11:1)..1 .
r :7 .3. .

a)!

rm... .1.

m. .3...

A?

. fur.

w???

-. “.anlMHuurm uh
{<2 :1).
I L\\ .r.
u ﬁn... ..

. . .vaWq.

.. sat. ;.
thus ; v.3...
:3....

.z

1.2.. 1
I . 21...):

 

. § V. .
up“. i a
3

. .. . . . - . . - . . - . . . ..A..”"I.w{ 5 LII.
.. . . . . . . . . . .Wcurr} Imam.)
- .. . . . . . , . . . .. . . .. JasmmmuLam Wan...
.ﬁvrnimmﬁf .. .. .. .. .. . . . .. . - . . _ .. . . .2 x. B
[3 v rumm.”~ .. . . \(2 u v . . . . . - . . - . _ . . . . .13! In; I . u
z- . z. 3...: . .. . . . . . . .. . .
g .35? y l. . . . . . . .
. . vuxmm. nu...»- . .
.. - a. azhti -
. u". 56:33:11... .
$45..» . .. i 311.
imam-3.1..
aux» . . . . .. . . - . -. . . . - .H .- . . -. ; . pm.” .23 ”34.. .
.34. . . . . . . . . .. .
{ruby

. .2.
.‘

 

.5- .-
INA}:

-. vs 94911.3; I
If...

'U‘

. 30K; 5 y

'1‘!

 

.5
. .5. 1.1.l :litfizY‘A
I. . 3|? alv‘...‘
é - 51.31:.

~ 35?.

L

39...!- 4 vii...

awamhﬂﬂtsgn
_ “(a

PI.
. I
l - I;

C“

.E t. , ,

.‘ﬁgﬂﬂv. .. .v .\ I . L (y . - t». v - D ‘ 'L...2l3.l.-13
; 1:5)? 2.1-7. . .a;h..a.f1r; .1 .ivrliﬁ. 5...”... LL‘- l: . 5. 54‘ . .I-h~(|w..oﬁ
. . . . - .- a. 3- -3 u...rnxﬂuhﬁnx:3.€ .. 3:33-33...

 

'9

j l!!!(WIHIIHWIIHHIll(“INN/ll!!!)NHIHUHHUHH

806018
700 F.)

 

LIBRARY

Mlchigan State
University

 

 

 

This is tO certify that the

thesis entitled

ORGANIC VAPOR PERMEABILITY OF TRANSPARENT HIGH
BARRIER FILMS UTILIZING A HIGH PERFORMANCE
THIN A1203 COATING LAYER

presented by
Hiroshi Suzuki
has been accepted towards fulﬁllment

Of the requirements for

M. S. degree in Science

Majorprmfsor

6394M. ﬂu»

Date December 9th 1999

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

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

 

 

~ ~ 2997

. I
‘ ’ A
:' \r

i;

I—‘ [‘1'1
r“ a»:
”:3
CD
3...;

 

 

 

 

 

 

 

 

 

 

 

 

 

moo animus-p.14

ORGANIC VAPOR PERMEABILITY OF TRANSPARENT HIGH
BARRIER FILMS UTILIZING A HIGH PERFORMANCE
THIN A1203 COATING LAYER

By

Hiroshi Suzuki

A THESIS
Submitted to
Michigan State University
in partial fulfillment Of the requirements
for the degree of

MASTER OF SCIENCE

School of Packaging
1 999

Professor Jack R. Giacin

ABSTRACT

The organic vapor (d-limonene and ethyl acetate) permeability of A1203-coated
PET ﬁlms and SiOx-coated PET ﬁlms was studied. Evaluation was made utilizing the
dynamic purge and trap / thermal desorption procedure.

A comparison of the permeability (and its temperature dependence) of A1203-
coated PET ﬁlm to various permeants (d-limonene, ethyl acetate, oxygen and water
vapor) was made. Results showed that water vapor behaved signiﬁcantly differently than
other permeants, in terms of the mechanism of mass transfer.

There were no major differences observed between the permeability of the A1203-
coated PET ﬁlm and that of the SiOx-coated PET ﬁlm which suggested that the
mechanism of mass transfer through both ﬁlms may be similar.

The effect of physical abuse on gas or vapor permeability was also evaluated for
A1203-coated PET ﬁlm. Barrier deterioration, caused by cracks of the ceramic surface,
occurred at less than 10 times ﬂexing abuse for both non-polar and polar permeant/
A1203-coated PET systems. For A1203-coated PET / LLDPE laminated ﬁlm samples, the
loss in barrier performance varied for non-polar and polar permeants.

A simple model has been prepared to account for the observed permeability
behavior for the A1203-coated PET ﬁlm. This model provides an explanation for the

permeability behavior of both normal (non-abused) and abused A1203-coated PET ﬁlms.

Dedicated to my daughter and my wife.

iii

ACKNOWLEDGMENTS

I would like to express my appreciation to Dr. Jack R. Giacin for his guidance,
understanding, and patience to complete this research. I would also like to thank Dr.
Randy Beaudry and Dr. Tee Downes for serving on my committee members.

In addition, I am very grateful to TOPPAN Printing Co., Ltd. for their ﬁnancial

supports during my studies in Michigan State University.

iv

TABLE OF CONTENTS

Page

List of Tables ................................................................................................................ viii
List of Figures ................................................................................................................. xi
Chapter 1. Introduction ................................................................................................. 1
Chapter 2. Literature Review ....................................................................................... 5
2-1. Ceramic-Coated Films .................................................................................................. 5
2-1-1. Evaporation Processes .................................................................................. 5

2-1-2. Development of Ceramic-Coated Film ......................................................... 8

2-1-3. SiOx-Coated Films ........................................................................................ 9

2-1-4. A1203-Coated Films .................................................................................... 13

2-1-5. Ceramic Mixture Coated Films ................................................................... 14

2-2. Properties of Ceramic-Coated Films .......................................................................... 15
2-2-1. Oxygen and Moisture Barrier ..................................................................... 15

2-2-2. Organic Vapor Barrier ................................................................................. 16

2-2-3. Flex Resistance ............................................................................................ 20

2-3. General Theory Related to the Permeation Process ................................................... 21
2-3-1. Permeation Steps ......................................................................................... 21

2-3-2. Sorption Mechanism ................................................................................... 22

2-3-3. Diffusion Mechanism .................................................................................. 22

2-3-4. Permeation Mechanism of Ceramic-Coated Films ...................................... 23

2-4. Factors Affecting to the Permeation Process ............................................................. 25
2-4-1. Nature of Permeants .................................................................................... 25

2-4-2. Measuring Temperature .............................................................................. 27

2-5. Permeability Measurement ......................................................................................... 28
2-5-1. Mathematical Approach for Permeation Process ....................................... 28

2-5-2. Isostatic Method ......................................................................................... 31

2-5-3. Dynamic Purge and Trap/Thermal Desorption Technique ......................... 32
Chapter 3. Materials and Methods ............................................................................ 34
3-1. Materials and Apparatuses ........................................................................................ 34

3-1-1. Films and Laminated Structures .................................................................. 34

3-1-2. Permeants .................................................................................................... 34

3-1-3. Solvents ...................................................................................................... 35

3-1-4. Apparatuses ................................................................................................ 36
3-2. Methods ..................................................................................................................... 38

3-2-1. Calibration Curve for Vapor Pressure ......................................................... 38

3-2-2. Calibration Curves for Dynamic Purge and Trap / Thermal Desorption

Procedure ................................................................................................... 39

3-2-3. Organic Vapor Permeability Test System ................................................... 41

3-2-4. Permeability Measurements ........................................................................ 43

3-2-5. Dynamic Purge and Trap / Thermal Desorption Procedure ........................ 46

3-2-6. Analysis of Test Permeant by Thermal Desoption

/ Gas Chromatography Procedure ............................................................... 47

3-2-7. Analysis of Ceramic Coating Surface by Optical Microscopy ................... 48
Chapter 4. Results and Discussions ......................................................................... 49
4-1. Estimation of the Detection Sensitivity Limit ............................................................ 49
4-2. Time to Reach Steady State ........................................................................................ 50
4-3. Comparison of Permeability of Non-Abused Films ................................................... 53

4-3-1. Comparison by Permeants .......................................................................... 53

4-3-2. Comparison of Ceramic Coatings ................................................................ 60
4-4. Dependency of Permeability on Temperature ........................................................... 62
4-5. Permeant Factors Affecting Permeation ..................................................................... 68

4-5-1. Permeant Size .............................................................................................. 68

4-5-2. Polan'ty of Permeant ................................................................................... 71

4-5-3. Combined Factor by Size and Polarity ....................................................... 73

4-6. Permeation Mechanism for Non-Abused ﬁlms .......................................................... 76

4-6-1. Permeation through Defects ........................................................................ 76

4-6-2. Generation of Attractive Defects ................................................................ 78

4-6-3. Ale3-Coated PET Film ............................................................................. 78

4-6-4. Ale3-Coated PET / LLDPE Laminated Film ............................................ 80

4-7. Eﬂ‘ect of Physical Damage .......................................................................................... 83

4-7-1. Non-Polar Permeants Permeability ............................................................ 83

4-7-2. Polar Permeants Permeability .................................................................... 86

4-7-3. Microscopic Observations .......................................................................... 93

4-8. Permeation Mechanism for Abused Films ................................................................. 98

4-8-1. A1203-Coated PET Film ............................................................................. 98

4-8-2. A1203-Coated PET / LLDPE Laminated Film .......................................... 101

Chapter 5. Summary and Conclusions .................................................................. 105

Chapter 6. Future Studies ......................................................................................... 107

Appendices .................................................................................................................... 108

Appendix A. Calibration Curve for Setting Vapor Pressure ................................ 108
Appendix B. Calibration Curve for Dynamic Purge and Trap

/ Thermal Desorption Procedure .................................................. 111

Appendix C. Comparisons of Vapor Pressure of Compounds ........................... 114

Appendix D. Dimensions ofPermeants .............................................................. 116

Bibliography .................................................................................................................. 120

vii

LIST OF TABLES

Page

Table 2-1 Pros and cons of evaporation methods ............................................................... 8
Table 3-1 Basic information of the permeants .................................................................. 35
Table 3-2 Basic information of the solvents ..................................................................... 36
Table 3-3 Initial conditions of gas chromatograph ............................................................ 38
Table 3-4 Temperature programming for gas chromatograph

for direct injection procedure ........................................................................... 38
Table 3-5 Conditions of thermal desorption unit ............................................................. 40
Table 3-6 Temperature programming for gas chromatograph

for thermal desorption procedure .................................................................... 40
Table 4-1 Time to reach steady state of permeation ........................................................ 50
Table 4-2 Permeability of Ale3-coated PET (12pm) to various permeants .................. 54
Table 4-3 Permeability of Ale3-coated PET (12um) / LLDPE (40um)

to various permeants ........................................................................................ 55
Table 4-4 Permeability of PET (12 pm) to various permeants ........................................ 56
Table 4-5 Permeability comparison between A1203-coated PET ( IZum)

and non-coated PET (12pm) ............................................................................. 57
Table 4-6 Permeability comparison between A1203-coated PET (12pm) /LLDPE

(40pm) and non-coated PET (12pm) /LLDPE (40pm) ................................... 59
Table 4-7 Calculated permeance at 60 °C through A1203-coated PET (12um) ............... 60
Table 4-8 Permeability comparison of ceramic coatings ................................................... 61
Table 4-9 Temperature effect on the permeability

of the A1203 PET /d-limonene system ........................................................... 63

viii

Table 4-10

Table 4-11

Table 4-12

Table 4-13

Table 4-14

Table 4-15

Table 4-16

Table 4-17

Table 4-18

Table 4-19

Table 4-20

Table 4-21

Table 4-22

Table 4-23

Table A-1

Table A-2

Table A-3

Temperature effect on the permeability

of the A1203 PET / ethyl acetate system ....................................................... 64
Comparison of activation energy for the permeation process

through A1203 PET using various permeants ................................................. 67
Dimensions of the permeants ......................................................................... 69
Dielectric constant of permeants (taken from CRC, 1987) ............................ 71
Combined factor (Size and polarity) of permeants ......................................... 73
Comparison of the various trends related to permeability ............................. 74
Flex resistance for oxygen permeability of A1203 PET

and A1203 PET /LLDPE ................................................................................ 84
Flex resistance for d-limonene permeability Of A1203 PET

and A1203 PET /LLDPE ................................................................................ 85
Flex resistance for water vapor permeability of A1203 PET

and A1203 PET /LLDPE ................................................................................ 87
Flex resistance for ethyl acetate permeability of A1203 PET

and A1203 PET /LLDPE ............................................................................... 88
Permeability comparison Of abused A1203 PET ﬁlm using

various permeants (permeance comparison) ................................................... 89
Permeability comparison of abused A1203 PET / LLDPE ﬁlm using

various permeants (permeance comparison) ................................................... 90
Permeability comparison of abused A1203 PET ﬁlm using

various permeants (deterioration rate comparison) ........................................ 9]
Permeability comparison of abused A1203 PET /LLDPE ﬁlm using

various permeants (deterioration rate comparison) ........................................ 92
Calibration data of d-limonene for setting vapor pressure ............................ 109
Calibration data of ethyl acetate for setting vapor pressure .......................... 110
Calibration data Of d-limonene for dynamic purge and trap / thermal

desorption procedure ..................................................................................... 112

Table A-4 Calibration data Of ethyl acetate for dynamic purge and trap / thermal

desorption procedure ..................................................................................... 113
Table A-5 Comparison of saturated vapor pressure of d-limonene ............................... 114
Table A-6 Comparison of saturated vapor pressure of ethyl acetate ............................ 1 15

LIST OF FIGURES

Page
Figure 2-1 Schematic Model of PVD .................................................................................. 6
Figure 2-2 Schematic Model of CVD ................................................................................. 7
Figure 2-3 Matrix of Si02 and SiOx ................................................................................. 10
Figure 2-4 Oxygen and water vapor barrier properties
of high barrier performance ﬁlms .................................................................... 17
Figure 3-1 Schematic diagram of dynamic purge and trap permeability system .............. 42
Figure 4-1 Transmission proﬁle of d-limonene and ethyl acetate
through A1203-coated PET ﬁlm ....................................................................... 51
Figure 4-2 Transmission proﬁle of water vapor
through A1203-coated PET ﬁlm ...................................................................... 52
Figure 4-3 Permeability of A1203-coated PET (12 um) to various permeants ................ 54
Figure 4-4 Permeability of A1203-coated PET (12pm) / LLDPE (40pm)
to various permeants ........................................................................................ 55
Figure 4-5 Permeability of PET (12 um) to various permeants ....................................... 56
Figure 4-6 Permeability comparison between AIZO3-coated PET (12pm)
and non-coated PET (12um) ............................................................................ 57
Figure 4-7 Permeability comparison between A1203-coated PET (12um) / LLDPE
(40pm) and non-coated PET (12pm) / LLDPE (40pm) .................................. 59
Figure 4-8 Permeability comparison of ceramic coatings .................................................. 61
Figure 4-9 Temperature effect on the permeability
of the A1203 PET / d-limonene system ........................................................... 63
Figure 4-10 Temperature effect on the permeability
of the A1203 PET / ethyl acetate system ...................................................... 64

Figure 4-11 Arrhenius plot for the permeance

of the A1203 PET / d-limonene system ......................................................... 65
Figure 4-12 Arrhenius plot for the permeance

of the A1203 PET / ethyl acetate system ...................................................... 65
Figure 4-13 Comparison of Arrhenius plots of various permeants

through A1203 PET ....................................................................................... 66
Figure 4-14 Comparison of activation energy for the permeation process

through A1203 PET using various permeants ................................................ 67
Figure 4-15 Relationship between permeability and penetrant Size ................................. 70
Figure 4-16 Relationship between permeability and polarity of permeants .................... 72
Figure 4-17 Relationship between permeability and combined factor .............................. 75
Figure 4-18 Defect leading permeation model ................................................................... 77

Figure 4-19 Generation of attractive defects in ceramic coating layers

and permeation mechanism through A1203 PET ........................................... 79
Figure 4-20 Barrier mechanism for non-polar permeants through A1203 PET ................ 81
Figure 4-21 Barrier mechanism for polar permeants through A1203 PET ........................ 82

Figure 4-22 Flex resistance for oxygen permeability of A1203 PET
and A1203 PET /LLDPE ............................................................................... 84

Figure 4-23 Flex resistance for d-limonene permeability of A1203 PET
and A1203 PET /LLDPE ............................................................................... 85

Figure 4-24 Flex resistance for water vapor permeability of A1203 PET
and A1203 PET /LLDPE ............................................................................... 87

Figure 4-25 Flex resistance for ethyl acetate permeability of A1203 PET
and A1203 PET / LLDPE ............................................................................... 88

xii

Figure 4-26 Permeability comparison of abused A1203 PET ﬁlm using
various permeants (permeance comparison) .................................................. 89

Figure 4-27 Permeability comparison of abused A1203 PET / LLDPE ﬁlm using
various permeants (permeance comparison) .................................................. 90

Figure 4-28 Permeability comparison of abused A1203 PET ﬁlm using
various permeants (deterioration rate comparison) ....................................... 91

Figure 4-29 Permeability comparison of abused A1203 PET / LLDPE ﬁlm using

various permeants (deterioration rate comparison) ....................................... 92
Figure 4-30 Surface observation for non-abused A1203 PET ........................................... 94
Figure 4-31 Surface Observation for 10 times ﬂexing A1203 PET .................................... 95
Figure 4-32 Surface observation for 100 times ﬂexing A1203 PET .................................. 96
Figure 4-33 Surface observation for 200 times ﬂexing A120 3 PET .................................. 97

Figure 4-34 Barrier mechanism for non-polar permeants through A1203 PET
(with Gelbo ﬂexing) ....................................................................................... 99

Figure 4-35 Barrier mechanism for polar permeants through A1203 PET
(with Gelbo ﬂexing) ..................................................................................... 100

Figure 4-36 Barrier mechanism for non-polar permeants through A1203 PET / LLDPE
(with Gelbo ﬂexing) ..................................................................................... 102

Figure 4-37 Barrier mechanism for polar permeants through A1203 PET / LLDPE

(with Gelbo ﬂexing) ..................................................................................... 104
Figure A-l Calibration data of d-limonene for setting vapor pressure ........................... 109
Figure A-2 Calibration data of ethyl acetate for setting vapor pressure ........................ 110

Figure A-3 Calibration data of d-limonene for dynamic purge and trap / thermal
desorption procedure .................................................................................... 112

Figure A-4 Calibration data of ethyl acetate for dynamic purge and trap /thermal
desorption procedure .................................................................................... 113

xiii

Figure A-5 Comparison of saturated vapor pressure of d-limonene .............................. 114

Figure A-6 Comparison of saturated vapor pressure of ethyl acetate ............................ 115
Figure A-7 Dimensions of water molecule ...................................................................... 116
Figure A-8 Dimensions of oxygen molecule ................................................................... 116
Figure A-9 Dimensions of d-limonene molecule (top view) ........................................... 117
Figure A-10 Dimensions of d-limonene molecule (side view) ........................................ 118
Figure A-l 1 Dimensions of ethyl acetate molecule ........................................................ 119

xiv

Chapter 1

INTRODUCTION

The importance of polymeric materials for packaging purposes has been
increasing because of their unique characteristics (e.g., light weight, high ﬂexibility, high
transparency ). Moreover, the nature of polymers allows the properties of plastic
materials to be controlled at will, utilizing various types of modiﬁcations. High barrier
plastic ﬁlms are a very important practical application for modiﬁed polymers. Ceramic

(e. g., silicon oxide (SiOx) and aluminum oxide (A1203)) coating processes for the surface

of plastic ﬁlms have received considerable attention as surface modiﬁcation techniques.
Silicon oxide (SiOx) coated polyethylene terephthalate (PET) ﬁlms were

introduced into packaging markets in the early 1980’s. These ﬁlms might be called the

“ﬁrst generation” of SiOx-coated PET. They were transparent barrier ﬁlms that afforded

good water vapor and oxygen gas barrier properties but were in limited use in the
packaging ﬁeld at that time.

