_ 3... _
flip-fl.“ . -- v1.41\
.3... L . “warns.
inequhaq 5
finanwnuhhahm fl.
. .. .n.s§.
.3... a .5

31h ._ wink ......P
rtsvh, ‘ $

1. o.

‘lt 3"
Q. . A is!!!
vnv' . .wr

, ‘Wdh. .
.11.; .5
» .§n_.....«|..,‘.a
.

. ‘ ..UV....MM»$I final
. «much. u. ‘ . ‘ ‘ $ .
u . . .
.P... .6 ..

h

{

.
Ft. , .5, .
...s.;...3m
..

”row

'“ii‘fiu
“ ”UH

.

6...»... v
u. s. z.
033:...

”"1

.Jb....i.r. .
...¢.>v.h..-H
3.1.- . .1
,t..fi.~ :v...f.l<. .
.3 c. J, . . bruvgk.

.nrt )| .v 1) I.

Q‘ at I Iri‘ll‘l.‘l’l

{514 .l ‘ .
.vvlvi‘ 1:37... . . .1 .
:On}...-)\.\ 1‘

V n
'0.
an...
I f. 117.
3

nufld. ‘

lav.

.M‘u
~13le

v3.0v..il .
.p».VIabkoL0.:‘l )

Iti..wn¢rrnfl ..\-..
.H I [5-5‘-l¢ .
{bit .105.
< fl). ‘9]. 3.1,
L\avv3...t.....
A. 13:
ufrx‘ s.
.3011??th
. :y . vv Ital.
..L... ..F...xa.i...
.03“. N. ..
71.34. 8.! .a

1
570'

.vu F. u. v.

15.: A
33. n4

. p xvn.kl
. O“1o>v|.l'~
. I . vvz. . . ‘
.IINS‘EQ‘TY.

h: ‘. . . ., ‘

.. .. v.
Hulrflt. .
asikw.

lltvoy... Al .

§

 

 

THESIS

HISGAN TSAT

II2IIII

             

III

 

 

 

 

 

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

01565 0934

 

LIBRARY

Michigan State
University

 

 

 

This is to certify that the

thesis entitled

DETERMINATION OF THE PARTITION DISTRIBUTION OF

VANILLIN AND ORTH-VANILLIN BETWEEN CORRUGATED-BOARD

AND HEADSPACE; AND BETWEEN LOW DENSITY POLYETHYLENE
AND AN AQUEOUS PHASE

presented by

LEI PEI

has been accepted towards fulfillment
of the requirements for

MASTER degree in PACKAGING

 

 

     

Major professor

Date—AERIL_29.._J.9.9J__

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

 

PLACE IN RETURN BOX to roman thh chookout from your rooord.
TO AVOID FINES return on or before date duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I II I_I
I L:

 

 

 

 

 

 

 

L__I:I

DETERMINATION OF THE PARTITION DISTRIBUTION OF
VANILLIN AND ORTHO-VANILLIN BETWEEN CORRUGATED-
BOARD AND HEADSPACE; AND BETWEEN LOW DENSITY
POLYETHYLENE AND AN AQUEOUS PHASE

By

Lei Pei

A THESIS

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

MASTER OF PACKAGING

School of Packaging

1997

ABSTRACT

DETERMINATION OF THE PARTITION DISTRIBUTION OF
VANILLIN AND ORTHO—VANILLIN BETWEEN CORRUGATED-
BOARD AND HEADSPACE; AND BETWEEN LOW DENSITY
POLYETHYLENE AND AN AQUEOUS PHASE

By

Lei Pei

Gas chromatograph, high performance liquid chromatograph, and
thermal stripper/thermal desorption procedures were employed to determine
the partition coefficient, (K) values, of vanillin and ortho-vanillin between
corrugated board and headspace, and between low density polyethylene
(LDPE) and an aqueous phase. The derived coefficient values were as follows:

1) The partition coefficients, KB,H values for vanillin between
corrugated board and headspace, were 179 at 23°C, 98 at 38°C, and 48 at
50°C, respectively. The KB,H values for ortho-vanillin were 663 at 23°C, 353
at 38°C, and 223 at 50°C, respectively.

2) For the aqueous phase/LDPE/ system, the partition coefficient,
Km were 32.7 at 23°C and 18.7 at 32°C for vanillin and 48.5 at 23°C as
compared with 28.7 at 32°C for ortho-vanillin.

3) The diffilsion coefficients (D) for vanillin and ortho-vanillin
through LDPE were determined according to Crank’s method (1976) and were
2.0 x 10'13 cmz/sec at 23°C and 3.9 x 1013 cmZ/sec at 32°C for vanillin; and 6.0
x 10'” cmZ/sec at 23°C and 1.5 x 10'13 cmZ/sec at 32°C for ortho-vanillin at

90% of equilibrium state.

To my parents

iii

ACKNOWLEDGMENTS

I would like to express my sincere appreciation and thanks to my
advisor, Dr. Jack R. Giacin, School of Packaging, Michigan State
University, for his advice, kindness, patience, encouragement, and
positive guidance.

I also wish to extend my sincere thanks to the committee
members, Dr. Ruben J. Hernandez, School of Packaging, and Dr. Alec
B. Scranton, Department of Chemical Engineering, for their valuable
and professional guidance and advice.

I offer final thanks to the graduate students in the School of
Packaging for their suggestions and encouragement throughout the

course of research.

iv

p—A
.0

TABLE OF CONTENTS

LIST OF TABLES ............................................................ viii
LIST OF FIGURES ............................................................ ix
INTRODUCTION ............................................................... 1
LITERATURE REVIEW ..................................................... 3
4.1 Vanillin and Ortho-vanillin ........................................ 3
4.2 Corrugated Board in Packaging ................................. 5
4.2.1 Construction of Corrugated Board ................... 6
4.2.2 Method of Corrugated board Making .............. 7
4.3 Plastics in Packaging ............................................... 10
4.3.1 Polyethylene ................................................... 12
4.3.2 Polypropylene ................................................ 15
4.3.3 Polyvinyl chloride (PVC) ............................... 17
4.3.4 Polyethylene Terephthalate (PET) ................. 18
4.3.5 Polyamides ..................................................... 20
4.3.6 Ethylene Vinyl Alcohol (EVOH) ................... 23
4.3.7 Morphology of Packaging Polymers ........... 23
4.4 Mass Transport phenomena ..................................... 25
4.4.1 Permeation ..................................................... 25
4.4.2 Solubility ........................................................ 28
4.4.3 Migration ....................................................... 32

4.5 Modeling the equilibrium distribution of vanillin
and o-vanillin between the paperboard shipper

and a LDPE primary packaging system ................... 33
MATERIALS AND METHODS ........................................ 35
5.1 Materials ................................................................... 35
5.2 Methods .................................................................... 37

5.2.1 Part 1: Determination of equilibrium
partition distribution of vanillin and
ortho-vanillin between corrugated board

V

5.2.2

5.2.3

and its headspace .................................. 37

5.2.1.1 Construction of calibration curve ...... 37
5.2.1.2 Initial vanillin and ortho-vanillin
levels in corrugated board ................. 39
5.2.1.3 Equilibrium partition between
corrugated board and headspace ....... 39
Part 2: Determination of the stability of

vanillin and ortho-vanillin solutions. 43
5.2.2.1 Determining calibration factors for

the HPLC procedure .......................... 43
5.2.2.2 Determining the stability of vanillin
and ortho-vanillin solutions .............. 46

Part 3: Equilibrium partition distribution of
vanillin and o-vanillin between low density
polyethylene and an aqueous contact phase .. 49

RESULT AND DISCUSSIONS ......................................... 53
6.1 The results of equilibrium partition distribution

of vanillin and ortho-vanillin between

corrugated board and its headspace ......................... 53

6.1.1

6.1.2

Initial vanillin and ortho-vanillin in

corrugated board ........................................... 53
Vanillin and ortho-vanillin equilibrium

partition coefficient between corrugated

board and headspace ...................................... 54

6.2 Stability of vanillin and ortho-vanillin
in aqueous phases ..................................................... 61
6.3 The results of uptake of sorbates by the LDPE as
function of storage time ........................................... 68

6.3.1

Determination of partition coefficient Km
and data analysis ............................................ 68

6.3.2 Estimation of diffusion coefficient

6.3.3

(D) of vanillin and ortho—vanillin

in low density polyethylene ........................... 79
The solubility parameter (6) of vanillin
and ortho-vanillin in LDPE ........................... 86

vi

7: SUMMARY AND CONCLUSIONS ................................. 95

8: APPENDIXES
APPENDIX A: Calibration factors for gas

Chromatography ............................ 98
APPENDIX B: Calibration factors for HPLC
process ........................................... 104

APPENDIX C: Time needed to reach a steady state
within a paperboard/headspace
sorption cell .................................. 107

9: BIBLIOGRAPHY ............................................................... 1 10

vii

LIST OF TABLES

Table 1- Initial concentration of vanillin and ortho-vanillin
in corrugated board ............................................... 51
Table 2- Partition coefficient KB,H between headspace and
paperboard for vanillin and ortho-vanillin as a

function of temperature .......................................... 5 8
Table 3- Storage stability of vanillin and ortho-vanillin in

aqueous solution (23 °C) ........................................ 62
Table 4- The data of partition distribution studies of

vanillin at 23 °C ...................................................... 70
Table 5- The data of partition distribution studies of

vanillin at 32 °C ...................................................... 71
Table 6- The data of partition distribution studies of

ortho-vanillin at 23 °C ............................................. 72
Table 7- The data of partition distribution studies of

ortho-vanillin at 32 °C ............................................. 73

Table 8- MM, vs. storage time of vanillin and ortho-vanillin.. 81
Table 9- The diffiision coefficient (D) values determined to

reach M/M,= 0.6, MJM.= 0.7, and M/M.= 0.8,

according to Eq. 14 ................................................. 81
Table 10- The solubility parameter (J/cm3)'/’ of water, LDPE,

vanillin, and ortho-vanillin .................................... 88
Table 11- The solubility parameter (J/cm3)'/’ difference

between water, LDPE, vanillin, and o-vanillin ...... 89
Table 12- Prediction of equilibrium concentration in the

paperboard and in an aqueous phase ...................... 93

viii

LIST OF FIGURES

Figure 1- Chemical structure of vanillin and ortho-vanillin ........ 3
Figure 2- Typical chromatogram of standard solution of

vanillin and vanillic acid .......................................... 4
Figure 3- Basic construction of corrugated fiberboard ................ 6
Figure 4- The flutes of corrugated fiberboard ............................. 6
Figure 5- Cross section of fibers .................................................. 9
Figure 6- Cellulose molecule, length of chain is from 100

to 3,000 units ........................................................... 9
Figure 7- molecular structures of polyethylene .......................... 14
Figure 8- PET is condensed from trrephthalic acid (PTA)

and ethylene glycol ................................................. 18
Figure 9- Oxygen transmission of clearcoat on 0.5 mil

PET vs. coating thickness ....................................... 19
Figure 10- Chemical structure of selected commercial nylons. 21
Figure 11- Chemical structure of MXD-6 Nylon ....................... 22
Figure 12- The permeation mechanism ...................................... 26

Figure 13- Ortho-vanillin generation system for preparation
of paperboard samples of known o-vanillin

concentration ......................................................... 40
Figure 14- Direct exposure method for preparation of

paperboard samples of known concentration ........ 41
Figure 15- Vacuum filter system ............................................... 45
Figure 16- Vacuum degas system .............................................. 48
Figure 17- Environmental controlled storage chamber ............... 51
Figure 18- Sorption liquid test cell ............................................ 52

Figure 19- Vanillin partition coefficient KB,H between
headspace and paperboard as a function of
temperature ............................................................ 59
Figure 20- Ortho-vanillin partition coefficient KB,H between
headspace and paperboard as a function of

ix

temperature ............................................................ 60
Figure 21- Storage stability of vanillin and ortho-vanillin in

aqueous phase (23 °C) .......................................... 64
Figure 22- Chromatograms of vanillin peak decrease

as a function of storage time ................................ 65
Figure 23- Chromatograms of vanillin and vanillin acid ......... 65
Figure 24- Chromatograms of ortho—vanillic ........................... 65

Figure 25- The sample concentration change in both polymer

and aqueous phase as a function of time,

including control sample and standard solution,

vanillin at 23 °C ................................................... 74
Figure 26- The sample concentration change in both polymer

and aqueous phase as a function of time,

including control sample and standard solution,

vanillin at 32 °C ...................................................... 75
Figure 27- The sample concentration change in both polymer

and aqueous phase as a function of time,

including control sample and standard solution,

ortho-vanillin at 23 °C ............................................ 76
Figure 28- The sample concentration change in both polymer

and aqueous phase as a function of time,

including control sample and standard solution,

ortho—vanillin at 32 °C ............................................ 77
Figure 29- The concentration ratio for both vanillin and

ortho-vanillin at different temperature vs.

storage time ............................................................ 78
Figure 30- The M/M. vs. time for sorption of vanillin

by LDPE at 23°C ................................................... 82
Figure 31- The MJM. vs. time for sorption of vanillin

by LDPE at 32°C ................................................... 83
Figure 32- The MJM. vs. time for sorption of ortho-vanillin

by LDPE at 23°C ................................................... 84
Figure 33- The M vs. time for sorption of ortho-vanillin

by LDPE at 32°C ................................................... 85
Figure 34- Typical storage and transportation environment of a

shipper/polymer packaging system ...................... 94
Figure 35- Standard calibration curve of vanillin ..................... 100
Figure 36- TS/TD calibration curve of vanillin ........................ 101

X

Figure 37- Standard calibration curve of o-vanillin ................. 102

Figure 38-TS/TD calibration curve of o-vanillin ..................... 103
Figure 39- Vanillin calibration curve of HPLC process .......... 105
Figure 40- O-vanillin calibration curve of HPLC process ...... 106
Figure 41- The time needed to reach a steady state with a
paperboard/headspace sorption cell .................... 109

INTRODUCTION

The transfer of volatile, low molecular weight organic compounds,
between product and package is always a concern with polymeric package
systems ( Mohney, 1988). This can lead either to a direct loss of product
quality or to damage of the package, which indirectly might affect the product
negatively (Nielsen, 1992). The phenomenon of transport involves sorption
which refers to the uptake by the packaging material of sorbate initially
contained by the product, and migration which is the reverse process by which
compounds originally contained in the package are transferred to the packaged
product. The sorption process is of paramount importance in avoiding the
effect of “flavor scalping” (Mohney, 1988), while migration is important with
regard to food safety issues (F oulke, 1991).

