sealing of the containers was carried out in sanitized environment. For this
phase of testing, it was carried out in nitrogenous environment. This allowed for the
purging of headspace Oz. Benzaldehyde is readily oxidized, removal of oxygen helped to
avoid any risk of benzaldehyde being oxidized. Besides the sealing environment, all other

parameters were exactly the same as with phase I testing.

3.7.2.4 Package Storage

After sealing, the containers were placed on corrugated partitions prepared for the
purpose and placed under controlled conditions. For this phase of testing, the containers
were kept under nitrogenous environment at 23 °C to avoid possible oxidation of
benzaldehyde. The enclosing chamber had a volume of approximately 21400 in3
(350,683 cc). The chamber was flushed with medical grade nitrogen at least once in 3
days with fresh nitrogen to remove permeated benzaldehyde and possibly diffused air.
Nitrogen stream was turned on every time when the chamber was opened to remove
containers for testing. Also, positive pressure was maintained inside the chamber. The

containers with product were stored in this condition for 3, 7, 14, 21 and 28 days.

47

3.7.2.5 Quantification of Benzaldehyde
Benzaldehyde quantification was carried out in exactly the same way as in Phase I

testing presented in section 3.7.2.4.

3.7.2.6 Sensory Testing with Trained Panel

Out of 15 panelists selected after screening, only 9 were used for actual testing for
logistical reasons. At least 8 of the panelists were present at each stage of the testing. The
panelists were presented with one sample from each type of container and a control
sample that did not have any benzaldehyde in it. Control sample was used as a reference
sample and the panelists were asked to rate the other two samples on a scale of 0-7, with
7 being strongest and representative of a benzaldehyde concentration approximately 50
ppm. The questionnaire is attached in the appendix F.3. They were allowed to rate the

samples equally if they perceived no difference in the samples.
3.8 Container Properties

3.8.1 Thickness Profile

Thickness of the containers was measured at 5 different spots on each container.
From each type, 5 containers were tested. Magnamike 8500 thickness tester
manufactured by Panametrics- NDT and provided by Olympus NDT Inc., Waltham, MA.
The probe was first calibrated with manufacturer’s instructions and steel ball size used

was 1/8”.

3.8.2 Water Vapor Transmission Rate

Water Vapor Transmission Rates (WVTR) of the containers were measured in

accordance to ASTM F 1249-06 [79] using a MoCon Perrnatran® 3/33 manufactured by

48

Modern Controls, Inc., Minneapolis, MN at 37.8 °C and 100% RH. Three samples of
each of the two types of containers were tested and the results were compared
statistically. The containers were sealed on a metal platform using Loctite Extreme
Repair adhesive manufactured by Henkel Corporation, OH. Metal platform was
connected to Permatran system with copper tubing to carry N2 into the container and back
to the detector. This system was sealed in a polyethylene pouch along with wet sponges
used to produce 100% RH inside. The test was continued until at least 10 steady state
points (<5% WVTR variation) were achieved and the average of those last 10 points were
taken to calculate the steady state permeation rate. Figure 8 explains the test set-up for

WVTR testing.

 

LDPE
Pouch

 

 

 

    

N2 to detector

Figure 8 Water Vapor Transmission Rate test setup

3.8.3 Oxygen Transmission Rate
Oxygen Transmission Rate (OTR) of the containers was determined in according

to ASTM D 3985-05 [80] using a Illinois Instruments 8000 series oxygen analyzer,

49

manufactured by Illinois Instruments, J ohnsburg, IL. Three containers of each type were
tested at 23 °C and 0% RH. The containers were sealed on a metal platform with Loctite
Extreme Repair Adhesive provided by Henkel Corporation, OH. The metal platform was
connected to the oxygen analyzer with copper tubing carrying N2 in and out of the
containers. The containers were tested with air as the permeant gas and the results were
compensated to 100% oxygen. The test was continued until at least 10 steady state points
(6% OTR variation) were achieved and the average of those last 10 points were used to
calculate the steady state permeation rate. Figure 9 explains the test set-up for OTR
testing.

Room air as permeant

   
   

f

Figure 9 Oxygen Transmission Rate test setup
3.8.4 Crystallinity
Crystallinity of the containers was determined using TA Instruments Q100

Differential Scanning Calorimeter manufactured by TA instruments, New Castle, DE.

50

The temperature calibration of equipment was performed in accordance with ASTM 13
967-03 [81] and the heat flow calibration was performed in accordance with ASTM E
968-02 [82]. The enthalpies of fusion and crystallinity were measured and calculated in
accordance with ASTM D 3418-03 [83]. Three containers of each type were tested. For
polyhedron containers, four samples were extracted from the walls and two samples were
extracted from the comers of the each polyhedron container. Four samples were cut from
each semi-spherical container. All samples weighed between 5-8 mg. Only one heating
cycle was used to measure the crystallinity. Samples were heated from 40 °C— 180 °C at a

rate of 10 °C/min.

3.9 Statistical Techniques

Statistical techniques were applied to design the experiments and later on analyze
the results. A type I error rate (or) of 5% was applied to all the experiments. For the phase
I sensory test involving consumer panel, number of consumers (40) were calculated
based on type I error rate (a) of 5%, power of test (l-B) of 80% and assumed proportion
of distinguishers (pd) of 30% [84]. To compare the variance and mean values of WVTR,
OTR and BATR of containers F-test and t-Test were applied respectively. F—test and t-
Test were also utilized to compare the ratings of the product from the two types of
containers as adjudged by the trained panelists. Tukey’s correction was applied to t-tests
when comparing crystallinity in different parts of the containers. The type I error rate was
divided by the number of possible comparisons and p—value was compared to this
modified type I error rate. The statistical tests were conducted using Microsoft Excel®

provided by Microsoft Corporation, Redmond, WA.

51

Chapter 4

Results

The results are presented in the following sections with container properties
coming first and shelf life studies coming after that. Finally, discussion is presented to
suggest the correlation between various factors and other possible causes of the results
and variability if any. Statistical inference is included with all the results with a type I

error rate of 5%.
4.1 Container Properties

4.1.1 Thickness

Thickness distribution of the two containers is profiled in figure 10 and figure 11.

8.91- 1.9 mil

    

10.81" 3.8 mil

11.9: 2.4 mil

+ .
12.2: 3.1 mil 10'4‘ 3‘1”“1

Figure 10 Thickness distribution of semispherical container
*1 mi]: 25.4 um

52

13.3: 2.9 mil

14,1: 23 mil 13.4: 2.3 mil

   
 

12: 2.9 mi]

12.3: 2.7 mil
Figure 11 Thickness distribution of polyhedron container
*1 mil: 25.4 pm
The overall thickness of semispherical containers was 10.7:I: 3 mil while that of

polyhedron containers was 13:1: 2.5 mil. Thickness of polyhedron containers was

statistically higher than that of semispherical containers at a type I error rate of 0.05.

53

4.1.2 Water Vapor Transmission Rate (WVTR)
The WVTR of semi-spherical and polyhedron containers is shown in Figure 12

 

0.025 weer 74—m—4

0.020 -

Hg,

 

0.015 -

0.010 -

WVTR (gms/pkg-day)

0.005 -

 

 

 

 

 

 

0.000
Semispherical Polyhedron

Figure 12 Water Vapor Transmission Rate (gms/ pkg-day) of Containers

*At 37.8 °C and 100% RH. “Different letters over the bars indicate statistically significant difference at
(1:005

WVTR of the semi-spherical container was 0.017 10.001 gms/ pkg-day and that
of polyhedron containers was 002110.002 gms/ pkg-day at 37.8 °C and 100% RH. The
WVTR of semi-spherical containers was 21% less than polyhedron containers. It is
visible that both polypropylene containers have very low water vapor transmission rate
but even at such low permeation rates, the differences between the two containers were
statistically different. One implication of the low permeation rates is that at 23 °C and
50% RH (actual storage conditions for shelf life studies), the water loss from the
simulated drink would be quite low. This will have very small or negligible effect on

changing the overall flavor balance of the product during the storage. Thus, the results in

54

shelf life studies are affected in a very limited manner because of the difference in
WVTR. But the above results show that geometry affects the WVTR in a significant
manner and for the packages with higher permeation rates, this can play critical role. The

moisture sensitiveness of the product would also be a factor to consider in such scenario.