In the high barrier ﬁlm market, there is no perfect ﬁlm in terms of the properties
and the cost. Even though ceramic-coated ﬁlms had superior barrier properties, in order to
expand the usage of ceramic-coated ﬁlms, there were three problems to be solved. (1)
These ﬁlms were relatively expensive, compared to other barrier ﬁlms such as
polyvinylidene chloride (PVdC)-coated PET. (2) Because of the nature of ceramics, the

ceramic layers on the ﬁlms were relatively brittle, even though they were less than 200

nm thick."(3)/,These ﬁlms were Slightly yellow in color, which could negatively inﬂuence
the visual appearance of the packaged product (Imai, 1998). For example, the ﬁlms could
not be used for the packaging Of bonito shavings, which are a traditional Japanese spice.
Fresh bonito Shavings are pink in color, but when they are oxidized or deteriorated, their
color turns amber. The slightly yellow packaging ﬁlm made the product appear to be
amber in color, even though the actual product color was pink.

Recently, however, some signiﬁcant efforts have resulted in the production of

advanced ceramic-coated ﬁlms. They include not only efforts to improve SiOx-coated
PET (the “second generation” of SiOx-coated PET) but also to develop new ceramic (i.e.
aluminum oxide (A1203)) coated PET ﬁlms. The A1203-coated PET ﬁlms are less costly
than the ﬁrst generation of SiOx-coated PET and comparably priced to PVdC coated PET
(Hoffmann et al., 1994; Kelly, 1994). In addition, the ﬁlms are less brittle and very clear.
Potentially, therefore, a number of ﬂexible barrier packages can be replaced with A1203-
coated PET based structures.

Although A1203-coated PET ﬁlms exhibit good water vapor and oxygen gas
barrier properties (Imai, 1998), the organic vapor (ﬂavor or aroma moieties) permeability
of A1203-coated PET ﬁlms has not been determined. For some products, a package
system is required which can prevent the loss of volatile organic compounds through the

package wall to the external environment. For instance, cosmetics in ﬂexible packages

Oﬁen need to be sealed against the loss of aroma compounds from the product.

IQ

The primary focus of this study, therefore, was to provide the ﬁrst data on the

barrier properties of Ale3-coated PET ﬁlm to organic compounds under varying

conditions. The organic permeants selected for this study are d-limonene and ethyl
acetate. Ethyl acetate was selected based on its common use as a solvent in the converting
industry and its presence as a residual solvent from laminating and printing processes. D-
limonene was selected to represent a major ﬂavor volatile for citrus oil and essence
(Indou, 1985).

In order to Simulate abuses of the packaging ﬁlm during packing or distribution,
this study also evaluated the effect of mechanical stresses, such as ﬂexing stress, on the

barrier properties of the A1203-coated PET ﬁlms and A1203-coated PET / LLDPE

laminated structures.
Another major focus Of this study was to investigate the mechanism of mass

transfer of gases or vapors through A120 3-coated PET ﬁlm. Water vapor exhibits
signiﬁcantly higher permeability through A1203-coated PET ﬁlm than oxygen, even

though these permeants have Similar molecular Sizes. Several investigations have
proposed a possible mechanism for permeability of oxygen and water vapor through
ceramic-coated ﬁlms (Barker et al., 1995; Tropsha, 1997), but the studies did not include
organic vapors. By using permeants of different characteristics, this study may provide

insight into the gas or vapor permeation mechanism of A1203-coated PET ﬁlms.

The speciﬁc objectives of this study are:

1) To evaluate the temperature effects on the organic vapor permeability of A1203-
coated PET ﬁlm using d-limonene and ethyl acetate.

2) To compare the organic vapor permeability of two different types of ceramic-coated
PET (SiOx and A1203).

3) To evaluate the organic vapor permeability of A1203-coated PET, which has been
damaged by ﬂexing at various abuse levels.

4) To evaluate the effect of permeability of a series of permeants of various
characteristics (oxygen, water vapor, d-limonene, and ethyl acetate) through A1203-

coated PET.

Chapter 2

LITERATURE REVIEW

2-1. Ceramic-Coated Films
2-1-1. Evaporation Processes

The technique to deposit a very thin coating layer on a substrate is one of the key
technologies to make modiﬁed materials in many industries. The evaporation process is
an effective way to manufacture the layer. It consists of two steps: step (1); vaporize the
source materials, and step (2); deposit the vaporized materials onto the substrate.
There are various evaporation processes, but they can be categorized into two general
methods: (i) physical vapor deposition (PVD), which includes vacuum evaporation,
sputtering, and molecular beam epitaxy; and (ii) chemical vapor deposition (CVD). The
most critical theoretical difference between PVD and CVD is the driving force for the
transition from step (1) to step (2) in the evaporation process. PVD uses the physical
energy of the evaporated material (e. g., the vapor pressure in the vacuum evaporation
process) to form deposits onto the surface of the polymer ﬁlm substrate. In contrast,
CVD uses a chemical reaction as the driving force to deposit the ceramics onto the ﬁlm
surface.

The PVD method (mainly, vacuum evaporation) is most widely employed in the
packaging industry to manufacture ceramic-coated films. The schematic model of the

process is shown in Figure 2-1. The source materials (i.e. silicon monoxide (SiO) or

aluminum (Al) etc.) are heated in a crucible and then evaporated. The evaporated
materials are deposited onto the surface of a ﬁlm substrate, which is cooled by a chill roll.
One of the most important advantages of the PVD method is its very high machine speed,
which can reduce production costs through greater efficiency. This method has a potential
drawback, however, which is that the ﬁlm can potentially be damaged in the high

temperature environment (Kelly, 1993).

PVD

Running direction

9—

l Substrate J

Ceramic coating

 

 
 
 
   

 

Driving force

Source material

Crucible

Figure 2-1 Schematic Model of PVD

The two most common heating systems for the evaporation process involved in
PVD method are the boat system and the electron beam gun (EB) system. In the boat

system, the crucible is heated by an electric power source, either by a resistor or

induction, and the deposition materials are heated by conduction energy. For the EB
system, an EB directly heats the material itself. Although the machinery cost for the EB
system is higher than that of the boat system, the EB system contains some
indispensable advantages, including (i) rapid heating; (ii) higher power; and (iii) precise
evaporation rate control (Hoffmann, 1994).

CVD, on the other hand, employs a different mechanism to deposit the ceramic
coatings onto the ﬁlm surface. The schematic model of the process is shown in Figure 2-
2. The precursor materials (ie. organosilicones etc.) are carried as gases, and react with

other material such as oxygen in the gas phase.

CVD

Running direction

<

 

 

 

 

 

 

Substrate
\ \ \ \ \ I
\I\I\I\ I I
' ' ’\ \ \ \’\/\/\ /\ \ \’ /\’\
Ceramic coating «1" («a «xx/‘2 xx x ,
I I I I I I I I IN \ \ D - - f
‘ * ‘~ *, * , , ,\ \ rivmg orce

x
x
I II II
xxxx
III I I II
xxxxx xx '
x x’x’x’x’x x’ x’x x x I I ' ‘
IIIIII I::IIII I Chemlcai reaCtIOII
xxxxx x
IIIIIWIIII
xxx xxxxx x

II IIIIII
\\\\\\
IWIIIII

#7 E

Source material

\I

Figure 2-2 Schematic Model of CVD

The ﬁnal reaction products then deposit onto the surface of the ﬁlm. To promote
the gas phase reactions, a plasma state is usually employed. Because CVD does not use a
high temperature, the ﬁlm will not be damaged. Also, CVD does not need an ultra vacuum
state (less than 10'8 torr), which is required in the PVD process. This can contribute to
reduced machinery cost. The machine or line speed, however, is not very high, and makes
the ﬁnal cost of the product high. Table 2-1 summarizes the advantages and the

disadvantage of these two processes (Charoudi, 1991).

Table 2-1 Pros and cons of evaporation methods

 

 

 

 

 

PVD CVD
Barrier @ 0
Heat Resistance A O
Flexibility A O
Transparency O (D
Speed © A
Machinery Cost A Q
Total Cost O X

 

© excellent, 0 good, A fair, X poor

2-1-2. Development of Ceramic-Coated Film

A ceramic-coated ﬁlm for packaging purposes was initially developed in 1964 by
DuPont de Nemours and Company (Dupont, 1964) as a silicon oxide (SiOx)—coated PET
ﬁlm. Even though it was an epoch making invention, the product did not succeed in the
packaging market because of its high cost. In the mid 19805, the Toppan, Toyo, and
Ajinomoto companies cooperated to create the future for ceramic-coated ﬁlm (Sakamaki,

1989). The joint venture aimed to develop a new ﬁlm, which was called by the trade name

of GL ﬁlm, for microwaveable pouches used for retortable foods. GL ﬁlm and several
similar products, however, enjoyed success in a relatively small market, which included
retort foods, laminated tubes, and liquid containers. However, because of the negatively
perceived yellowish color of the ﬁlm, the niche market was limited to less than $100
million per year (Charoudi, 1991).

These ﬁlms, which can be called the “ﬁrst generation” of SiOx-coated ﬁlms, had
some essential problems such as high cost, the yellowish color and inadequate barrier
properties. To solve these problems, two major efforts emerged in research and
development ﬁeld. One aim was to reduce the cost by using PVD, and the other was to

remove the yellowish color by using CVD (Kelly, 1993). The resultant modiﬁed SiOx-

coated ﬁlms can be called the “second generation” of SiOx-coated ﬁlms.

2-1-3. SiOx-Coated Films

Using the PVD method, SiOx coatings are generally produced by evaporation of
silicon monoxide (SiO), either by the boat system or by the EB system. The vapor of SiO
is oxidized in a controlled reactive atmosphere to achieve an average degree of oxidation
between x = 1.5 and 1.8. The reason the evaporated material has excess Si atoms in the
matrix is that if the matrix has been made to be stoichiometrically correct (i.e. silicon

dioxide (Si02)) by the PVD method, the matrix will have a low packing density (Figure 2-

3), which leads to high permeation rates (Hlavac, 1983). The stoichiometrically

Si atom

O atom

Si02 (crystalline) d=2.65 g/cm 3
(stoichometrically correct)

 

Si02 (amrphous) d=2.20 g/cm 3
(still stoichometrically correct)

  
   

Si atom
with dangling bond

SiOx (actual) d=2.24 g/cm 3
(stoichometrically incorrect)

Figure 2-3 Matrix of Si02 and SiOx

10

insufﬁcient structure results in Si atoms with dangling bonds in the coating structure. The
presence of the dangling bonds in the matrix causes the slightly yellowish appearance of
the resultant SiOx-coated structure (Kaihou, 1989).

The properties of the ﬁrst generation of SiOx coating are strongly inﬂuenced by
the layer thickness and the chemical composition (Charoudi, 1991). To maintain good
barrier properties, the thickness of the SiOx-layer needed to be over 1600 A for a retort
grade SiOx-ﬁlm.

Second generation SiOx-coated ﬁlms have succeeded in providing transparent
barrier ﬁlms for the packaging market. Signiﬁcant efforts have been made to improve the
coating technique, and the thickness of the SiOx-layer has been reduced to 400 A or less.
This has resulted in a clearer ﬁlm, and a higher speed operation, which has reduced the
cost of producing the SiOx-coated ﬁlms.

For the substrate of the PVD SiOx coatings, PET is preferred, and the resultant
SiOx-coated PET ﬁlms show excellent barrier characteristics to oxygen and water vapor
(Imai, 1998). Nylons and biaxially oriented polypropylene (BOPP) have also been
evaluated as substrates, but have not as yet afforded good barrier properties, due to their
lower thermal stability compared to PET.

Since the late 1980's, there has been a signiﬁcant effort to improve the clarity of
SiOx-coated ﬁlms, especially by using the plasma enhanced chemical vapor deposition
(PECVD) process (Nelson, 1993). Organosilicones like 1,1,3,3-tetramethyldisiloxane

(TMDSO) or hexamethyldisiloxane (HMDSO), as non-toxic liquid process monomers,

ll

are generally employed as a source material. The monomers are evaporated into an electric
ﬁeld to dissociate and ionize. These reactions result in the deposition of an Si02 coating
onto the substrate surface, while various gas phase byproducts of the reactions are
pumped away.

This method provides a number of advantages over the PVD method. Because
chemical reactions are occurring in this process, the deposited materials are packed tightly
and are stoichiometrically correct . Because of these factors, the ﬁlm is clearer than the
PVD processed ﬁlm. The resistance to elongation is also better than that of the PVD
processed ﬁlm (Nelson, 1993). Since PECVD employs a lower temperature process,
various substrate ﬁlms (i.e. PET, BOPP, Nylons, etc.) can be utilized (Y amamoto, 1998).
Although this process has these advantages, slow machine speed makes the ﬁlm
expensive and it remains commercially unsuccessful.

During recent years, the second generation of SiOx-coated ﬁlms has become
established as a high quality, transparent barrier ﬁlm. However, compared to the entire
high barrier packaging market, the market share of SiOx-coated ﬁlms is still not very high.
There are several reasons for this, but the main problems are that the SiOx-coated ﬁlms

still have a higher cost than PVdC-coated ﬁlms and they still have a yellowish color, even

though the color is much better than the ﬁrst generation's (Hoffmann, 1994).

2-1-4. Ale3-Coated Films

One of the latest efforts to improve the barrier characteristics and the clarity of

ceramic-coated ﬁlm is the development of aluminum oxide (alumina = A1203)-coated ﬁlm.
Ale3-coated ﬁlm may provide a solution to these problems. Either a direct

evaporation system or a reactive evaporation system, which are analogous to PVD

methods, can be employed to evaporate the ceramic coating. In A1203 coating, direct
evaporation A1203 is evaporated from an A1203 solid source. On the other hand, reactive

evaporation uses aluminum and oxygen as source materials and requires their reaction
either in the gas phase or at the surface of the substrate (Schiller et al., 1993) .
For the direct evaporation method, the EB system is necessary, due to the high

vaporizing temperature of A1203, which is about 2400 K. The substrate to be coated

passes a cooling drum arranged above the evaporator crucible. Although this process

appears to simply involve evaporation of A1203, the stoichiometry of the deposited
material is still a concern, because part of the A1203 will split up into suboxides, which

will make the ﬁlm darker during the evaporation. To obtain transparent or
stoichiometrically correct ﬁlm, additional admission of oxygen between the crucible and
substrate is required.

Because of its relatively low evaporation temperature, aluminum can be
evaporated either by the boat system or the EB system for reactive evaporation. The
oxygen stream, which reacts with the evaporated aluminum, must remain constant over

time and area. It was originally thought that the process involved two sequential

13

reactions, condensation of aluminum on the substrate followed by surface oxidation
(Kelly, 1994). Recently, some kinetic and surface topography studies have suggested that
the oxidation reaction occurs primarily in the gas phase, and the oxidized particles are
then deposited onto the substrate (Kelly, 1994). The stoichiometry of the deposited
material (i.e. oxygen : aluminum ratio) can be varied. A small improvement in the oxygen

barrier is observed with increasing aluminum content.

2-1-5. Ceramic Mixture-Coated Films

Not only single oxide (SiOx , A1203 etc.) coated ﬁlm, but also multi-oxide coated
ﬁlms have been prepared to improve barrier properties (Phillips, 1993). PVD evaporated
SiOx coatings have a contradictory issue, which is that if SiOx-coated ﬁlm is to maintain
high barrier properties, the ﬁlm will have a yellowish color. On the other hand, if it has a
clear appearance, the ﬁlm will have higher permeation rate, because of the lower density
packing formation (Hlvac, 1983).

To solve this contradiction, aluminum oxide, yttrium oxide, tin oxide, and a
mixture of Si02 and MgO have been proposed as nucleation layers for Si02, whose color
is water clear. The idea was to provide a preferable surface for Si02 deposition, so it
would have a dense formation. These materials have not been launched commercially, but

they have a lot of potential for success in the transparent high barrier ﬁlm market

(Charoudi, 1991 ).

l4

2-2. Properties of Ceramic-Coated Films
2-2-1. Oxygen and Moisture Barrier

Excellent oxygen and moisture barrier characteristics are ﬁmdamental properties
for clear, high barrier ﬁlms. PVdC coating onto plastic ﬁlms has been widely used as a
mainstay to increase the barrier properties of plastic ﬁlms. However, because PVdC
coatings contain chlorine, which causes acid rain and dioxin formation when incinerated,
there is an ongoing trend in both Europe and Japan to phase out PVdC-coated ﬁlms
(Kelly, 1994). In fact, some supermarkets in Europe have already banned the use of
PVdC packaging materials (Kelly, 1993).

With increasing pressure from consumers and government regulations, the food-
packaging industry has been seeking new materials with environmental safety in mind.
The most favorable choices for alternative chlorine-free barrier materials are either
ceramic-coated ﬁlms or EVOH ﬁlm. Both ﬁlms can provide excellent oxygen barrier
properties at low humidity conditions. EVOH, however, loses its barrier properties in a
high-humidity environment. Ceramic-coated PET does not show a loss of barrier
performance at elevated humidity levels (Imai, 1998).

The oxygen barrier properties of ceramic-coated PET ﬁlms compare favorably
with PVdC-coated PET or EVOH ﬁlms. As a transparent high barrier packaging material,
this barrier level allows a long shelf life for most processed foods.

