W; 1.3 1 $2.23. 1. 223‘ 1:... . a 5. a 3. .11... . limb rjrwwmm .I 17 1'1 189 —) W‘Iti‘fi’i‘ .ui .. Id.“ 1d: i. » ‘ i It... ’- . “I: quxa‘ 12:314.. ‘lti‘fl .xUA . ultx.‘ v. , iii, . 11-31).. x 1.1.? tih‘usiin‘ I. V .II :1 .A!!. {3‘ 3th. Wmmtfiknflvrlfl...}..ur”flufvui . ‘ . ‘ .. V ‘ . a . . . , . 1.5m....6...n Tammpmaflmh 2.1 “LA.” 14m. iackmLL .. . W ‘. .. ,. bi... new»mfiaggwmwmfiwvmedfi. .' .. 1 {.1. .il. YntStS SAT Ilmmllll "Illl‘lllllllllllfllll 3 1293 01389 7735 LIBRARY Mlchlgan State Unlverslty This is to certify that the thesis entitled PERMEABILITY OF ORGANIC VAPORS THROUGH A PACKAGED CONFECTIONERY PRODUCT WITH A COLD SEAL CLOSURE: THEORETICAL AND PRACTICAL CONSIDERATION presented by CHAI-HUNG LIN has been accepted towards fulfillment of the requirements for MASIER degree in PACKAGING Date_.IAN11ARX_l.9_._l.9_9.§_ 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE OI RETURN BOX to remove We checkout from your record. TO AVOID FINES return on or betore dete due. DATE DUE DATE DUE DATE DUE mm -C:l___l -| :- l l I I MSU le An Afflrmettve Action/Equel Opportunity Inetlttnlon W ”3-9.1 PERMEABILITY OF ORGANIC VAPORS THROUGH A PACKAGED CONFECTIONERY PRODUCT WITH A COLD SEAL CLOSURE: THEORETICAL AND PRACTICAL CONSIDERATION by Chai—Hung Lin A THESIS Submitted to Nfichigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE School of Packaging 1996 ABSTRACT PERMEABILITY OF ORGANIC VAPORS THROUGH A PACKAGED CONPECTIONERY PRODUCT WITH A COLD SEAL CLOSURE: THEORETICAL AND PRACTICAL CONSIDERATION by Chai-Hung Lin The permeability of toluene and ethyl acetate were determined for the following commodity films: (i) oriented polypropylene; (ii) polyethylene; (iii) glassine; (iv) saran coated oriented polypropylene; (v) acrylic coated oriented polypropylene; and (vi) metallized polyethylene terephthalate/oriented polypropylene laminate. Parameters evaluated included the permeant vapor activity and temperature. For each temperature, three vapor activity levels were evaluated. The temperature dependency of the transport processes was found to follow well the Arrhenius relationship. The extent of toluene vapor uptake by a confectionery product packaged in the respective commodity packaging structures and sealed by a cold seal closure was also evaluated and the predicted and experimentally determined levels of toluene uptake compared. The toluene uptake values for the six package/product systems were compared statistically, with the oriented polypropylene package system being assigned as the control. Based on the statistical analysis, there was no statistically significant difference between the performance of the control package system and the comparison package systems, with respect to toluene uptake, except for the glassine package. To my parents iii ACKNOWLEDGMENTS I would like to express my deepest thanks and appreciation to my major advisor, Dr. Jack R. Giacin, for his encouragement, advice and thoughtful guidance throughout my graduate program. I would also like to express my appreciation to Dr. Bruce Harte and Dr. Ian Gray for serving on my guidance committee members. A special appreciation to Dr. D. Gage for helping me to operate the equipment in the Department of Biochemistry - Mass Spectrometry Laboratory, Nuchigan State University. Also, I would like to thanks all my friends, Shu-Shin Chou, Shu-Jung Huang, Shaun C. Chen, Kuo-Chung Yin, and all my friends in School of Packaging, for their help and encouragement through my research study. Finally, a special thanks will give to my dear parents in Taiwan, for their love, encouragement and support throughout my life and education. iv TABLE OF CONTENTS LIST OF TABLES ........................................ viii LIST OF FIGURES ......................................... x INTRODUCTION ............................................ 1 LITERATURE REVIEW Permeation mechanism for the Transfer of Organic Vapor Through Polymeric Packaging Materials ........ 6 Permeation theory ............................. 10 Permeation measurement ........................ 13 Isostatic method ......................... 15 Quasi-isostatic' .......................... 17 Taints and Off-flavors in Foods and Confectionery Products ........................................... 18 Confectionery packaging ....................... 19 Source of taints .............................. 20 Threshold for volatile compounds .............. 21 Techniques of Analysis of Flavors .................. 26 Distillation/Extraction methods ............... 27 Headspace technique ........................... 28 Adsorbent trapping - thermal desorption technique ..................................... 