4.1.3 Oxygen Transmission Rate (OTR)

Figure 13 shows the OTR values for semispherical and polyhedron containers.

 

 

 

 

 

 

 

 

8 R, “1- _:
b
6 .
5
l? a
03
8
n:
'—
O
2 .
0 I
Semispherical Polyhedron

Figure 13 Oxygen Transmission Rate (cc/ pkg-day) of Containers
*At 23 °C and 100% 02 concentration. "Different letters over the bars indicate statistically significant
difference at (1:005

OTR measured with air as permeant was compensated to 100% oxygen
concentration. OTR of polyhedron containers compensated to 100% oxygen was found to
be 6.81:1.18 cc/ pkg-day and that of semi-spherical containers was 4.22:0.52 cc/ pkg-
day at 23 °C. Thus, the OTR of semi—spherical containers was 60% less than that of
polyhedron containers. These results demonstrate that the geometry of the container

affects the oxygen transmission rate significantly. Benzaldehyde is known to readily

55

oxidize at room temperature. Since the product was stored in nitrogenous environment
for the phase H studies, these effects were minimized for this phase. However, there is a
possibility that this could have affected the phase I study, both of which are explained
later. Since, the oxygen permeability of containers is affected by the geometry, it
becomes essential to understand the phenomenon in order to improve the barrier. In
commercial environment, products are usually not stored under nitrogen, and besides the
flavor compounds, other components of food systems like lipids and vitamins also tend to

oxidize.

4.1.4 Crystallinity

In earlier studies it has been shown that the extent of crystallinity affects the
barrier properties of the films against organic compounds[4]. Thus, to see the difference
in crystallinity, DSC tests were run on the polymer extracted from different locations in
each container. Crystallinity of polypropylene in semi-spherical containers was found to
be 30.7:l.1%. Crystallinity in walls of polyhedron containers was 30.3:0.5% and in the
corners of the polyhedron containers 29.3:0.9%. A graphical comparison is shown in the

figure 14.

56

35-

 

30- _L

25-

Crystallinity, %

 

 

 

 

o l
Semi Walls Poly Walls Poly Corners

 

Figure 14 Crystallinity in different regions of containers

*Different letters indicate statistically significant difference at ot=0.05

There is a statistically significant difference in the average values of crystallinity
in walls of polyhedron containers and the polyhedron comers. Semi-spherical container
walls are not statistically different as compared to corners or walls of polyhedron
containers with a type I error rate of 0.05. Thus, it is difficult to say that crystallinity is
playing a role for the difference found in the barrier properties of the two types of
containers. It is suggested that two other factors could have played a synergistic role in
determining the difference in the barrier properties. Differential orientation possibly
caused by difference in geometries can change the barrier properties [13, 85]. The other
factor of significance might be the distribution of crystallites. Drifting and realignment of
crystals due to variable stresses caused can make the permeant path relatively more
tortuous, ultimately affecting the barrier properties of the package. This study did not

include measurement of these properties but these might be important avenues to look

57

into for future studies to enhance understanding of the changes caused by the differential

geometry.

4.2 Shelf Life Studies

Sensory and HPLC studies could establish that the differences in barrier
properties discussed earlier might have a significant effect on the perceptibility of the
cherry flavor by the consumers, if the differences are large enough to be measurable both
by passive and human instruments. The above mentioned measured and suggested
variables were affecting the flavor of the product as much as to bring it in the spectrum of

practical significance.

4.2.1 Sensory Results

4.2.1.1 Consumer Panel

Consumer panel was used as a precursor for the final testing. At this stage, the
product was sealed and stored under room conditions at 23 °C and 50% RH. Tested with
triangle test, the consumers could see the difference in the products stored in two
different types of containers as soon as 7 days. But it could not lead to a conclusive result
because these results could not be verified with HPLC analysis which showed no
presence of benzaldehyde in the product. This led to believe that the consumers are
seeing difference in a different constituent of product because all of the benzaldehyde had
probably oxidized completely or permeated completely. Though it did not undermine the
fact that the two types of containers were different in terms of barrier properties, but with
this, it can not be concluded that their barrier properties against the permeation of

benzaldehyde were different. As discussed earlier the Oxygen transmission rates of the

58

two containers were different and that might have played a role because the containers
were stored in the room environment at this stage and not in nitrogenous environment.
The learnings taken from this stage were to develop a trained panel that would test only
for the differences in benzaldehyde presence and to store the product in a san-oxygen

environment

4.2.1.2 Trained Panel

Presented here are the results obtained with the trained panel that was utilized
after making appropriate corrections to the overall methodology and training the panel.
Appendix K presents the performance improvement of the trained panel in the 5 weeks of
training. Figure 15 shows the average rating by the trained panel to the products stored in
the two types of containers. A rating of 0 represents no perceptible benzaldehyde
presence in the product and a rating of 7 corresponds to a concentration of 45 ppm or

higher.

59

 

 

 

7 -3 W W i , , —- j
_ i
6 a§ ‘
a 1 a
5 - a I I a5 t
i I I .
a l
4 - a I l
E b
g E
3 ~ b
2 " I l
b l
1 .
o I 1 l l l
3 7 14 21 28
Time, days

Figure 15 Benzaldehyde concentration rating

*D‘ifferent letters over the dots indicate statistically significant difference at 0:005 (o) semispherical, ( " )
polyhedron. §Average rating for five panelists from two containers.

The panel could detect differences significantly after day 21 and results were
verified with day 28 samples. The average ratings for the products from the semispherical
and polyhedron containers were 5 :l: 0.13 and 3.38 i 0.13 at the end of 21 days and 4.75 :l:

0.11 and 2 i 0.07 at the end of 28 days, respectively.

4.2.2 Benzaldehyde Concentration Determination

During the testing with the trained panel, the product was simultaneously tested
with HPLC, whose results are presented here. Figure 16 shows the residual concentration
of benzaldehyde in the product stored in the two types of containers after different time

storages.

6O

50am ,2 _1 ~ A4

 

 

 

 

 

 

 

E
a 40 -
c“
.9
E
‘3'; 30 -
O
C
O
0
(D
a 20 ~
.2
(D
2
(U
E‘
a, 10 -
m
0 1 l 1 l l
0 3 7 14 21 28
Time, days

Figure 16 Residual Benzaldehyde concentration of product in two containers

*Different letters over the dots indicate statistically significant difference at 0L=0.05, (o) semispherical, (it )
polyhedron. §Average of six injections from two containers.

The benzaldehyde strength was found to be significantly more in product stored in
semispherical containers than in polyhedron containers at the end of 21 and 28 days. The
benzaldehyde concentration in semispherical containers was 40.39 i 0.34 ppm and 38.84
i 0.76 at the end of 21 and 28 days respectively while it was 19.06 i 9.95 and 0.05 i
0.025 in the polyhedron containers.

As shown in Figure 17, the sensory panel results and benzaldehyde concentration
measured with HPLC technique are in good agreement with each other and follow the

same trend.

61

A

 

Panel Rating

 

 

 

0 3 7 14 21 28
Days

Figure 17 Benzaldehyde panel rating and concentration measured by HPLC
(A) semispherical concentration, (A) polyhedron concentration,

(I) semispherical rating, (I) polyhedron rating

8

Benzaldehyde Concentration, ppm

 

S

8

 

:50

u—t
O

This confirms that the two tools devised to study benzaldehyde presence in the

considered an aberration caused by outliers.

‘

drink were individually correct and collectively veritable. One deviation observed was on
day 14. On day 14, it was found that product from polyhedron container was rated higher
than semispherical container. But in contrast to the ratings given by panelists,
benzaldehyde strength in the products from two containers was not statistically different,
though the trend seemed to follow. Thus, this difference in the ratings can be considered

an aberration caused by external reasons. But it could not be repeated, and it was

This study confirms that there is a relationship between package configuration

62

and barrier properties of containers causing difference in the sensory flavor perception.

This goes on to demonstrate that with application of design principles in appropriate
manner, the barrier properties can be customized and the shelf life can be improved

whenever mass transfer is concerned.

63

Chapter 5

Conclusions

5.1 Outcomes from the study

Restating the null and alternate hypothesis
Ho:

Two polymeric containers made with same material and same surface area but in
different shapes have the same barrier properties and the product stored will have the
same sensory flavor perception.

Ha:

Two polymeric containers made with same material and same surface area but in
different shapes have different barrier properties and the product stored will have
different sensory flavor perception.