The existence of a polyoleﬁn sealant layer , which provides a relatively good

moisture barrier, with PVdC-coated ﬁlms or EVOH based barrier structures makes these

15

ﬂexible packaging structures good moisture barriers (i.e. water vapor transmission rate
(WV TR) < 3 g/m2.day. However, it is very hard to attain a WVTR of 1 g/m2.day or less,
which is required for dry food packages using existing clear ﬂexible materials. The
ceramic-coated PET structures have a WVTR below this level, at any and all conditions.
Oxygen and moisture barrier properties of typical barrier ﬁlms are summarized in Figure

2-4 (Imai, 1998).

2-2-2. Organic Vapor Barrier

Flavors are complex systems that consist of a number of different classes of
volatile organic compounds. These compounds may be sorbed by the packaging material
or may permeate through it. These interactions may cause a decrease in ﬂavor
components and can result in an unbalanced proﬁle or undesirable ﬂavor (Kail, 1984).

There are three main phenomena which occur between ﬂavor or interactive
compounds and packaging materials that may result in a change in the ﬂavor proﬁle:
scalping, permeation, and migration. Scalping is the loss of ﬂavor components due to their
sorption by the packaging material as a result of those components being soluble in the
packaging material. Permeation is the movement of volatile compounds through the
packaging material, while migration is the movement of low molecular weight components
from the packaging material itself into the product. Among these mass transfer processes,
scalping generally plays a major role in ﬂavor deterioration for products in ﬂexible
packaging, because a large quantity of aroma compounds can readily dissolve in the

sealant material of the ﬂexible package system (Ikegami et al., 1988; Rogers, 1959).

16

WVTR (g/mz.day@37.8 °c, 90%RH)

 

 

Q . . 93:33:93} ::;.... ' ' -GL'AE,GL-E (lZum) (see Section 3

0 I 1 2 3 4
AI-foil (lSum)

P T (12p.m)/EVOH (15pm)/LLDPE (30pm)

Ceramic coated PET (12pm)
(First Generation)

I

   
 
   

 

. . .
.-..-.. -.-..- .....

- nu:
.......

PVdC- oated PET

.........

Meta- ' PVdC-coated BOPP
(1 (20pm)

iiiii , . , Ceramic coated PET (12um)
(Second Generation)

   
  

Ll-l)

 

CTR (cc/m?day.atm@24.8 °C, 0%RH)

Figure 2-4 Oxygen and water vapor barrier properties
of high barrier performance ﬁlms

17

S

Even though scalping has been the primary area of study of ﬂavor changes due to
loss of ﬂavor components, losses resulting from permeation cannot be neglected.
Moreover, because the ceramic layers of the ceramic-coated PET are very thin (less than
400 A), scalping by the ceramic layer is not likely to be a major problem. Since the
ceramic layers are so vital to barrier properties for ceramic-coated ﬁlms, studies
concerning permeation of organic penetrants through such barrier structures should be
given more attention.

Unlike simple gas or moisture vapor barriers measurements, a ﬂavor's intensity is
subjective. For this reason, sensory evaluations have been carried out to determine ﬂavor
qualities (Kail, 1984; Allison, 1985). These evaluations, however, require well-trained
panelists.

Using an objective instrumental approach, there had been no standard test method
to measure the organic vapor permeability of polymer membranes, until 1997. In
addition, it was difficult to measure the barrier properties of each component of a ﬂavor
or aroma proﬁle due to a lack of appropriate instruments and the low level of permeation
through the high barrier materials.

For these reasons, the oxygen barrier properties of barrier ﬁlms have been used to
estimate the ﬂavor barrier properties in the packaging ﬁeld. The general consensus is that
ﬁlms with good oxygen barrier properties will also have good ﬂavor barrier properties,
but this statement is not correct in some cases. For example, polystyrene (PS) has a high
oxygen transmission rate. In the case of aromas, however, PS shows better barrier

properties than even poly-vinilydene chloride (PVdC). Because PS is a highly glassy

18

polymer, larger permeants, such as organic vapors, tend to be obstructed the permeation
by the glassy matrix (DeLassus, 1993). This illustrates the danger of estimating ﬂavor
barrier properties on oxygen barrier properties alone. Therefore, it is essential to develop
direct methods for measuring ﬂavor barrier properties.

In order to measure organic vapor permeability of high barrier ﬁlms,
improvements in the measuring systems have been necessary. One early study employed
a quasi-isostatic procedure to determine the ethyl acetate permeability of SiOx-coated
ﬁlm (Sajiki, 1991). In this procedure, the lower concentration chamber of the permeation
cell is initially ﬁlled with a carrier gas and completely closed. After introduced the gas or
vapor into the higher concentration chamber, the permeants then accumulates in the lower
concentration cell chamber and is monitored at predetermined time intervals by gas

chromatography. This early study showed that SiOx-coated PET exhibited excellent
barrier properties (e. g. permeance = 1.9 x 10'19 kg.m2.sec.Pa@65°C) for ethyl acetate.

This barrier level compares favorably with that of ethylene vinyl alcohol copolymer
(EVOH) ﬁlm. In addition, the results suggested that the temperature effect on
permeability of SiOx-coated PET was much less than that of EVOH.

Recently, rapid evaluation of organic compound permeation through polymer
membranes was made possible with the introduction of two commercially available
systems. These systems are the MASZOOOTM Organic Permeation Detection System
(Testing Machines Inc, Amityville, NY), and the AROMATRAN TM Permeation Test

System (Modern Controls Inc., Minneapolis, MN). Both systems follow a newly

l9

introduced standard (ASTM F1769-97) (ASTM, 1997a) and employ an isostatic test
procedure. The details of these systems will be discussed in Sections 2-4-2 and 2-4-3.
Using the MASZOOOTM, a series of commodity packaging ﬁlms have been investigated for
their organic vapor barrier properties (Huang and Giacin, 1998). Chang (1996) also used

the MASZOOOTM and studied the permeability of a-pinene through PET / SiOx-coated
PET laminated ﬁlms. Results showed the SiOx-coated PET-based laminated ﬁlms had
better barrier properties (e. g. permeance = 1.8 x 10'17 kg,m2.s.Pa@60°C) than EVOH-

based laminated ﬁlms.

2-2-3. Flex Resistance

For most products in ﬂexible packages, the ability of the package to withstand
ﬂex cracking is indispensable. Even though the packaging ﬁlm may have excellent barrier
properties as a ﬂat ﬁlm, if the ﬁlm has no durability for ﬂex cracking, which can occur
during ﬁlling, packaging, or distributing, the barrier properties as a ﬂat ﬁlm are
meaningless (Oliveira, 1997). One way of assessing resistance to ﬂex cracking is by the
Gelbo ﬂex test (ASTM F3 92-74) (ASTM, 1987).

As the instrument cycles in the Gelbo ﬂex test, its movable head compresses the

sample, while rotating one end 440 °. It can be adjusted to simulate any degree of ﬂex

abuse by changing the ﬂex cycles.
The oxygen barrier properties of aluminum metallized ﬁlms and the ﬁrst

generation of SiOx-coated ﬁlms dropped signiﬁcantly when they were abused by the

20

Gelbo ﬂex tester. In contrast, the moisture barrier properties did not deteriorate
signiﬁcantly (Marzolf, 1981). These results could be due to the generation of pin-holes
during ﬂexure, but the speciﬁc cause of barrier property loss has not been resolved.

A number of efforts have been made to improve the ﬂex resistance of ceramic-
coated ﬁlms. For example, using the CVD method, highly ﬂex resistant ﬁlms have been
made (Nelson, 1993). The mechanism of this improvement is still not clear, but one
possibility is that the contamination of carbon elements from the gas source may make
the ceramic layer ﬂexible. Improvements and developments in new evaporation
techniques have also been employed for the PVD method, and the resultant PVD-
processed ceramic-coated layers have shown much better ﬂex durability than the ﬁrst

generation ﬁlms.

2-3. General Theory Related to the Permeation Process

2-3-1. Permeation Steps

Permeability is often referred to as the ease of transmission of gasses or vapors
through a resisting material, which has no macroscopic pores (Rogers, 1985). The
transport of a gas or vapor through polymeric ﬁlms commonly used in packaging
typically involves the activated diffusion process. The process involved three steps:
(1) absorption of the permeating species, in which the gas or vapor dissolves into the
polymer matrix at the high penetrant concentration surface,

(2) diffusion through the polymer wall along a concentration gradient, and

(3) desorption from the surface at the lower concentration (Rogers, 1985).

2-3-2. Sorption Mechanism

The steps of adsorption and desorption can be categorized as having the same
mechanism and may be described by the sorption process. Whenever a gas is in contact
with a solid, there will be an equilibrium established between the molecules in the gas
phase and those which are bound to the surface of the solid. The amount of gas molecules
sorbed onto the polymer ﬁlm surface will depend on the pressure of the gas above the
ﬁlm surface and the temperature of the system. Increasing the gas pressure increases the
collision rates of the sorbate molecules into the polymer membrane and also increases the
amount sorbed. On the other hand, higher temperatures tend to increase the internal
energy of the sorbate molecules, thus decreasing their sorption by the polymer membrane

(Rogers, 1965).

2-3-3. Diffusion Mechanism

Some domains of a polymer are a randomly arranged mass of macromolecule
chains, which consist of a network containing voids or holes. The diffusion process is the
result of polymer molecules having a kinetic agitation or thermal motion. In other words,
it will depend on the number, size, and distribution of pre-existing holes, and it also

depends on the ease of hole formation (Rogers, 1965).

The diﬂirsion of permanent gases (i.e. hydrogen, oxygen, nitrogen), which have a
much smaller molecular size than the monomer unit of a given polymer, occurs readily
since the rotational oscillation of one or two monomer units will provide sufficient cross
sectional area for the diffusant molecule. This mechanism follows both Fick's ﬁrst law
and Henry's law, and the diffusion coefficient is independent of concentration (Barrer,
1939)

On the other hand, the diffusion mechanism for molecules larger in size than the
monomer unit of a polymer requires a cooperative movement by the micro-Brownian
motion of several monomer units in order to take place. Water vapor or organic vapor,
which interact strongly with the polymer, are categorized as this type of permeant
molecule. As a result, their diffusion coefﬁcient is primarily controlled by the mobility of
the polymer segmental unit, which is directly proportional to the permeant concentration

and temperature (Meares, 1965).

2-3-4. Permeation Mechanism of Ceramic-Coated Films

The mechanism of permeation of gases and vapors through ceramic-coated ﬁlms is
not well known. Two different mechanisms, however, have been proposed to explain the
permeation process. The basic concept of both mechanisms is “the defects leading
permeation” model, where defects in the ceramic coating result in the permeation of gas or
vapor. In other words, if there is no defects, the ceramic coating will provide a perfect

barrier to permeants.

One of the proposed permeation mechanisms requires the existence of macro
defects, such as pinholes in the ﬁlm, which leads permeation. In fact, the permeation path
of a permeant through aluminum metallized ﬁlm is primarily determined by existing
pinholes in the aluminum coating (Jamieson and Windel, 1983) . The size of the pinholes

varies, but most pinholes of 1 to 2 pm in size which are associated with the presence of

dust particles on the ﬁlm surface. When the ﬁlm is metallized, dust particles create
shadows on the ﬁlm, thus when the dust particle falls, it creates a un-metallized area (i.e.
pinhole) on the ﬁlm surface. The area of the total pinholes controls the rate of permeation
in this mechanism.

For ceramic-coated ﬁlms, several studies have been carried out to apply the

“pinhole theory”. Activation energies of transport of oxygen (A Ep(02)) through SiOx-

coated ﬁlms were determined to describe the permeation mechanism (Sajiki, 1991;
Tropsha, 1997). Tropsha suggested that if the ceramic layer is perfectly defect-ﬂee, the

A Ep(oz) between ceramic-coated material and the non—coated substrate should be
different. In these studies, however, the A Ep(02) of both coated and non-coated materials

showed similar values. This result suggested oxygen mainly permeated through the PET
matrix via defects or “non-continuous” areas in the oxide layer.

The other route proposed to explain the permeation mechanism for ceramic-
coated ﬁlms is through imperfections in the network of the ceramic coating at the

molecular level (micro-defects) (Norton, 1953). For example, the Si02 matrix is, in

theory, comprised of a crystal lattice structure, which is composed of hexagon type in

crystal plane. Since Si atoms and oxygen atoms have atomic radii of 0.41 A and 1.4 A
respectively, the size of the voids within the matrix is not large enough to accommodate
gases (e. g., oxygen molecules or even helium atoms which have atomic radius of 1.1 A). In
reality, however, the matrix has similar or larger voids or holes (Figure 2-3) than an ideal
crystal lattice. To conﬁrm this micro-defect mechanism, the effect of permeant size has
been studied (Norton, 1953). The size of the permeant was found to be directly
proportional to the permeation rate for a series of inert gases. For instance, the
transmission of helium and neon gases through silica glass, which is an amorphous
structure and has micro-defects which could be measured. Argon gas, however, did not
afford any measurable transmission rate through the silica glass. The atomic diameter of

argon is 3.2 A, which is much smaller than macro-defects which are estimated to be
between 1 to 2 pm in size.

These proposed mechanisms were based on the permeation of simple gases.
However, the permeation of moisture vapor or organic vapors may have a more
complicated mechanism, due to the polarity of the permeant and the potential interaction

of the permeant with the ceramic layers. However, few studies were reported in this ﬁeld.

2-4. Factors Affecting the Permeation Processes
2-4-1. Nature of Permeants

There are three signiﬁcant factors, which affect the permeation processes, related

to the nature of the permeants. (1) The size of a permeant molecule. In general, an

25

increase in penetrant size results in a increase in the solubility coefficient (S) value,
because larger molecule has higher cohesive force to substrate surface. In same time,
however, the diffusion coefficient (D) value tends to decrease with increasing the
penetrant size. Kosinowski (1986) studied the effect of n-alkane permeant size on mass
transfer processes through low density polyethylene. The diffusion coefﬁcients
decreased with increasing numbers of carbon atoms.

Because the permeability is decided by the product of these two parameters (D
and S), the effect of permeant size is canceled out and is much less than what its
expected. Even though, the effect of permeant size is not signiﬁcant, the effect still exists
in most permeation processes. The uncertainty of molecule diameter by different deciding
methods, however, may scatter the effect of permeant size to its permeability.

(2) Shape of the penetrant is other factor which will effect permeability. Berens
and Hophenberg (1982) reported that the anisometric molecules may permeate through
polymers along their long axes or smallest cross sectional area. Rogers (1985) found a
spherical molecule has lower permeability than a straight chain molecule, if they have
similar chemical nature.

(3) The third major factor of how the permeant nature can effect permeability is
the polarity of the penetrant. In general, if the polymer and the permeant have similar
chemical composition or polarity, the permeability is expected to be high. For example,
since polyethylene (PE) is a non-polar polymer, the permeation rate of PE to non-polar

gases (e. g., oxygen, nitrogen) is high (Rogers, 1985).

2-4-2. Measuring Temperature

Many factors related with the measuring conditions can contribute to effect the
permeation process. The temperature of the test cell, however, can be one of the most
important factors.

Since permeability is the product of the solubility coefficient (S) and the diffusion
coefficient (D), the effect of measuring temperature on these coefficients have been
widely studied (Rogers, 1985; Huang and Giacin, 1998). In general, the solubility
coefﬁcient (S) decreases or increases with increasing temperature, depending on the
physical state of the permeants. The diffusion coefficient (D), however, increases with
increasing temperature, as an increase in temperature provides the energy for the
segmental motion and hole formation, which increases the free volume within the polymer
bulk phase (Crank, 1956) .

The Arrhenius equations can be introduced to explain the temperature dependence

of solubility and diffusivity.

 

 

-AH
5m = SOxexp( 3) Eq. 2.11
(-ED)
D(T) = R0 ex E .2-2
x P RT q

where ABS and ED are heat of solution and diffusion activation energy, respectively. So

and D0 are pre-exponential terms, R is the gas constant and T is the absolute temperature.

The Arrhenius expression can also express the temperature dependence of permeance.

27

(-ER)

R(r) = Rox exp
RT

 

Eq. 2-3

where ER is the activation energy for permeation, R0 is the pre-exponential term, R is the
gas constant and T is the absolute temperature.

The permeation behavior is generally quite different below and above the glass
transition temperature (Tg). Above the Tg of the polymer, enough energy is provided to
produce the micro-Brownian motion and the chains can experience segmental motion,
while below the Tg, the polymer chains are ﬁxed in a speciﬁc conformation. Therefore, in

general, at temperatures above Tg, the permeability coefficient is more temperature

dependent, but below Tg, it is less temperature dependent.

2-5. Permeability Measurement
2-5-1. Mathematical Approach for Permeation Process

The transmission rate or ﬂux (F) of a permeant through a polymer membrane can
be deﬁned as the amount passing through a surface of unit area normal to the direction of
ﬂow during unit time (Rogers, 1985).

F = Q/At Eq. 2-4

where Q is the total amount of permeant which has passed through the area (A) during
time (t). The transfer of diﬁ’usant through a unit area can be expressed as being
proportional to the negative gradient of concentration at any point in the polymer. This

can be described by Fick's ﬁrst law of diffusion (Crank, 1956).

28

F = -D (dC/dx) Eq. 2-5
where D is the diffusion coefficient with units of (length)2(time)", x is the length in the
direction in which transport of the permeant occurs, and C is the concentration of the
permeant in the polymer. Therefore, dC/dx deﬁnes the concentration gradient of the
permeant across a thickness dx. This law can be applied to the diffusion in the steady
state. Under steady state conditions, a gas or vapor diffuses through a polymer at a
constant rate, if a constant concentration or partial pressure difference is maintained
across the polymer (Crank, 1956).

Fick's second law describes non-steady state diffusion, where the concentration of
the diffusing substance is changing with time. The mathematical treatment of Fick's
second law is described by Eq. 2-6 (Crank, 1956).

dC/dt = D (dzc/dxz) Eq. 2-6
There are a number of solutions to this equation, which have been derived for various
boundary conditions (Crank, 1956).

The permeation of a gas or vapor through a polymeric material is usually
described by the permeability coefﬁcient (P), which can be quantiﬁed from knowledge of
the diffusion coefficient (D) and solubility coefficient (S), as described by Eq. 2-7.

P = D x S Eq. 2-7

The diffusion coefﬁcient is a kinetic parameter and is a measurement of how

rapidly the transport process will occur and indicates the ease with which a penetrant

molecule moves within the polymer matrix.