The diffusion coefficient, D, is used to provide information related to the
time required to attain a constant rate of sorption or migration, since both
processes tend to reach equilibrium to equate their chemical potential values
(Gavara, 1992). The amount of a specific volatile substance sorbed by a
packaging material can be described by the partition coefficient, value K,
which is defined as the concentration of the sorbate in the polymer phase
divided by the concentration of the sorbate in the contact phase (Nielsen,
1992). In a two-phase food/polymer system, the transfer of a component
(sorbate) from one phase to the other, proceeds until the system attains

thermodynamic equilibrium. The equilibrium concentration value of the

1

2

sorbate in each phase is a function of the initial concentrationsorbate solubility
in both contact and polymer phases, and temperature (Gavara, 1992).

In this regard, corrugated board or paperboard is another packaging
concern, especially for recycled paper products, if they contact food directly,
such as the food contained by corrugated boxes during transportation. Paper
is recycled at a fairly low temperature, and contaminants could remain. Some
recycled products could be made of newspaper, which may contain dioxins
and furans, and a complex mixture of related compounds that are formed in
trace amounts when chlorine or chlorine derivatives are used as the primary
bleaching agents in making white paper products. The most toxic dioxin is 2,
3, 7, 8-tetrachlorodibenzo-p—dioxin (TCDD), which has been shown to cause
cancer in laboratory test animals (Foulke, 1991).

In the present study, two equilibrium partition systems were considered,
namely, (1) corrugated board/vanillin or ortho-vanillin/headspace; and (ii) low
density polyethylene-vanillin or ortho-vanillin—aqueous phase. These served
as a model system to address the problem, where volatilization of organic
moieties from corrugated board and their subsequent permeation through a
primary package, which is a polymeric structure, can result in contamination
of the contained product. The partition coefficients (K) of vanillin and ortho-
vanillin between corrugated board and headspace; and LDPE and an aqueous
phase were determined by a series of analytical methods, including gas
chromatography, high performance liquid chromatography, and a thermal
stripper / thermal desorption (TS/I'D) procedure. Statistical analysis of the
partition coefficient (K) data, the vanillin and ortho-vanillin diffusion

coefficient values, and their solubility in LDPE were also performed.

LITERATURE REVIEW

4.1 Vanillin and Ortho-vanillin

Vanillin, the common name for 3-methoxy-4-hydroxy-benzaldehyde,
occurs in nature as a glucoside which hydrolyzes to vanillin and sugar. Ortho-
vanillin is a byproduct from the manufacture of vanillin from guaiacol which

is present in wood tars (Van Ness, 1983). Their chemical structures are shown

in Figure 1.
CH0 CHO
OH
OCH3 O
OH OCH,
Vanillin Ortho-vanillin

Figure 1: Chemical structure of vanillin and ortho-vanillin

The aroma and taste of vanillin have been recognized because of its wide
use in food as a flavoring. The use of vanillin as a flavor can be traced back to
about 1520 when Cortez, the Spanish conqueror of Mexico, was served a
chocolate drink flavored with vanilla, which contains vanillin, by the Aztec
Indians. Cortez took the flavor back to Europe which started its spread around
the world (Van Ness, 1983). The major quantity of vanillin used today comes
mainly from lignin, which is a waste product from the manufacture of paper

pulp. A smaller amount is made synthetically from guaiacol (2-
3

4

methoxyphenol) (Van Ness, 1983). Besides being a very popular flavor,
vanillin is also used in the synthesis of drugs. The largest single use for
vanillin as a starting material for the manufacture of the antihypertensive drug
called Aldomet, which is manufactured by Merck. Aldomet is l-3-(3,4-
dihydroxy phenyl)-2-methylalanine. L-dopa is another drug made from vanilla
and is used for the treatment of Parkinson’s disease ( Ainscough and Brodie,
1990)

A liquid chromatographic method for determination of vanillin, vanillic

acid, etc. was described by Didney (1989). The liquid chromatograph was

peak high \

5 1O 15 20
retention time

 

 

 

 

Figure 2: Typical chromatogram of standard solution of vanillin and vanillic acid.

operated at flow rate of 2.5 ml/min, with UV detection at 254 nm. The column
was stainless steel, 25 cm x 4.6 mm id, packed with C18 on 10 um totally
porous irregular particles, LiChrosorb or equivalent. Mobile phase was 10%

(v/v) methanol and 90% (v/v) LC grade water. A typical chromatogram of a

5

standard solution of vanillin and vanillic acid is shown in Figure 2. Peak 1 is
vanillin with a retention time of 11 min and peak 2 is vanillic acid with a

retention time of 13 min.

4.2 Corrugated Board in Packaging

By virtue of its remarkable utility, exceptional strength characteristics,
light weight, and low cost, corrugated and fiberboard packaging materials are
extensively used in a variety of industrial packaging functions. About 67
million tons of paper are produced annually in the United States, and this
tonnage is about equally divided between paper (0.012 in thick or less) and
paperboard (over 0.012 in) (Handbook of Packaging Engineering, 1992).
4. 2. 1 Construction of Corrugated Board

Two categories of packaging paperboard are used today: one used for
cartons and the other used for corrugated board. The corrugated fiberboard is
produced basically for the construction of single-faced corrugated fiberboard,
single-wall corrugated fiberboard, double-wall corrugated fiberboard, and
triple-wall corrugated fiberboard, as shown in Figure 3. Single-faced
corrugated fiberboard is used principally for wrapping, and for interior
packing. Single-wall corrugated fiberboard, is made by gluing flat facings to
each side of the corrugated member. It is manufactured in various kinds and
grades of paperboard and in varying thickness. The archetype construction,
which gives strength and rigidity to the fiberboard, also provides cushioning
and insulation characteristics. Double-wall corrugated fiberboard consists of
three flat facings and two corrugated members. Like single-wall board, it is

manufactured in various types and grades of materials and in varying

Figure 3: Basic construction of corrugated fiberboard

”E" flute
mm

“C' flute

m

"B' flute

m:

'A' flute

AW

 

Figure 4: The flutes of corrugated fiberboard

7

thicknesses, but it offers greater strength or cushioning than that provided by
the single-wall structure. The triple-wall corrugated fiberboard consists of four
flat facings and three corrugated members and is most applicable for packages
requiring exceptional rigidity and high puncture resistance.

The flutes or arches of corrugated board are specified by size A, B, C,
or E, as shown in Figure 4 (Friedman, 1977). A is the largest. C is next
followed by B. The “E” flute has the smallest flute (Friedman, 1977). Listed

below are the dimensions for each flute.

A Flute 36 :l: 3 flutes per lineal foot 1/ ” in thickness
B Flute 50 i 3 flutes per lineal foot 1/8” in thickness
C Flute 42 i 3 flutes per lineal foot 3/ 16” in thickness
E Flute 94 :l: 3 flutes per lineal foot 3/64” in thickness

For double-wall and triple-wall board, the flute sizes for each sheet of
corrugated material are usually different (Fibre Box Association, 1992). This
technique combines their advantages, and prevents any weakness that might
result if the flutes were perfectly aligned. For instance, the B and E flute have
better printing character, and A and C flute have relatively better cushioning
ability. The CB flute double wall board will give the user better cushion and
compression ability, at the same it will have better printing quality.

4. 2.2 Method of Corrugated board Making

The facings of corrugated board are usually made from spruce or pine
chips digested by an “incomplete cook” process which preserves the long
fibers of the original wood cellulose. This process results in a high yield and
a product of exceptional tearing strength ( Friedman, 1977). “Softwood” is the

term used in the paper industry for coniferous or needle-bearing trees, whose

8

fibers will run up to 1/4 inch in length, as compared to less than 1/ 10 inch for
that of “Hardwood” fibers. A fiber in cross section looks somewhat like a
hollow tube of irregular wall made up of a lamellae of fibrils, as shown in
Figure 5. These fibrils are made up of microfibrils that are composed of chains
of cellulose molecules, about 3 million in each microfibril, along with short-
chain hemicellulose molecules and other polymeric residues (see figure 6)
(Handbook of Packaging Engineering, 1992). In papermaking, the wood is
cooked with chemicals to remove the lignin and carbohydrates that enclose the
lamellae of fibrils. The krafi process, also known as the sulfate process, is used
on soft-woods to produce the strongest pulp product which is called natural
kraft paper. Wood that has been converted into small chips is cooked for
several hours in digester, which is followed by washing and bleaching. The
slurry for making paper is then put into beaters, where the bundles of fiber are
broken up and hydration of the fibrils takes place; that is , the fibrils combine
chemically and physically with water to become thoroughly “wetted” so that
they swell. From there the mixture of pulp is ready to go through the
papermaking machine. Two basic machines are used today: the fourdrinier
machine and the cylinder machine. (Handbook of Packaging Engineering,
1992). The paperboard above 12 points can be made on either type of
machine.

Corrugating medium is made by a semi-chemical process in which
virgin hardwoods and recycled corrugated containers are used as the raw
materials. A soaking in caustic soda or neutral sodium sulfite, to soften the

lignin and carbohydrates that bind the fibers together, is followed by grinding

Lumen
Lamellas

  

Figure 5: Cross section of fibers

OH OH H20H

CH20H OH OH

Figure 6: Cellulose Molecule
Length of chain is from 100 to 3,000 units

10

in a diskrefiner (Handbook of Packaging Engineering, 1992). This method
results in a low-cost pulp because of the raw materials and process, but it is
hard to bleach. The common basis weights of corrugating medium are 26 and
33 lb./msf (Fibre Box Association, 1992).

Paper is a kind of material that can be recycled. There is a growing
industry, government and consumer interest in recycling materials in the
United States. However, paper is recycled at relatively low temperature, and
contaminants could remain. Some recycled products are made of newspaper
which may contain dioxins and furans, and a complex mixture of related
compounds that are formed in trace amounts when chlorine or chlorine
derivatives are used as the primary bleaching agents in making white paper
products. The most toxic dioxin is 2, 3, 7, 8-tetrachlorodibenzo-p—dioxin

(TCDD), which has been shown to cause cancer in laboratory test animals

(Foulke, 1991).

4.3 Plastics in Packaging

Plastics are now one of the fastest growing materials in the packaging
industry. The principal resins used in packaging are polyethylene,
polypropylene, polyvinyl chloride (PVC), polyethylene terephthalate (PET),
polyamides, and ethylene vinyl alcohol copolymer (Bonis, 1994). Among
them, polyethylene, particularly low-density polyethylene (LDPE), is widely
used throughout the packaging industry.

The choice of a plastic for use in a particular application will be
governed by a number of considerations. There are first, the environmental

conditions of packing, shipment, and storage; second, physical characteristics

11

of the packaging materials; and third, chemical characteristics including the
compatibility between packed product and packaging. The following listing
presents several important properties of plastics that are related to the
packaging industry. The numbers in parenthesis are the ASTM standards
describing the method for their evaluation.
Chemical Chemical composition and microstructure
Molecular weight and molecular weight distribution
Morphology and crystallinity
Physical Density (ASTM D-1505)
Coefficient of linear thermal expansion, Length/Length
(ASTM D-696)
Deflection temperature (ASTM D—648)
Thermal conductivity, cal-cm/sec. sq cm 0C (ASTM
C- l 77)
Melting temperature Tm
Glass transition temperature Tg
Melt Flow Index (ASTM D-123 8)
Mechanical Tensile strength, psi (ASTM D-63 8)
Elongation, length/length (ASTM D-63 8)
Bursting strength (ASTM D-774)
Impact strength (ASTM D-3420)
Tear strength (ASTM D-1004)
Coefficient of friction (ASTM D-1894)
Heat sealability
Gloss (ASTM 2457)

12

The International Union of Pure and Applied Chemistry (IUPAC)
defined a polymer as a substance composed of molecules characterized by the
multiple repetition of one or more species of atoms or groups of atoms
(constitutional units) linked to each other in amounts sufficient to provide a set
of properties that do not vary markedly with the addition or removal of one or
a few of the constitutional units (also called a macromolecule). Polymers
include wood, protein, and plastic. The molecular weight range of a synthetic
polymer is between 10,000 and 10 million. It has been found that in many
cases the polymer molecules have the form of linear or slightly branched
chains (Mark, Atlas, 1995). For synthetic polymers or plastics, the chemical
composition and microstructure, molecular weight distribution, and the
morphology and crystallinity properties are important, since they affect many
of the characteristic physical and mechanical properties of a polymer
(Billmeyer, 1972).

4. 3. 1 Polyethylene

Ethylene is the simplest of all monomers, as far as chemical structure is
concerned. Through a chemical reaction, the double bond can be opened and
two ethylene molecules combined into a dimer. Further addition of many

repeating units yields the polymer, polyethylene (Giacin, 1994).

E— O— I
II
E— O— I

l3

Polyethylene is a family of addition polymers based on ethylene , which

is essentially a straight-chain compound:

H H II H H
I I I I I

—C—C-w:—C—c—
I I I I I
H H r1 H H

|
m—O—m

l
m—O—m

|
m—O—m

with hundreds, and sometimes thousands of carbon atoms in each chain.
Polyethylene can be linear or branched, a homopolymer or copolymer. A
segment of a chain branching interferes with the close packing of the
molecules and results in a soft and lower density material. Less branching, on
the other hand, allows the molecules to pack together and makes the polymer
tend to crystallize because of the structural regularity, which produces a stiffer,
more dense, and less permeable material (Hernandez, 1994). The glass
transition temperature (Tg) is about -125 °C , and Tm is 135 °C for high density
and 110 °C for low density, respectively (Mark and Atlas, 1995).

There are basically two variants of polyethylene, differing in structure
and properties as well as in the manufacturing process. High density
polyethylene (HDPE), whose density is between 0.95 and 0.970 gram/cc, has
a percentage of crystallinity from 70 to 90%. Low density polyethylene
(LDPE) with density from 0.910 to 0.950 gram/cc, typically has a crystallinity
ranging between 40 to 60% (Hernandez, 1994) .

The process for making high density polyethylene was developed by
Prof. Karl Ziegler of Germany and several others by using a catalyst of

titanium tetrachloride and triethyl aluminum in the middle 19503’, notably

14

Phillips Petroleum Company (Handbook of Packaging Engineering, 1992).
This process works at normal atmospheric pressure and room temperature, so
it is sometimes called low-pressure polyethylene. The molecular structure of

high density polyethylene is essentially linear with relatively few side branches

Are-d M

Low-density polyethylene High-density polyethylene

Figure 7: Molecular structures of polyethylene

as compared to LDPE. This is shown in Figure 7. Thus, the molecules are
permitted to line up nearly parallel and close together to crystallize and provide
a hard, stiff, impermeable material for inj ection-molding containers.
Long-chain branched low-density polyethylene is normally produced by
high-pressure processes, in autoclave or tubular reactors by means of a free-
radical polymerization procedure (Manders, 1994) which is carried out at
approximately 50,000 psi and temperatures of about 300 °F (Handbook of
Packaging Engineering, 1992). This process produces ramifications in the main
backbone chain resulting in a branched polymer that is actually a mixture of
large molecules with different backbone lengths, various side chain lengths and
with various degree of side-chain branching. The applications of LDPE can

be as both film and sheet, injection molding, coating, and blow molding

15

because of their flexibility, moisture protection, and light weight. In the United
States, about 55% of the total LDPE is made into films with thickness as less
than 12 mils (300 microns) (Handbook of Packaging Engineering, 1992).