This work established that the null hypothesis can be safely rejected with a type I
error rate of 0.05. Thus, the shape of the container affects its barrier properties enough so
as to cause difference in sensory flavor perception by the consumer.

It was demonstrated that the water vapor transmission rates and oxygen
transmission rates are significantly different for the two differently shaped containers
with a type I error rate of 0.05. Average WVT R of semi-spherical containers was 20%
less than that of polyhedron containers, and average OTR of semi-spherical containers
was 60% less than that of polyhedron containers.

The crystallinity in the walls of polyhedron container were found to be

statistically different than the crystallinity in the comers of the container. The

crystallinity in the semispherical container walls was different than the crystallinity in the
comers of polyhedron container. Though the difference was statistically significant, in
absolute terms, the walls of both semispherical and polyhedron containers did not have a
considerable difference in crystallinity in comparison to crystallinity in the comers of the
polyhedron container. In this scenario, the difference in barrier properties may be
attributed to the possible difference of orientation in the two shapes and the distribution
of crystallites.

The difference in the barrier properties of the two containers against
benzaldehyde was found to be significantly affecting the sensory perception of the
product stored in as soon as 21 days. The residual benzaldehyde in the polyhedron
containers was less than that in the semi-spherical container as established by HPLC
analysis of the product. Thus, sensory and HPLC results were in agreement establishing
that the barrier properties are significantly different to affect the sensory perception of the

product stored.

5.2 Recommendations for Future work

Though this work established that the container shape affects the barrier
properties significantly, there is still scope of further work in this direction. Based on the
limitations and results of this work, a need to follow up this study is recognized in
following areas:

1. Orientation, distribution of crystallites and many other factors that define the
morphology of a polymer need to be studied in such differently shaped containers as used
in this study in order to precisely determine the reasons of differences in the barrier

performance.

65

2. A similar study can be conducted with other kinds of processing like injection
molded and blow molded components. This will help understand if the specific processes
bring more differences in the shapes and other processes do not. In such a case, these
particular processes would need to be given special attention in a commercial production
facility.

3. Another avenue for further research would be with other kinds of higher and
lower barrier materials than PP and with other kinds of flavor compounds, individually
and a mixture of the flavor compounds. This will help in demonstrating if this
phenomenon is limited to particular polymer, permeant pairs or if the phenomenon is
more general. When using more than one flavor compound, such a study will relate better

to the commercial situation where the product usually have many flavor compounds.

66

Appendices

67

Appendix A Approval from Institutional Review Board (IRB)

Punks] Kumar

From: "lRB" <lRB@ora.msu.edu>

Date: Thursday, July 05, 2007 10:57 AM

To: "Pankaj Kumar“ <kumarpan@msu.edu>
Subject; RE: About IRB# X05-337

Dear Pankaj Kumar,

Thank you for your inquiry regarding your project titled “Correlation between Sensory Aroma Perception and
Package System Configurations.” Our records indicate you have been listed as an additional investigator since
this projects inception. Please give this correspondence to the graduate office. if this does not suffice and you
need an official letter please let me know.

Thank you.

Judy McMillan

lRB Staff

Human Research Protection Programs

Biomedical and Health Institutional Review Board (BIRB)
Community Research institutional Review Board (CRIRB)
Social Sciencei'BehavioraI.-’Education institutional Review Board (SIRB)
Office of Regulatory Affairs

202 Olds Hall

Michigan State University

East Lansing, MI 48824-1046

Phone: (517) 355-2180

Fax: (517) 4324503

Email: irb@msu.edu

Website: wwwhumanresearchmsuedu

68

MICHIGAN STATE Initial IRB

 

u N I v E R s I T Y Application
Am 14. 2005 Approval
To: Rafael Auras

140 Packaging
Re: leases-337 Category: EXEMPT 1-6

Approval Date: April 14. 2005
Expiration Date: April 13, 2006

Title: CORRELATION BETWEEN SENSORY AROMA PERCEPTION AND PACKAGE SYSTEM
CONFIGURATIONS .

The University Committeeon Research Involving Human Subiects (UCRIHS) hascornpleted their review of
yourprolect. Iampleesedtoadvieeyouthatyourprolecthssbsenspprowd.

Thecommiueehasfoundthetyomreeearchproiectisappropnatehdeslgn,protectstherightsendwelereot
bananubiects.mdmeetsfinmqukemmbdMSUsFedaelWideAssmmdtheFedwflGddeines
(45CFR46and21CFRPartSO). Theprotectiondhumariwbiedslnreseeroliisapamershipbetweenthe
lRBandtheinvestigalors. We looklorward toworkingwtthyou aswebothfulfiilourreaponsibiiltles.

Renewals: UCRll-iSapprovetisvalidmtlltheexpiratlondelelistedebove. ityouarecontlnungyour project.
ywmustsuhnimAppflcedmlorReMepplicefimatleaetmemuflhbebmexpkafion. lltheprojectls
completed, pleasesubmitanApptlcstlonforPenmnentCIosm.

Revisions: UCRIHSnustreviewanychangeelntheproiect,priorto inttiationofmechanoe. Please submitan
WWWontohaveyouchangesreviewed. Itchengesaremedeatthetimeotrenewelmbese
Include an Application IorRevlsIonwlth the renewal application.

Problems: ltIssuesshouldeneeduringtheconductottheresearch. suchasmenticipatedprobiems. adverse
events, or any problem that may increase the risk to the lumen subjects. notify UCRIHS promptly. Forms are
avallebietoreport these Issues.

PleaseusethelRBnumberllstedaboveonenytonnssubmittedwhichrolatetothlsproiect.oronany
corremndencewithUCRIHS.

Good luck in your research. if we can be of further assistance. please contact us at 517—355-2180 or via email

at QCRIH§Qmfledu Thank you for your cooperation.
Sincerely.

age/jg»:

Peter Veslienko. PhD.
UCRIHS Chair

c: Janice HARTE
114 Trout FSHN Bldg.

69

School of Packaging. Michigan State University
ansumer Panel Consent {form

Evaluation of Packaging Cherry/Fruit Drink Applications

Dear Participant:

Michigan State University researchers are investigating packaged fruit/cherry flavored drinks and
asking that panelists participate in a storage study of packaged cherry drink stored in plastic packages
(Poly Propylene). We are asking for volunteers, over the age of 18, taste these samples. if you have a known food
allergy to any of the following possible FDA approved food ingredients: sugar, citric acid and artificial fruit
flavors, stabilizers/gums please do not volunteer for this study. Packaging materials are approved by the
FDA for contact with food. Each evaluation will take about 10 minutes after you receive your samples.
You will be given a coupon or food neat worth $2 or less as an appreciation for your participation and
completion of the questionnaires.

Though none is anticipated, if you have a problem upon sniffing and tasting the samples, notify the on-
site sensory evaluation coordinator and/or principle investiptor immediately. You will be released from
participating in this study. Please note if you are injured as a result of your participation in this research
project, Michigan State University will assist you in obtaining emergency care, if necessary, for your
research related injuries. If you have insurance for medical care, your insurance carrier will be billed in
the ordinary manner. As with any medical insurance, any costs that are not covered or in excess of
whatever are paid by your insurance, including deductibles, will be your responsibility. Financial
compensation for lost wages; disability, pain or discomfort is not available. This does not mean that you
are giving up any legal rights you may have. Your response is confidential and we will protect your
confidentiality to the full extent of the law.

If you have any questions during your reading this consent form, or during or after your participation,
please do not hesitate to contact the on-site sensory evaluation leader and/or the principal investigator, Dr.
Rafael Auras, via phone at 517-432-3254, or regular mail at 130 Packaging Building, East Lansing MI
48824. He can also be reached via email at aurasraf@gsu.edg for any inquiry you might have related to
your participation in the study. In case you have questions or concerns about your role and rights as a
research participant, please feel free to contact Dr. Peter Vasilcnko, Michigan State Universitfs Chair of
University Committee on Research involving Human Subject (UCRIHS) by phone: (517) 355-2180, fax:
(517) 432-4503,cmail: ucrihsfizzmsuedg or regular mail: 202 Olds Hall, East Lansing, MI 48824.

I voluntarily agree to participate in the study.