29

The solubility coefficient is a thermodynamic parameter and is a measurement of
the concentration of penetrant molecules sorbed in the polymer matrix. The solubility
coefﬁcient is an equilibrium partition coefficient describing the distribution of the
penetrant between polymer matrix and vapor phase.

Eq. 2-5 can be integrated, where D is independent of concentration, to give:

F=D(C1-C2)/L Eq. 2-8
where C1 and C2 are the steady state concentrations of the permeant at the two surfaces
of the ﬁlm and L is the thickness of the ﬁlm.
Eq. 2-8 can also be rewritten by substituting for F, using Eq. 2-4, to give Eq. 2-9.
Q=D(C 1 -C2)At/L Eq. 2-9
This enables calculation of the quantity of permeant diffusing through a polymer of area
A in time t.

For gases and vapors, the concentration of the permeant in the polymer, at the
high concentration surface, is proportional to the concentration or partial pressure of
permeant in the surrounding gas phase. This relationship can be expressed by Henry's
law:

C = Sp Eq. 2-10
where S is the solubility coefficient of the permeant in the polymer and p is the partial
pressure of the gas or vapor. By combining Eq. 2-9 and 2-10, it gives:
Q=DS(p1-p2)At/L Eq. 2-11
Since the relationship between P, D and S is expressed as Eq. 2-7, Eq. 2-11 can be

rewritten as Eq. 2-12 which enables determination of P from experimental data.

30

P = QL/At(p1-p2) Eq. 2-12

2-5—2. Isostatic Method

In the isostatic method, a penetrant ﬂows continually through the high
concentration cell chamber, and an inert carrier gas ﬂows continually through the low
concentration cell chamber. Therefore, the total gas pressure on the two sides of the test
ﬁlm is essentially equal. The partial pressure gradient of the permeant provides a driving
force for permeation. The isostatic method allows for the continuous monitoring of the
transmission rate of organic vapors through test ﬁlms from the initial time zero until
steady state condition.

The permeability coefﬁcient P is calculated from the transmission rate at steady
state by the expression:

P = [C]fL/A(pl-p2) Eq. 2-13
where [C] is the steady state concentration of permeant conveyed to the detector, in mass
per unit volume; and f is the rate of carrier gas ﬂow in the low concentration cell, in
volume per unit of time.

If the ﬁlm is essentially free of penetrant, the diffusion coefficient D can be
calculated by:

D = MN. l99.t1/2 Eq. 2-14
where tm is the time required to reach a transmission rate value that is equal to half of

that at the steady state, in time units.

31

Using these principles, two commercial instruments, the MAS2000TM and the
AROMATRAN TM series, are available for performing organic vapor permeability
measurements. Since these units have become commercially available, the ease of
measurement of organic vapor permeability has been improved, dramatically.

Both systems employ a ﬂame ionized detector (F ID), which provides sensitivity
levels in the low parts per billion region, to quantify the level of penetrant which has
permeated. Both systems allow rapid evaluation of the diffusion, solubility and

permeability coefﬁcients of organic vapors through packaging materials (Huang, 1996).

2-5-3. Dynamic Purge and Trap / Thermal Desorption Technique

The dynamic purge and trap / thermal desorption technique is one of the most
effective approaches to determine the permeation rate of high barrier ﬁlms. The technique
involves the use of absorbents to trap and to concentrate the permeant. The vapors
concentrated in the sorbent trap are subsequently recovered by a desorption system and
transferred directly to a gas chromatograph for analysis. The amount recovered per
trapping time is treated as the permeation rate.

The application of a dynamic purge and trap / thermal desorption procedure
coupled with the MAS2000TM Permeation Test System was developed and performed
(Chang, 1996). The permeance of a-pinene vapor through a series of high barrier

composite ﬁlms, including the ﬁrst generation of SiOx-coated PET, which could not be

tested by a normal isostatic procedure, was determined. The lowest detection sensitivity

32

of the dynamic purge and trap/thermal desorption procedure was found to be 0.2 ng/hr
which is three to four orders of magnitude less than the continuous ﬂow isostatic
procedure (Chang, 1996). The increased detection sensitivity of the method provides the
ability to determine the permeation rate of aroma / ﬂavor permeation rates of high barrier

ﬁlms.

33

Chapter 3

MATERIALS and METHODS

3-1. Materials and Apparatuses
3-1-1. Films and Laminated Structures

In this study, 2 different ﬁlms and one laminated structure were prepared to
investigate their permeability. All ceramic-coated ﬁlms was manufactured by the physical
vapor deposition method (see Section 2-1-1). The laminated structure was manufactured
by a solvent-lamination method using an urethane type adhesive.
1) Films:
a) Ale3-Coated PET ﬁlm (12 um):
GL-AE ﬁlm (TOPPAN PRINTING Co., Ltd, Tokyo, Japan)
b) SiOx-Coated PET ﬁlm (12 um):

GL-E ﬁlm (TOPPAN PRINTING Co., Ltd.)

2) Laminated structure:

a) N203-Coated PET ﬁlm (12 um) / LLDPE (40 um) laminated ﬁlms:
A1203-Coated PET; GL-AE ﬁlm (TOPPAN PRINTING Co., Ltd.)
LLDPE; (Tocelo Co., Ltd, Tokyo, Japan)

Adhesive; Urethane type

3-1-2. Permeants
D-limonene and ethyl acetate were selected as the organic vapor permeants. The

basic information of the permeants is summarized in Table 3-1.

34

Table 3-1 Basic information of the permeants
id-Limonene (Aldrich Chemical Co., Milwaukee, WI)

 

 

 

 

 

v Molecular structure (C10H16)
> Density at 25 °C 0.840 g/cc
Molecular weight 136.24
Boiling range 175.5-176 °C
Molar volume 162 cc/mole
Purity 97% _
Ethyl Acetate (Aldrich Chemical Co., Milwaukee, WQ—
. Molecular structure (CH3C02C2H5) ,
Density at 25 °C 0.894 g/cc
Molecular weight 88.11 »
Boiling range 77.1 °C
' Molar volume 98.56 cc/mole
Purity 99.9%

 

3-1-3. Solvents

Three different solvents were employed to prepare standard solutions. For the d-

limonene standard solution to calibrate both vapor pressure of permeation test and the

purge and trap / thermal desorption procedure, carbon tetrachloride was used. For the

ethyl acetate standard solutions used to calibrate the vapor pressure of the permeation

test, acetonitrile was used. For the ethyl acetate standard solutions used to calibrate the

purge and trap / thermal desorption procedure, dichlorobenzene was used. The basic

information of the solvents are summarized in Table 3-2.

35

Table 3-2 Basic information of the solvents
Carbon Tetrachloride (Mallinckrodt, IncLParis, KY)

 

 

 

Molecular structure (CCl4)
Density at 25 °C 1.585
Molecular weight 153 .84
Boiling range 76.3-76.8 °C
Purity 99.9%
Acetonitrile (EM Science, Gibbstown, NJ)
Molecular structure (CH3CN)
Density at 20 °C 0.786 g/cc
Molecular weight 41.05
Boiling range 81.6 °C
Purity 99.8%

 

Dichlorobenzene (Aldrich Chemical Co., Milwaukee, WI)

 

Molecular structure (C6C4C12)
Density at 25 °C 1.551 g/cc
Molecular weight 147
Boiling range 179 °C
Purity 99 %

 

 

3-1-4. Apparatuses

In this study, various apparatuses were employed. The types and suppliers of

these apparatuses are listed below:

(1) Thermal Desorption Apparatus:

a) Dynatherm 890/891 thermal desorption unit (Dynatherm, Kelton PA)

b) CarbotrapTM 300 multi-bed thermal desoption tubes; 6 mm OD. x 4 mm ID. x 11.5
cm length (Supelco Inc., Bellefonte, PA)

(2) Gas Chromatograph:

Hewlett Packard model 5890A interfaced with a HP 3395 integrator (Avondale, PA)
(3) Gas Chromatography Column (Fused Silica Capillary Column):

a) SPBTMS (non-polar bounded stationary phase) 30 m long, 0.32 mm ID, 1.0 pm ﬁlm

thickness (Supelco Inc., Bellefonte, PA)

36

b) SupelcowaxTM 10 (polar bounded stationary phase) 60 m long, 0.25 mm ID, 1.0 pm
ﬁlm thickness (Supelco Inc., Bellefonte, PA)

(4) Permeation Test Apparatus:

a) For organic vapor; (Measuring system follows ASTM F 1769-97 (ASTM, 1997a))
MASZOOOTM Organic Permeation Detection System (Testing Machines Inc., Amityville,
NY)

b) For oxygen; (Measuring system follows ASTM D 3985-81 (ASTM, 1981))
MOCON Ox-Tran 200 Permeability Tester (Modern Controls Inc., Minneapolis, MN)
c) For water vapor; (Measuring system follows ASTM F 1770-97 (ASTM, 1997b))
MOCON Perrnatran W 3/31 Permeability Tester (Modern Controls Inc.)

(5) Water Bath:

Endocal RTE-100 (NESLAB Instruments, Inc, Portsmouth, NH)

(6) Bubbler:

25 ml Standard Midget Bubbler (ACE Glass Incorporated, Vineland, NJ)

(7) Needle Valves:

Nupro M-Series (Nupro Co., Willoughby, OH)

(8) Fittings:

Swagelok Fitting (Supelco Inc., Bellefonte, PA)

(9) Electronic Mass Flow Meter:

Model Top-Trak 821 (Sierra Instruments, Carmel Valley, CA)

(10) Syringes:

a) 500 ml gas-tight syringe (Hamilton Co., Reno, NV)

b) 5 ul syringe (Hamilton Co., Reno, NV)

(1 1) Flex tester:

Gelbo Flex Tester Model # 5000 (Research & Testing Co., Inc, Hoboken, NJ)

(12) Optical microscope:

Olympus BH-2 (Olympus Optical Co., Ltd. , Tokyo, Japan)

37

3-2. Methods

3-2-1. Calibration Curve for Vapor Pressure

The concentration of the permeant vapor evaluated with the MASZOOOTM Organic
Permeation Detection System was determined by gas chromatography analysis. Standard
solutions of the compound in solvent (carbon tetrachloride for d-limonene, acetonitrile for
ethyl acetate) were prepared and a calibration curve for the permeant was constructed
according to the analytical conditions. Initial conditions and temperature programming of
gas chromatography are summarized in Table 3-3 and 3-4, respectively.

Table 3-3 Initial conditions of gas chromatograph

 

 

 

 

Compound d-limonene ethyl acetate
Column SPBTMS Supelcowaxmlo
Injection temperature(° C) 220 220
Detector temperature (°C) 250 250

Head pressure (psi) 10 ' 20

Total ﬂow port (split vent) (ml/min) 27.8 28.8
Septum purge (purge vent) (ml/min) 2.76 2.5
Helium ﬂow rate (ml/min) 1 1

 

Table 3-4 Temperature programming for gas chromatograph
for direct injection procedure. .

 

 

Compound d-limonene ethyl acetate -
Initial oven temperature (°C) 50 40

Initial time (min) 2 1

Rate (°C/min) 7 5

Final temperature (°C) 110 200

Final time (min) 0 10
.Rate.A‘.“’(°C/min) 3o -

Final temp A“) (° C) 200 -

Final time A“) (min) 3 -

Total run time (min) 16.58 43.00

 

 

 

(a) the second temperature programming cycle

38

 

The above conditions gave a retention time for d-limonene of 8.94 min and for
ethyl acetate 8.06 min. The quantity of permeant detected was determined by

multiplying the standard concentration (v/v) times the volume injected (1 pl), which is

then multiplied by the density of the permeants. The quantity injected plotted versus the
corresponding area response gave the calibration curve, which established the linearity
and sensitivity of the analysis for the respective permeant. The calibration proﬁles

obtained for compounds are shown in Appendix A.

3-2-2. Calibration Curves for Dynamic Purge and Trap /

Thermal Desorption Procedure

The Carbotrap 300TM adsorbent tube was selected for the present study because
its multi-bed adsorbent design allows trapping of various organic compounds of different
size and functionality.

The standard calibration curve for the test compound was obtained by the

following procedures. A 1 pl sample of a standard solution of the compounds in solvent

(carbon tetrachloride for d-limonene, dichlorobenzene for ethyl acetate) was directly
injected onto the sorption tube. The sorption tube was then inserted into the heating
chamber of the thermal desorption unit, which is directly interfaced to the column of the
gas chromatograph. SPBTMS fused silica capillary column was chosen for this analysis.
The test compound was thus desorbed by heating and then separated by GC. The test
conditions of the thermal desorption procedure and gas chromatography analysis are

summarized in Table 3-5 and 3-6, respectively.

39

Table 3-5 Conditions of thermal desorption unit.

 

Tube desorption chamber temp. (° C) 370
Valve compartment temp. (° C) 250
Transfer line temp. (° C) j _ 250
Tube preparation chamber temp. (0 C) 350
Desoption time (min) 8

Preparation time (min) 30
Desorption carrier gas ﬂow rate at ﬂow check port (ml/min) 9

Preparation carrier gas ﬂow at side port (ml/min) 15

 

 

Table 3-6 Temperature programming for gas chromatograph
for thermal desorption procedure.

 

 

 

 

Compound d-limonene ethyl acetate
Initial oven temperature (°C) 35 40
Initial time (min) 2 5

Rate (°C/min) , ‘ 20 . 30.
Final temperature (°C) 200 200
- Final time (min) 10 10
Total run time (mim 20.25 - 20.33

 

The above conditions gave a retention time of 6.75 min for d-limonene and 2.31
min for ethyl acetate. The quantity injected plotted versus the corresponding area
response gave the calibration curve, which established the linearity and sensitivity of the
analytical procedure. The calibration proﬁles obtained for the compounds by the thermal

desorption procedure are presented in Appendix B. After sample desorption, the sorbant

tubes were conditioned at 350 °C for 30 minutes for re-use.

3-2-3. Organic Vapor Permeability Test System

Permeability studies were carried out with the MASZOOOTM Organic Permeation
Detection System, which was modiﬁed with a device for trapping permeated organic
vapors. MAS2000TM Organic Permeation Detection System is based on an isostatic
permeation test procedure. This system allows for the continuous collection and

measurement of the permeation rate of the organic vapor through a polymer membrane.

The cell temperature can be accurately controlled from ambient to 200 °C.

Even though the MASZOOOTM Organic Permeation Detection System has an
incorporated detection system, this study required a sensitivity for the measurements
which was signiﬁcantly greater than that provided by the original isostatic system. In
order to obtain higher sensitivity, the test system employed a dynamic purge and trap
technique, which allowed accumulation of the permeated vapor. The sensitivity of this
system is two orders of magnitude greater than that of the MAS2000TM original isostatic
system (Chang, 1996). A bypass line was installed to convey the permeated vapor to the
sorption trap. The trapping system ensured that the low concentration cell chamber was
continuously ﬂushed with carrier gas and the permeated vapor was conveyed directly to
the trapping tube attached. The sorption trap (CarbotrapTM 300) was connected to the
exit port of the bypass line, which is incorporated onto the instrument chassis, via a 1/4”
thumb wheel Swegelok Fitting for easy removal. Figure 3-1 provides a schematic of the

permeation test and trap system.

41

83?? 3:53:53 nab 23 “~qu 383% mo aﬁwﬂc oﬂmEosom Tm PawE

53an

+

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.802
88%.: :6 sates E; as;
. . _ il‘l
. _ «3% L . M £5
1 .885
963 wcﬂoﬁsm tom
28an
oz II +
95 8:88 3256 N2
\\
00 1d

 

 

 

HOE—awed

42

3-2-4. Permeability Measurements
The permeability studies were carried out within a temperature range of 55 to 80

°C. Prior to initiating a test run, the test ﬁlm was conditioned at selected temperatures for

at least 6 hours to desorb residual monomer and other volatiles from the ﬁlm. For each
test run, a sample ﬁlm was cut, mounted on a paperboard ﬁlm holder with tape. Then the
sample was placed in the permeability cell with the PET side facing the sample gas
chamber (ﬁont cell) side. Therefore, the direction of permeation process was ﬁ'om PET
through the ceramic-coated layer (or LLDPE). The area of the test ﬁlm was 0.0081 m2.

A constant concentration of permeant vapor for the high concentration cell
chamber was produced by bubbling nitrogen through the liquid permeant. The ﬂow rate
of the line was 30 ml/min. The liquid permeant was contained in a 25 ml standard midget
bubbler which was placed in a temperature controlled water bath to generate a saturated
vapor pressure of desired level. While the organic vapor was generated at the selected
temperature @oint 1), the ﬁlm permeability was evaluated with the test cell maintained at
a higher temperature (point 2). The parameters of partial pressure, temperature,
concentration and volume for the organic vapor generator and test conditions are denoted
as p1, p2, T1, T2, v1 and v2, respectively (Huang and Giacin, 1998).

Assumptions:
1. At points 1 and 2, the mass ﬂows are equal. Therefore,
M1=M2=M Eq.3-1

where M is the mass ﬂow (mass/time).

43

2. The organic vapor pressure is usually very low, and therefore the mass ﬂow consists
primarily of canier gas (nitrogen). For this case, it can therefore be assumed that the

organic vapor behaves as an ideal gas, therefore:

[:2 '- F1 XI; Eq. 3-2
T1
and by deﬁnition:
0 = M/F Eq. 3-3

where F is the gas ﬂow rate (volume/time).
So, the penetrant vapor concentration at point 1: (Cl) is equal to M1/F1 and at point 2:
(C2) is equal to M2/Fz, therefore:

cl/cz = T1/T2 Eq. 3-4

When expressing c or p by the ideal gas law, at point 1:

 

 

P1=nXRlea R xml’dl: R xclel Eq.3-5
V1 MW V1 MW
and
32.42112 Eq_3_6
P1 €1XT1

where c = m/v = mass of permeant vapor per unit volume.

It follows therefore, that:

 

l T 1
pZ-plx{c X IXTZX )=pl Eq3-7
\ 72 C1XTI

So, at points 1 and 2:

* temperature are different,

* mass ﬂows (mass/time) are equal,
* ﬂow rates (volume/time) are different, and
* partial pressures are equal.
In order to conﬁrm the vapor pressure of a permeant, a gas sampling port was
installed between the bubbler and the test cell. To determine the speciﬁc vapor

concentration, a 50 ul sample was withdrawn from the sampling port and injected

directly into the gas chromatograph (GC) for quantiﬁcation. The GC analysis conditions
were the same as that for determining the saturated vapor pressure. The experimentally
determined saturated vapor pressures were compared with the interpolated values
obtained from Perry’s Chemical Handbook (1984) (see appendix C). In this comparison,

the following equation was used to calculate the saturated vapor pressure.