29 Identification and Quantification of Trace Volatile Compounds by Gas Chromatography/Mass Spectrometry .. 32 Characteristics of Packaging Seal .................. 36 Heat seal ..................................... 38 Cold seal ..................................... 4O MATERIALS AND METHODS Part I: Permeability of Organic Volatiles Through Packaging Film Roll Stock .................. 44 Packaging materials ...................... 44 Permeation test .......................... 45 Part II: Permeability of Toluene Through a Packaged Part III: Confectionery Product with A Cold Seal Closure .............................. 50 Permeation chamber design ................ 51 Storage condition ........................ 54 Toluene quantification apparatus for trapping of volatiles .................... 54 Flavor analysis procedure ................ 58 Dynamic purge and trap procedure .... 58 Gas Chromatograph - Mass Spectrometric analysis .............. 62 Leak Testing ............................. 67 RESULTS AND DISCUSSION Part I: Permeation Study of Packaging Films ........ 69 The effect of vapor activity and temperature on ethyl acetate permeability ............ 69 The effect of vapor activity and temperature on toluene permeability .................. 87 Estimation of permeance values at vapor activity levels impractical to measure experimentally .......................... 108 Part II: Permeability of Packaged Confectionery Product Systems with a Cold Seal Closure.. 115 Predicted uptake levels of toluene vapor by the packaged confectionery product as a function of time (24%3 ................. 118 vi Experimental uptake levels of toluene vapor by the confectionery product as a function of time (24%3 ................. 121 Part III: Seal Integrity Evaluation of the Cold Seal Closure for Packaged Confectionery Product Systems ......................... 133 SUMMARY AND CONCLUSIONS ................................ 138 FUTURE STUDIES ......................................... 141 Appendix 1: Calibration Curve for ethyl acetate and toluene by Gas Chromatograph ....................... 143 Appendix II: Calibration Curve for toluene by Gas Chromatograph - Mass Spectrometry ......... 145 Appendix III: Standard toluene mass spectrum ........... 146 Appendix IV: The analysis of variance table of permeance for toluene vapor* ......................... 147 Appendix V: Estimation of toluene uptake by the packaged confectionery product by a permeation mechanism .................................. 149 Appendix VI: The statistical calculation for the comparison of toluene uptake by the packaged confectionery product ............ 151 Appendix VII: Leak test values and corresponding toluene uptake levels as a function of time .................................. 152 BIBLIOGRAPHY ........................................... 154 vii LIST OF TABLES Tables Pages 1. The partition coefficient (K91) of organic sorbates at 23%: ............ ' ......................... 24 2. Sorbate sensory threshold in chocolate ............... 25 3. Permeance constant of ethyl acetate through oriented polypropylene as a function of vapor activity on temperature .......................................... 70 4. Permeance constant of ethyl acetate through high density polyethylene as a function of vapor activity on temperature ....................................... 71 5. Permeance constant of ethyl acetate through glassine as a function of vapor activity on temperature .......... 72 6. Permeance constant of ethyl acetate through saran coated oriented polypropylene as a function of vapor activity on temperature ....................................... 73 7. Permeance constant of ethyl acetate through acrylic coated oriented polypropylene as a function of vapor activity on temperature .............................. 74 8. Activation energy for the permeation of ethyl acetate through polymer membranes ............................ 86 9. Permeance constant of toluene through oriented polypropylene as a function of vapor activity on temperature .......................................... 88 10. Permeance constant of toluene through high density polyethylene as a function of vapor activity on temperature ......................................... 89 11. Permeance constant of toluene through glassine as a function of vapor activity on temperature ...... 9O 12. Permeance constant of toluene through saran coated oriented polypropylene as a function of vapor activity on temperature ...................................... 91 viii 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Permeance constant of toluene through acrylic coated oriented polypropylene as a function of vapor activity on temperature ...................................... 