SIGN ED DATE

 

UCRIHS APPROVAL FOR
THiS protect EXPIRES:

APRIBZDOS

“gaunt” a“
DATETO I

70

MICHIGAN STATE

UNIVERSITY

 

C‘)
.lr‘I-r .'-'7 lb .7- a :‘flfliilll
April 6, 2006 " 'I.,,|,'l. 29".: 21:1. " .__'lll i'.
To: Rafael Auras I it .. . lull
14o Packaging g...
’62. """ IrI I'5I. ii";
Re: lRB # X05337 Category: Exempt
Approval Date: March 29, 2006

Title: Correlation between Sensory Aroma Perception and Package System Configurations

The Biomedical and Heath institutional Review Board (BIRB) has completed their review of your
renewal application. lam pleased to advise you that your project has been continued as
exempt and approved under our new exempt policy.

The lRB has found that your research project meets the criteria for exempt status and the criteria
for the protection of human subjects in exempt research. Under our exempt policy the
Principal investigator assumes the responsibilities for the protection of human subjects in
this project as outlined in the assurance letter and exempt educational material. The other; has
received your signed assurance for exempt research. A copy of this signed agreement is
appended for your information and records.

Renewals: Exempt protocols do mt need to be renewed. It the project ls complamd, please
submit an Application for Permanent Closure.

Revisions: Exempt protocols do _ngt require revisions. However, if changes are made to a
protocol that may no longer meet the exempt criteria. a new initial application will be required.

Problem: If issues should arlse during the conduct of the research. such as unanticipated
problems, adverse events, or any problem that may increase the risk to the human subjects and
change the category of review. notlfy BIRB promptly. Any complaints from participants regarding
the risk and benefits of the project must be reputed to the BIRB.

Follow-up: if your exempt project is not completed and closed after three raersygg ,BlRB will
contact you regarding the status of the project and to verify that no changes have occurred that
may affect exempt status.

Please use the [RB number listed above on any forms submitted which relate to this project, or on
any correspondence with UCRIHS.

Good luck in your research. If we can be of further assistance, please contact us at 517-355-
2180 or via email at immsuadu. Thank you for your cooperation.

Sincerely

Mi.

Peter Vasilenko, PhD.

Biomedical and Health Institutional Review Board Chair

C: Janice Harte
114 Trout FlSN Bldg.

71

 

NamcofPrincipallnvestigatorzltafaelAmas ENE ““30”
Title of Project Correlation between Sensory Aroma Perception and Package System Configtuations

[RB axes-3% X0§p337

The University Committee on Research Involving Human Subjects (UCRIHS) has deemed this project as caermt.
inaccordwiththe federal regulations fortheprotectionofhmnan subjects. AsaneotenptprojecttheIRB will not
be fin'therinvolvedwith thercvieworcontinuedreviewoftheproject, as longsstheprojectmaintainsthe
properties that make it exenpt

Since theIRB isnblongerinvolved inthe reviewandcontinuedreviewofthisproject. itisthe Principal
Investigator who assumes the responsibilities for the protection ofhuman subjects' in this project and
ensues ties the project is perfumed with integrity and within accepted ethical standards- particularly as
oufined by the Belmont Report (see exempt educational materials).

The Principal Investigator assumes theresponsibility for ensuringthat research subjects beinformed of the
researchtlroughadocumentedormidoctnnentedccnaentrmcess, ifappropriate.

The Principd Investigator mes theresponsibility tomaintain confidentiality of the subjectaandthe
demandnnintamtheprivacyoffliembjectsandprotectionofthedatathroughappropriatemeans. Ifdata

iammymoimdieinvesfigatomwinrmkemanenptwidenfifyanyindividmls.

The Principal Investigator assumes responsibility for asaining that human subjects will be selected
equitably, so that risks and benefits of research are justly dimbuwd.

The Principal Investigator assumes the responsibility tint co-investigators and otha members of the
research teamadhere totbeappropriate policies fortheprowctionoflannan subjects, maintain
confidentiality and privacy, and adhere to accepted ethical standards.

Ifthe Principal Investigator adds additional investigators to an exempt project, hdshe may inform UCRIHS
of the additions. This may be of particular importance to graduate students if the MSU Graduate School
requires proof of IRB approval.

Any complaints from participants regardingthe risks andbenefits of the project mustbe reported to
UCRII-IS. Any irrformation,1mexpectedoradverseeventsthatvvould increasetherisktohmnansubjects
andcausethecategoryofreviewtobeupgradedtonpeditedm-fiillkeviewrmstalsobereportedto
UCRIHS.

Since the Principal hvestigator and co-investigators are charged with human subject protection and
adhering to ethical principles in exempt research, it is Wm that investigators be trained in human
subject principles. The Principal Investigator and all members of the research team are required to
complete MSU IRB educational requirements or equivalent.

Any change in the project that may raise the project from exempt to an expedited or full review category
must be pmnted to UCRIHS. Ifthere is any question about a change in the project, the Principal
lnvesti gator should consult the Chair of UCRIHS. Failure to submit changes that misc the project out of
the exempt category will be considered non-compliance and will be subject to investigation and action by
UCRIHS.

By signing below, the Principal Investigator assures that he/she will abide by the terms of this assurance and the
UCRIHS exempt policy.

jg » oslquec

Signm Principal Investigator Date

6/2 7/05

72

Appendix B Process Flows

B.1 Drink composition

i
i
l

V

 

Conceive Simulant System

 

 

 

i
Y
Find literature value for
threshold of Benzaldehyde

 

 

 

 

 

 

Identity appropriate
concentration steps for
threshold testing

 

 

 

 

 

Prepare the drink with
various concentration steps

 

 

 

 

 

Run Sensory Testing

 

 

 

 

 

Conduct analysis

 

 

 

 

 

Finalize appropriate
concentration levels for
Benzaldehyde

 

 

 

i
i

Figure B-i Process Flow for drink composition

73

B.2 Making containers

V Y

Find optimal conditions for
Cast Sheet Extrusion

i l
l

 

 

Define Design Considerations

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

v 5
Create 3-D Designs in i
Rhinoceros 3.0
Calculate required number of
containers

1
Define process parameters;
convert designs in machine

ready format
Prepare renwood blocks; mill Extrude Polypropylene
the molds Sheet

 

 

 

 

 

 

 

 

Drilling vacuum holes

Find optimal
thermoforming conditions

 

 

 

 

 

 

 

 

 

 

‘

 

Thermoform containers

l
+

Remove defective
containers

Figure B-ii Process Flow for making containers

 

 

 

 

 

 

 

74

Appendix C Stability testing of drink

The product was tested under N2 and 02 for stability and presence of
benzaldehyde. Product was stored under air and N2 headspace. For this purpose 60 ml
glass bottles were utilized and 6 ml of product was stored in these bottles. Half of the
bottles with each type of headspace were exposed to light and the other half were covered
with foil to protect from light. The samples were stored at room temperature. Comparison
between air and N2 headspace was carried out up to 4 days. The samples were tested with
HPLC technique described in the main body of the thesis. Starting concentration was
46.63 i: 0.45 ppm. The concentrations of benzaldehyde with air in headspace in dark and
light exposed conditions at the end of 4 days were 41.35 :1: 0.07 ppm and 41.10 :l: 0.09
ppm respectively. The concentration at same time with N2 in headspace and dark and
light exposed conditions were 39.54 :1: 0.51 and 36.77 ppm, respectively. This suggests
that there was not a considerable difference in the product stability with N2 or air
headspace since the degradation rate was not significant. It was also observed that the
initially the concentration went down faster, most probably owing to the evaporation of
benzaldehyde and then it became stagnant. To assess the effects of light further with N2
in headspace, the experiment was repeated with only N2 headspace for up to 21 days. At
the end of 18 days, the concentrations in the dark and light exposed conditions were
15.68 i 6.46 and 8.82 :l: 7.3 ppm, respectively. At the end of 21 days, these
concentrations were 20.24 i 8.5 and 9.02 :l: 0.96 ppm. No definitive effect could be
assessed on benzaldehyde concentration because of presence or absence of light with

these results.