Currently, Polyethylene continues to experience significant
technological developments Wanders, 1994). Over the years, production of
polyethylene has developed a high degree of control over such polymer
properties as molecular-weight distribution and degrees of short- and long-
chain branching. Molecular weight can be controlled by chain-termination
agents and reactor conditions, such as with linear low density polyethylene
(LLDPE).

LLDPE is characterized by a narrow distribution of MW and a linear
structure with very short branches, due to the presence of butene, hexene or
octene as a comonomer (Hernandez, 1994).

Butene HC = CHICHZCH3
Hexene HC = CH(CH2)3CH3
Octene HC = CH(CH2)5CH3

LLDPE has a higher degree of crystallinity and is stiffer than LDPE
because of the linearity of its molecules, and it has higher tensile strength,
puncture resistance, tear properties and elongation than LDPE.

4. 3.2 Polypropylene

Polypropylene (PP) is a group of thermoplastic polymers based on the
polymerization of propylene monomer, CH2=CH-CH3. It is produced under
controlled conditions of heat and pressure in the presence of Ziegler Natta

catalysts (Bailey and Brauer, 1994).

16

I' 9H3
H—fC— C+n
I I
H H

There are three basic types of polypropylene: isotactic, atactic, and
syndiotactic (Giacin, 1994). Isotactic polypropylene is the most common
commercial polypropylene, whose molecular structure allows it to assume a
helical and crystalline configuration, which offers a enough material stiff
enough for a wide range of commercial applications. It has a density of 0.90-
0.91 gram/cc, and a melting point of 165 °C. The melt-flow index for this
material ranges from under 0.3 gram/10 min. for extruded sheets to greater
than 800 gram/ 10 min. for some fiber grades (Bailey and Brauer, 1994). Other
advantages of the resin include resistance to most organic solvents, with the
exception of very strong oxidizing agents such as fuming nitric acid or sulfuric
acid. Atactic polypropylene, which has a random molecular pattern, is
commonly used as an adhesive, or mixed with other materials, such as asphalt,
for roofing applications. Syndiotactic polypropylene offers a lower melting
point, better impact resistance and more clarity than isotactic homopolymer
polypropylene (Hernandez, 1994).

Random copolymers of polypropylene typically contain several percent
of ethylene as comonomer (Hernandez, 1994). The copolymers with a melt
flow index of 35 gram/ 10 min. or above find applications in thin-wall parts,
recently used for inj ection-molded food packaging (Bailey and Brauer, 1994).

l7

4. 3.3 Polyvinyl chloride (PVC)

PVC is a homopolymer of vinyl chloride, CH2=CHC1. The vinyl chloride
is produced by the thermal dehydrochlorination of ethylene dichloride, which
in turn is manufactured by either the direct chlorinating of ethylene, or the oxy-
chlorination of ethylene with hydrochloric acid and oxygen (Hernandez, 1994).

C1 H Cl H Cl H Cl H

I I I I I I I |
—C—C—C—C—C—C—C—C—

II II II II
H H H H H H H H

Vinyl chloride is converted to PVC via a free-radical addition
polymerization, using one of four basic processes: Suspension, emulsion, bulk
or solution (Maresca, 1994). Approximately 90% of the world production of
PVC is produced by the suspension polymerization method and initiated by
peroxide free-radical catalysts. In this process, the microscopic droplets of
vinyl chloride are suspended in water by vigorous agitation. After
polymerization, the residual vinyl chloride monomer is removed by heating
the resin under vacuum. PVC is then centrifiJged and dried to remove the
water. The molecular weight of PVC is controlled largely by the
polymerization temperature, and to a much lesser extent by the initiator
concentration. Higher reaction temperatures generally produce lower-
molecular—weight resins.

PVC has a wide range of properties depending on the manufacturing
process used. In addition to good physical properties, the polymer has

excellent transparency, chemical resistance, long-term stability, flammability

18

resistance, weatherability, flow and electrical properties. The versatility of
PVC is a direct result of its blending compatibility with a wide variety of
plasticizer, fillers, pigments, and other additives. The PVC resin can be
processed by all the standard thermoplastic fabrication techniques including:
(1) extrusion, (2) calendering, (3) injection molding, (4) blow molding, (5)
powder molding, (6) blown film, and (7) coating (Maresca, 1994).

4. 3.4 Polyethylene Terephthalate (PEI)

PET resins are produced by the reaction of purified terephthalic acid
(PTA) and ethylene glycol as shown in figure 8. It is a condensation, linear,
semicrystalline, thermoplastic homopolymer that has found great acceptance
as carbonated beverage packaging material due to its properties of toughness,
clarity, recyclable, and the development of high speed bottle processing
technology. PET has a Tg of approximately 69 °C and Tm of 265 °C. (Mark,
1995)

Although the vast majority of PET resin ends up in blow-molded
containers (Mitchell, 1994), PET based glass coated films have been used

COOH

H2CfOH
CF) / + H2COH
COOH

Terephthalid acid Ethylene glycol

O

0
II 1|
HOIC< Q >c—o—CH,CH,o "I,

Polyethylene Terephthalate

Figure 8: PET is condensed from terephthalic acid (PTA) and ethylene glycol

19

120

g—A
O
O

on
C

02 Permeation (cc/sq.m.24hrs)
A O\
O O

N
O

 

O L 13“.: - L—__4__>
50 100 150 200 250 300

ThicknessIA)

Figure 9: Oxygen transmission of Clearcoat on 0.5 mil PET vs. coating thickness
Source: Mastat Izu, High Performance Ovonic Clearcoat Barrier Film Technology, 1993

20

commercially and have been available at development pricing since mid-1993,
for incorporation into packaging systems on an everyday basis for high barrier
to permit microwaveability in food packaging (Brody, 1993). The chemical
composition of the coating ranged from SiO to SiO2 (Silicon dioxide). The
coating has been given the designation SiOx rather than silica’s SiO2 because
the coating is a mixture of SiO + SiO2 to achieve barrier coatings. Glass coated
packaging materials usually have significantly higher oxygen barrier properties
than their base plastic substrate films, due to the properties imparted by the
coating. Figure 9 shows a plot of the oxygen transmission rate vs. coating
thickness data for 0.5 mil PET films coated with Ovonic Clearcoat, which is
a technology developed by Energy Conversion Devices, Inc.(Izu, 1993). The
figure illustrates that a coating of approximately 120 A is sufficiently thick to
maximize oxygen barrier properties. To date, deposition of 810X coatings has
been mainly onto PET substrates. Since the bulk of the work has been on PET,
the lowest transmission rates have been achieved on it.(Felts, 1993)
4. 3.5 Polyamides

Polyamides or Nylons, are condensation, linear, thermoplastic polymers
that contain the amide group as a recurring part of the chain. At low humidity,
polyamide exhibit strong inter and intramolecular force of atraction, due to the
presence of H—bonding between the HN— group and the polar carbonyl group
-C=O. The molecular structure thus gives Nylons the properties of toughness,
good puncture resistance, impact strength and temperature stability. For
instant, Nylon-6,6 has the melt point of 269 °C. Polyamides also typically
exhibit good chemical resistant and are a good barrier to gas, oil, and aromas.

It, is, however, moisture sensitive, due to H-bonding disruption of the polymers

 

21

NHZ COOH
I |
n (CH2)6 + n (CH2)4 =’
| I

NH2 COOH
Hexamethylene diamine Adipic acid
H H H O O

I | | I I
H [- N-(CH2)6- N-C-(CH2)4 - (3-] n0H+(2n-1)H,o

Nylon 6,6

H O

I II
n H2N(CH2),0COOH - H[ N- (CH2),0,C ],H+(n-1)H,o

ll-amino undecanoic acid Nylon 11

Figure 10: Chemical structure of selected

commercial nylons

22

naturaly which results in oxygen permeability increasing with moisture content
(Hernandez, 1994). Nylon can be made by the condensation of diamines and
dibasic acids or the condensation of amino-acids (See figure 10).

A diverse degree of crystallinity is obtained depending on the
temperature and quenching rate ( Hernandez, 1994). When the cooling rate is
increased, a less crystalline nylon is obtained, since the polymer was not given
sufficient time to form crystallites. For most packaging applications, nylons are
used in film form. It can be combined with other materials that add moisture
barrier and heat sealability such as LDPE, ionomer, or EVA. The degrees of
crystallinity can be controlled during the film production.

A new polymer was developed by Toyobo Co., Ltd. that is called MXD-
6 Nylon (Watanabe, 1987). The MXD-6 nylon is a polycondensation of meta-

xylene diamine (MXDA) and adipic acid as shown in Figure 11.

H,Nll,C©Cll,Nl-l, + nooci CH, ~I4C00H
Mcta-xylylene diamine Adipic acid

a H INHCH, @ CH,NHCf CH)4 CI- OH

Figure l 1: Chemical structure of MXD-6 Nylon.

23

MXD-6 Nylon provides a good barrier not only to gases but to moisture as
well. The gas barrier properties are better than Nylon-6 and PET at all
humilities, and better than ethylene / vinyl alcohol copolymer (EVOH) at
100% RH, due to the existence of the benzene ring in the MXD-6 polymer
chain( Watanabe, 1987). The benzene ring in the MXD-6 polymer chain results
in its having a higher tensile modulus than Nylon-6; and the MXD-6 Nylon
film has good dimensional stability at the high temperatures and tensions
found in printing and laminating processes. (Watanabe, 1987)

4. 3.6 Ethylene Vinyl Alcohol (E VOH)

Ethylene / vinyl alcohol copolymer are produced by a controlled
hydrolysis of ethylene / vinyl acetate copolymer. The presence of OH
functional groups on the backbone chain has a dramatic effect on the
polymers properties. If the percent of vinyl alcohol in the ethylene / vinyl
alcohol copolymer is in the range of 20% to 50%, the copolymers obtained
combine the processability and water resistance of polyethylene and the gas
and odors barrier characteristics of polyvinyl alcohol. The most important
characteristic of EVOH is the outstanding oxygen and odor barrier properties
(Hernandez, 1994). EVOH also provides a very high resistance to oils and
organic vapors. For packaging applications, EVOH can be used in a flexible
film structure or fabricated into a rigid container. It can also be coextruded in
combination with PE or PP, and laminated or coated to various substrates
including PET, PE, and Nylons, etc. (Hernandez, 1994).

4. 3. 7 Morphology of Packaging Polymers
The increasing use of plastics, which contain traces of solvent residues

from processing, in food packaging has led to the need for more information

24

about interactions between plastic packaging materials and foods (Halek and
Chan, 1994). At the same time, sorption is also one of the major problems of
concern with the packaging of foodstuffs (Hotchkiss J. H., 1988). In general,
these are related to the morphology within the polymer, such as structural
regularity of the chains, free vibrational and rotational motion of chain
segments, so that different conformations can be assumed, presence of specific
groups which produce strong lateral intermolecular bonds, and the absence of
bulky side groups, which can inhibit the vibrational and rotational mobility of
an intrinsically flexible chain (Mark 1995).

The polymer molecule experiences torsional oscillation and / or
rotation around covalent single bonds, which is called micro-brownian motion.
The size and frequency of “hole” formation between the molecules are quiet
important to the diffusion, solubility, and permeability of permeant molecules.
Temperature is another factor that can influence the mass transfer process.
Above the glass transition temperature (Tg), the polymer chains experience
segmental mobility, and as the temperature is increased, the size and frequency
of ‘hole’ formation increases, as a result of increased segmental mobility.
Thus, creating a more accessible pathway for permeant molecules to diffuse

through the polymer bulk phase.

25

4.4 Mass Transport phenomena
4. 4. 1 Permeation

The major advantages of plastics films are flexibility, the indeep draw
capability, heat sealability, and the fact that they can be either translucent or
opaque, as required. Their barrier properties vary and are dependent on the
individual polymers chemical and physical properties. No polymer is an
absolute barrier. However, recent developments have made available high
barrier plastics packaging materials, particularly in the form of multi-layer
structures and with the use of coatings. These developments led to the
necessity to investigate the specific permeability of plastic films to various
gases, vapors and liquids. The transfer of gases, vapors and liquids through
films is considered under the title of permeation.

The permeation mechanism actually involves three steps: (1) absorption
and dissolution of the penetrant into the polymer surface exposed to the high
penetrant concentration; (2) diffiision of the penetrant through the polymer;
and (3) desorption and evaporation of the penetrant from the low concentration
surface of the polymer, as shown in figure 12.

Presented in Figure 12 is a generalized permeation model. If it is
assumed that (1) the concentration gradient of the permeant gas within the

polymer film or slab (dc/dl=constant) is constant at steady-state; and (2) there
is no interaction of penetrant with the polymer, the permeability constant P

can be related to the diffusion coefficient D and the solubility coefficient S
of the penetrant in the polymer by the following equation (Mohney, 1988):

P=DS (n

26

 

 

<1) i I"
P1
(5)
c. ’ ’ >
P2
C2
Where: P1 and P2 = Partial pressure of permeant gas at the

two surfaces of the film or slab.
C1 and C2 = Concentration of permeant gas at the
two surfaces of the film or slab.
L = Film or slab thickness

Figure 12: The permeation mechanism

In this case, the permeability constant P is a function of both the
diffusion and solubility coefficients. The diffusion process can be described by
the Fick’s first and second laws of diffusion (Crank, 1975)

27

F = -D (dc/dx) (2)

dc/dt = d[ D (dc/dx)]/dx (3)

Where, F : The flux or the rate of transfer of penetrant

per unit area, expressed as mass of diffusant
per unit area per time.

c: the concentration of the penetrant in the
film, expressed in the same units of mass of
diffusant per unit of volume or mass of the
polymer.

D: the mutual diffusion coefficient, in
(length)2/time.

t: time

x the length in the direction in which transport

of the penetrant molecules occurs.

To obtain the flux ( F ) or the diffusion coefficient ( D ) from equations
(1) and (2), the initial and boundary conditions associated with the
experimental method are needed, and the expressions solved to give the
desired values (Hernandez, 1986).

Only approximated values will be obtained when the diffusion
coefficient is calculated using these equations. If permeation involves

interacting organic vapors, the relationship between P, D, and S is more

28

complex than that indicated by equation (1). For such cases, diffusion
processes may be concentration as well as time dependent that results in
swelling of the polymer matrix and a non-ideal Henry’s Law relationship
(Crank, 1975).
4.4.2 Solubility

Sorption is a phenomena describing the dissolution or transfer of a
substance from the contacting phase to a polymeric packaging materials. The
solubility of essential flavors in polymer structures is of paramount importance
in avoiding the effect of “flavor scalping” or loss due to sorption. The
solubility coefficient can be calculated from sorption experiments by

substitution into the expression (Hernandez, 1986):

 

Moo
S = (4)
w - b
Where: S: solubility coefficient

Moo: total amount of mass absorbed
w: weight of the polymer tested

b: driving force (concentration or pressure)

The solubility coefficient is expressed as mass of vapor sorbed at
equilibrium per mass of polymer per driving force concentration or penetrant
partial pressure. The amount of penetrant (Mt) sorbed by the film during time
't is given by following expressions (Crank, J, 1975).