_CFxAUxRxT
’ wav

 

p Eq. 3-8

where:

p = partial pressure (Pa)

CF = calibration factor (g/AU)

AU = area unit response from integrator (AU)

R = gas constant, 6.236x107(mmHg.u1/ moleK)
T = temperature (K)

MW = molecular weight (g/mole)

V = gas sample volume (ul)

The accuracy of the vapor pressure values between experiment and literature was

within 15 % for d-limonene and 40 % for ethyl acetate. (see Appendix C)

45

3-2-5. Dynamic Purge and Trap / Thermal Desorption Procedure

Once the operational parameters of gas ﬂow rates, temperature, and vapor
pressure became stable, the permeation tests were started. After introducing permeant gas
into the high concentration cell, the switching valve was activated and the system was
operated in the bypass mode. In conducting a permeability run, the test ﬁlm is initially
exposed under isostatic conditions for a period of 72 hours at the required test
temperature and vapor pressure, during which time it is assumed that a concentration
gradient is established within the ﬁlm and a steady state transmission rate is attained.
Following exposure of the ﬁlm for a 72 hour period, the level of sorbant accumulated in
the sorption tube for a predetermined time interval was determined. For high barrier ﬁlm

like A1203-coated PET ﬁlm, a long trapping time was employed (3 hours for d-limonene,

and 1 hour for ethyl acetate). The sorbant tube was then removed and replaced by a new
trapping tube and permeated vapor again accumulated for quantiﬁcation. Sampling of the
trapping tube was conducted at least twice a day.

The sorbant tube removed from the permeability test system was then
immediately transferred to a thermal desoption unit (Dynatherm 890/891 (Dynatherm,
Kelton PA)), which thermally desorbs any organic volatiles from the sorbant tube and
transfers them to the gas chromatograph for quantiﬁcation. The sorbed volatiles were

desorbed by heating for 8 minutes at 370 °C, with the valve and transfer line held at 250
°C to maintain the desorbed compounds in the vapor phase, while being transferred to the

GC. Helium was used as a carrier gas through the thermal desorption unit, at a ﬂow rate

of 9 ml/min for 40 psi. After sample desorption, the sorbant tubes were conditioned at

350 °C for 30 minutes to re-use. The trapping and subsequent thermal desorption of

volatiles allows their effective release, undiluted, and allows monitoring of otherwise
undetectable levels of permeant concentration.

This procedure was repeated for at least 3 continuous days. The results of the
detection levels were then compared, to insure the system was at steady state.

After each test run, the permeation cell, switching valve and bypass line of the

MASZOOOTM were heated for at least 3 days at 100 °C to desorb up any residual volatiles

in the system.

3-2-6. Analysis of Test Permeant by Thermal Desorption
/ Gas Chromatography Procedure

GC analysis was carried out with a Hewlett-Packard Model 5890 gas
chromatograph, equipped with a ﬂame ionization detector and interfaced to a Hewlett-
Packard Model 3395 integrator, for quantiﬁcation of permeated vapor. The GC condition
were the same as that for determining the calibration curve. The permeance was

determined by substitution into the following equation.

CF xAU

- Eq. 3-9
I x A x Ap
where;
R = permeance (kg/seem2 .Pa)
AU = area unit response from integrator (AU)
A = exposed area of the ﬁlm
CF = calibration factor for dynamic purge and trap/thermal desorption (g/AU)
t = trapping time (sec)

47

Ap = vapor pressure gradient (Pa)

3-2-7. Analysis of Ceramic Coating Surface by Optical Microscopy
Optical microscopic observations were conducted on the ceramic coating surface
to investigate the effect of physical abuse on the integrity of the ceramic coatings, by
using an optical microscope (Olympus BH-2 (Olympus Optical Co., Ltd. , Tokyo,
Japan». Specimens (1 x 2 cm in size) were cut from the center of the abused ﬁlm. The
samples were mounted with the ceramic coating surface in contact with the glass
microscope slide, with light transmission from the underside. Photographs of the ceramic
coating surface were taken with Polaroid 52 PolaPan 4 x 5 instant sheet ﬁlm at 200 and

500 times magniﬁcations.

Chapter 4

RESULTS and DISCUSSIONS

4-1. Estimation of the Detection Sensitivity Limit

The lowest detection limit for the dynamic purge and trap / thermal desorption
procedure depends upon the trapping time and the gas chromatographic output signal
from the permeant. Applying the dynamic purge and trap / thermal desorption procedure,
the lowest signal output from the gas chromatograph was assumed to be around 5,000
area units. In this study, the longest trapping time was 3 hours for d-limonene. Thus, the

minimum measurable transmission rate of d-limonene was calculated as follows;

5000/1U 2.07 10'14 _ _ k
x x g =3.45x10 ”£- =9.58x1018-g-
3hr AU hr sec

 

The 2.07x10'l4 g/AU value is the calibration factor for d-limonene, determined by the
thermal desorption procedure (Appendix B).
According to previous studies, the minimum measurable transmission rate

estimated by operating the MAS2000TM Organic Permeation Detection System in the

isostatic mode was 6.86 x 10‘14 kg/sec (Laoharavee, 1997). Therefore, the dynamic purge
and trap / thermal desorption procedure has a sensitivity 4 orders of magnitude greater

than that of the original isostatic procedure.

49

4-2. Time to Reach Steady State

When the permeation behavior of a gas through a ﬁlm reaches an equilibrium state,
the permeability rate also attains an equilibrium state, which is called the steady state.
For conducting typical permeability measurement, N203-coated PET ﬁlm was initially
exposed under isostatic conditions for a period of 72 hours, and then the sorption tube
was set to accumulate the penetrants (see Section 3-2-5). The transmission rate was
assumed to reach steady state within 72 hours.

In order to verify this assumption, the permeability measurements were
conducted at intermediate time periods (less than 72 hours), and the transmission proﬁles
were obtained (Figure 4-1 for d-limonene and ethyl acetate, and Figure 4-2 for water
vapor). Utilizing these proﬁles the time to reach steady state was determined. The results

are summarized in Table 4-1.

Table 4-1 Time to reach steady state of permeation

 

 

Time (hours)
d-Limonene 12*
Ethyl Acetate 20*
Water 38**

 

 

* measured at 60 OC
** measured at 37.8 °C

These results indicate that the conditioning period (72 hours) is sufﬁcient to reach
steady state transmission rate for these permeant-ﬁlm systems. Unfortunately, because

of the relatively slow permeation behavior and the nature of the purge-trap

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.0E-16
o Ethyl Acetate
A d-eronene Steady State
1.5E-16 » ‘
A o ' .
5: 0
§
"' 10E-16 ~
8 ' o
R)
E Steady State
. A 1A A A 1%
5.0E’17 ‘ A
A
A A
0.0E+00 i V 1
O 10 20 30 40

Time (hours)

 

 

Figure 4-1 Transmission proﬁle of d-limonene and ethyl acetate

through Ale3-Coated PET Film

51

 

 

R (kg/m2.sec.Pa)

7.5E-l3

 

 

I

Steady State

 

 

 

 

 

0 2O 4O 6O 80

Time (hours)

 

Figure 4-2 Transmission proﬁle of water vapor

through Ale3-Coated PET Film

52

 

system, the diffusion coefficient (D) could not be estimated from these graphs. However,

the trend of diffusivity at 60 °C can be expressed in a qualitative manner as:

(low) Ethyl Acetate < d-Limonene (high).

4-3. Comparison of Permeability of Non-Abused Films
4-3-1. Comparison by Permeants

The permeability of Ale3-coated PET ﬁlm (Table 4-2, Figure 4-3), A1203-
coated PET-LLDPE laminated ﬁlms (Table 4-3, Figure 4-4), and PET (Table 4-4, Figure
4-5), using d-Iimonene, ethyl acetate, oxygen, and water vapor, were measured. Table 4-5

and Figure 4-6 summarize the permeance values for these respective permeants through

the Ale3-coated PET ﬁlm.

For d-limonene, ethyl acetate, and oxygen as permeants, the improvement in

barrier properties are signiﬁcant when comparing PET to A120 3-coated PET. For water

vapor, however, the improvements are limited.

53

Table 4-2 Permeability of Ale3-coated PET (12pm) to various permeants (a) (b)

 

 

 

 

 

Perrneance (kg/m2.sec.Pa)
Oxygen W Water (d) d-Limonene (e) Ethyl Acetate (t)
A1203 PET 1.09 x 10'17 7.62 x 10'13 5.78 x 10‘17 1.21 x 10'16
Std.Dev. 2.49 x 10‘18 1.25 x 10"3 3.23 x 10'17 8.41 x 10'18

 

 

 

 

(a) All values are average of replicate runs.

(b) All values were obtained after 72 hours sample conditioning.
(c) Measured at 23 °C (cell temp), 101.3 kPa (vapor pressure).
(d) Measured at 37.8 °C (cell temp), 6.2 kPa (vapor pressure).
(e) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).
(1) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

 

 

 

 

 

1E-10
:
113-12 : a
Q E
91 E
o I:
a) i
a; p
N - r.-
8 IBM
E)
5 E
9‘ E
1E-16 E g
E D
1E-18 i
Oxygen Water d-Limonene Ethyl Acetate
Permeants

 

 

 

Figure 4-3 Permeability of Ale3-coated PET (12pm) to various permeants

Table 4-3 Permeability of Ale3-coated PET (12pm) / LLDPE (40pm)

to various permeants (a) (b)

 

 

 

 

 

Penneance (kg/m2.sec.Pa)
Oxygen W Water (d) d-Limonene (c) Ethyl Acetate (f)
A1203PET/LLDPE 1.14x10’l7 5.87x10'l3 1.11x 10'16 5.96x10'l7
Std.Dev. 2.82 x 10'18 1.22 x 10'13 1.05 x 10'17 6.44 x 10'19

 

 

 

 

(a) All values are average of replicate runs.

(1)) All values were obtained after 72 hours sample conditioning.
(c) Measured at 23 °C (cell temp), 101.3 kPa (vapor pressure).
(d) Measured at 37.8 °C (cell temp), 6.2 kPa (vapor pressure).
(e) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).
(1) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

 

 

 

 

 

lE-lO :
E
lE-lZ : U
I; i
13., _
8 :
”l
N - r
8 1E 14
E) :
5 5
n4
- L 6
1E 16 6
n
lE-18
Oxygen Water d-Limonene Ethyl Acetate
Permeants

 

 

 

Figure 4-4 Permeability of Ale3-coated PET (12pm) / LLDPE (40pm)

to various permeants

55

Table 4-4 Permeability of PET (12pm) to various permeants

(a) (b)

 

 

 

 

 

Perrneance (kg/m2.sec.Pa)
Oxygen W Water (d) d-Limonene (e) Ethyl Acetate (0
PET 2.52 x 10‘14 9.27 x 10'11 2.36 x 10‘14 2.98 x 10'”
Std.Dev. 7.47 x 10''6 1.23 x 10'12 5.03 x 10'15 4.02 x 10’15

 

 

 

 

(a) All values are average of replicate runs.
(b) All values were obtained after 72 hours sample conditioning.
(c) Measured at 23 °C (cell temp.),101.3 kPa (vapor pressure).
(d) Measured at 37 .8 °C (cell temp), 6.2 kPa (vapor pressure).
(e) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).
(1) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

 

lE-OS

lE-lO

lE-12

R (kg/m2.sec.Pa)

1E-14

lE-l6

 

 

rm]

 

 

 

Oxygen

Water d-Lirnonene Ethyl Acetate

Permeants

 

 

L————_—_-._—.-——

Figure 4-5 Permeability of PET (12pm) to various permeants

56

Table 4-5 Permeability comparison between A1203 PET (l2um)
and non-coated PET (12pm)

 

 

 

Permeance (kg/m2.sec.Pa)
Oxygen (a) Water (b) d-Limonene (c) Ethyl Acetate (d)
741203 PET 1.09 x 10'17 7.62 x 10'13 5.78 x 10'17 1.21 x 10'“
PET 2.52 x 10'14 9.27 x 10‘” 2.36 x 10'” 2.98 x 10'”

 

 

 

 

 

(3) Measured at 23 °C (cell temp.),101.3 kPa (vapor pressure).
(b) Measured at 37.8 °C (cell temp), 6.2 kPa (vapor pressure).
(c) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).
((1) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

 

 

 

 

 

 

 

 

 

1.00E-O9
A1203 PET
I PET
g3 10013-12
8
"8
R
65
ed 1.00E-15
1.00E-18
02 H20 Limo EA
Permeants

 

 

 

Figure 4-6 Permeability comparison between A1203 PET (12um) and
and non-coated PET (12m)

57

To provide a clear comparison between non-coated PET and A1203-coated PET,
the theoretical permeability of the PET-LLDPE laminated structure can be estimated by
using the conventional equation below (Rogers, 1985):

m = .1251. Amaze. Eq 4,]
Rfaminatc RPEr RLLDPE

where; llaminatea lpET, and lLLDpE are the thickness of the laminate structure, PET and
LLDPE, respectively. Rlamjmte, RpET, and RLLDpE are the permeance of the laminate
structure, PET and LLDPE, respectively.

Table 4-6 and Figure 4-7 summarize the permeance values for these permeants
through A1203-coated PET / LLDPE laminated ﬁlms. Because of the higher permeation
rate of LLDPE to d-limonene, the contribution of the LLDPE layer was minimized. For
ethyl acetate and water vapor, however, the LLDPE layer reduced the overall
permeability of the laminated structures.

Because of apparatus limitations, the temperatures of test were varied. However,

the permeance values were standardized at a constant temperature (60 °C), utilizing the

equation below (Rogers, 1985):

__1_

513(3-
R472

R2= R1 X 6 Eq. 4'2
where R1 and R2 are permeance at temperature 1 (T1) and temperature 2 (T2),

respectively. ER is an activation energy (for oxygen 9.72 kcal/mole, for water vapor 39.92

kcal/mole (see Section 4-4 )), and R is a gas constant. Table 4-7 shows a comparison of

58

Table 4-6 Permeability comparison between A1203 PET (12pm) / LLDPE (40pm)
and non-coated PET (12pm) / LLDPE (40pm)

Permeance (kg/m2.sec.Pa)

 

 

 

 

 

 

Oxygen “0 Water (b) d-Limonene (c) Ethyl Acetate (d)
A1203 PET/LLDPE 1.14x10'l7 4.12x 10'13 1.11x10'16 2.45x10-17
LLDPE 6.52 x 1043‘” 3.00 x 10'” “l 2.93 x 109‘“) 7.75 x 10'11 1")
PET/LLDPE “’ 9.65 x 10'14 3.55 x 10"1 1.02 x 10'13 1.29 x 10'13

 

(a) Measured at 23 °C (cell temp), 101.3 kPa (vapor pressure).

(1)) Measured at 37.8 °C (cell temp), 6.2 kPa (vapor pressure).

(c) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).

((1) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

(e) Measured at 24.8 °C (cell temp), 101.3 kPa (vapor pressure) (T ocelo 1995).
(1) Measured at 37.8 °C (cell temp), 6.2 kPa (vapor pressure) (T ocelo 1995).
(g) Measured at 45 °C (cell temp.) (Kobayashi 1995).

(h) Measured at 22 °C (cell temp), 4.11 kPa (vapor pressure) (Barr 1997).

(i) Caluculated by conventional equation (eq. 4-1).

 

 

 

 

 

 

 

 

 

1.00E-O9
A1203 PET / LLDPE
I PET/LLDPE
E 1.00E-12
g
"E
H)
65
a 1.00E-15
1.00E-18
02 H20 Limo EA
Permeants

 

 

 

Figure 4-7 Permeability comparison between A1203 PET (12pm) / LLDPE (40pm)
and non-coated PET (12um) / LLDPE (40m)

59

the standardized permeability. The trend of permeability performance through A1203-
coated PET ﬁlm at 60 °C for the respective permeants can be expressed qualitatively as:

(lowest) d-Limonene < 02 S Ethyl Acetate << H20 (highest).

Table 4-7 Calculated permeance at 60 °C through A1203-coated PET (12 um)

Permeance (kg/m2.sec.Pa)
Oxygen Water d-Limonene Ethyl Acetate

A1203PET 8.08x10‘l7 5.67x10‘ll 5.78x10'17 1.21x10’l6

 

 

 

 

 

 

 

 

4-3-2. Comparison of Ceramic Coatings

Two different ceramic-coated PET ﬁlms were evaluated for both the d-limonene
and ethyl acetate permeability. The results are summarized in Table 4-8. and presented
graphically in Figure 4-8.

The results showed that even though the ceramic matrices are different, their
permeability characteristics were quite similar. These results suggest that the mechanism

for mass transfer through both A1203-coated PET and SiOx-coated PET ﬁlms was

similar.

60

Table 4-8 Permeability comparison of ceramic coatings

 

 

 

 

Permeance (R); (kg/mzsecPa x 10'”)
d-Limonene (c) Ethyl Acetate (d)
A1203 PET SiOx PET A1203 PET SiOx PET
Ave. 5.78 10.20 12.06 12.96
Std.Dev. 3.23 6.65 0.84 1.86

 

 

 

 

 

(a) All values are average of replicate runs.

(b) All values were obtained after 72 hours sample conditioning.

(c) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).
((1) Measured at 60 °C (cell temp), 2.67 kPa (vapor pressure).

 

R (kg/m2.sec.Pa)

 

 

 

 

1.00E-15
1.00E—16 . 5 §
1.00E-17 —
1.00E-18
A1203 PET SiOX PET A1203 PET SiOx PET
d-Limonene Ethyl Acetate

 

Figure 4-8 Permeability comparison of ceramic coatings

61

 

 

4-4. Dependency of Permeability on Temperature
The effect of test temperature on the organic vapor (d-limonene and ethyl acetate)

permeability of A1203-coated PET ﬁlm was evaluated at ﬁve permeation cell temperature

levels. The results are summarized in Table 4-9 and shown graphically in Figure 4-9 for d-
limonene and in Table 4-10 and Figure 4-10 for ethyl acetate.

The results show there was a slight increase in the permeation rates as a ﬁrnction
of temperature from 55 to 70 °C in both the d-limonene and ethyl acetate systems. As
shown, the permeation rates then increased markedly, this was observed for both
systems at around 75 °C, which is the T8 of PET.

Utilizing these permeance values and equation (Eq. 3-2), the activation energy
(ER) was calculated for both the d-limonene (Figure 4-11) and ethyl acetate (Figure 4-12)
systems, below Tg. For comparison, ER values for the oxygen and water vapor
permeability of A1203-coated PET ﬁlm were also calculated by using literature data

(Imai, 1998). A comparison of Arrhenius plots is shown in Figure 4-13. Table 4-11 and

Figure 4-14 summarize the activation energy of these permeants through A1203-coated
PET ﬁlm, below Tg.