75

Based on these experiments, where huge headspace was provided to the product,
and based on the concentrations seen after day 18 and day 21, it can be said that the
product is shelf stable at room conditions under both N2 and air headspace. In these
experiments, detection wavelengths used were 201 nm and 265 nm, and at these
wavelengths, most of the compounds can be seen. No additional peak was detected at any
of the stages, suggesting that all the loss of benzaldehyde was entirely due to evaporation
and not due to degradation. Evaporation of benzaldehyde was slow and thus, attaining
equilibrium between headspace and dissolved benzaldehyde was slow too. This is
confirmed when partial pressure of benzaldehyde is calculated and compared using
Antoine equation and ideal gas law at 23 °C. Partial pressure calculated from Antoine
equation (Pa) was 11.5 Pa. The ideal gas equation yielded a partial pressure (P,) of 2.43 x
107 Pa. As is apparent that Pi>>>Pa, the evaporation of benzaldehyde would keep on
taking place for very long time since the driving force is available. This also means that

equilibrium between headspace and liquid phase will take a long time to achieve.

76

Appendix D Sensory promotional flyer

Participate in a quality evaluation of
Packaged Fruit/Cherry Drink!

J c‘ >
.,
-

 

We are seeking panelists to participate in a consumer research study

that will involve evaluation of packaged cherry drink. Each evaluation
will take about 10 minutes or less.

where? 102 Trout FSHN, MSU
When? March 16, 2007 9 AM- 1 PM

What? Packaged Fruit/cherry Drink packaged in FDA
approved packages

Food treat coupons will be provided following tasting
the drink and completion of the survey.

Any questions? Contact (517-432-3254), Dr. Rafael Auras in
the School of Packaging, aurasraf@msu.edu

77

Appendix E Consent Forms

E.l Consumer Panel Consent form

Evaluation of Packaging Configuration for Packaged Cherry Drink
Applications

Dear Participant:

Michigan State University researchers are investigating packaged cherry drink
and asking that panelists participate in a storage study of packaged cherry drink stored in
plastic packages (Poly Propylene). These materials are approved by the FDA for contact
with food. Each evaluation will take about 10 minutes or less after you receive your
samples. You will be given a coupon or food treat worth $2 or less as an appreciation for
your participation and completion of the questionnaires.

Though none is anticipated, if you have a problem upon sniffing and tasting the
samples, notify the on-site sensory evaluation coordinator and/or principle investigator
immediately. You will be released from participating in this study. Please note if you are
injured as a result of your participation in this research project, Michigan State University
will assist you in obtaining emergency care, if necessary, for your research related
injuries. If you have insurance for medical care, your insurance carrier will be billed in
the ordinary manner. As with any medical insurance, any costs that are not covered or in
excess of whatever are paid by your insurance, including deductibles, will be your
responsibility. Financial compensation for lost wages; disability, pain or discomfort is
not available. This does not mean that you are giving up any legal rights you may have.
Your response is confidential and we will protect your confidentiality to the full extent of

the law.

78

If you have any questions during your reading this consent form, or during or
after your participation, please do not hesitate to contact the on-site sensory evaluation
leader and/or the principle investigator, Dr. Rafael Auras, via phone at 517-432-3254, or
regular mail at 130 Packaging Building, East Lansing, MI 48824. He can also be reached
via email at gmsrafébmsuedu for any inquiry you might have related to your
participation in the study. In case you have questions or concerns about your role and
rights as a research participant, please feel free to contact Dr. Peter Vasilenko, Ph.D.,
Director of Human Research Protections, by phone: (517) 355-2180, fax: (517) 432-

4503, email: irb@msu.edu or regular mail: 202 Olds Hall, East Lansing, MI 48824-1047.

I voluntarily agree to participate in the study.

SIGNED DATE

79

E.2 Trained Panel Consent Form
(Page 1)

Evaluation of Packaging Configuration for Packaged Cherry Drink
Applications

Dear Participant:

Before you decide to sign this consent form and continue to participate in this
study, please read this document carefully for the information related to the study,
ingredients, packaging material and procedures used in the study. Potential risks and
benefits from your study, assurance of your privacy and your rights as a human subject in
our study are also listed.

If you have any questions during your reading this consent form, or during or
after your participation, please do not hesitate to contact the on—site sensory evaluation
leader and/or the principle investigator, Dr. Rafael Auras, via phone at 517—432-3254, or
regular mail at 130 Packaging Building, East Lansing, MI 48824. He can also be reached
via email at gurasraf@msu.edu for any inquiry you might have related to your
participation in the study. In case you have questions or concerns about your role and
rights as a research participant, please feel free to contact Dr. Peter Vasilenko, Ph.D.,
Director of Human Research Protections, by phone: (517) 355-2180, fax: (517) 432-
4503, email: irb@msu.edu or regular mail: 202 Olds Hall, East Lansing, MI 48824-1047.

PLEASE NOTE THAT UPON YOUR SIGNING THIS CONSENT FORM,
YOU VOLUNTARILY AGREE TO PARTICIPATE IN THIS STUDY. YOUR
SIGNATURES INDICATE YOU HAVE READ ALL THE INFORMATION

PROVIDED IN THIS CONSENT FORM AND THAT YOU HAVE HAD AN

80

ADEQUATE OPPORTUNITY TO DISCUSS THIS STUDY WITH THE PRINCIPLE
INVESTIGATOR AND HAVE HAD ALL YOUR QUESTIONS ANSWERED TO
YOUR SATISFACTION. A COPY OF THIS CONSENT FORM WITH YOUR
SIGNATURE FOR YOUR RECORDS CAN BE PROVIDED UPON YOUR
REQUEST.

I voluntarily agree to participate in the study.

SIGNED DATE

 

81

Trained Panel Consent Form continued (Page 2)
Invitation to Participate: You are invited to participate in the study that assesses
the effects of container configurations on the flavor barrier properties of packages for a

simulated cherry drink.

Purpose of the study: We are investigating packaged cherry drink and asking
that panelists participate in a storage study of packaged cherry drink stored in plastic
packages (Poly Propylene). This study would help establish if the container shape affects

the barrier properties of the package in preserving the flavor of cherry drink.

Procedure of the study: Each panelist would be served simulated cherry drink
stored in different containers and would be asked to rank the sample after tasting it. Each
sample would be coded with a unique 3—digit code. We are asking that panelists
participate in this study that will be conducted over a 2-month period. Training will
consist of approximately 3—4 sessions of 30-60 minutes. After training, approximately 6

evaluations will be scheduled. Evaluations should last about 30 minutes or less.

Sample Preparation: All the materials used in this study are approved by the
FDA for contact with food. The product would be stored in the containers made with
food grade plastic (polypropylene) at ambient conditions. Following is the composition of
the product:

Water, Sugar, Citric Acid (for tartness), Vitamin E, Gum, Benzaldehyde (Cherry

flavor)

82

Potential Risks: Since all the ingredients are FDA approved food grade materials
in this study, these samples pose no adverse health risk, providedthe subject has not been
identified as being susceptible to an allergic reaction to the previously listed sample ingredients.
Though none is anticipated, if you have a problem upon sniffing and tasting the
samples, notify the on-site sensory evaluation coordinator and/or principle investigator
immediately. You will be released from participating in this study. Please note if you are
injured as a result of your participation in this research project, Michigan State University
will assist you in obtaining emergency care, if necessary, for your research related
injuries. If you have insurance for medical care, your insurance carrier will be billed in
the ordinary manner. As with any medical insurance, any costs that are not covered or in
excess of whatever are paid by your insurance, including deductibles, will be your
responsibility. Financial compensation for lost wages; disability, pain or discomfort is
not available. This does not mean that you are giving up any legal rights you may have.
Your response is confidential and we will protect your confidentiality to the full extent of

the law.

Expected Benefits: This study would enable the researchers understand that the
phenomenon of flavor barrier better in formed containers. This in turn will help the food
packaging industry devise better strategy to preserve flavor of the product and elongate

sensory shelf life at an optimum cost.

Assurance of confidentiality: Any information obtained in connection with this
study that could be identified with you will be kept confidential by ensuring that all
consent forms and response sheets are securely stored. All data collected and analyzed

will be reported in an aggregate format that will not permit associating subjects with

83

specific responses or findings. Your privacy will be protected to the maximum extent
allowable by law.

Withdrawal from the study: Participation in this study is voluntary. Your
decision to refuse participation or discontinue participation during this study will be

honored promptly and unconditionally.

84

Appendix F Sensory Questionnaires

F.1 Questionnaire for measuring sensory threshold of benzaldehyde

 

3-AFC Threshold Test

 

Name Date

Sample: Packaged Cherry Drink # panelist number

 

Instructions

Five independent sets (A, B, C, D, E) of samples will be presented sequentially.
Each set has 3 different samples; out of which two are same, determine the odd one.
If no difference is apparent, you must guess.