 

29

 

 

M oo 8 -(2n + 1)2n2Dt
— = l - ex 5
Moo A (2n + 1)21t2 p [ £2 ] ( )
Where, Moo: the amount of penetrant sorbed by the film at
equilibrium.

M,: the amount of penetrant sorbed by the film during
time t.

i : the thickness of the film.

If the time to reach M,/Moo = 0.5 is t,,2 , which means the time required
to reach a sorption level equal to half the equilibrium value, Mac, the diffusion

coefficient (D) can be determined from equation (5) as follows:

0.049 £2
D = —— (6)

ha

For film samples, the diffusion coefficient (D) is determined by
substitution into equation (6) for sorption studies carried out until equilibrium
is reached (Mohney, 1988).

The solubility and diffusivity of liquids and gases in polymers is
dependent upon chemical composition, polymer molecular structure, and
polymer morphology. For sorption, if the polymer and sorbate have equivalent

solubility parameters, they should be mutually soluble or, substances with

30

similar chemical structures should be quite likely to dissolve in each other
(Holtyzer and Krevelen, 1976). The solubility parameter, 6, for a particular
polymer or sorbate can be obtained from the respective cohesive energy Ewh,
which is defined as the increase in internal energy U per mole of substance, if
all the intermolecular forces are eliminated in a condensed state (Hoftyzer and

Krevelen, 1976).

 

Eco
" )“2 (7)

5:
(V

Where, Ecoh: The cohesive energy
6 : solubility parameter

V: Molar Volume

Formally, the cohesive energy may be divided into three compoments,
corresponding to the following three types of interaction forces, dispersion
forces, polar forces, and the contribution of hydrogen bonding. The solubility
parameter components may be predicted from these group contributions, using

the following equations (Hofiyzer, P. J. and Van Krevelen, 1976):

 

 

2 F d.
5d = (8)
V
5 = (9)

V

31

 

5 2 F di

= 10
h V ( )
5 = (6‘12 + 6p2 + 6h2)l/2 (11)

Solubility is dependent upon the chemical composition of both the
packaging materials and penetrant, polymer molecular structure, and polymer
morphology. The storage temperature also affects the sorption significantly
(Nielsen, 1992). In general, with an increase in temperature there is a decrease
in solubility. The temperature dependence of the solubility coefficients can

also be described by an Arrhenius type equation (Maekawa, 1994):
S :80 exp -( Hs/RT) (12)

Where, S: solubility coefficient
S0: pre-exponential constant

H,: the heat of condensation

The heat of condensation, H,, can be positive or negative. For the
condensable vapors such as an organic compound, it is negative due to the
relatively large heat of condensation. Therefore, the solubility coefficient of a

vapor decreases with increasing temperature.

32

4. 4.3 Migration

Migration is the passage of molecules originally contained by the
package itself into the product. In food packaging regulations, the term
migration is used to describe the transfer of substances from the package to the
foodstuff. A distinction is usually made between Global and Specific
migration. Global migration refers to the total transfer or migration of all
migrating species, whether they are toxic or not, from the packaging system
into the packaged food, whereas specific migration relates to one or more
identifiable substances that is a constituent of the packaging material. In
principle, the measurement of total or Global migration from a packaging
materials is quite simple. It consists of placing a sample of the material of
known surface area in contact with an appropriate food simulant or solvent
under defined time-temperature conditions (FDA Guidelines for Chemistry and
Technology Requirements of Indirect food Additive Petitions, 1976) and
determining total transfer. With respect to specific migration, limits are
designed to restrict the transfer into food of those compounds (i.e. monomer)
known or deemed potentially hazardous to human health, but do not take into
account the total quantity of other migrants transferring into the contact phase
(Giacin, 1994).

Both sorption and migration involve the partitioning of low molecular
weight organic species between the polymeric packaging material and the

contact phase, quantitatively described by the partition coefficient ( K )

z [(2.]
[(3.]

K

 

(13)

33

Where, Cp: substance concentration in polymer

Cc: substance concentration in contacting phase

The partition coefficient K is defined as the ratio of the equilibrium
concentration of a component in the contacting polymeric material over the
equilibrium concentration of the component in a fluid phase. Hence, a large
partition coefficient means that the substance has a higher affinity for the
plastic than for the contacting phase, and larger amounts are absorbed by the

packaging material.

4.5 Modeling the equilibrium distribution of vanillin between the
paperboard shipper and a LDPE primary packaging system

A model of equilibrium distribution of vanillin between the paperboard
shipper and a LDPE primary packaging system was described by Maekawa
(1994). Vanillin and ortho-vanillin were proposed as model compounds for
this study. A schematic diagram of the mechanism of vanillin and ortho-

vanillin volatilization or permeation and associated equilibria is as follows:

Corrugated board Container Pol eth lene Aqueous
shipping container ‘— headspace“ (IDE ‘—phase

Gas chromatographic analysis processes, High performance liquid
chromatographic analysis processes, and electrobalance analysis were involved
in Maekawa’s study to determine the partition coefficient of vanillin or ortho-

. vanillin between paperboard and headspace, LDPE film and the aqueous phase.

34

All the studies were carried out at 23°C. The average partition coefficient
values for vanillin and ortho-vanillin between paperboard and headspace were
21.6 and 77.5, respectively at 23°C. The average partition coefficient values for
vanillin and ortho-vanillin between LDPE film and an aqueous phase were 3.8
and 574, respectively.

Migration of volatiles such as vanillin and ortho-vanillin fiom a
corrugated paperboard shipper was considered a mass transfer process under
defined test conditions to include:

1. Time

II. temperature

111. the nature and volume of the respective phases associated with

the system equilibrium
The driving force for migration is the concentration gradient, where the
volatile species of interest transfer from a region of higher concentration (i.e.
corrugated paperboard) to the region of lower concentration. The results from
this study show that the times to reach equilibrium in the corrugated
container/headspace/LDPE packaging system are on the order of 103 days for
a 20 mil wall thickness LDPE ampule (Maekawa, 1994).

MATERIALS AND METHODS

5-1 Materials
The following items and equipment were used in this study:
1: Packaging material:
- Kraft corrugated board. D/F 200 lb specification, C-flute,
(Midwest Paperrnart, Lansing MI)
0 Low density polyethylene, (Dey Laboratories, Inc. Napa,
CA).
2: Volatile sorbates:
- 99% purity vanillin and ortho-vanillin (Aldrich Chemical
Co. Milwaukee, WI).
3: Solvent:
- Chromatography grade acetonitrile (Aldrich Chemical Co.
Inc. Milwaukee, WI).
0 HPLC grade water (Aldrich Chemical Co. Inc. Milwaukee,
WI).
4: Vials and Cans:
- 40 ml screw cap amber glass vials with Teflon coated
silicone septa (Supelco Inc., Bellefonte, PA).
0 Uncoated two-piece steel cans (American National Can,

La Porte, IN)

35

5:

10:

36

Gas Chromatography:
Type: HP 5890A (Hewlett Packard, Avondale,
PA)

Column: Fused silica capillary column
(30m x 0.32mm ID) intermediate
polarity stationary phase SPB-S
(Supelco Co., Bellefonte, PA)
Dynamic Thermal Stripper System, Model 1000 (Dynatherm
Analytical Instruments Inc., Kelton, PA)
Thermal Desorption System, Model 890 (Dynatherm Analytical
Instruments Inc., Kelton, PA)
High Performance Liquid Chromatography, Model 150-C ALC/GPC

with Waters 486 tunable absorbance detector

Thermal ovens
Analytical balance

Vacuum filter system

37

5.2 Methods
There are three principal experimental phases in this study which are

described in detail below:

5.2.1 Part 1: D-t-111'1. 1'1 1 '1.-"'1_1-111.-1"111't'11-'111 1-1' li1
1.11. 0111- .1. .111 “1 .ll-:1t‘! 11, 1.1-11. ‘11. .1-. ‘
5.2.1.1 Construction of Calibration Curve

Gas chromatography analysis required construction of two (2)
calibration curves, based on sample handling. A standard calibration curve was
developed by direct injection of vanillin and ortho-vanillin standard solutions
of known concentration into the gas chromatograph. Calibration curves for
both vanillin and ortho-vanillin, based on the thermal stripper / thermal
desorption (TS/I'D) procedure, were also constructed. The procedure for
developing the TS/TD calibration curve involved injecting a predetermined
quantity of vanillin or ortho-vanillin standard solution directly onto the
carbotrap 300 sorption tube, and then desorbing the vanillin and ortho-vanillin
through the Dynatherm thermal desorption system (model 890) to the gas

chromatograph for quantification.

Test conditions for gas chromatography analysis:

Type: HP 5890A, Hewlett Packard, Avondale, PA.

Column: Fused silica capillary column (30m x 0.32 mm ID)
intermediate polarity stationary phase SPB-S (Supelco
Co., Bellefronte, PA)

Injector: 5 ul liquid syringe, 500 111 gas tight syringe (Hamilton)

38

Flow rate: Helium 7.0 ml/min.
Air 400 ml/min.
Hydrogen 40 ml/min.
Nitrogen 30 ml/min.
Temp: Initial 60 °C
Final 200 °C
Rate 7.5 °C/min.

The conditions of DT/ST system:

Type: Model 1000, Dynatherm Analytical Instruments Inc.,
Kelton, PA.
Dynamic thermal stripper:

Temperature: oven 80 °C
block 105 °C
sorption tubes 50 °C

Purging time: 46 min

Thermal desorption unit:

Temperature: transfer line 230 °C
desorber 340 °C
valve 230 °C

Desorption time: 8 min

Peak areas and retention time were recorded by the use of a computer
integrator (HP3395, Hewlett-Packard, Farmington, MI). The retention times
of vanillin and ortho-vanillin observed in the determination of the calibration
factor, by direct on column injection, were 14.73 and 12.88 minutes,
respectively. Retention times for both vanillin and ortho-vanillin, based on the
thermal stripper / thermal desorption (TS/TD) procedure, were 15.3 and 13.2
minutes, respectively. The data is shown in appendix A. Calibration factors
(C.F.) were determined from the slope of the linear calibration curve.The

I calibration curves are shown in Figure 19, 20, 21, and 22.

39

5.2.1.2 Initial vanillin and ortho-vanillin levels in corrugated board
The initial levels of vanillin and ortho-vanillin in corrugated board were
determined by the thermal stripper / thermal desorption method. In this
procedure, two small pieces of paperboard, about 3 x 60 mm in dimension, for
a total surface area of 3.6 cm2 were placed into the sparging tube (Supelco Co.,
Bellefonte, PA) of the thermal stripper instrument (Model 1000, Dynatherm
Analytical Instrument, Inc. Kelton, PA). The sparging tubes were then
mounted in the oven of the thermal stripper instrument (Model 890,
Dynatherm Analytical Instruments, Inc., Kelton, PA) and connected to the
sorption tubes positioned outside the oven. The sorption tubes, containing
Carbotrap 300 multi-bed materials (Supelco Co., Bellefonte, PA), were
covered with sleeve heaters. After sample preparation, the carbotrap sorption
tubes containing the trapped volatiles were transferred to the tube chamber of
the thermal desorption unit (Model 890, Dynatherm Analytical Instrument, Inc.
Kelton, PA). The sorbed volatiles were then desorbed by heating for 6 minutes
at 340 °C with the valve and transfer line heated to 230 °C, so as to maintain
the desorbed compounds in the vapor phase and allow their transfer to the gas
chromatograph. The operating conditions for the TS/TD procedure are the
same as the conditions described above for construction of the required
calibration curve.
5.2.1.3 Equilibrium partition between corrugated board and headspace
Two different techniques were employed to provide corrugated board
samples of known migrant concentration levels. For the ortho-vanillin studies,
samples were prepared with an ortho-vanillin generation system as shown in

. Figure 13. Here, an aliquot of 500 pl was withdrawn from the headspace of the

“(B

 

 

 

 

 

 

40

Glass tube
egulator Sample port

To hood

__iu__lL‘/

 

 

 

    
  

‘3

2.7:“: ..

      
      

refill

llr|
all.

    
   
 

:1

 

   
 

‘51

 

 
 

vél:

V w
A

 

~14-‘4
k65£

vs,

 

 

 

 

 

 

 

 

 

A

 

 

Figure 13: Ortho-vanillin generation system for preparation

of paperboard samples of known o-vanillin concentration

41

/,/Sample port

 

Exposure cell
'1" /

 

/Papcrboard

/

 

 

../

 

 

 

 

 

 

 

 

 

 

Vanillin wder
K P0

 

 

 

 

 

 

 

 

Figure 14: Direct exposure method for preparation of

paperboard samples of known vanillin concentration

42

stainless steel cell and injected directly into the gas chromatograph to
determine the ortho-vanillin vapor pressure, which was controlled through the
adjustment of the nitrogen gas flow rate. Corrugated board samples were
analyzed at predetermined time intervals by the thermal stripper-thermal
desorption method to provide corrugated board samples of known ortho-
vanilllin concentration levels. For vanillin migration studies, samples were
prepared by a direct exposure technique, as shown in Figure 14.

The procedure employed to determine the partition coefficient for the
corrugated board / headspace / vanillin and ortho-vanillin is as follows:

42 strips of paperboard samples, 1 x 6. 5 cm in dimension, for a total
surface area of 273.0 cm2 ,were cut from paperboard which had been exposed
to vanillin or ortho-vanillin vapor, to provide corrugated board samples of
known vanillin and ortho-vanillin concentration levels. The strips of
paperboard were placed into uncoated two-piece steel cans, 300 x 406 mm in
diameter and height, respectively (American National Can, La Porte, IN). Prior
to sealing, a small hole (1 mm, diameter) was punched in the can lid to provide
a sampling port, which was covered (interior surface) with Teflon pressure
sensitive tape (Supelco Co., Bellefronte, PA). A self sealing silicone rubber
septum was also affixed to the outer surface of the sampling port (Mocon,
Minneapolis, MN). The cans were seamed with a hand seamer (Model 23 H,
Dixie Canner Equipment Company, Athens, GA).

The sorption cells were then stored at 23, 38 and 50 °C respectively for
7 days to allow system equilibration and then sampled (Maekawa, 1994). The
equilibrium concentration of vanillin and ortho-vanillin in the headspace was

determined by removing a 500 111 sample from the headspace with a gas-tight

 

43

syringe (Hamilton Co.) through the sampling port and analyzing for vanillin
and o-vanillin levels by gas chromatography. The vanillin and ortho-vanillin
levels in the paperboard were determined by taking paperboard samples of
approximately 0.3 gram and analyzing for sorbate concentration levels by the

thermal stripper-thermal desorption method.