For d-limonene, ethyl acetate and oxygen, the ER values are relatively small,
which suggests that the permeation process through A1203-coated PET ﬁlm is not
affected much by changing temperature, below the Tg of PET. On the other hand, ER of

water vapor is much higher than the other permeants. This behavior agrees with the

62

Table 4-9
Temperature effect on the permeability of the A1203 PET / d-limonene system

(Permeance (R) (kg/m2.sec.Pa x 1047))

(a). (b), (c)

 

Temperature (°C)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

55 60 65 70 75 80
Ave. 4.81 5.62 7.34 8.34 24.53 47.47
Std.Dev. 0.88 1.62 0.60 1.14 2.17 12.65
(a) All values are average of replicate runs.
(b) All values were obtained after 72 hours sample conditioning.
(c) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).
100
80 ~
"’2
x 60 7 -
cc:
91
U
8 o
if, 40 -
33 _
a:
E
20 ~
9 U n
O I 1 1
50 60 70 80
Temperature (°C)

 

 

 

Figure 4-9 Temperature effect on the permeability of the A1203 PET /

d-limonene system

63

Table 4-10
Temperature effect on the permeability of the A1203 PET / ethyl acetate system
(Permeance (kg/m2.sec.Pa x 1046))

(a), (b), (C)

 

 

 

 

 

 

 

 

Temperature (°C)
55 6O 65 70 75 80
Ave. 1.12 1.21 1.25 1.27 1.95 2.09
Std.Dev. 0.22 0.08 0.13 0.10 0.37 0.04

 

 

(a) All values are average of replicate runs.
(b) All values were obtained after 72 hours sample conditioning.

(c) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

 

 

 

 

 

5
4 2
"'2
>< 3 i
«3
91
8
8.2— o 0
$5 -
m -
0 § 2 2
1 * _
O 1 1 1 1 1 1
50 55 60 65 70 75 80 85
Temperature (°C)

 

 

 

Figure 4-10 Temperature effect on the permeability of the A1203 PET /

ethyl acetate system

 

-35 _

 

    

 

 

 

0

-36 ~ E
a:
5 y = -4.3162x - 24.423

_ R2=0.9824
-37 r
_38 L 1
2.80 2.90 3.00 3.10

1000/T (IOOO/K)

 

 

 

Figure 4-11 Arrhenius plot for the permeance of the A1203 PET / d-limonene system

 

 

 

 

 

 

 

 

-35 1
-36 a
0 o
m -
5 T § + 5 t
— ‘0
'37 . y=-O.9322x-33.87 _
R2=O.9219
_38 1 A
2.80 2.90 3.00 3.10
1000fT(IOOOﬂ()

 

 

 

Figure 4-12 Arrhenius plot for the permeance of the A1203 PET / ethyl acetate system

65

 

 

 

 

 

 

 

 

 

 

 

0 Oxygen I
I Water Vapor
. I
A d-eronene I
o Ethyl Acetate I
-30 _
Tg of PET
at
c:
._1
-35 ~ A
O
O I O Q
A A O 0
° 0
_40 L 1 1 1 1
2.50 2.70 2.90 3.10 3.30 3.50 3.70
1000/T (1000/K)

 

Figure 4-13 Comparison of Arrhenius plots of various permeants
through A1203 PET

66

 

Table 4-11 Comparison of activation energy for the permeation process

through A1203 PET using various permeants

 

 

 

ER
kcal/mol
Oxygen 9.72
Water Vapor 3 9.92
d-Limonene 19.66
Ethyl Acetate 4.25

 

 

 

Activation Energy (kcal/mole)

 

 

 

: ‘—
4 8
g >
I—
a
«I
3

d-eronene
Ethyl Acetate

 

 

 

Figure 4-14 Comparison of activation energy for the permeation process
through A1203 PET using various permeants

67

signiﬁcant difference observed between the water vapor permeability and that of the
other permeants (d-limonene, ethyl acetate and oxygen). These facts imply that the
permeation mechanism (or factors effecting permeation) for water vapor and the other

permeants may differ.

4-5. Permeant Factors Affecting Permeation

The mechanism of permeation of gases and vapors through ceramic-coated ﬁlms is
not well known. Two different mechanisms, however, have been proposed to explain the
permeation process (see Section 2-3-4). The basic concept of both mechanisms is the
“defects leading permeation” model, where defects in the ceramic coating result in
permeation of gas or vapor. Since defects may provide the driving force for the
permeation through ceramic-coated ﬁlms, interactions between the defects and the

penetrants may effect the mass transfer process.

4-5-1. Permeant Size

Permeant size is, therefore, a major factor effecting the permeability of gases or
vapors through A1203-coated PET ﬁlm. Hence the dimensions of the permeants were
estimated from atomic radii and bond lengths (Appendix D). These ﬁgures represent the
smallest cross sectional area of the permeants. Table 4-12 summarized the calculated

dimensions of the permeants.

68

Table 4-12 Dimensions of the permeants.

 

 

 

 

 

 

 

Dimensions
(10'10 m) 10'20 m2
w * h ** d *** w x h h x (1
Oxygen 2.71 1.48 1.48 4.01 2.19
Water 2.17 1.64 1.48 3.56 2.43
d-Limonene 8.73 4.68 4.61 40.86 21.57
Ethyl Acetate 7.32 3.88 2.53 28.40 9.82

 

 

 

* Longest axis of the permeant
** Second longest axis of the permeant
*** Shortest axis of the permeant

The effect of permeant size was discussed in Section 2-4-1 and as pointed out the
smallest cross-sectional area of a permeant may be the crucial factor in determining the
effect of penetrant size on permeability. In this study, the same concept was employed
and the smallest cross sectional area (h x d) was calculated.

In an attempt to determine a relationship between the size of the permeant and
permeability, the smallest cross sectional area was plotted as a function of the log of the
permeance for the respective test penetrants (Figure 4-15).

This plot shows no linear relationship between size of permeants and the log of
the permeance. Water vapor had a signiﬁcantly higher permeance value, relative to its
size, as compared to the other permeants evaluated. Guttman (1990) suggested water

vapor may have a different diffusion mechanism in SiOx-layer (see Section 4-6-2). The

same mechanism may be operating in the A1203-coated PET ﬁlm / water vapor system.

69

 

 

 

 

 

 

Size (h x d) (m2 x 102°)

1E-9
O
1E-11 — H20
1?
9-1
8
NE 1E-13 P
H)
$3
08
113-15 2 O Ethyl Acetate ,
2 Limonene
O
. O
1E-17 1 1 1
10 15 20

25

 

 

Figure 4-15 Relationship between permeability and permeant size

70

4-5-2. Polarity of Permeant
Another possible contributing factor is the polarity of the permeants. Dielectric

constant (8) values can be used to represent the polarity of gases or vapors. Table 4-13

summarizes the dielectric constants for each permeant.

Table 4-13 Dielectric constant of permeants (CRC, 1987)

 

 

. s

A Oxygen 1 .51
Water 80.37

' d-Limonene 2.30
Ethyl Acetate 6.02

 

The order is shown below:

(low) 02 < d-Limonene < Ethyl Acetate << H20 (high).

In order to evaluate the relationship between the polarity of the permeant and
permeability, permeance values were plotted as a function of permeant dielectric constant
(Figure 4-16). This plot shows relatively proportionate relationship between polarity of
permeant and permeability. However, because of the broad range of dielectric constant
values for the permeants evaluated (from 1.5 (oxygen) to 80.37 (water)), it is difficult to
establish a precise relationship without determining the permeability values for a series of

other permeants, representing the full range of dielectric constants

71

3585—3 mo btﬁoa ES bzﬁmogon c6953 aEmcoEw—om 07¢ oSwE

 

 

 

 

 

 

 

 

 

 

 

30$ :3.—mm 32> ES
3 8:23 E 85.8
S m o 2: 8 o
. :9 . :-m_
0
Dams—OS: 1 1 me:
.5 m \\\. m
0 v... ‘\ x.
w w
o o 1 2-1”: .sz - ME: .8
N 9 x
O .3 .0
888,433 m d
( w
of 1 :-m:
0
:1”: 69m:
.iL

 

 

 

 

 

 

72

4-5-3. Combined Factor by Size and Polarity

The permeation mechanism of gases or vapors through A1203-coated PET ﬁhn is
assumed to have complicated process. Therefore, the factors effected the mass transfer
mechanism may also be interrelated.

In this study, an attempt to establish a relationship between permeance and
permeant characteristics, the reciprocal of the smallest cross-sectional area and dielectric

constant were combined, which lead to a new factor (8 x (h x d)‘1) (Table 4-14).

Table 4-14 Combined factor (size and polarity) of permeants. ‘

 

 

 

 

e x (h x d)"

(m‘2 x 102°)
Oxygen 0.69
Water 33. 1 l
d-Limonene 0.1 l
Ethyl Acetate 0.56

 

The order of the combined factors is shown below;

(small) d-Limonene < Ethyl Acetate S 02 « H20 (large).

In order to make a clear comparison, Table 4-15 summarizes the order of
permeability and possible permeability deciding factors. The permeability and combined
factor showed similar trends, while the size and polarity factor does not show good

agreement with the permeability trend.

73

Table 4-15 Comparison of the various trends related to permeability.
_Permeability low Limo < 02 5 EA << HzO high

 

 

_Size large Limo >> EA >> H20 > 02 small
Polarity low 02 < Limo < EA << H20 high
Size x Polarity largex low Limo < EA 5 02 << H20 small x high
02, Oxygen; H20, Water; Limo, d-Limonene; EA, Ethyl Acetate

 

 

 

 

 

 

In Figure 4-17, permeance is plotted as a function of the combined factor. The
plot suggests a linear relationship between permeance and the combined factor. However,
additional data is necessary, over the entire range of combined factor values, to establish
such a linear relationship.

The permeant parameters effecting to permeability may not so simple. However,

this introduced factor (8 x (h x d)"1 ) can be used for simple estimation of gases or vapors

permeability of A1203-coated PET ﬁlm.

74

88mm 3:588 23 b25883 5353 aimeouﬂom :4 823m

32> REE

tsoe,=:m

 

 

Age 5. 48 is x e x 6
885m wocBEoU
m.o ed v.0 No Q

 

 

4 _ i _ h — 1N0. —
25.5856 H
NO . m
o - 8_aav_ ma
6 3
(am my
me

 

me—a:

 

 

A22 8. 48 is x a x .8
88mm woﬁnEoU

 

 

ow cm om o_ O
_ _ 4 hﬁrmﬁ
--\\ - m_am_
T\\\. 56
inﬁxx MW
- m;-m_ .z
S
m
"a
Aw
Aoma . _;-m_
0
momm_

 

 

 

 

 

 

 

75

4-6. Permeation Mechanism for Non-Abused Films

In order to better understand the permeability behavior of gases or vapors through

A1203-coated PET ﬁlm, a simple model is proposed.

4-6-1. Permeation through Defects

Before the modiﬁed model is mentioned, a general model (defect leading
permeation model) of gas permeation through ceramic-coated ﬁlms is described below.

The ceramic coating layer may have some micro defects (matrix irregularities) or
macro defects (pinholes) by nature. When permeants are supplied, the permeants start to
diffuse through the substrate (i.e., PET). After a certain time period, the diffusion reaches
an equilibrium state or steady state, and a concentration gradient of the permeant is
established within the substrate. Thus, the permeation behavior through defects in the
ceramic coating layer will be dependent upon the permeant concentration gradient,
permeant characteristics (e. g., size, shape, polarity), and the status of the defects. Figure

4-18 shows a schematic representation of this concept.

76

 

 

 

 

mags/1am

 

l] Permeant supply

Permeant

 

 

.84/W

 

 

 

‘LReach steady state

 

Through defect

PET
A1203

Diffusion through PET

Equilibrium pennant
concentration gradient

 

high

 

- low

 

Concentration

 

 

Figure 4-18 Defect leading permeation model

77

4-6-2. Generation of Attractive Defects

Guttman (1990) introduced a model of water vapor permeation through SiOx-
layer. He suggested that hydrolysis of the Si-O-Si bond created a pathway for water
vapor diffusion through the SiOx-layer.

In this study, the Guttman (1990) model was applied with the defect leading
permeation model, and it lead to a new complex model with the following proposed

hypothesis. The hypothesis is: “Water vapor in the atmosphere reacts with A1203

around the defects, and the defects then become a hydrophilic regions (or polar attractive
defects)”. The hydrophilic regions can attract polar molecules and may promote increased
permeability. This hypothesis (Attractive Defects (AD) model) is schematically shown

in Figure 4-19.

4—6-3. A1203-Coated PET Film

For non-polar permeants, the contribution of attractive defects to increased
permeability may be minimized. Moreover, the defects can even obstruct the permeation
of non-polar permeants, because the reaction with water vapor may decrease the diameter

of the defects. Thus, for non-polar permeants permeability of A1203-coated PET ﬁlm

can be expected to be very low (excellent).
For polar permeants, the contribution of attractive defects can be signiﬁcant. Polar

permeants may be attracted to the defects, and the resulting permeability of polar

78

 

Mechanism of attractive region generation

 

PET
A1203

 

 

 

 

No attractive region
H20 (Hydrophobic defect)

 

 

 

: ££££££ﬂ£€SSIKBSSiK’aﬁﬂ’Is'a'a'IA

Attractive region for polar permeants
(Hydrophilic defect)

Barrier mechanism

For non-polar permeants For polar permeants

    

awzzaaz888431.131;
‘E’n'b'

No effect for
non-p013r permeants Attract

 

polar permeants

Figure 4-19 Generation of attractive defects in ceramic coating layers
and permeation mechanism through A1203 PET

79

permeants through A1203—coated PET ﬁlm can be expected to be signiﬁcantly higher than

that of non-polar permeant.

4-6-4. A1203-Coated PET / LLDPE Laminated Films

The AD model can apply not only to A1203-coated PET ﬁlm but also A1203-

coated PET / LLDPE laminated ﬁlms and was ﬁt to the permeability data obtained
(Tables 4-5,6 and Figures 4-6,7). Figure 4-20 and 4-21 show schematic expressions for
non-polar and polar permeant, respectively.

For non-polar permeants, since LLDPE and the permeants have the same non-
polar nature, the contribution of LLDPE to the permeant barrier is minimized. Therefore,

for non-polar permeants (d-limonene and oxygen) permeability of A1203-coated PET ﬁlm
and A1203¢coated PET / LLDPE laminated ﬁlms may be quite similar.

On the other hand, for polar permeants, the contribution of LLDPE to the
permeant barrier may be signiﬁcant. Therefore, improvements in barrier properties for

polar permeants (ethyl acetate and water vapor) can be expected.

80

Defect by nature

A12 03 PET (existing attractive defect)

  

PET
Ab03

        
 
    
   

   

V 5.. zéSAﬁAgfrA
"

 

No effect for
non-polar permeants

A1203 PET/LLDPE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Similar
PET barrier
A1203 properties
Adhesive
/
0\ LLDPE
Poor non-polar permeant
mm; high protection by LLDPE
low
Concentration

 

 

 

Figure 4-20 Barrier mechanism for non-polar permeants through A1203 PET

81

Defect by nature
A1203 (existing attractive defect)

  
 

” :ngT
3m: § A120
willlltllﬁ illillllil 3

 

 

Attract
polar permeants

 

 

 

A1203 PET/LLDPE
Different
.. . U PET bam'er
am 2: "‘7'" A1203 properties
V Adhesive
\ I LLDPE

 

\

Good polar permeants
2§2§2Es=is§2§2is§s§222§2 high protection by LLDPE

 

 

..........

..............

 

 

low

 

Concentration

 

 

 

Figure 4-21 Barrier mechanism for polar permeants through A1203 PET

82

4-7. Effect of Physical Damage

Packages meet with various types of physical damage during the packing and
distributing processes. In order to simulate this physical damage, Gelbo ﬂex tests were
conducted, and the permeability of related gases and vapors was measured through the

abused ﬁlms.

4-7-1. Non-Polar Permeants Permeability

Table 4-16 and Figure 4-22 show the effect of physical damage on oxygen

permeability of A1203-coated PET ﬁlm and A1203-coated PET/LLDPE laminated ﬁlms.

Signiﬁcant (greater than 10 times) deterioration in barrier characteristics occurred in both

A1203-coated PET ﬁlm and A1203-coated PET/LLDPE laminated ﬁlms after 10 ﬂexures.

Catastrophic (greater than 50 times) deterioration in barrier properties occurred in both

A1203-coated PET ﬁlm and A1203-coated PET/LLDPE laminated ﬁlms, at less than 100

ﬂexures.

Similar behavior was observed for d-limonene permeability of A1203-coated PET
ﬁlm and A1203-coated PET/LLDPE laminated ﬁlms systems (Table 4-17 and Figure 4-

23). Signiﬁcant deterioration occurred at less than 100 ﬂexures, and catastrophic

deterioration occurred at less than 150 ﬂexures for the A1203-coated PET ﬁlm and less

than 200 ﬂexures for the A1203-coated PET/LLDPE laminated ﬁlms.

83

Table 4-16 Flex Resistance for oxygen permeability (permeance (kg/m2.sec.Pa))
of A1203 PET and A1203 PET / LLDPE (a), M (‘0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample F lexing Cycle
0 10 100 150 200
A1203 PET 1.14 x10.17 7.34 x 10'16 1.06 x 10'15 1.14 x 10'15 2.38 x 10‘1““)
A1203 PET / LLDPE 8.15 x 10‘18 3.28 x 10’16 1.12 x 10'15
(a) All values are average of replicate runs.
(b) All values were obtained after 72 hours sample conditioning.
(C) Measured at 23.0 °C (cell temp), 101.3 kPa (vapor pressure).
((1) Measuring limitation (Some results showed "Not Dated")
113-1 2
O A1203 PET
1E-13 P \
1:1 A1203 PET / LLDPE T
O
”56‘ 113-14 P (or higher)
‘31
§
“é lE-l 5 ~ . i Q
E El
Dd 113-1 6 —
1E-17 l5
113-1 8 1 1 1 1
0 .50 100 150 200
Flexing Cycle (times)

 

 

 

Figure 4-22 Flex resistance for oxygen permeability (permeance (kg/m2.sec.Pa))

 

Table 4-17 Flex resistance for d-limonene permeability (permeance (kg/m2.sec.Pa))
of A1203 PET and A1203 PET / LLDPE (a), (b), (‘0

 

 

 

 

 

Sample Flexing Cycle
0 10 100 150 200
A1203 PET 5.78x10'l4 2.10x 10'16 1.10x10.15 4.45xlO'15 3.35x10.12
A1203PET/LLDPE 5.21 x 10'17 1.20x 10'16 8.21 x 10'16 125x 10-13

 

 

 

 

(a) All values are average of replicate runs.
(b) All values were obtained after 72 hours sample conditioning.
(c) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).