Smell and taste samples from left to right.
You may spit the sample out after judging (foam cup is provided).

Which one is the odd one? (Check the box)

 

 

 

 

 

 

 

 

 

 

 

Set A D 328 El 481 U 869

Set B U» 298 El 966 El 543

Set C D 769 D 849 El 228

Set D E] 488 El 125 El 252

Set E El 747 El 885 El 526
COMMENTS:

 

 

85

F.2 Questionnaire for consumer panel sensory testing

 

Triangle Test

 

Name Date

Sample: Packaged Cherry Drink # panelist number

 

 

Instructions

There are 3 different samples; out of which two are same, determine the odd one.

If no difference is apparent, you must guess.

Smell and taste sampr from left to right. Rinse the pallet with crackers and] or
water between samples.

You may spit the sample out after judging (foam cup is provided).

Which one is the odd one? (Check the box)

 

 

 

Sample El 385 El 496 El 647

 

 

 

 

COMMENTS:

 

 

 

86

F.3 Trained panel rating questionnaire

 

Name: Date:

 

Type of Sample: Cherry (Benzaldehyde) Flavor

Characteristic Studied: Strength of Cherry Flavor in various packages after storage

 

Instructions:
1. Receive the samples and note each sample code in the next section according to

its position on the tray.

2. Taste the samples from left to right and note the degree of strength of cherry
flavor. Wait at least 30 seconds or more as appropriate between samples and rinse
palate as required.

3. The sample coded with the letter “C” is control sample without any concentration
of benzaldehyde. It is for your reference. You do not have to rate that sample in
this questionnaire.

4. Rate the samples according to the following key for the cherry flavor profile.

Slightly perceptible

UJN

Imperceptible

WC

Moderately perceptible Strongly Perceptible

LII-b
\10

 

Code:

 

 

 

Rating:

 

 

 

 

 

 

 

Comments:

87

Appendix G Panelist Response for Benzaldehyde threshold study

Table G.1 Number of correct responses for different benzaldehyde concentrations

 

 

Concentration Correct Responses
4.9 ppm 23
14.7 ppm 22
44.1 ppm 26
132.3 ppm 25
396.9 ppm 28

 

Individual Threshold was calculated by the finding geometric means of the
highest concentration that the panelist evaluated incorrectly and the next higher
concentration. If the panelist evaluated all samples correctly, then it was assumed that the
concentration one step lower than the least concentration (4.9 ppm/3: 1.5 ppm) would
yield an incorrect response and the geometric mean of 1.5 ppm and 4.9 ppm was taken as
individual threshold value. If the panelist gave an incorrect response at the highest served
concentration, then it was assumed that the next higher step (396.9 ppm x 3: 1190.7
ppm) would yield a correct response. In this case, the geometric mean of 396.9 ppm and
1190.7 ppm was treated as the individual threshold.

Group threshold was calculated by taking geometric mean of the individual
threshold values. It was found that the Group threshold for benzaldehyde in this drink

system was 15.25 ppm.

88

Appendix H Codes for creating designs in Rhinoceros® 3.0

H] Semispherical Mold

A rectangular block of 12.7 cm x 12.7 cm x 12.7 cm was drawn. A sphere was
drawn with its center coinciding with the top surface of the rectangular block. The sphere
was added to the rectangular block to provide the unified structural design for the semi-
sphere mold.

Command: File> New

Select ”Centimeters" as the template.
Command: Solid> Box> corner to corner, height
Parameters:

First corner of Base: 0,0,0

Other corner of Base: 12.7, 12.7, 0

Height: 5

Command: Solid> Sphere> Center, Radius
Parameters:

Center: 6.35, 6.35, 5

Radius: 3.63

Command: Solid> Union

First click on rectangular block, hit enter,

click on sphere, hit enter

H.2 Polyhedron Mold

A rectangular block of 12 cm x 12 cm x 5 cm was drawn using box tool. A
hexagon was drawn with sides measuring 4.80 cm with the center coinciding with the
center of top surface of the rectangular block. The hexagon was extruded with a draft

angle of 45° and height of 3.03 cm to provide a bottom edge dimension of 1.30 cm to the

89

polyhedron. The polyhedron was added to the rectangular block to provide unified
structural design for the polyhedron mold

Command: File> New

Select ”Centimeters" as the template.

Command: Solid> Box> corner to corner, height
Parameters:

First corner of Base: 0,0,0

Other corner of Base: 12, 12, 0

Height: 5

Command: Curve> Polygon> Center, Radius

Click on NumSide and enter 6

Click on circumscribed

Parameters:

Center: 6, 6, 5

Midpoint of Polygon Edge: 10.15, 6, 5

Command: Surface> Extrude> Tapered

Click on Draft Angle, enter —45 and hit enter
Enter 3.03 for Extrusion Distance and hit enter
Command: Solid> Union 1

First click on rectangular block, hit enter,

click on polyhedron, hit enter

90

Appendix I Deskproto Milling Parameters

Milling Tool: 3/8” diameter flat tip milling tool

Operation Parameters:
[1 Ge n e r a l

a. Distance between tool paths: 0.00778"
b. Stepsize along tool paths: 0.00778”
c. Feed rate= 10 inch/ min
d. Spindle speed= 1000 rpm
[:1 Strategy
a. Main= block
b. Detail Settings, block= inside out
[:1 Roughing
a. Roughing: use
b. Layer Height: Custom, 0.04”
c. Protect Vertical Surfaces
El Segment
a. Use segment of part
[1 Borders
a. Size of the borders: No extra
El Movement
a. Milling direction= Conventional
b. Free movement height= 0.0250”
c. Feed rate for plunge movement: 100% of federate
d. Feed rate for high chip loads= 50% of feedrate
El Advanced
a. Ambient skipping= skip total ambient
b. Vertical surfaces: Do not check for vertical
surfaces
El Simulation

a. Level of detail: High

91

Note: Two major problems faced once the designs were created and processed
with Deskproto®. The available version of CPmill® software (milling machine software)
could only be used with Windows 95® only, and because of that there were unexpected
problems. Every stroke performed by the machine was 4-times the expected stroke length
as defined by the design and the parameters. The problem was resolved by scaling the
drawing down to 1/4th of original size and defining all the parameters again as 1/4th of
- original. All of these corrections were carried out in Deskproto®.

CPmill® runs on Windows® 95 where it could not execute more than 36,000
command lines when running a milling program. Windows® would give an error message
after executing 36,000 lines and would close the software. For a workable milling
operation with enough fineness, the command program had around 92,000 command
lines. The problem was resolved by manually creating 3 parts of the command program.
With this, it was required to bring back the mill tool to the original starting point called

“home” at the start of each program.

92

Appendix J Calibration curve for benzaldehyde quantification

Figure J-i presents the calibration curve for quantification of benzaldehyde
concentration in the product. The benzaldehyde was prepared in different concentrations

in methanol and the area response was measured in the HPLC.

18° " y = ze-osx
160 « R2 = 0.9993

 

 

 

E 140 J
o.
‘f 120 «
c r
g 1004
g i
i a”
l
c 50f
s ,0,
20
0 .____ -_._.i._-_._-___l___,_- .1 -__ __ .__L._ __ _n
0 2 4 6 8 10 12

Area Response x 10‘, AU

Figure J-i Benzaldehyde Calibration Curve for HPLC

93

Appendix K Panelist Performance Charts

Figure K-i shows the panelist performance in rating the benzaldehyde strength on
a scale of 0-7 when presented with three different samples of varying strength. Various
lines correspond to the different panelists. It is visible that the panelists were not on the

same scale and their responses were not proportional or consistent to the intended scale.

Panelist Performance Week 1

 

 

 

 

 

 

 

 

 

 

 

 

 

8 RAE h Lf._,_
6 ..
— AP
........ E
4 ~ _-_ A
CED) _ ........ O
E — — -- D
a: — —-—. Au
2 *' ——— R
— Pr
0 .-
A B C
Sample Name

Figure K-i Response in training week 1
where A=0, B=2, C=7

94

Figure K—ii shows the performance of panelists in week when presented with two
samples to rate. It is apparent the extent of improvement that panelists achieved with
most of them rating the samples in same direction though with limited variability in
rating the overall strength. Important development was that they were able to rate the

samples consistently as was observed with Week 5 training as well.