5.2.2 Part 2: D- “1111...! o -1: -1 -_ m o 1-1, 1 Ll. o 1 “-1.11'
W
5. 2. 2. 1 Determining calibration factors for the HPLC procedure
Approximately one gram of vanillin or ortho-vanillin was weighed,
transferred to a 100 ml volumetric flask and diluted to volume with
acetonitrile. Both vanillin and ortho-vanillin solutions were then diluted by a
serial dilution technique, to give standard solutions of 0.5, 1.0, and 2.0 ppm
(wt/volume), respectively. The standard solutions were filtered with a
Micromate syringe (Popper & Sons, Inc. New Hyed Park, NY) with a
Millipore Millex filter HV4, 0.45 pm (Waters Chromatography Division,
Marlborough, MA), then transferred to 4 ml screw cap clean glass vials with
a hole cap (Supelco, Inc, Bellefonte, PA) and a PFTE septum for the HPLC
automatic sampling injection system. The standard solutions were analyzed by
the HPLC analysis procedure. The mobile phase was prepared by a filter-degas
system as shown in Figure 15. The filter was an organic replacement filter, 0.5
pm, from Waters Chromatography Division. The mobile phase was 40%
methanol with 60% HPLC grade water (v/v percent).

44

The calibration curves were plotted as an average area respond for each

standard concentration vs. the quantity of vanillin or ortho-vanillin injected.

The calibration factors were then determined based on the slope of the curves.

The conditions for HPLC analysis was as follows:

Model:
Detector:

Column:
Mobile phase:
Temperature:
Flow rate:

Inj . Vol.:

WatersTM 150—C ALC/GPC

WatersTM Turnable Absorbance Detector
Wavelength 275 um

Delta-PakTM, C18 300A, 3.9 x 150 mm

40% Methanol & 60% HPLC grade water (v/v %)
23 °C

1 mI/min.

10 ul

45

  
  

Organic replacement filter

To vacuum system

 

 

iltercd solution

 

Figure 15: Vacuum filter system

46

5. 2. 2.2 Determining the stability of vanillin and ortho-vanillin solutions

Two sample solutions were evaluated in this study. For the first sample,
approximately 0.001 gram of vanillin was weighed, transferred to a 1000 ml
volumetric flask and brought to volume with fresh HPLC grade water. An
accurate initial concentration was determined by the HPLC analysis procedure.
A 35 m1 sample of the standard solution was then added to 40 ml screw cap
amber glass vials with teflon coated silicone septa (Supelco, Bellefonte, PA)
and the sample stored at ambient conditions (23 °C, 50% RH).

For the second solution, 0.001 gram of vanillin was weighed, transferred
to a 1000 ml volumetric flask and brought to volume with HPLC grade water
which was degassed. A vacuum degass system, Figure 16, was designed to
outgas the aqueous phase solution. Degassing was carried out for about 30
minutes. The solution was degassed again for about 20 minutes, and then 35
ml sample solution was transferred to 40 ml screw cap amber glass vials with
teflon coated silicone septa (Supelco, Bellefonte, PA).

The headspace of the vials were flushed with pure N2 gas to displace the
headspace air and the vials capped. All the glassware used in this process was
covered by aluminum film to protect vanillin from light because of its light
sensitive property. The vials containing sample solutions were stored in an
environmental chamber, Figure 17 (page 54), while pure N2 gas continuously
flowed through the chamber to maintain the samples under a nitrogen
atmosphere, and effectively excluded oxygen. The storage temperature was 23
°C. At predetermined time intervals, the sample vials stored at both ambient
condition and in the environmental chamber were assayed for vanillin levels.

Ortho-vanillin sample solutions of known concentration were also

47

prepared as described above, degassed and stored in the environmental

chamber, and assayed at predetermined time intervals as well.

 

48

rubber sealer

  
  

to vacuum system

sample solution

_ water

/

........... . ultrasonic cleaner
- / (vibration & wave)

 

 

 

 

 

Figure 16: Vacuum degas system

 

The procedure was designed to determine sorbate uptake by the LDPE,

as a function of storage time, to provide equilibrium as well as non-steady
state uptake values for both vanillin and o-vanillin. The level of uptake by the
LDPE was determined by the difference method (Kashtock, Giacin, and
Gilbert, 1980)

14 Round polymer disks, 23 mm diameter, for a total surface area of
approximately 8.3 cm2 were cut from LDPE film with a punch. The surface to
volume ratio is the same as found with a designed LDPE form-fill seal
ampoule. The films of approximately 5 mils thickness were made from the
LDPE ampoule resin by a compression molding procedure. The polymer disks
were weighted and threaded onto a stainless steel wire frame with 2 mm high
and 4 mm diameter glass rings used to separate the polymer disks. The wire
fi'ames, with the mounted disks, were then placed into 40 ml screw cap amber
glass vials (28 mm diameter, 98 mm high) with teflon coated silicone septa
(Supelco, Bellefonte, PA) and 35 ml of the respective sorbate solutions that
had been degassed were added to the vials. A schematic design of the test
system is shown in Figure 18. Concentrations of 0.520 ppm (wt/volume ) for
the aqueous solutions of ortho-vanillin, and 0.434 ppm (wt/volume) aqueous
solutions of vanillin were evaluated in the sorption studies. The sorbate
solutions were prepared by dissolution of the individual sorbates in HPLC
grade water that had been degassed.

The sorption vials were stored in the environmental chamber shown in

Figure 17 at temperatures of 23 and 32 °C, respectively, and sampled every

50

two days until equilibrium sorption conditions had been obtained. Aliquots of
the aqueous solution were removed from the sorption vials and the
concentration of vanillin and ortho-vanillin determined by HPLC analysis.
Control samples were also prepared and stored under the same
conditions. The control samples consisted of degassed vanillin or ortho-
vanillin standard solutions, stored in the 40 m1 screw cap amber glass vials in
the absence of polymer disks. These solutions were assayed at predetermined
time intervals to insure that there was no degradation or concentration change
in the sample solutions that was not attributed to sorbate uptake by the LDPE

film samples.

51

Nitrogen outlet

./ / Nitrogen
i/ , ; tank

 

 

 

 

/ Regulator

// vials

< I Nitrogen

cell

 

[:::n 1:”,
\ll

 

 

 

 

 

< Themal

\ } 0V6“

 

 

 

 

 

 

 

 

 

Figure 17: Enviromental controlled sample storage chamber

52

 

Figure 18: Sorption liquid test cell

RESULT AND DISCUSSIONS

6.1 The results of equilibrium partition distribution of vanillin and
ortho-vanillin between corrugated board and its headspace
6.1.] Initial vanillin and ortho-vanillin levels in corrugated board

The measurement of vanillin and ortho-vanillin levels in corrugated
board was carried out by the thermal stripper/thermal desorption technique. A
series of preliminary analyses with corrugated board samples were carried out
to determine the optimum temperature and purge time needed for the
quantitative desorption of residual vanillin and ortho-vanillin. Based on the
results of preliminary studies, an oven temperature of 80 °C and a purge time
of 30 minutes afforded conditions which allowed for the desorption of over
90% of the initial residual vanillin and ortho-vanillin in the paperboard. The
initial vanillin and ortho-vanillin concentrations (weight/weight) are
summarized in Table l. The following equation is used to calculate the
concentration:

A.U. x OF.
[C] = (14)
w

 

Where: A.U.: area response from chromatograph (A.U.)
C.F.: calibration factor (g/A.U.)
W: sample weight (g)

53

54

Table 1: Initial concentration of vanillin and ortho-vanillin in corrugated board

 

 

 

 

 

 

 

 

sample o-vanillin vanillin vanillin
(sample from stack (sample from stack (sample from stack
center) center) top)
ppm (wt/wt) ppm (wt/wt) ppm (wt/wt)

1 0.3000 1.0594 0.2838

2 0.1923 0.8587 0.1891

3 0.3131 0.7928 0.3101

4 0.3255 1.1376 0.2348
average 0.2829 0.9591 0.2545
std.dev $0.061 $0.163 $0.054

 

 

Measurements were carried out on four (4) replicate samples. The initial
concentration levels of vanillin and ortho-vanillin in the paperboard, which
was located in the middle of the board stack, were 0.96 ppm with a standard
deviation of $0.163 and 0.2829 ppm with a standard deviation of $0.061,
respectively. The vanillin concentration for the board at the stack top had a
lower value, which was 0.25 ppm with $0.054 as the standard deviation. The
reason for this finding may be that the vanillin molecules, which were in the
paperboard placed on the top of the stack, were volatilized into the surrounding
air during shipping and storage after production.

6.1.2 Vanillin and ortho-vanillin equilibrium partition distribution between
corrugated board and headspace

The partition coefficient KB,H is defined here as the equilibrium
concentration of vanillin or ortho-vanillin in the paperboard divided by the
equilibrium concentration in the headspace. Partition equilibrium experiments
were carried out at three temperatures, namely 23, 38, and 50 °C, to determine

. the influence of storage temperature on KB,H values.

55

To determine the time needed to attain a steady state concentration level
in the sorption cells, a 500 pl aliquot was withdrawn from the headspace of the
sorption cell following 2 days storage, and the sample injected directly into gas
chromatograph for quantitation. This process was repeated every two days
until the headspace concentration of vanillin and o-vanillin reached a steady
state level. A different sorption cell was sampled for both vanillin and ortho-
vanillin analysis. The results showed that the system had equilibrated within
7 days period (See Appendix C). To get a material balance, the levels of
vanillin and o-vanillin retained by the corrugated board were determined by the
thermal stripper/thermal desorption method. All analyses were carried out in
duplicate. The partition coefficient (KW) values were calculated by substitute
into Equation (15), and the results are summarized in Table 2. The test
conditions for the headspace studies for both vanillin and ortho-vanillin are

listed below.

The test conditions for the headspace studies of ortho-vanillin:

Initial ortho-vanillin concentration: 3.4 ppm (wt/wt)
The weight of paperboard which

was sealed in every sorption cell: 16.6 grams

The surface area of paperboard

sealed in every sorption cell: 270 cm2

Initial ortho-vanillin content

in paperboard: 56.44 ug

Time of storage: 7 days

56

The test conditions for headspace studies of vanillin:

Initial vanillin concentration: 3.17 ppm (wt/wt)
The weight of paperboard which

was sealed in every sorption: 16.6 grams

The surface area of paperboard

sealed in every sorption cell: 273 cm2

Initial vanillin content

in paperboard: 53 ug

Time of storage: 7 days

A series of sorption cells were prepared to follow the equilibrium partition
distribution, with each cell being sampled only one time. This was done to
insure no losses of volatiles through the sampling system. The relationship
between the partition coefficient KB,H and temperature is illustrated graphically
in Figures 19 and 20. As shown, the equilibrium partition coefficient KB/H
decreased with an increase in storage temperature. This is not unexpected, as
the solubility of vanillin and o-vanillin in the paperboard was found to

decrease with an increase in temperature.

 

[Ca]
KM, = (15)
[CH]
Where: [CB]: vanillin or o-vanillin concentration in corrugated
board at equilibrium (,ug/g).

[CH]: vanillin or o-vanillin concentration in headspace at
equilibrium (pg/g).

57

The partition coefficients (KBM) determined for vanillin and ortho-
vanillin were 179 and 663 for the corrugated board and headspace,
respectively, at 23 °C. The KB,H values were reduced to 48 and 223 for vanillin
and ortho-vanillin respectively, when the storage temperature was 50 °C. Thus,
at the higher storage temperature, higher levels of volatile are lost from the
paperboard. The storage temperature thus affected the partition coefficient
significantly in this study.

58

Table 2: Partition coefficient KB,H between headspace and paperboard

for vanillin and ortho-vanillin as a function of temperature.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vanillin
Vanillin in headspace Vanillin in paperboard
Temp. conc. content cone. cone. content conc. Km,
°C ppm ug avg & ppm #8 avg &
(wt/v) std. (wt/wt) std.
(us/CC) xl0'3 (fig/g)
x10‘3
23 16.22 7.18 16.69 3.10 51.70 2.99 179
17.16 7.60 “(1.0.67 2.88 48.03 std.0.16
38 25.79 11.42 24.31 2.56 42.75 2.39 98
22.83 10.12 “(12.09 2.22 37.07 “(1,024
50 42.46 18.81 38.90 1.77 29.59 1.86 48
35.34 15.66 std.5.04 1.95 32.57 “(1,0,1 3
ortho-vanillin
O-vanillin in headspace O-vanillin in paperboard
Temp. cone. content cone. conc. content conc. Km,
°C (wt/V) 11g avg & ppm 11g avg &
(pg/cc) std. (wt/wt) std.
Ppm xl0’3 (us/g)
x10‘3
23 4.62 2.05 5.01 3.23 53.58 3.32 663
6.39 2.83 “(1,125 3.41 56.64 “(10.13
38 7.89 3.50 8.88 2.90 48.14 3.14 353
9.88 4.37 std. 1 .39 3.38 56.06 std.0.33
7
50 14.86 6.58 12.55 2.83 46.91 2.81 223
10.24 4.54 std.3.27 2.78 46.21 std.0.03

 

 

 

 

 

 

 

 

 

59

 

 

 

 

180
<
160 _
; I40 _
:2
U
as [20 1
C
.9
,2 100 .
E
a.
80
60
4o 1 l l l 1
23 28 33 38 43 48 53
Temperature °C

Figure 19: Vanillin partition coefficient K3,], between headspace and paperboard

as a function of temperature.

6O

 

Partition coefficient Knm

 

 

 

 

53

Temperatm'e °C

Figure 20: Ortho-vanillin partition coefficient KB,” between headspace
and paperboard as a function of temperature

61

6.2 Stability of vanillin and ortho-vanillin in an aqueous phase

To determine the stability of vanillin and ortho-vanillin in an aqueous
phase, standard solutions of known concentration were prepared and
transferred into 40 ml screw cap amber glass vials. One set of samples had
been degassed prior to filling the test solution and was stored in a nitrogen
flushed chamber maintained at ambient temperature (See Figure 17). The other
set of samples was simply prepared with non-degassed HPLC grade water and
stored at ambient conditions. The aqueous samples were analyzed every 3 days
by the HPLC process described in the previous section. A standard solution
made of vanillin in acetonitrile was used to monitor the stability of the HPLC
procedure. The data for this study is summarized in Table 3, and presented
graphically in Figure 21, where the concentration of vanillin and o-vanillin in
the respective solutions is plotted as a function of storage time.

As shown in Table 3, the concentration of vanillin in the samples simply
prepared with non-de gassed HPLC grade water and stored at ambient condition
degraded significantly over time and after 9 days storage, the concentration of
vanillin tended to zero. This concentration change was found to be the result
of vanillin oxidation to yield a more polar product, as evidenced by the
detection of an additional peak in the chromatograph prior to the retention

time of vanillin.