 

 

lE-12

 

O A1203 PET

”3'13 1:1 A1203 PET/LL

l

 

 

 

r.“
A
1

[3

1E-15

R (kg/mzsecPa)
:71

a

[30

113-17

T

 

 

 

1E'18 1 1 1
0 50 100 150 200

Flexing Cycle (times)

 

 

 

Figure 4-23 Flex resistance for d-limonene permeability (permeance (kg/m2.sec.Pa))
of A1203 PET and A1203 PET / LLDPE

85

These results indicate that new defects (possibly pinholes) were created in the

A1203 coating layer by ﬂexing abuse on both A1203-coated PET ﬁlm and A1203-coated

PET/LLDPE laminated ﬁlms.

4-7-2. Polar Permeants Permeability

In contrast, for water vapor (Table 4-18 and Figure 4-24) or ethyl acetate (Table
4-19 and Figure 4-25), deterioration in barrier properties between A1203-coated PET ﬁlm
and A1203-coated PET/LLDPE laminated ﬁlms appeared to be quite different.

For both water vapor and ethyl acetate, A1203-coated PET ﬁlm showed

signiﬁcant deterioration at less than 10 ﬂexures and catastrophic deterioration at less than
200 ﬂexures. However, no catastrophic deterioration was observed at over 200 ﬂexures

for both water vapor and ethyl acetate permeation through A1203-coated PET / LLDPE

laminated ﬁlms.
These results are summarized in Tables 4-20 to 4-23 and Figure 4-26 to 4-29.

Permeability comparisons for A1203-coated PET ﬁlm are shown in Table 4-20 and Figure
4-26. For A1203-coated PET / LLDPE laminated ﬁlms, permeability comparisons are

shown in Table 4-21 and Figure 4-27. In order to make clear comparisons, Table 4-22 and

Figure 4-28 indicate deterioration rates for Al203-coated PET ﬁlm compared to non-
abused ﬁlm. Also, Table 4-23 and Figure 4—29 indicate deterioration rates for A1203-

coated PET / LLDPE laminated ﬁlms compared to non-abused ﬁlm.

86

 

Table 4-18 Flex resistance for water vapor permeability (permeance (kg/m2.sec.Pa))
of A1203 PET and A1203 PET / LLDPE (8” W ‘°’

 

 

 

Samme Flexing Cycle
0 10 100 150 200
A1203 PET 6.56x10‘l3 8.28x10‘12 1.63x10‘ll 1.99x10‘ll 2.25x10’10(d)
A1203PET/LLDPE 4.68x10‘l3 5.81x10‘13 1.12x10’12 1.31x10‘12 1.12x10’12

 

 

 

 

 

 

(a) All values are average of replicate runs.

(b) All values were obtained after 72 hours sample conditioning.

(c) Measured at 37 .8 °C (cell temp), 6.2 kPa (vapor pressure).

((1) Measuring limitation (Some results showed "Not Dated")

 

 

 

 

 

 

 

 

 

 

1E-O7
o A1203 PET
1E-08 * DAJZO3 PET/LLDPE
g 1E-09 _ T
Q
8 0
NE" lE-lO e (orhigher)
E 9
Z lE-ll ~ .
1E-12 é ‘3 D D
1:1
lE-l3 1 1 ‘ J
0 50 100 150 200
Flexing Cycle (times)

 

 

Figure 4-24 Flex resistance for water vapor permeability (permeance (kg/m2.sec.Pa))
of A1203 PET and A1203 PET / LLDPE

87

 

Table 4-19 Flex resistance for ethyl acetate permeability (permeance (kg/m2.sec.Pa))
0f A1203 PET and A1203 PET / LLDPE W (bl (°)

 

 

 

 

 

 

 

Sample Flexing Cycle

0 10 100 150 200
A1203 PET 1.20x10'l6 2.57x 10'15 2.35 x 10‘15 1.48 x 10'14 3.14x10'l4
A1203 PET / LLDPE 5.96 x 10'17 7.52 x 10'16 8.15 x 10‘" 6.43 x 10'16

 

 

(a) All values are average of replicate runs.

(b) All values were obtained after 72 hours sample conditioning.

(c) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

 

 

 

 

 

 

 

 

 

1E-12
o A1203 PET
”3'13 * DAIZO3 PET/LLDPE
0
g 1E—14 . °
8 0 o
30
'5,
a; 1E-16$
1E-l7 ~
1E_18 1 1 1 1
0 50 100 150 200
Flexing Cycle (times)

 

 

 

Figure 4-25 Flex resistance for ethyl scetate permeability (permeance (kg/m2.sec.Pa))

88

 

Table 4-20 Permeability (Permeance (kg/m2.sec.Pa)) comparison of abused

A1203 PET ﬁlms using various permeants

(comparison based on permeance)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Permeant Flex cycles
0 10 100 150 200
Oxygen W 1.14 x 10'17 7.34 x 10‘16 1.06 x 10'15 1.14 x 10'15 2.38 x 1014‘”
Water ‘b’ 6.56 x 10‘13 8.28 x 10'12 1.63 x 10'“ 1.99x10'11 2.25 x 101°“)
d-Limonene ‘°’ 5.78 x 10'14 2.10 x 10‘16 1.10 x 10‘15 4.45 x 10'15 3.35 x 10‘12
Ethyl Acetate (d) 1.20x 10‘16 2.57 x 10‘15 2.35 x 10’15 1.48 x 10'14 3.14 x 10‘14
(a) Measured at 24.8 °C (cell temp), 101.3 kPa (vapor pressure).
(b) Measured at 37.8 °C (cell temp), 6.2 kPa (vapor pressure).
(c) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).
((1) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).
(e) Measuring limitation (Some results showed "Not Dated")
1E-O9 L OOxygen IWater T
A d-Limonene O Ethyl Acetate I
' (or higher)
I
A I ' A
(6
On
8' 1E-12 I
018' T
B) . 9
id“ A (or higher)
M O .
1E-15 r o O O
A
lE-l 8 1 1 1 A
0 50 100 150 200
Flex Cycles (times)

 

 

 

Figure 4-26 Permeability comparison of abused A1203 PET ﬁlms

using various permeants (comparison based on permeance)

89

 

Table 4-21 Permeability (Permeance (kg/m2.sec.Pa)) comparison of abused
A1203 PET / LLDPE ﬁlms using various permeants

(comparison based on permeance)

 

Permeant

Flex cycles

 

0

10

100

150

200

 

Oxygen (a)

Water (b)

d-Limonene (C)

Ethyl Acetate (d)

 

8.15 x 10'18
4.68 x 10'13
5.21 x 10'17
5.96 x 10'17

 

3.28x 10'16
5.81 x 10'13
1.20x 10'16
7.52 x 10'16

1.12x10'15
1.12x10'12
8.21 x 10'16
8.15 x 10‘16

 

 

1.31 x 10'12

 

1.12x 10'12
1.25 x 10'”
6.43 x 10'16

 

(a) Measured at 24.8 °C (cell temp), 101.3 kPa (vapor pressure).

(b) Measured at 37.8 °C (cell temp), 6.2 kPa (vapor pressure).

(0) Measured at 60 °C (cell temp), 0.15 kPa (vapor pressure).

((1) Measured at 60 °C (cell temp), 2.7 kPa (vapor pressure).

 

1E-O9

1E-12

R (kg/m2.sec.Pa)

lE-lS

lE-18

 

 

 

l

 

0 Oxygen

I Water

A d-Limonene
O Ethyl Acetate

 

 

l

l

[>00

 

 

 

100

150

Flexing Cycle (times)

200

 

 

Figure 4-27 Permeability comparison of abused A1203 PET / LLDPE ﬁlms

using various permeants (comparison based on permeance)

9O

 

Table 4-22 Permeability (permeance (kg/m2.sec.Pa)) comparison of abused

A1203 PET ﬁlms using various permeants

(comparison based on deterioration rate (R(abuscd ﬁlm) / R(non-abuscd ﬁlm)»

 

 

 

 

 

 

 

 

Permeant Flex cycles
0 10 100 150 200
Oxygen 1 64 93 100 2086 (a)
Water 1 13 25 30 343 (a)
d-Limonene 1 4 19 72 57899
Ethyl Acetate 1 22 20 124 263

 

(a) Measuring limitation (Some results showed more deterioration)

 

 

 

 

 

 

 

 

 

100,000
A
0 Oxygen
I Water
IO’OOO _ A d-Limonene
8 C Ethyl Acetate (01‘ higher) I
a:
‘3: 1,000 l
g Catastrophic deterioration (or higher) a
.‘g‘ /
100 -
Q """""""""""""""" '- """"""
0 I
10 .. -l. ________________________ Signlflzcaat. 499998999.
- A
1 i L m 1
O 50 100 150 200
Flexing Cycle (times)

 

 

 

Figure 4-28 Permeability comparison of abused A1203 PET ﬁlms
using various permeants

(comparison based on deterioration rate (Rmbused ﬁlm) / R<mmabuscd ﬁlm)))

91

 

Table 4-23 Permeability (permeance (kg/m2.sec.Pa)) comparison of abused

A1203 PET / LLDPE ﬁlms using various permeants

(comparison based on deterioration rate (Rama, ﬁlm) / Rmombused ﬁlm)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Permeant Flex cycles
0 1 O l 00 1 50 200
Oxygen 1 4O 13 7
Water 1 1 2 3 2
d-Limonene 1 2 1 6 23 96
Ethyl Acetate 1 1 3 14 1 1
10,000
0 Oxygen
I Water A
1 000 _ A d-Limonene
Q ’ O Ethyl Acetate
E
c:
.g 0
g 100 ” Catastrophic deterioratiori
E - .0 .......................................
8 5 Signiﬁcant deterioration
10 e ... ...... 9 ............................ n -
A I I I
1 I . 1 1 1 1
O 50 100 1 50 200
F lexing cycle (times)

 

 

 

Figure 4-29 Permeability comparison of abused A1203 PET / LLDPE ﬁlms
using various permeants

(comparison based on deterioration rate (R(abused ﬁlm) / R(non-abused ﬁlm»:

92

 

These comparisons indicate that deterioration of barrier characteristics for A1203-
coated PET ﬁlm occurred similarly for both non-polar and polar permeants. However, for
the A1203-coated PET / LLDPE laminated ﬁlms, signiﬁcant loss of barrier properties
occurred only for the non-polar permeants. For polar permeants, the laminated structure
plays a signiﬁcant role in preventing deterioration by physical abuse, even though the

abuse created new defects on the structures.

4-7-3. Microscopic Observations
Optical microscopic observations were recorded in order to identify
morphological differences between the various degrees of abused A1203-coated PET

ﬁlms. To make a clear comparison, the surface of a non-abused A1203-coated PET ﬁlm

was observed (Figure 4-3 0). There was no cracks on the non-abused ﬁlm surface.

The surface of the A1203 coating induced cracks after 10 ﬂexures (Figure 4-31).
After 100 ﬂexures, propagation of the number of cracks was observed (Figure 4-32).
Aﬁer 200 ﬂexures, scattered crystal-like particles were observed (Figure 4-33). These
particles may be particles of A1203 coating peeling from the substrate.

These observations can explain the size dependency for the permeability on
different permeants. After 10 ﬂexures, permeability of abused ﬁlms shows a trend of
permeant size dependence (Table 4-22 and Figure 4-28). The barrier performance for
smaller permeants (e. g., oxygen) tended to deteriorate signiﬁcantly, at this point. With

increasing ﬂexing cycles, the trend of size dependence is reduced.

93

 

 

Magniﬁcation (x 500)

Figure 4-30 Surface Observation for non-abused A1203 PET

 

Magniﬁcation (x 200)

      

‘ A

Figure 4-31 Surface observation for 10 times ﬂexing A1203 PET

95

 

 

o .- . O V-,-,.,.

 

 

 

 

 

7 W ,_ ._ i . , 21
Magniﬁcation (x 500)

Figure 4-32 Surface observation for 100 times ﬂexing A1203 PET

96

 

 

Magniﬁcation (x 200)

 

Magniﬁcation (x 500)

Figure 4-33 Surface observation for 200 times ﬂexing A1203 PET

97

 

With light abuse (10 ﬂexures), the width of cracks may be very small and cause
size dependence. Under heavy abuse (200 ﬂexures), the cracks may cause peel-off of

A1203 particles from substrate, and the defects may no longer cause size dependence,

because the defects are too large.

4-8. Permeation Mechanism for Abused Films
The same model (AD model) can be applied to both abused A1203-coated PET
ﬁlm and A1203-coated PET / LLDPE laminated ﬁlms and was ﬁt to the permeability data

obtained (Tables 4-16 to 4-19 and Figures 4-22 to 4-25).

4-8-1. A1203-Coated PET Film

Figure 4-34 shows schematic expressions, for non-polar permeants through
abused A1203-coated PET ﬁlm. When the ﬁlm was abused, new defects could be
introduced in the A1203-layer, and the defects may react with water vapor in the
atmosphere. This reaction can make polar attractive defects on the A1203 coatings.

However, since the permeants have a non-polar nature, “attractive” or “non-attractive”
defects may not cause signiﬁcant permeability differences. Therefore, the non-polar
permeants permeability of abused ﬁlm can deteriorate proportionately simply with
increasing numbers of pinholes.

The permeation model for polar permeants through abused A1203-coated PET

ﬁlm is shown in Figure 4-35. New defects may turn into polar attractive defects on the

98

 

No ﬂex Defect by nature PET
(already be attractive defect) A1203

 

anaezaazzaaazzzzae
Physical
Damage Pinhole

With ﬂex by ﬂex

\‘5'

 

1+ HzO Become

React with HzO to attractive defect

     

No effect for
non-polar permeants

Barrier deterioration

Concentration

 

 

Figure 4-34 Barrier mechanism for non-polar permeants
through A1203 PET (with Gelbo ﬂexing)

99

 

Defect by nature PET
(already be attractive defect)

    

No ﬂex

  
     

Physical
Damage

Pinhole

 

 

+ HzO
(frn environ.) Become

React With H20 to attractive defect

    

 
 

 

 

{can ' ' '

 

 

    

Attract
polar permeants

 

 

Barrier deterioration

 

 

Concentration

 

Figure 4-35 Barrier mechanism for polar permeants
through A1203 PET (with Gelbo ﬂexing)

100

 

A1203 coatings in a manner similar to the non-polar permeants case. These defects may

attract polar permeants. Therefore, the polar permeants permeability of abused ﬁlm can

also deteriorate proportionately simply with increasing numbers of pinholes.

4-8-2. A1203-Coated PET / LLDPE Laminated Films

Figure 4-36 shows a schematic diagram for the proposed mass transfer process,

involving non-polar permeants through abused A1203-coated PET / LLDPE laminated

ﬁlms. New defects could be introduced by physical damage, but they may not react with
water vapor because the LLDPE layer can prevent the penetration of water vapor
towards the new defects.

Even though the new defects could not result in attractive defects as deﬁned

above, the permeability of non-polar permeants through A1203-coated PET / LLDPE
laminated ﬁlms may exhibit increases similar to those observed for the A1203-coated PET

ﬁlm. As was discussed in Section 4-8-1, for non-polar permeants, “attractive” or “non-
attractive” defects may not result in permeability differences. In addition, the
contribution of the sealant layer (LLDPE) to the barrier properties would be minimized,
since LLDPE has higher permeation rates for non-polar permeants. Therefore, the
permeability of abused ﬁlms to non-polar permeants may show little or no differences

between A1203-coated PET ﬁlm and A1203-coated PET / LLDPE laminated ﬁlms.

101

Defect by nature

N0 ﬂex (existing attractive defect)

 

.

222?2222&rvvvvv"v
V,

 

PET
A1203

 

/\¢/\

 

 

 

‘Y
Physical Adhesive LLDPE
Damage Pinhole
With ﬂex by ﬂex

 

 

 

 

 

 

+ H20
No reaction
with H20

(fm environ.)

 

Remain
as normal defect

 

 

 

 

[Y’k

 

 

 

No effect to
non-polar permeant
permeability

 

high
low

Poor non-polar permeant
protection by LLDPE

 

Barrier deterioration

 

 

Concentration

 

Figure 4-36 Barrier mechanism for non-polar permeants
through A1203 PET/LLDPE (with Gelbo ﬂexing)

102

The schematic diagram for the proposed mass transfer process of polar permeants

through A1203-coated PET / LLDPE laminated ﬁlms is also shown in Figure 4-37. In this

system, polar permeants may not be attracted by the non-attractive defects, and the
contribution of the sealant layer to the barrier properties may be maximized, because
LLDPE has good polar permeants barrier properties. Therefore, the permeability of the
laminated ﬁlms / polar permeant systems do not undergo catastrophic deterioration even
under heavy physical damage. In other words, the permeability of abused ﬁlms to polar

permeants may be quite different between A1203-coated PET ﬁlm and A1203-coated

PET / LLDPE laminated ﬁlms.

103

 

Defect by nature PET

(existing attractive defect)

   
   
 
   

No ﬂex

 

 

 

 

Adhesive LLDPE
Physical
Damage Pinhole
by ﬂex

With ﬂex

 

+ H20
No reaction (fm environ.) Remain
with H20 as normal defect

  
 
    

.1; ' ‘9' Laémarga
V

W /’i‘.\

 

 

 

 

 

high No attraction to GOOd polar permeant
polar permeant protection by LLDPE
low per [new
Concentration

 

 

 

No signiﬁcant deterioration

Figure 4-37 Barrier mechanism for polar permeants
through A1203 PET/LLDPE (with Gelbo ﬂexing)

 

Chapter 5

SUMMARY and CONCLUSIONS

The permeability of A1203-coated PET and SiOx-coated PET ﬁlms to organic
vapors (d-limonene and ethyl acetate) was studied. Permeability data were generated
utilizing the highly in sensitive dynamic purge and trap / thermal desorption procedure,
with the MASZOOOTM Permeation Test System. For both d-limonene and ethyl acetate,

the permeability of A1203-coated PET ﬁlm was expectedly low.
A comparison of the permeability of A1203-coated PET ﬁlm to a series of

permeants (d-limonene, ethyl acetate, oxygen and water vapor) was made. In order to
obtain the activation energy for the permeation processes, the temperature dependence of
the permeability was also evaluated. Results from these studies showed that water vapor
behaved signiﬁcantly differently than the other permeants evaluated, in terms of the mass
transfer process. The trend of permeability could not be explained by either the size or
the polarity of the permeants tested. The combined factor of permeant size and polarity,
however, was found to provide a better explanation for the trend observed. Smaller and

more polar permeants have a greater capacity to permeate through A1203-coated PET

ﬁlm. Even though the proposed size / polarity hypothesis is simple in principle, this

concept may help to estimate the organic vapors permeability of A1203-coated PET ﬁlm.