Panelist Performance Week 4

 

 

 

 

 

 

8r- ~r ~~
7»
—Ap
6- ........ E
-—- A
._ ........ 0
5...
.E’ -—_ D
a — ----- Au
m4. _—_ R
-——-Pr
........ Pl
3P
2_ a
1 I l J

 

 

Sample

Figure K-ii Response in training week 4
where A: 4, B: 7

95

References

96

10.

Caskey, T.L., Dynamic Gas Chromatographic Method for Measuring Gas
Permeability of Film. Modern Plastics, 1967 (December): p. 148-155.

Holden, P.S., Orchard, G.A.J., and Ward, I.M., A Study of Gas Barrier Properties
of Highly Oriented Polyethylene. Journal of Polymer Science: Polymer Physics
Edition, 1985. 23: p. 709-731.

Keller, L.E., Barrier Packaging - Protective Properties of BOPP Films. Journal
of Plastic Film & Sheeting, 1987. 3 (October): p. 299-311.

Meares, P., Transient Permeation of Organic Vapors through Polymer
Membranes. Journal of Applied Polymer Science, 1965. 9: p. 917-932.

Rubino, M. and Tung, M., Permeation of oxygen, water vapor and Limonene
through printed and unprinted biaxially oriented polypropylene films. Journal of
Agriculture and Food Chemistry, 2001. 49: p. 3041-3045.

Sajiki, T. and Giacin, J .R., Permeation of Ethyl Acetate Vapor through Silica
Deposited Polyethylene Terephthalate Film and Composite Structures. J ourna] of
Plastic Film & Sheeting, 1993. 9: p. 97-111.

Zobel, M., Measurement of Odour permeability of Polypropylene Packaging
Films at Low Odourant Levels. Polymer Testing, 1982. 3: p. 133-142.

Affinito, J .D., et al., A new method for fabricating transparent barrier layers.
Thin Solid Films, 1996. 290-291: p. 63-67.

Lange, J. and Wyser, Y., Recent innovations in barrier technologies for plastic
packaging - a review. Packaging Technology and Science, 2003. 16(4): p. 149-
158.

Nilsson, T., Mazzone, R., and Frandsen, E. (1991) Oxygen barrier properties of
pet containers. PLM AB USA Patent No. 5034252.

97

ll.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Hernandez, R.J., Selke, Susan E.M, Culter, John D., Plastics Packaging
Properties, Processing, Applications and Regulations. First ed. 2000, Cincinnati:
Hanser Gardner Publications, Inc.

Kalay, G. and Bevis, M.J., Processing and physical property relationships in
injection-molded isotactic polypropylene. 2. morphology and crystallinity. Journal
of Polymer Science Part B Polymer Physics, 1997. 35 (2): p. 265-291.

Landoll, L.M., Hiltner, A., and Baer, E., Effect of Orientation on the Free Volume
and Oxygen Transport of a Polypropylene Copolymer. Journal of Polymer
Science: Part B: Polymer Physics, 2005. 43: p. 1230-1243.

Zhang, X., Ajji, A., and Jean-Marie, V., Processing-structure-properties
relationship of multilayer films. 1. Structure characterization. Polymer, 2001. 42
(19): p. 8179-8195.

Peterson, R.J. (1996) Flavor composition for an oral electrolyte rehydration
solution. Firrnenich Incorporated (Plainsboro, NJ) United States Patent No.
346355.

Teranishi, R., Flath, R.A., and Sugisawa, H., Flavor Research Recent Advances.
1981, New York: Marcel Dekker, Inc. p. 1.

Teranishi, R., Flath, R.A., and Sugisawa, H., Flavor Research Recent Advances.
1981, New York: Marcel Dekker, Inc. p. 3.

Fayoux, S.C., Seuvre, A.M., and Voilley, A.J., Aroma transfers in and through
plastic packagings: orange juice and d-limonene. A review. Part 11: overall

sorption mechanisms and parameters—a literature survey. Packaging Technology
and Science, 1997. 10(3): p. 145-160.

Widen, H., Leufven, A. and Nielsen, T., Migration of model contaminants from
PET bottles: influence of temperatures, food simulant and functional barrier.
Food Additives and Contaminants, 2004. 21 (10): p. 993-1006.

Brown, W.E., Plastics in Food Packaging: Properties, Design, and Fabrication.
1992: Marcel Dekker.

98

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

Crank, J ., The Mathematics of Diffusion. 2nd ed. 1975, Oxford: Oxford University
Press. p. 424.

Pick, A., Ueber Difi‘usion. Annalen der Physik, 1855. 170(1): p. 59-86.

Flory, P.J., Principle of Polymer Chemistry. 1953, Ithaca: Cornell University
Press.

Freeman, B., Yampolskii, Y., and Pinnau, 1., Materials Science of Membranes for
Gas and Vapor Separation. 2006, Wiley & Sons, Ltd.

Barr, C.D., Giacin, J .R., and Hernandez, R.J., A determination of solubility
coefiicient values determined by gravimetric and isostatic permeability
techniques. Packaging Technology and Science, 2000. 13(4): p. 157-167.

Comyn, J ., Polymer Permeability. 1994: Springer.

Yasuda, H. and Rosengren, K., Isobaric measurement of gas permeability of
polymers. Journal of Applied Polymer Science, 1970. 14(11): p. 2839-2877.

Niebergall H, Humeid, A., and Blochl, W., The Aroma Permeability of Packaging
Films and its Determiantion by Means of a Newly Developed Measuring
Apparatus. Lebensmittel Wissenschaft und Technologie, 1978. 11 (l): p. 1-4.

Apostolopoulos, D. and Winters, N., Measurement of permeability for packaging

films to d-limonene vapour at low levels. Packaging Technology and Science,
1991. 4: p. 131-138.

Duncan, B., Urquhart, J ., and Roberts, S. Review of Measurement and Modelling
of Permeation and Diffusion in Polymers. 2005 [cited August 10, 2007];
Available from:
http://www.npl.co.uk_/Ln_aterials/thermoI/pdf/depc%5Fmpr%5F012.pdf.

Kim, J .K., Lee, J .H., and Ha, Y.S., Measurement of Permeability of Food
Packaging Polymer Films to Organic Vapors. Journal of Food Science and
Nutrition, 2001. 6 (4): p. 247-252.

99

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

Lange, J ., et al., Water Vapour Transport through Large Defects in Flexible
Packaging: Modeling, Gravimetric Measurement and Magnetic Resonance
Imaging. Packaging Technology and Science, 2000. 13: p. 139-147.

Qin, Y., et a1. Simultaneously Gravimetric Assessment of Organic and Moisture
on Containers. in Antec 2006. 2006. Charlotte, NC.

Moylan, C.R., Best, M.E., and Ree, M., Solubility of Water in Polyimides: Quartz
Crystal Microbalance Measurements. Journal of Polymer Science: Part B:
Polymer Physics, 1991. 29: p. 87-92.

Smith, A.L., Ashcraft, J .N., and Hammond, P.T., Sorption isotherms, sorption
enthalpies, difi’usion coefficients and permeabilities of water in a multilayer
PEO/PAA polymer film using the quartz crystal microbalance/heat conduction
calorimeter. Thermochirnica Acta, 2006. 450 (1-2): p. 118.

Smith, A.L., Mulligan, SR, and Shirazi, H.M., Determining the Efiects of Vapor
Sorption in Polymers Using the Quartz Crystal Microbalance/Heat Conduction
Calorimeter. Journal of Polymer Science: Part B: Polymer Physics, 2004. 3906.

Wibawa, G., et al., Solubilities of I 1 Polar Organic Solvents in Four Polymers
Using the Piezoelectric-Quartz Sorption Method. J. Chem. Eng. Data, 2002. 47: p.
1022-1029.

Wong, C.H., Campbell, S.W., and Bhethanabotla, V.R., Sorption of benzene,
toluene and chloroform by poly(styrene) at 298.15 K and 323.15 K using a quartz
crystal microbalance. Fluid Phase Equilibria, 1997. 139: p. 371-389.

Zhang, C., et al., Glassy Polymer-Sorption Phenomena Measured with a Quartz
Crystal Microbalance Tecnique. Journal of Polymer Science: Part B: Polymer
Physics, 2003. 41: p. 2109-18.

Balik, CM. and Simendinger, W.H., An attenuated total reflectance cell for
analysis of small molecule diffusion in polymer thin films with F ourier-transform
infrared spectroscopy. Polymer, 1998. 39 (20): p. 4723-4728.