62

Table 3: Storage stability of vanillin and ortho-vanillin in aqueous solutions (23 °C)

 

vanillin concentration change (ppm wt/v)

 

 

 

 

 

 

 

 

 

 

 

 

storage time 0 days 3 days 6 days 9 days 12 days
non-degassed 1.01 0.95 0.42 0.00 0.00
degassed 1.07 1.07 1.09 -- 1.08
ortho-vanillin concentration change (ppm wt/v)
storage time 0 days 5 days 10 days
degassed 0.46 0.46 0.46

 

 

 

 

 

Figure 22 shows that vanillin has a retention time of about 2.46 minutes
under the assay conditions employed. Further, the peak at 2.46 minutes
decreased gradually as a fimction of storage time, while at the same time a new
peak (retention time of 1.00 min.) appeared in the chromatograph following
3 days storage and this peak at 1 minute continued to increase as a function of
storage time. Comparing the retention time of the newly developing peak with
that of vanillic acid, showed similar retention time values, indicating that the
vanillin was undergoing oxidation under the storage condition selected.
Perestelo (1989) reported that vanillin undergoes oxidation to vanillic acid.
Therefore, the newly developing peak was assumed to be vanillic acid, the
oxidation product of vanillin, (See Figure 23).

From the ortho-vanillin storage stability studies, a comparison of the
respective chromatograms showed similar results, as evidenced by the
development of a new peak in the chromatograph at a retention time of 0.86
minutes, as compared to a retention time of 3 minutes for the ortho-vanillin
sample. This is illustrated in Figure 24, where the chromatograms obtained at
time equal 0 and 9 days storage are presented.

For both vanillin and ortho-vanillin solutions prepared by the degassing

63

procedure and stored in the nitrogen environmental chamber, no evidence of
oxidation was observed over a 12 day storage period. This indicated that if
properly degassed of oxygen, solutions of vanillin and o-vanillin were quite
stable. Therefore, all sorption studies carried out to determine the partition
distribution of vanillin and o-vanillin between an aqueous phase and LDPE,
the solutions were degassed, as described in the Materials and Methods
Section, to insure that measured changes in sorbate concentration were due to

their uptake by the polymer phase, and not the result of oxidation.

64

 

 

   
  
  

 

E + Degassed vanillin solution
—0— Non-degassed vanillin solution
; + Stamhrd solution

i + Degassed O-vanillin solution

-J------

 

 

 

 

Concentration levels (ppm wt/v)

 

 

 

 

 

 

 

 

 

 

Storage time (days)

Figure 21: Storage stability of vanillin and ortho-vanillin in aqueous phase (23 0C)

 

 

65

 

 

 

 

 

 

INJECT Initial vanillin peak
Initial peak E “2.53 .
‘The ”a". ‘3’ ”9‘” PM?“ Vanillin peak
from vanillin degradation height reduced
fi=:=:=——— 2’ 47
3 days (2.

‘ New peak
K Vanillin peak

hei ht reduc d
g 0.92 2 e ‘7
6 days " 2.38

 

New peak increased
*4 ' 0.99

 

9 days _
K Vanillin peak
disappeared

Figure 22

Process of vanillin degradition shown by the chromatography peak change

66

INJECT
The peak of vanillin acid

1.00

 

a“:

 

The peak of new product

[from o-vanillin degradation
? 00

Figure 23 . Both vanillin acid peak and the peak from vanillin

degradition have the similar retention time.

 

INJECT

I

 

.
q

l
l

 

LA‘A‘A h“—
“flv

_‘

 

67

The peak of new product
from o-vaniIlin degradation

. 0.8 6'
9 days \lr— '

o-vanillin peak
disappeared

—-- Initial o-vanillin peak

l
Initial peak <15:—
11

Figure 24 The oxidition of ortho-vanillin yielded

a more polar product which has the retention time of 0.86 minute

68

6.3 The uptake of sorbates by the LDPE as a function of storage time
6. 3. 1 Determination of partition coeflicient values (K m) and data analysis

In these studies, the uptake of vanillin and o-vanillin from an aqueous
phase was determined as a function of storage time to provide equilibrium as
well as non-steady state uptake values for both vanillin and ortho-vanillin.
Partition distribution studies were carried out at 23 and 32 °C, respectively.
There were two replicate samples in these studies and every sample was
assayed three times by the HPLC procedure. The samples were analyzed every
two days over a 14 days period to determine the non-steady state uptake
values, as well as the equilibrium state values. Individual vials were sampled
just one time. For one concentration of vanillin or o-vanillin at one storage
condition, 42 data points were available for their analysis.

The results of the partition distribution studies for vanillin and ortho-
vanillin at 23°C and 32 °C are summarized in Tables 4, 5, 6, and 7,
respectively. For better illustration, the results are also presented graphically
in Figure 25, 26, 27, and 28, where the sample concentration change in both
polymer and aqueous phases are plotted as a function of storage time. The
results of the control samples and the vanillin- and o-vanillin solutions
(acetonitrile standards) were also plotted as a reference to monitor the stability
of the aqueous solutions of vanillin and o-vanillin as a function of time and
temperature, and the consistory of the HPLC procedure. Here, the partition
coefficient values (KP/A) are defined as the concentration of vanillin or ortho-
vanillin in the LDPE phase divided by their aqueous phase concentration levels
(equation 12).

The results of the sorption studies are presented graphically in Figure 29,

69

where the concentration ratio for both vanillin and ortho-vanillin at the two
temperatures of test (i.e. 23°C and 32 °C) were plotted as a function of storage
time to equilibrium. The initial concentrations of vanillin and ortho-vanillin in
the aqueous phases, which were 0.434 ppm (wt/wt) and 0.520 ppm (wt/wt)
respectively, decreased while their concentration in the polymer phase
increased until equilibrium concentration levels were attained. These
equilibrium concentration levels were then used to determine the respective
Km values. The storage times needed to reach steady state levels for vanillin
in this polymer / aqueous phase system was four (4) days at 32°C and six (6)
days at 23°C. For ortho-vanillin, the system attained equilibrium after six (6)
days storage at 23°C and eight (8) days storage at 32°C.

70

Table 4: The data of partition distribution studies of vanillin at 23°C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

storage time weight of polymer avg. conc. avg. conc. Rm“ avg. R m
(days) (gram) in aqueous in polymer and Std-
(ppm w/w) (ppm w/w)
2 0.73 0.42 0.93 2.24 2.8
0.68 0.41 1.34 3.27 std. 0.5
4 0.73 0.29 6.80 23.23 23.9
0.59 0.31 7.55 24.63 std. 0.7
6 0.71 0.26 8.17 33.78 34.6
0.58 0.27 9.68 34.47 std. 0.8
8 0.64 0.26 9.70 37.65 35.0
0.63 0.27 8.87 32.35 std. 2.7
10 0.66 0.28 8.15 29.07 31.0
0.60 0.28 9.16 33.00 std. 2.0
12 0.56 0.28 9.80 35.41 32.7
0.63 0.28 8.43 29.90 std. 2.8
14 0.65 0.27 8.17 31.94 30.3
0.69 0.28 7.94 28.60 std. 1.7
Rm: Concentration Ratio, defined as sorbate concentration in the polymer

phase divided by the aqueous phase concentration at time t.

 

71

Table 5: The data of partition distribution studies of vanillin at 32°C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

storage time weight of polymer avg. cone. avg. cone. Rm“ avg. R m
(days) (gram) in aqueous in polymer and Std.
(ppm w/w) (ppm w/w)
2 0.54 0.42 0.94 2.24 2.9
0.63 0.41 1.46 3.58 std. 0.7
4 0.64 0.33 5.74 17.37 16.8
0.66 0.33 5.37 16.13 std. 0.6
6 0.67 0.32 5.76 17.80 17.8
0.69 0.32 5.72 17.78 std. 0.0
8 0.66 0.31 6.46 20.62 19.4
0.65 0.32 5.92 18.20 std. 1.2
10 0.64 0.32 5.98 18.42 17.9
0.70 0.32 5.61 17.41 std. 0.5
12 0.68 0.32 5.97 18.71 20.0
0.60 0.32 6.78 21.30 std. 1.3
14 0.67 0.33 5.70 17.50 18.43
0.64 0.32 6.22 19.35 std. 0.9
Rm: Concentration Ratio, defined as sorbate concentration in the polymer

phase divided by the aqueous phase concentration at time t.

 

72

Table 6: The data of partition distribution studies of ortho-vanillin at 23°C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

storage time weight of polymer avg. conc. avg. conc. Rm“ avg. R m
(days) (gram) in aqueous in polymer and Std-
(ppm w/w) (ppm w/w)
2 0.75 0.47 2.48 5.30 4.8
0.70 0.48 2.09 4.36 std. 0.5
4 0.66 0.39 7.18 18.63 19.2
0.58 0.39 7.76 19.77 std. 0.6
6 0.67 0.35 8.76 24.83 25.1
0.67 0.35 8.88 25.37 std. 0.3
8 0.61 0.28 13.55 47.89 47.1
0.58 0.29 13.59 46.21 std. 0.8
10 0.74 0.25 13.03 52.83 51.1
0.61 0.28 13.86 49.30 std. 2.0
12 0.64 0.27 14.00 52.69 50.1
0.56 0.30 13.99 47.43 std. 2.7
14 0.61 0.28 13.53 47.60 45.8
0.63 0.29 12.73 43.91 std. 1.8

 

 

 

 

 

 

Rm: Concentration Ratio, defined as sorbate concentration in the polymer
phase divided by the aqueous phase concentration at time t.

 

73

Table 7: The data of partition distribution studies of ortho-vanillin at 32°C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

storage time weight of polymer avg. cone. avg. cone. Rm" avg. R m
(days) (gram) in aqueous in polymer and Std-
(ppm w/w) (ppm w/w)
2 0.67 0.49 1.55 3.15 3.1
0.62 0.49 1.51 3.07 std. 0.1
4 0.67 0.42 5.44 13.03 11.1
0.58 0.45 4.15 9.20 std. 0.7
6 0.62 0.36 9.30 26.06 26.3
0.56 0.37 9.72 26.55 std. 0.3
8 0.60 0.35 9.79 27.68 26.0
0.69 0.35 8.56 24.30 std. 1.7
10 0.60 0.35 9.68 27.35 26.7
0.68 0.35 8.93 25.82 std. 0.8
12 0.70 0.33 9.55 29.05 31.8
0.59 0.33 11.40 34.62 std. 2.8
14 0.60 0.34 10.38 30.27 29.0
0.62 0.35 9.70 27.76 std. 1.3
Rm: Concentration Ratio, defined as sorbate concentration in the polymer

phase divided by the aqueous phase concentration at time t.

 

74

 

 

 

 

 

 

 

 

15 0.6
12 5 x __x.___ x 35 I 0 5
i A A
A — r
E I0 l n 0 4 3‘
E F l I 3
a s
8 1?
I: 7.5 0.3 3
~ i
'3 v e e e E
E §
§ 5
= +Standard solution (right scale) 0
(3 5 _ 4 0.2
+ Control sample solution (right scale)
+ Vanillin cone. in aqueous phase (riy'lt scale)
2.5 _ _._ Vanillin cone. in polymer (left scale) l 0.I
0 l 1 l l l J 0
0 2 4 6 8 10 12 14
Storage time (days)

Figure 25: The sample concentration change in both polymer and aqueous phase as a
function of time, including control sample and standard solution.
vanillin at 23°C

75

 

 

 

 

   
 

 

 

 

IS , 0.6
12 5h X __u__i x g I 0 5
4 s e
A 10 _ 0.4 A
E ‘ g
3 s.
g z?
- 1 .9.
a 7.5 _ v v v A_1, 0.3 ..
.2 5
8
g I I 5
g 5 - 0.2
+ Standard solution (rim scale)
+ Control sample solution (right scale)
2'5 + Vanillin cone. in aqueous phase (right scale) “ 0']
+ Vanillin cone. in polymer (left scale)
0 l l l l l 0
0 2 4 6 8 10 12 14

Storage time (days)

Figure 26: The sample concentration change in both polymer and aqueous phase as a
function of time, including control sample and standard solution.
vanillin at 32°C

76

 

 

 

 

  

 

 

 

15 0.6
n r l _
12.5% + g u o s
A
E
E ‘0 ~ _ 0.4 ’3;
E
g 5“
5 7 s 1 03
E ' . . - . i
I:
I: 8
3 a
g L)
Q 5 _ _ 0.2
+ Standard solution (right scale)
.5. Control sample solution (right scale)
2-5 .9. Ortho-vanillin cone. in aqueous phase (right scale) 0.I
+ Ortho-vanillin cone. in polymer (Iefl scale)
0 l l l l 1 0
0 2 4 6 8 IO 12 14

Storage time (days)

Figure 27: The sample concentration change in both polymer and aqueous phase as a
function of time, including control sample and standard solution.
ortho-vanillin at 23 °C

77

 

 

 

 

  

12.5 A x- g 0.5
"3
E 10 0.4 9
I
s ' 3
e 9 L E.
s 0 £3”
‘5 7.5 _ . 0.3 .2
g E
C
3 §
8 8
s l 0.2
+ Standard solution (right scale)

4. Control sample solution (right scale)
2.5 .4. Ortho-vanillin cone. in aqueous phase (right scale) 0.I

+ Ortho-vanillin cone. in polymer (left scale)

 

 

 

0 1 l 1 1 0
0 2 4 6 8 10 12 14
Storage time (days)

Figure 28: The sample concentration change in both polymer and aqueous phase as a
fimction of time, including control sample and standard solution.
ortho-vanillin at 32 °C

Concentration Ratio

78

60

 

so_ ‘ ,

o-vanillin at 23°C A

 

40 _
vanillin at .23“C

o x
30 o-vanillin at 32 °

 

 

 

 

 

20 {anillin at 32 °C I
I
10
0 l l l
0 2 4 6 8 10 12 14
Storage Time (days)

Figure 29: The concentration ratio for both vanillin

and ortho-vanillin at different temperature vs. storage time.

79

6. 3.2 Estimation of difi‘usion coefiicient (D) of vanillin and ortho-vanillin in

low density polyethylene

For the aqueous phase/LDPE system, sorbate equilibrium is shown
experimentally when the liquid phase sorbate concentration remains constant
after a given contacting time, such as the case for determination of the
equilibrium partition coefficient values (KM, ) in the present studies. In
Figures 30 to 33, the fiaction or percent of equilibrium reached is plotted as a
fimction of contact time and indicates that the phase equilibrium was reached
after 6 - 8 days, as shown by the vanillin and ortho-vanillin concentration
levels in the aqueous phase remaining constant after this contact period (See
Table 8).