Gas and water vapor permeability measurements were conducted following Gelbo

ﬂex testing to evaluate the effect of physical abuse on the barrier characteristics of A1203-

105

 

coated PET ﬁlm. Signiﬁcant deterioration in barrier properties occurred at less than 10
ﬂexures abuse for oxygen, water vapor, and ethyl acetate / A1203-coated PET systems.
This deterioration likely resulted from the generation of cracks, the presence of which

were conﬁrmed by microscopic observations.

For abused A1203-coated PET / LLDPE laminated ﬁhns, however, considerable

barrier property difference was observed for polar permeants and non-polar permeants.

For non-polar permeants, a similar deterioration trend occurred in both simple A1203-

coated PET ﬁlm and the laminated ﬁlms. However, for polar permeants, the extent of
deterioration of barrier properties observed was different for the two ﬁlms (i.e., no
catastrophic deterioration observed in the ﬁlms laminated with LLDPE). The laminated
structure thus played an important role in preventing barrier deterioration for polar
permeants.

The Attractive Defects (AD) model was introduced here to account for the
permeability behavior for A1203-coated PET ﬁlm. The model is derived from the
hypothesis that is “the defects by nature reacted with water in the environment and thus
became hydrophilic (or attractive) regions”. The attractive regions (or defects), then, may
attract polar permeants and cause barrier deterioration. This model can provide an

explanation for the permeability behavior of normal (non-abused) and abused A1203-

coated PET ﬁlms. The prevention of the reactions between defects and water in the
environment may be one of the keys to improve polar components barrier properties

through A1203—coated PET ﬁlm.

106

 

Chapter 6

FUTURE STUDIES

In order to improve the level of barrier properties of A1203-coated PET to that of
aluminum foil, understanding of the permeation mechanism is crucial. The major permeant
factors effecting permeability may be the size and polarity of the penetrant. In this
study, d-limonene and ethyl acetate were selected as permeants, because of their common
usage in related industries. However, determining the permeability using a series of
penetrants, which have similar structural characteristics may be useful for evaluating
molecular size, polarity, or other factors related to the mass transfer mechanism through
ceramic-coated substrates. For example, a series of saturated alcohols (methyl alcohol,
ethyl alcohol, etc.) may be a good candidate for study.

In this study, the Attractive Defects (AD) model was proposed to explain

permeability behavior through A1203-coated PET ﬁlm. The model proposed that reaction

of non-attractive (or normal) surface defects with water vapor may form attractive
defects on the A1203 coatings, with a concomitant deterioration in barrier properties.
Thus, the presence of water vapor as a co-permeant may effect permeability of A1203-

coated PET ﬁlm. The AROMATRANTM can introduce water vapor with a permeant to
simulate humid conditions, which may occur in actual distribution and storage conditions.

Utilizing this system, studies investigating humidity effects on A1203-coated PET ﬁlm

can be carried out to establish the validity of the AD model.

107

 

APPENDICES

 

Appendix A

Calibration Curve for Setting Vapor Pressure

In order to obtain conversion factors (CF) of vapor pressure for these
compounds, calibration curves were plotted by using standard solutions. For d-limonene,
5, 20, 40, 100, and 200 ppm (v/v) standard solutions were prepared by diluting with
carbon tetrachloride. For ethyl acetate, 1000, 10000, and 20000 ppm (v/v) standard
solutions were prepared by diluting with acetonitrile. A 1 ul of these solutions was
directly injected into the gas chromatograph and the area response was recorded.

Table A—1 and Figure A-l presents the results for the d-limonene standard
solutions. Table A-2 and Figure A-2 presents the results for the ethyl acetate standard

solutions.

108

 

 

Table A-1 Calibration data of d-limonene for setting vapor pressure

 

 

 

 

 

 

 

 

     
   

Concentration ppm (v/v) 5 20 40 100 200
Mass ofCompound (gx10'8) 0.42 1.68 3.4 8.4 16.8
AreaResponce (AleOS) 0.4 2.05 3.8 10.59 21.43
25
20
"‘2 y = 1.27E+00x
>< R2=9.99E-01
: 15 1
:5
8
C‘.
O
9..
E 10 —
§
<
5
Calibration factor = 7.79 x 10'14 g/AU
O 1 1 1

 

 

 

 

O 5 10 15 20

Mass of Compound (g x 10'8)

 

 

Figure A-l Calibration data of d-limonene for setting vapor pressure

109

 

Table A-2 Calibration data of ethyl acetate for setting vapor pressure

 

 

 

 

 

 

 

 

 

   
    
   

 

 

Concentration ppm (v/v) 1000 10000 20000
Mass of Conyound (& 107) 9 90 180
Area Responce (AUx104) 212.56 1679.72 3082.68
3500
3000 ~
”g 2500 4
>< y = 17.446x
: R2 = 0.9955
:5 2000 —
8
c: 0
EL 1500 ~
Q)
ad
(3
g 1000 t
500 7 CF = 5.73 x10'13 g/AU
O 1 1 1
0 50 100 150
Mass of Compound (g x 10'7)

 

200

 

Figure A-2 Calibration curve of ethyl acetate for setting vapor pressure

110

 

Appendix B

Calibration Curve for Dynamic Purge and Trap
/ Thermal Desorption Procedure

In order to obtain conversion factors (CF) of dynamic purge and trap / thermal
deposition procedure for these compounds, calibration curves were plotted by using
standard solutions. For d-limonene, 1, 5, 20, and 100 ppm (v/v) standard solutions were
prepared by diluting with carbon tetrachloride. For ethyl acetate, 5, 20 and 100 ppm (v/v)

standard solutions were prepared by diluting with dichlorobenzene. A 1 ul of these

solutions was injected into a thermal desorption tube. The tube, then, plugged into a
thermal desorption unit, which was directly connected into gas chromatograph and the
area response was recorded.

Table A-3 and Figure A-3 shows the result of d-limonene standard solutions.

Table A-4 and Figure A-4 shows the result of ethyl acetate standard solutions.

111

 

 

Table A-3 Calibration data of d-limonene
for dynamic purge and trap / thermal desorption procedure

 

 

 

 

 

 

 

 

 

  

 

 

Concentration ppm (v/v) 1 5 20 100
Mass ofCompound (gx 108) 0.084 0.42 1.68 8.40
Area Responce (AleOS) 0.30 1.61 7.53 40.68
50
40
“’93
X
:5
g 30 y=4.83E+00x
g R2=1.OOE+00
8.
§ 20
113’
<
10
CF = 2.07 x 10'14 g/AU
0 l l

 

O 2 4 6 8 10

Mass of Compound (g x 10'8)

 

 

 

Figure A-3 Calibration curve of d-limonene
for dynamic purge and trap / thermal desorption procedure

112

 

Table A-4 Calibration data of ethyl acetate
for dynamic purge and trap / thermal desorption procedure

 

 

 

 

 

 

 

 

    

 

 

 

Concentration ppm (v/v) 5 20 100
Mass of Compound (g x 108) 0.45 1.8 9.0
Area Responce (AU x 104) 14.73 52.84 158.25
200
VA 150 ~
9.
x
D
S
Q)
g 100 —
o
a
8
a:
(8
8
<3 50 ~
CF = 5.53 x10.14 g/AU
O
O l I l l
O 2 4 6 8 10
Mass of Compound (g x 10'8)

 

 

 

Figure A-4 Calibration curve of ethyl acetate
for dynamic purge and trap / thermal desorption procedure

113

 

Appendix C

Comparisons of Vapor Pressure of Compounds

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table A-5 Comparison of saturated vapor pressure of d-limonene
Temperature (°C) 10 15 20
Area response (AU) 3,407,210 5,067,699 8,218,672
Saturated vapor pressure From experimant 91.66 138.74 228.91
(Pa) From literature 106.51 147.96 203 .95
% difference 13.9 6.2 -12.2
250
O
A O From experimant
e: 200 o From literature
8
51
§ 150 o
:3.. O
S
g o
>
U 100 .
8
‘3
a
a 50
O L 1 1
5 10 15 20 25
Temperature (°C)

 

 

Figure A-S Comparison of saturated vapor pressure of d—limonene

114

 

Table A-6 Comparison of saturated vapor pressure of ethyl acetate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Temperature (°C) 5 24
Area response (AU) 8,875,820 27,676,368
Saturated vapor pressure From experimant 2667.77 8887.12
(Pa) From literature 4224.28 11638.42
% difference 36.8 23.6
14000 ~
0 From experimant

12000 . oFrom literature
1; O
E:
Q)
a 10000 ~
§ 0
S 8000 ~
0
a.
g
1: 6000 _
8
8
+3 4000 — 0
(I)

O
2000 _
O 1 1 1 1 1
0 5 10 15 20 25 30
Temperature (0 C)

 

 

Figure A-6 Comparison of saturated vapor pressure of ethyl acetate

115

 

 

 

Appendix D

Dimentions (A = 10.111111) of Permeants

HZO HzO
(longer axes) (shorter axes)

2.17 1.48

1.64 f 1.64 /

 

 

 

 

 

 

 

 

 

Figure A-7 Dimentions of water molecule

Oxygen Oxygen
(longer axes) (shorter axes)
2.71 1.48

 

 

 

 

 

 

 

 

 

/
\/

Figure A-8 Dimentions of oxygen molecule

 

 

 

 

 

 

 

 

116

 

d-Limonene
(longer axes)

5.07

8.73 3

 

 

 

 

 

A

 

 

 

 

 

 

 

 

 

 

 

 

/\

 

 

 

 

 

Q 3

Figure A-9 Dimentions of d-limonene molecule (top view)

 

 

 

117

 

d-Limonene
(shorter axes)
5.07

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure A-lO Dimentions of d-limonene molecule (side View)

118

Ethyl Acetate
(longer axes)

7.13

 

3.88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ethyl Acetate
(shorter axes) 2-53

3.88

 

 

 

 

 

 

 

 

 

Figure A-ll Dimentions of ethyl acetate molecule

119

BIBLIOGRAPHY

BIBLIOGRAPHY

Allison, H. L., "High-Barrier Packaging," Packaging, March, pp. 25-29 (1985)

ASTM, " Standard Test Method for Evaluation odf Solubility, Diffusivity, and
Permeability of Flexible Barrier Materials to Water Vapor," ASTM F 1770-97,
(1997b)

ASTM, "Standard Test Method for Flex Durability of Flexible Barrier Materials,"
ASTM F 392-74, (Reapproved, 1987)

ASTM, "Standard Test Method for Measurement of Diffusivity, Solubility, and
Permeability of Organic Vapor Barriers Using a Flame Ionization Detector," ASTM F
1769-97, (1997a)

ASTM, "Standard Test Method for Oxygen Gas Transmission Rate through Plastic
Film and Sheeting Using a Coulometric Sensor," ASTM D 3985-81, (1981)

Baker, C. P. et al. "Atomic Force Microscopy and Permeability Study of Stretching-
Induced Gas Barrier Loss of Ale Layers," Thin Solid Films, 259, pp. 46-52 (1995)

Barr, C. D., "A Comparison of Solubility Coefﬁcient Values Determined by
Gravimetric and Isostatic Permeability Techniques," Master's thesis, Michigan State
University, East Lansing, MI, (1997)

Barrer, R. M., Permeation, Diffusion and Solution of Gases in Organic Polymers,
Trans. Faraday Society, 35, pp. 628, (1939)

*‘ Berens, A. R. and H. B. Hopfenberg, "Diffusion of Organic Vapors at Low

Concentrations in Glassy PVC, Polystyrene, and PMMA," J. of Membrane Science,
10, pp. 283-303 (1982)

Chang, Y., "Organic Vapor Permeability of High Barrier Polymer Membranes By a
Dynamic Purge and Trap / Thermal Deposition Procedure," Master‘s thesis, Michigan

State University, East Lansing, MI, (1996)

Charoudi, D., "Barrier Coating Advance with Plasma Technologies," Paper, Film &
Foil Converter, October, pp. 40-42 (1991)

120

 

Crank, J ., The Mathematics of Difﬁision, Clarendon Press, Oxford, UK, pp. 414
(1956)

CRC, Handbook of Chemistry and Physics 67th ed., CRC Press Inc., Boca Raton, FL
(1987)

DeLassus, P. T., "Permeation of Flavors and Aromas through Glassy Polymers, "
Proceeding of Polymers, Laminations & Coatings Conference, TAPPI, pp. 263-266
(1993)

Dupont, US. Patent 3.442.686, by J. W. Jones, E. I. Dupont de Nemours and
Company, Wilimington, DE (1964)

Guttman, L., "Ring Structure of the Crystalline and Amorphous Forms of Silicon
Dioxide," J. of Non-Crystalline Solids, 116, pp. 145-147 (1990)

Hlavac, J., The technology of Glass and Ceramics, Elsevier Scientiﬁc Publishing
Company, New York, NY, (1983)

f - _.>Hoffmann, R. et al. "Development of Clear Barrier Films In Europe," Proceedings of

I _

36th Annual Technical Conference, Society of Vacuum Coaters, pp. 312-317 (1993)

Huang, S. J. and J. R. Giacin, "Evaluating the Effect of Temperature and Vapor
Concentration on the Organic Vapor Barrier Properties of Polymer Membranes by an
Isostatic Procedure," Plastic Film & Sheeting, Vol. 14, pp. 308-333 (1998)

Huang, S. J ., "Evaluation Isostatic and Quasi-isostatic Procedures for Determining the
Organic Vapor Barrier Properties through Polymer Membranes," Master's thesis,
Michigan State University, East Lansing, MI, (1996)

Ikegami, T. et al. "Sorption of Volatile Compounds by Polyethylene Film," Nippon
Shokuhin Kogyo Gakkaishi, Vol. 35, No. 7, pp. 457-463 (1988)

p Imai, N., “New Transparent Barrier Film,” JPI Journal, Vol. 36, No. 9, pp. 939-944

(1998)

Indou, M., Kouryou no Jissai Tishiki, 2nd ed, Toyo Keizai Shinpou-shya, Tokyo,
Japan (1985)

Jamieson, E. H. H. and A. H. Windle, " Structure and Oxygen -Barrier Properties of
Metallized Polymer Film," J. of Materials Science, 18, pp. 64-80, (1983)

121

Kaihou, K., "Glass Coated Films," in K. Kondou et al. ed., Kinousei Syokuhin-
Housou Handbook, Science Forum, Tokyo, Japan, pp. 114-120 (1989)

Kai], J. A. E., "Flavor barrier evaluation enhances material selection," Packaging,
September pp. 68-70 (1984)

» -- ,.>Kelly, R. S. A., "Development of Aluminum Oxide Clear Barrier Films," Proceedings
of 37th Annual Technical Conference, Society of Vacuum Coaters, pp. 144-149 (1994)

Kelly, R. S. A., "Development of Clear Barrier Films In Europe, " Proceedings of 36th
Annual Technical Conference, Society of Vacuum Coaters, pp. 312-317 (1993)

Kobayashi, M. et al. "Permeability and Diffusivity of d-Limonene Vapor in Polymeric
Sealnt Films," J. of Food Science, Vol. 60, No. 1, pp. 205-209 (1995)

Kosinowski, J ., "Diffusion and Solubility of n-Alkanes in Polyoleﬁns," J. Applied
Polymer Sci, 31, pp. 1805 (1986)

Laoharavee, T., "The Permeability of Binary Organic Vapor Mixtures through
Polymer Membranes by a Dynamic Purge and Trap / Thermal Deposition Procedure,"
Master's thesis, Michigan State University, East Lansing, MI, (1997)

Marzolf, T., and P. T. Deiassus, "Comparison of Abuse Resistance in Barrier
Packaging Laminates," Paper, Film & Foil Converter, February, pp. 43-48 (1981)

Meares, P., Polymers: Structure and Bulk Properties, Van Nostrand & Co., Ltd,
London, UK, (1965)

Nelson, R. J ., "Plasma Enhanced Chemical Vapor Deposition of SiOx Coatings for Gas

Diffusion Barrier Enhancement of Flexible Polymerized Compounds, " Proceeding,
Society for the Research of Polymerized Compound Surface, Fukui, Japan, July
(1993)

Norton, N. J ., "Helium Diffusion through Glass," J. of the American Ceramic Society,
Vol. 36, No. 3, pp. 90-96 (1953)

Oliveira, L. M., "Mechanical Stress Resistance of Barrier Packaging Films,"
Proceedings of 10th World Conference on Packaging, IAPRI, (1997)

Perry's Chemical Engineers' Handbook, 6th edition, McGraw-Hill (1984)

122

‘1)

l

Ill/7’ Phillips, R. W., "Evaporated Dielectric Colorless Films on PET and OPP Exhibiting
High Barriers Toward Moisture and Oxygen," Proceedings of 36th Annual Technical
Conference, Society of Vacuum Coaters, pp. 293-301 (1993)

Rogers, C. E., "Permeation of Gases and Vapors in Polymers," in J. Comyn, ed.,
Polymer Permeability, Elsevier Applied Science Publishers, New York, NY, pp. 11-74
(1985)

Rogers, C. E., "Solubility and Diffusivity," D. Fox et al. ed., Physics and Chemistry of
the organic Solid State, Interscience Publishers, New York, NY, pp. 510-627 (1965)

Sajiki, T., "Permeation of Ethyl Acetate Vapor through Silica Deposited Polyethylene
Terephthalate Film and Composite Structures," Master's thesis, Michigan State
University, East Lansing, MI, (1991)

Sakamaki, C. and M. Kano, "Application Study of High Barrier Ceramics Deposited
Film," Proceedings of Barrier Pack 89, Chicago, IL, May (1989)

4”

.\\

,3-.. ’7" Schiller, S. et al. "How to produce A1203 coatings on Plastic Films," Proceedings of
7th International Conference, Vacuum Web Coating, Miami, FL, November (1993)

Theodorou, E., and J. S. Pail, "Effect of Organic Vapor Interaction on Permeation Rate
through Polymer Films, " Packaging Technology and Science, Vol. 5, pp. 21-25 (1992)

Tocelo, Tocelo Handbook, Tocelo Co., Ltd, Tokyo, Japan.

:3.: Tropsha, Y. G., "Combinational Barrier Effect of the Mulilayer SiOx Coatings on
Polymer Substrates," in Proceedings of Society of Vacuum Coaters, 40th Annual
Technical Conference, pp. 64-69 (1997)

Yamamoto, H., "Inorganic Barrier Film '[B-Film'," Convertec, 11, pp. 1-3 (1998)

123