Morrissey, P. and Vesely, D., Accurate measurement of difiusion rates of small
molecules through polymers. Polymer, 2000. 41 (5): p. 1865-1872.

100

42.

43.

45.

46.

47.

48.

49.

50.

51.

52.

Sammon, C., Yarwood, J., and Everall, N ., A FT lR-ATR study of liquid diffusion
processes in PET films: comparison of water with simple alcohols. Polymer,
2000. 41: p. 2521-2534.

Schmidt, P., Dybal, J ., and Trchova, M., Investigations of the hydrophobic and
hydrophilic interactions in polymer-water systems by ATR FT IR and Raman
spectroscopy. Vibrational Spectroscopy, 2006. 42 (2): p. 278-283.

Yossef A. Elabd, et al., Time-resolved Fourier transform infrared/attenuated total
reflection spectroscopy for the measurement of molecular diffusion in polymers.
Journal of Polymer Science Part B Polymer Physics, 2003. 41 (22): p. 2794-2807.

Csemica, J ., Baddour, RF, and Cohen, R.E., Gas permeability of a polystyrene-

polybutadiene block copolymer with oriented lamellar domains. Macromolecules,
1987. 20 (10): p. 2468-2471.

Jones, R.L., et al., Chain conformation in ultrathin polymer films. Nature, 1999.
400 (6740): p. 146-149.

Auras, R., Solubility of Gases and Vapors in Polylactide Polymers, in
Thermodynamics, Solubility and Environmental Issues, Editor T.M. Letcher.
2007, Elsevier Science. p. 342- 368.

Shastri, R., et al. Permeability of Competitive Oxygen-Barrier Resins. in ACS
Symposium #423. 1990.

Gremli, H., Flavor changes in plastic containers: A literature review. Perfumer
and Flavorist, 1996. 21 (6): p. 1-8.

Taub, LA. and Singh, R.P., Food Storage Stability. 1998: CRC Press.

Chuang, W.Y., et al., The efiect of polymeric additives on the structure and

permeability of poly( vinyl alcohol) asymmetric membranes. Polymer, 2000. 41
(15): p. 5633-5641.

Charara, Z.N., et al., Orange flavor absorption into various polymeric packaging
materials. J ourna] of Food Science, 1992. 57 (4): p. 963-966.

101

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

Auras, R., Harte, B., and Selke, S.E.M., An Overview of Polylactides as
Packaging Materials. Macromolecular Bioscience, 2004. 4 (9): p. 835-864.

Baner, A.L., Kalyankar, V, Shound, L.H, Aroma Sorption Evaluation of Aseptic
Packaging. Journal of Food Science, 1991. 56 (4): p. 4.

Nielsen, T]. and Giacin, J .R., The sorption of limonene/ethyl acetate binary
vapour mixtures by a biaxially oriented polypropylene film. Packaging
Technology and Science, 1994.7 (5): p. 247-258.

Hernandez, R.J., Eflect of water vapor on the transport properties of oxygen
through polyamide packaging materials. Journal of Food Engineering, 1994. 22
(1-4): p. 495-507.

Beidler, L.M., Flavor Research and Food Acceptance. 1958, New York: Reinhold
Publishing Corporation. p. 3.

Fenaroli, G., Fenaroli's Handbook of Flavor Ingredients. 3rd ed, ed. G.A.
Burdock. Vol. I. 1995, Ann Arbor: CRC Press. p. 3.

Bauer, K., Garbe, D., and Surburg, H., Common Fragrance and Flavor Materials.
1997, New York: Wiley- VCH. p. 2.

Odor and Flavor Detection Thresholds in Water. 2005 [cited Nov 15, 2004];
Available from: http://www.leffingwellcom/odorthrehtm ].

Acree, T. and Arn, H. F lavomet. 2004 [cited Nov 15, 2004]; Available from:
http://wwwflavornetorg/flavomethtml.

Baldwin, E.A., Fruit and Vegetable Flavor, in Citrus and Subtropical Products
Laboratory handbook. 2000, USDA: Winter Haven, Fl. p. 1-18.

Charalambous, G. and Inglett, G.E., Flavor of Foods and Beverages: Chemistry
and Technology. 1978: Academic Press.

Contis, E.T., Food flavors .' formation, analysis, and packaging influences :
proceedings of the 9th International Flavor Conference, the George

102

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

Charalambous Memorial Symposium, Limnos, Greece. Developments in food
science 1998, Amsterdam ; New York: Elsevier. p. 797 p.

Coppen, J .J .W. and Food and Agriculture Organization of the United Nations,
Flavours and fragrances of plant origin. Non-wood forest products, 1. 1995,
Rome: Food and Agriculture Organization of the United Nations. p. 101.

Crocker, E.C., Flavor. lst ed. McGraw-Hill series in food technology. 1945, New
York; London; McGraw-Hill Book Company, Inc. p. 172.

De Rovira, D.A., The dictionary of flavors : and general guide for those training
in the art and science of flavor chemistry. 1999, Trumbull, Conn.: Food &
Nutrition Press. p. 548.

Deibler, K.D. and Delwiche, J ., Handbook of flavor characterization : sensory
analysis, chemistry, and physiology. 2004, New York ; Basel: Marcel Dekker. p.
493.

European Chemoreception Research Organization, Chemical senses & flavour.
1978, Published by Information Retrieval Limited in association with the
European Chemoreception Research Organization.

Fenaroli, G. and Burdock, G.A., F enaroli 's Handbook of Flavor Ingredients.
1995: CRC Press Boca Raton.

Heath, H.B., Source Book of Flavors. 1981: Kluwer Academic Publishers.

Merory, J ., Food flavorings: composition, manufacture, and use. 2nd ed. 1968,
Westport, Conn.,: Avi Pub. Co. p. 478.

Reineccius, G.A., Source Book of Flavors. 1999: Aspen Publishers.

Stevens, R., Common fragrance and flavor materials. 3rd ed, ed. K. Bauer,
Garbe, D., and Surburg, H. 1997: Wiley-VCH, Weinheim. p. 278.

Kochi, M., Esumi, S., and Takeuchi, S. (1988) Pharmaceutical composition for
retarding and reducing cachexia. Kaken Pharmaceutical Co., Ltd. (Tokyo, JP);
Rikagaku Kenkyusho (Tokyo, JP) United States Patent No. 902293.

103

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

ASTM E679-04 Standard Practice for Determination of Odor and Taste
Thresholds By a F orced-Choice Ascending Concentration Series Method of
Limits. ASTM Book of Standards. Vol. 15.08. 2004.

ASTM D3078-02 Standard Test Method for Determination of Leaks in Flexible
Packaging by Bubble Emission. ASTM Book of Standards. Vol. 15.10. 2002.

ASTM E460-04 Standard Practice for Determining Effect of Packaging on Food
and Beverage Products During Storage. ASTM Book of Standards. Vol. 15.08.
2004.

ASTM F1249-06 Standard Test Method for Water Vapor Transmission Rate
Through Plastic Film and Sheeting Using a Modulated Infrared Sensor, in ASTM
Book of Standards. 2006.

ASTM D3985-05 Standard Test Method for Oxygen Gas Transmission Rate
Through Plastic Film and Sheeting Using a Coulometric Sensor. ASTM Book of
Standards. Vol. 15.10. 2005.

ASTM E967-03 Standard Practice for Temperature Calibration of Differential
Scanning Calorimeters and Differential Thermal Analyzers. ASTM Book of
Standards. Vol. 14.02. 2003.

ASTM E968-02 Standard Practice for Heat Flow Calibration of Differential
Scanning Calorimeters. ASTM Book of Standards. Vol. 14.02. 2002.

ASTM D34 I 8-03 Standard Test Method for Transition Temperatures and
Enthalpies of Fusion and Crystallization of Polymers by Differential Scanning
Calorimetry. ASTM Book of Standards. Vol. 08.02. 2003.

Meilgaard, M.C., Civille, G.V., and Carr, E.T., Sensory Evaluation Techniques.
3rd ed. 1999: CRC Press.

Boersma, A., Mobility and solubility of antioxidants and oxygen in glassy
polymers. 1. Concentration and temperature dependence of antioxidant sorption.
Journal of Applied Polymer Science, 2003. 89 (8): p. 2163-2178.

104