In most cases, the diffusion coefficient of a sorbate in the polymer
determines the kinetics of the process and , therefore, the time to reach

equilibrium. According to Crank (1975), the dynamics of the sorption process

 

 

can be described by:
Mt no 20: (a + l) -Dq,,2 t
— = 2 exp. ( ) (14)
M, 1 + a + (aqn)2 12
where, Mt/M,: the fraction of equilibrium reached at time t
a: the ratio of volumes of polymer film and
solution divided by partition coefficient

D: the diffusion coefficient of sorbate in LDPE
l: the half thickness of the polymer film

qn: non-zero positive roots of tan q, =-aq,,

80

In this study, the fraction of equilibrium reached at time t was
determined from the experimental data. The results are summarized in Table
8 and presented graphically in Figure 30 to 33. The time needed to reach a
certain value of the fiaction of equilibrium was obtained directly fi'om Figures
30 to 33, for vanillin and o-vanillin in the aqueous phase/LDPE system at 23
and 32°C storage conditions, respectively. Diffusion coefficient values for
vanillin and o-vanillin in LDPE were calculated for equilibrium fraction values
of 0.6, 0.7, and 0.8 respectively by solution of Equation 14. The results are
presented in Table 9.

The values in Table 9 were calculated from Equation 14 with enough
terms to produce an accuracy of > 99% using values of KM, and Mt/Moo
determined in the present study. The average diffusion coefficients values (D)
determined at 60%, 70%, and 80% of equilibrium for vanillin at 23°C and 32°C
were 4.4 x 10'16 mz/sec, 5.4 x 10'16 m2/sec, respectively, while the average
diffusion coefficient values for o-vanillin were 2.5 x 10‘16 m2/sec and 2.9 x
10'16 mz/sec, respectively. Both vanillin and ortho-vanillin had lower diffusion
coefficients values at 23°C storage temperature, as compared with the
diffusion coefficient values at 32°C storage temperature. The effect of
temperature is not unexpected, since temperature plays an important role in
the equilibrium partition process. The higher the temperature, the larger the
expected diffidsion coefficient (Crank, 1975).

81

Table 8: MW» vs storage time of vanillin and ortho-vanillin

 

 

 

 

 

 

 

 

 

storage time Mt/Moo of Mt/Moo of Mt/Moo of Mt/Moo of
(days) vanillin vanillin ortho-vanillin ortho-vanillin
at 23°C at 32°C at 23°C at 32°C
2 0.129 0.197 0.169 0.157
4 0.817 0.914 0.552 0.492
6 1.017 0.944 0.651 0.975
8 1.057 1.018 1.002 0.941
10 0.986 0.953 0.993 ‘ 0.954
12 1.038 1.049 1.034 1.074
14 0.917 0.980 0.969 1.030

 

 

 

 

 

Table 9: The difiusion coefficient (D) values determined to reach
Mt/Moo= 0.6, Mt/Moo=0.7, and Mt/Moo=0.8, according to Eq.14

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vanillin Ortho-vanillin
Temp. Mt/Moo time Diffusion Mt/Moo time Diffusion
°C (day) coef. (day) coef.
(mz/s) (mZ/s)
0.6 3.2 7.1 x 10'16 0.6 4.5 4.5 x 10'16
23
0.7 3.3 4.0 x 10‘16 0.7 5.1 2.1 x 10‘16
0.8 3.4 2.0 x 10‘” 0.8 5.7 1.0 x 10''6
avg. 4.4x 10'“ avg. 2.5x [0'16
std. 2.1 x 10'16 std. 1.5 x 10'16
0.6 3.1 8.6 x 10''6 0.6 4.2 4.6 x 10"6
32
0.7 3.2 5.1 x 10''6 0.7 4.6 2.9 x 10‘16
0.8 3.3 2.6 x 10''6 0.8 4.8 1.2 x 10"6
avg. 5.4x 10'" avg. 2.9x 10'"
std. 2.5 x 10‘16 std. 1.4 x 10'”

 

 

 

 

Mt/Moo

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

82

 

 

 

 

 

 

- I I
_I
n
Vanillin at 23 °C
6 8 10 12 14
Storage time (day)

Figure 30: The Mt/Moo vs. time for sorption of vanillin by LDPE at 23°C

83

 

0.9 _

0.8 _
0.7 p
0.6 Vanillinat32°C
0.5 d
0.4 L
0.3 __

0.2 _

0.1 _

 

 

O 1 l 1 l m l

2 4 6 8 10 12

 

Storage time (day)

Figure 31: The Mt/Moo vs. time for sorption of vanillin by LDPE at 32°C

14

84

 

0.9 u

0.8 u

0.7 a

   

0.6 h Ortho-vanillin at 23°C

Mt/Moo

0.5 _

0.4 a

0.3 1

0.2 _

0.1 a

 

 

l L 1 l

2 4 6 8 10 12 14

 

Storage time (day)

Figure 32: The Mt/Moo vs. time for sorption of ortho-vanillin by LDPE at 23°C

85

0.9

0.8

0.7

0.6 Ortho-vanillin at 32°C

Mt/Moc

0.5

0.4

0.3

0.2

0.1

 

 

 

1 l l

2 4 6 8 10 12 14

 

Storage time (day)

Figure 33: The Mt/Moo vs. time for sorption of ortho-vanillin by LDPE at 32°C

86
6. 3.3 The solubility parameter (b) of vanillin and ortho-vanillin in LDPE
Hoy’s method was employed to determine solubility parameter values,
6, in this study. The equations to be used in Hoy’s system for estimation of the

solubility parameter and its components are given as follows (van Krevelen,
1990)

For low-molecular Liquids (Solvents):

Fd+B

 

 

 

o, = V (B= 277) (15)
6 6( 1 F” )V’ (16)
p “ Fd+B
-1 1/2
a. = 6.( ) (17)
a
6 = (od2+op2+o,2)% (18)

log a = 3.39 (0.567 + AT - ATZ) - 0.1585 - log V (19)

For amorphous Polymers:

 

Fd+ B/fi
o, = — (B= 277) (20)
V
o o ( 1 F" )V’ (21)
p d at 0’) Fd +B/fi
am-l %
6h = 51( ) (22)

 

87

a = (6.2 + 5,2 + 6.21/2 (23)
at (P) = 777 up» / v (24)
fi = 0.5 / 11,0» (25)

F d is the molar attraction function, F p is its polar component: V is the

molar volume of the solvent molecule or the structural unit of the polymer. or
(P) can be derived for A, , which is the Lyderson correction for non-ideality,

used in the auxiliary equations. a is the molecular aggregation number and fl

is the number of repeating units per effective chain segment of the polymer. B
is a constant base value. All the values of F, , Fp , V, A, , and A10” for the
monomer unit of a group of polymers or solvents were given by van Krevelen
(1990). A computer program was created by F ayoux (1997) to calculate the
solubility parameters for organic volatiles and polymers. The program was
employed to calculate the solubility parameters for water, LDPE, and the
sorbates vanillin and o-vanillin, which were evaluated in the present study. The

results are summarized in Table 10:

88

Table 10: The solubility parameter (J/cm3)'/a of water,
LDPE, vanillin, and o-vanillin

 

 

 

 

 

 

water LDPE vanillin o-vanillinWl
o, 48.1 17.57 16.44 15.94%
6, 14.3 0 13.68 13.66
6,, 31.3 0 15.43 15.14
6 34.4 17.57 26.38 25.88

 

 

 

 

 

 

The solubility parameters (6) for vanillin and o-vanillin are 26.38 and
25.88 (J/cm3)"’, respectively. Water and LDPE have solubility parameter values
of 34.34 (J/cm3)'/’ and 17.57 (J/cm3)'/’, respectively. According to solubility
theory developed by Hildenbrand (Van Krevelen, 1990), the enthalpy of

mixing is given by:

Where:

Equation (26) predicts that A Hm = 0, if 6,: 62 , so that two substances
with equivalent solubility parameters should be mutually soluble or, substances
with similar chemical structures should be quite likely to dissolve in each other
(Van Krevelen, 1990). Equation 27 provides an estimation of the solubility or

compatibility of a sorbate molecules in a polymer, when dispersion, polar, and

A Hm: ¢1¢2(91'92)

A Hm:
4’1 and 8’23
61 and 62:

(26)

enthalpy of mixing per unit volume
volume fraction of components

solubility parameters of the components

hydrogen bonding contributions are considered (van Krevelen, 1990).

89

A6 = [(6dp ' 6ds)2 +(6pp ' 6ps)2 + (611p ' 511.3)2 1% (27)

For high compatibility or solubility of a sorbate molecule in a polymer
matrix, 116 should be less than 5. The 186 values for the respective
LDPE/vanillin or ortho-vanillin in water systems were calculated by
substitution of the parameters summarized in Table 10 into Equation 27, and
the results summarized in Table 11. As shown, the 86 value between LDPE
and vanillin or ortho-vanillin is about 20 (J/cm3)'/’, and between water and

vanillin or ortho-vanillin is 35 - 36 (J/cm3)'/’.

Table 11: The solubility parameter (J/cm’)” difference
between water, LDPE, vanillin, and o-vanillin

 

water/vanillin water/o-vanillin LDPE/vanillin LDPE/o-vanillin l

 

 

 

 

 

 

A5 35 36 20 20 I

 

Since the 116 for the respective sorbate/polymer, sorbate/aqueous phase
system are greater than 5, it is difficult to draw a conclusion regarding the
relative affinity or partitive distribution of vanillin and o-vanillin between
LDPE and aqueous phase, respectively.

6.3.4 Modeling the equilibrium distribution of vanillin or ortho-vanillin in a
paperboard/headspace/LDPE/an aqueous phase system

Paperboard may contain organic molecules, such as vanillin and ortho-
vanillin volatiles, ink constituents from printing, and other residual organic
compounds from the production process, such as the toxic dioxin (Foulke,

1990). Figure 34 shows a typical storage and transportation environment for

90

a shipper/polymer packaging system. Based on the following assumptions: (1)
the aqueous phase does not react with the volatiles, (2) the initial concentration
of the volatile in the system is initially present only in the paperboard, and (3)
volatilization of vanillin or ortho-vanillin from corrugated board to headspace
is inward, expressions to allow estimation of the equilibrium concentration of
volatiles in the paperboard, shipper headspace, polyethylene ampule, as well
as the aqueous phase were derived by Maekawa (1994) according to Cramer’s

rule (Kreyszig, 1979). The derived expressions are summarized below:

 

 

 

 

 

 

 

 

W13 ° C113
on = (28)
vs s-p-vs s-p-vs
wB + + +
KB/H K3711 KB/H ° Km
We ° CiB
CeB = (29)
s - 13 -v,.
WB'KB/H+VH+S °B'VL+
Km
W13 ° CiB
CeB = (30)
W13 ' K3711 Vn VA
VL + + +
S ° [3 S ' I3 KP/A
We ° C113
CeB = (3 0)

 

KB/H ° KP/A 'WB KP/A'VL
VA + Vt. ' KUA + +
S ' l3 S ' l3

 

where:

C113:

CeB, CeH, CeL, CeA:

We, V11, VL, VA:

KB/ll:

KP/A I

91

 

initial concentration of volatile in the shipper
paperboard.

equilibrium concentration of volatile in the
paperboard (ug/ g), in the shipper headspace (,ug/cc),
in the LDPE container (pg/g), and in the aqueous
phase (pg/cc), respectively.

weight or volume of the paperboard, of the shipper
headspace, of the LDPE container, and of the
aqueous phase, respectively (g, cc).

solubility coefficient between sorbate vapor and
LDPE.

partition coefficient between paperboard and
headspace.

partition coefficient between LDPE and aqueous

phase.

In this study, values for C113, Ksm, Km, for vanillin and ortho-vanillin

were determined as a function of temperature. Maekawa (1994) reported the
solubility coefficient of vanillin in LDPE was 0.3 (kg/m3Pa) at 23 °C and 0.23
(kg/m3Pa) at 32 °C, and for ortho-vanillin, the solubility coefficient was 11.8

'(kg/m3Pa) at 23 °C, respectively. The parameters for vanillin and ortho-vanillin ,

92

in the present study are as follows:

vanillin Clo:
Kit/a:
Km:
o-vanillin Cts:
KB/HI
Km:

For a designed packaging system shown in Figure 34, that was provided
by Dey laboratories, Inc (Napa, CA) with weight of the paperboard, We =
90.3 g, a volume of the headspace Vn= 1797 cc. In this case, the corrugated
shipper was unitized to 120, 5 cc LDPE ampules to give a total volume of
LDPE VL = 200 cc, and aqueous phase VA = 600 cc, respectively. The
equilibrium concentrations of vanillin and ortho-vanillin in the packaging
system was determined by substitution of the parameters above into Equations
28 to 30. The results are summarized in Table 12. For vanillin in a
paperboard/LDPE/aqueous phase packaging system, the higher the storage

temperature, the greater the possibility of organic molecules migrating from

0.96 ppm (wt/wt)

179 at 23 °C and 125 at 32 °C
33 at 23 °C and 19 at 32 °C
0.28 ppm (wt/wt)

663 at 23 °C and 450 at 32 0C
48 at 23 °C and 28 at 32 °C

the paperboard to an aqueous phase product.

93

Table 12: Prediction of equilibrium concentration in
the paperboard and in an aqueous phase.

 

 

 

 

 

 

 

 

 

vanillin 23°C, vanillin 32°C o-vanillin 23°C
C83 0.82 (pg/g) 0.77 (pg/g) 0.17 (pg/g)
CeH 4.56 X 10'3(ug/cc) 6.24 x 10'3(ug/cc) 2.25 x 10403gkc)
CeL 2.20 x 10'2(ug/g) 2.40 x lo'zmg/g) 4.59 x 10'2(ug/g)
CeA 6.70 x lO‘Oxg/cc) 14.33 x lO‘Qg/cc) 1.42 x 10’3(pg/cc)

 

 

* Corrugated board/Headspace/LDPE packaging system
WWW...

Corrugated board ”mm
odier

 

 

 

mplecnlesflum
toxic dioxin

/Headspace

Polymeric structure

An ueous hase
/ 3‘1 p

_
‘ Storage stacks

{/5 Packaging system

   

// Pallet

AAAAAAAAAAAAAAAAAAAAAAAAAA

Figure 34: Typical storage and transportation environment of a shipper/polymer

packaging system.

SUMMARY AND CONCLUSIONS

The partition coefficients, K values, of vanillin and ortho-vanillin
between corrugated board and headspace, and between low density
polyethylene (LDPE) and aqueous phase were investigated. The diffusion
coefficients (D) and solubility parameter of vanillin and ortho-vanillin in
LDPE were studied as well.

1) The partition coefficient, K values, for vanillin and o-vanillin
between corrugated board and headspace study was carried out with sorption
cells and the concentrations of vanillin or ortho-vanillin in headspace and in
corrugated board were determined by a gas chromatograph and thermal
stripper/thermal desorption procedure. The pa1tition coefficients, K values, of
vanillin were 179 (ug/ g / ug/v) at 23°C, 98 (ug/ g / ug/v) at 38°C, and 48
(ug/ g / ug/v) at 50°C, respectively. The K values of ortho-vanillin were 663
(pg/g / ug/v) at 23°C, 353 (pg/g / ug/v) at 38°C, and 223 (pg/g / ug/v) at
50°C, respectively. The K value that defined as the equilibrium concentration
in the paperboard over the equilibrium concentration in the headspace
decreaseed for both vanillin and ortho-vanillin while the storage temperature
increased. The higher temperature the storage condition was, the more volatile
the paperboard lost which means that the more volatile went into headspace.

2) The partition coefficients, K values, between low density
polyethylene and contacting aqueous phase were determined by a HPLC

95

96

procedure. Data showed that the steady state could be reached after 6 days
storage time for vanillin and 8 days storage time for ortho-vanillin in the
designed sorption cells. The storage temperature affected the result
significantly during the absorption process. The K values were approximately
twice as bigger at 23°C compared with 32°C for both vanillin and ortho-
vanillin, which were 32.72 (ug/ g / ug/v) at 23°C compared with 18.71 at
32°C (ug/ g / ug/v) for vanillin and 48.49 (ug/ g / ug/v) at 23°C compared
with 28.36 at 32°C (ug/ g / ug/v) for ortho-vanillin.

3) The diffusion coefficients (D) of vanillin and ortho-vanillin in
LDPE were determined according to Crank’s method (1976). They are 2.0 x
10'13 cmZ/sec at 23°C and 3.9 x 10‘13 cm2/sec at 32°C for vanillin; and 6.0 x 10'
‘4 cmZ/sec at 23°C and 1.5 x 10‘13 cmZ/sec at 32°C for ortho-vanillin. The
diffusion coefficients tend to be bigger when the temperature increase.
Therefore, the time needed to reach the equilibrium state during the experiment
was shorter if the samples stored in the higher temperature condition. The
storage times needed to reach equilibrium state for vanillin in this polymer /
aqueous phase system was four (4) days at 32°C and six (6) days at 23°C. For
ortho-vanillin, six (6) days needed at 23°C and eight (8) days needed at 32°C.

4) Hoy’s method (1976) was employed to determine the solubility
parameter, 6, in this study. The results show that water has the solubility
parameter of 34.4 (J/cm3)"2, LDPE 17.57 (J/cm3)”2 , vanillin 26.38 (J/cm3)“2
, and ortho-vanillin 25.88 (J/cm3)"2. According to solubility theory developed
by Hildenbrand, two substances with equivalent solubility parameters should
by mutually soluable or, be quite likely to dissolve in each other. The solubility

, parameter difference between LDPE and vanillin is 8.81 (J/cm3)”2 , between

97

LDPE and ortho—vanillin is 8.31 (J/cm3)”2. Therefore, the LDPE and ortho-
vanillin system is more equivalent than that of LDPE and vanillin system. That
explained that the K values of ortho—vanillin is larger than that of vanillin at the

same storage temperature.

APPENDICES

APPENDIX A

APPENDIX A

Calibration factors for gas chromatography

Gas chromatography analysis required construction of two kinds of
calibration curves based on sample handling, standard calibration curves and
thermal stripper/thermal desorption ( TS/TD) procedure calibration curves. The
vanillin and ortho-vanillin standard solutions were made by a serial dilution
technique to give the known concentration in acetonitrile. The condition of gas
chromatography was as follows:

The condition of gas chromatography:

Type: I-IP 5890A, Hewlett Packard, Avondale, PA.
Column: Fused silica capillary column (30m x 0.32 mm ID)
intermediate polarity stationary phase SPB-S (Supelco
Co., Bellefronte, PA)
Injector: 5 pl liquid syringe, 500 pl gas tight syringe (Hamilton)

Flow rate: Helium 7.0 ml/min.
Air 400 ml/min.
Hydrogen 40 ml/min.
Nitrogen 30 ml/min.
Temp: Initial 60 °C
Final 200 °C
Rate 7.5 oC/min.

The conditions of DT/ST system:
Type: Model 1000, Dynatherm Analytical Instruments Inc.,
Kelton, PA.

Dynamic thermal stripper:

98

99

Temperature: oven 80 °C
block . 105 °C
sorption tubes 50 °C

Purging time: 46 min

Thermal desorption unit:

Temperature: transfer line 230 °C
desorber 340 °C
valve 230 °C

Desorption time: 8 min

Peak areas and retention time were recorded by the use of a computer
integrator. Calibration factors (C.F .) were determined by the slope of the linear
calibration curve. The calibration factors are summarized in following table,

and calibration curves are shown in Figure 35, 36, 37, and 38.

Standard and TS/TD process calibration factors

 

Samples Process Calibration Factor
Vanillin standard 5.598 x 10'13
TS/TD 7.757 x 10'13
Ortho-vanillin standard 5.921 x 10"3
TS/TD 1.327 x 10'13

 

 

 

 

 

 

 

 

 

Area Units

60000

50000

40000

30000

20000

10000

100

R = 0.999 or. = 5.598 x 10''3 (gram/AU.)

 

 

 

 

/

 

/(

 

 

 

 

 

 

 

 

 

 

 

S 10 15 20

Quantity ( 10" gram)

25 30 35

Standard calibration data of vanillin

 

 

 

 

Quan. Inj. (ug) Area Units
0 0
8 15671
16 27651
24 41451
32 58568

 

Figure 35: standard calibration curve of vanillin

 

 

Area Units

3500000 R = ”-999

101

CE = 7.7375 x 10"‘(gmn/A.U.)

 

3000000

 

2500000

/

 

2000000

/

 

1500000

/,

 

1000000

/

 

500000 /

/

 

 

 

 

 

 

 

 

0.05

0.1 0.15 0.2
Quantity( 10‘3 gram)

TS/TD calibration data of vanillin

 

 

 

Quan. Inj. (ug) Area Units
0 0
0.125 1682271
0.188 2366638
0.250 3267948

 

 

Figure 36: TS/TD calibration curve of vanillin

0.25

Area Units

102

60000 R=0.998 C.F.=5.921x10"3(gram/A.U.)

5.... /

40000

 

 

 

30000 /

20000 //

/

 

 

 

 

 

 

 

 

 

 

 

 

5 10 15 20 25 30 35

Quantity( 10" gram)

Standard calibration data of ortho-vanillin

 

 

Quan. Inj. (u g) Area Units
0 0
8 11061
16 23177
24 37600
32 53966

 

 

 

 

Figure 37: standard calibration curve of ortho-vanillin

Area Units

103

800 R = 0999 C.F. = 1.327 x 10“3(gram/A.U.)

 

6000000

 

 

4000000 /

 

2000000

 

 

 

 

 

 

 

 

0.2 0.4 0.6 0.8 1.0
Quantity( 10'3 gram)

TS/T D calibration data of ortho-vanillin

 

 

 

 

Quan. Inj. (pg) Area Units
0 0
0.50 3602843
0.75 5663324
1.00 7507108

 

 

Figure 38: TS/TD calibration curve of ortho-vanillin

APPENDIX B

APPENDIX B

Calibration factors for HPLC process

The calibration factors of vanillin and ortho-vanillin for HPLC
procedure were determined by the analysis of 0.5, 1.0, and 2.0 ppm
(wt/volume) standard vanillin— and ortho-vanillin- acetonitrile solution. The

conditions for HPLC analysis was as follows:

Model: WatersTM 150—C ALC/GPC

Detector: Waters1M Tuable Absorbance Detector
Wavelength 275

Column: Delta-Pakm, C18 300A, 3.9 x 150 mm

Mobile phase: 40% Methanol & 60% HPLC grade water (v/v %)

Temperature: 23 °C

Flow rate: 1 ml/min.

Inj. Vol.: 10 ul

The calibration factor for vanillin and ortho-vanillin was 4.5 x 10'14

gram/A.U. and 6.7 x 10'” gram/A.U., respectively. The curves are shown in
Figure 39 and 40.

104

105

2.0 x 104 R=0.999 C.F.= 4.5 x 10'“ (gram/AU.)

 

1.5x10" /

 

Quantity mjected

1.0x10" _/

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5.0 x 10" /
. M
0 100000 200000 300000 400000
Areaunits
Area units Quantity injected
0 0
102070 5.0 x 10'9
209128 1.0 x 10-8

 

 

 

 

Figure 39: Vanillin calibration curve of HPLC process

106

2 0 x 104 R=0.999 c.1=.= 6.7 x 10'” (gram/AU.)

 

 

 

 

 

 

 

 

 

 

 

 

1.5 x 10“ A
3
2%
g 1.0 x 10"
3
5.0 x 10"
0
0 75000 150000 225000
Area units
Area units Quantity injected
0 0
102070 5.0 X 10'9
437636 2.0 X 10-8

 

 

 

 

300000

Figure 40: Ortho-vanillin calibration curve of HPLC process

APPENDIX C

APPENDIX C

Time needed to reach a steady state

within a paperboard/headspace sorption cell

To determined the time needed to attain a steady state concentration
level in the sorption cells, a 500 pl aliquot was withdrawn from the headspace
of the sorption cell following 2 days storage, and the sample injected directly
into gas chromatograph for quantitation. This process was repeated every two
days until the headspace concentration of vanillin and o—vanillin reached a
steady state level. Every sorption cell could just be withdrawn one time. For
both vanillin and ortho-vanillin samples, the results showed that the system
had equilibrated within 7 days period. The results were summarized in
following Table. The concentration levels change as a function of storage time

was illustrated graphically in Figure 41 as well.

107

108

The concentration change as a function of storage time until

 

 

 

 

 

 

 

 

 

 

 

 

 

a steady state was reached.

Storage vanillin E avg. O-vanillin avg.
time cone. in (ppb) conc. in (ppb)
(days) headspace headspace

(Ppb) (ppb)
3 10.5 12.40 3.5 3.20
14.3 2.9
5 19.7 18.80 4.6 5.10
17.9 5.6
II 7 18.4 17.65 6.1 5.85
16.9 5.6
fl 9 16.7 18.40 6.0 5.50
99 1 s n

 

 

 

 

 

109

 

 

 

 

 

 

 

 

 

20
I
1 5
“2
ii I vamllm' '
21‘:
- . o-vanillin
,, 3‘1 1 0
<2
{3
(1)
t9
5 ——-I__
l
3 5 7
Storage time (day)

Figure 41: The time needed to reach a steady state within
a paperboard/headspace sorption cell

BIBLIOGRAPHY

Aaron L. Brody, Glass-coated flexible film for packaging, Submitted
for publication in Eaekagingleehnelggmdfingineefing, 1993.

A. Keith Mitchell, Soaring demand and materials substitutions fuel a
boom, W, Page 82-86.

Crank, J., IhLMathematicLQLDifmsign, 2nd ed. Clarendon press,
Oxford, England. 1975.

Eric W. Ainscough and Andrew M. Brodie, The Determination of
Vanillin in Vanilla Extract, leumaleLChemieaLEdueatien, Volume 67,
No. 12, December 1990.

Fibre Box Association, MW , 1992, Page 7-8

Fred W. Billmeyer, SynthetiefiolxmersfiuildingthefiiantMeleeule
1972.

FDA Guidelines for Chemistry and Technology Requirements of
Indirect food Additive Petitions, 1976.

G. W. Halek and A. Chan, Partitioning and Absolute Flavor
Threshold Interactions of Aliphatic Food Packaging Solvent
Homologs in High-Fat Cookies, WWW,

Volume 59, No. 2, 1994. Page 420-422.

Herman F . Mark, Sheldon Atlas, Wee
1995, Page 7.

110

10:
11:

12:

13:

14:

15:

16:

17:

18:

19:

20:

21:

111

HandbeekeffiaekaginaEngineering, 1992, Page 2-2, 2-7, 8-5, 8-25.
Hoftyzer, P. J. and Van Krevelen, REID—9W, 2nd Edition,
Chapter7, 1976.

Hotchkiss J. H., An Overview of Food & Food Packaging

Interactions. WW5, American Chemical
Society, 1988, Page 1-10.

Jack R. Giacin, WW, 1994,
Chapter 2

John T. Felts, Transparent Barrier Coatings Update: Flexible
Substrates, mmmfliefiflmflheefing, Vol. 9-April, 1993.
Judith Foulke, Recycled Plastics, W , November 1991,
Page 11.

Laszlo J. Bonis, Plastics outflank traditional materials in grth and

market share, ModemPlastiesEnexelepediaL‘M, Page 28

Louis M. Maresca, Ashok C. Shah, W,
Page 52-58.

Masat Izu, High Performance Ovonic Clearcoat Barrier Film

Technology, WWW

Confereneeenioedflaekaginglnnexatiens. 1993.
Michael S. Bailey and Donald Brauer, Versatility and ease of

processing hike industrial and consumer demand, Modemflasjjes
EnexelepediaLQA, Page 48-52.
Peter W. Manders, New technologies are opening the door to wider

markets, MedemlllastiesEnexelepediafl Page 45-47
Rafael Gavara, Ruben J. Hernandez, Jack R. Giacin, Methods to

22:
23:

24:

25:

26:

27:

28:

29:

112

Determine Partition Coefficient of Organic Compounds in

Water/Polystyrene Systems, WW, Volume 61, No.
5, 1992, Page 947-952.

R. J. Hernandez, MW, 1994, Second Set

Ruben J. hemandez, Jack R. Giacin, A. L. Baner, The Evaluation of
the Aroma Barrier Properties of Polymer Films, Prepared for .LQmml
WW.1986.

Satoshi Maekawa, Volatilization of Vanillin and Ortho-Vanillin from a
Corrugated Paperboard Shipper and Sorption By a Low Density
Polyethylene Primary Container, 1994.

Sidney Kahan, Liquid Chromatographic Method for Determination of
Vanillin and Related Flavor Compounds in Vanilla Extract:
Collaborative Study, WW
Chemists, Vol. 72, No. 4, 1989, Page 614-618.

S. M. Mohney, R. J. Hernandez, J. R. Giacin, B. R. Harte, and J.
Miltz, Permeability and Solubility of d-Limonene Vapor in Cereal
Package Liners, Wee, Volume 53. No. 1, 1988.
Takehiko Watanabe, Packaging Evaluations of MXD-6 Nylon (Meta-
xylylene Diamine/adipic acid), Journal of WW, Vol.
3-July 1987.

Tim J, Nielsen, I Margaretha Jfigerstad and Rickard E. liste, Study of
Factors Affecting the Absorption of Aroma Compounds into Low-

Density Polyethylene, W, 60, Page 337-381, 1992.

Van Ness, J . H. Enexelepediaeffliemiealleehnelegy Vol 23, 1983,
page 704-717.

113

30: Walter F. Friedman, WW , 1977, Page 65-77.

 

MICHIGAN srnre UNIV. LIBRARIES
lllllillllllllllllllllllllllllll”Illlllllllllllllllllll
31293015650934