92 Activation energy for the permeation of toluene through polymer membranes .......................... 105 Estimated permeance values at packaging structures for toluene at 249C (based on Arrehenius equation)...109 Vapor activity dependence of the permeance for toluene at 24%: .................................... 112 Statistics comparison of toluene permeance for polymer structures ................................. 114 Estimated uptake time to exceed toluene sensory threshold levels in a packaged chocolate product ... 120 Total toluene uptake by chocolate for OPP, HDPE and Glassine based structures as a function of time (Experimental and Calculated data) ................. 122 Total toluene uptake by chocolate for Saran coated OPP, Acrylic coated OPP and Metallized PET/OPP based structures as a function of time (Experimental and Calculated data) ................................... 123 Analysis of variance (ANOVA) for toluene uptake for product/package system ......................... 125 Statistical comparison of toluene uptake for six package/product systems ............................ 127 The average values for leak testing by using Nikka Packaging Leak Tester (PLT-3501) ................... 134 Analysis of variance for toluene permeance constant ........................................... 147 Leak test values and corresponding toluene uptake levels as a function of time for OPP, HDPE and Glassine based packaging structures ................ 152 Leak test values and corresponding toluene uptake levels as a function of time for Saran OPP, Acrylic OPP and Metallized PET/OPP based packaging structures ......................................... 153 ix LIST OF FIGURES Figures Pages 1. Schematic diagram of test system for determining the permeability of packaged product ..................... 53 2. Photograph of vapor generator system and permeability test chambers for packaged product ................... 55 3. Schematic diagram of dynamic purge and trap system.... 57 4. The apparatus of trapping cell ....................... 59 5. Photograph of dynamic purge and trap system .......... 6O 6. Schematic of complete and desorption mechanism for the Short Path Thermal Desorption Unit (Scientific Instrument Services, Ringoes, NJ) .................... 63 7. Short Path Thermal Desorption Unit operation with the gas Chromatograph .................................... 64 8. The effect of ethyl acetate vapor activity on permeance for oriented polypropylene ........................... 75 9. The effect of ethyl acetate vapor activity on permeance for high density polyethylene ........................ 76 10. The effect of ethyl acetate vapor activity on permeance for Glassine ........................................ 77 11. The effect of ethyl acetate vapor activity on permeance for saran coated oriented polypropylene ............. 78 12. Temperature dependence of the permeance for ethyl acetate in oriented polypropylene ................... 81 13. Temperature dependence of the permeance for ethyl acetate in high density polyethylene ................ 82 14. Temperature dependence of the permeance for ethyl acetate in glassine ................................. 83 15. Temperature dependence of the permeance for ethyl acetate in saran coated oriented polypropylene ...... 84 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. Temperature dependence of the permeance for ethyl acetate in acrylic coated oriented polypropylene .... 85 The effect of toluene vapor activity on permeance for oriented polypropylene .......................... 94 The effect of toluene vapor activity on permeance for high density polyethylene ....................... 95 The effect of toluene vapor activity on permeance for glassine ........................................ 96 The effect of toluene vapor activity on permeance for saran coated oriented polypropylene ............. 97 Temperature dependence of the permeance for toluene in oriented polypropylene ........................... 99 Temperature dependence of the permeance for toluene in high density polyethylene ....................... 100 Temperature dependence of the permeance for toluene in glassine ........................................ 101 Temperature dependence of the permeance for toluene in saran coated oriented polypropylene ............. 102 Temperature dependence of the permeance for toluene in acrylic coated oriented polypropylene ........... 103 The effect of toluene vapor activity on permeance for different polymer structures at 24%: ............... 110 Consistency of toluene vapor driving force in permeation test chambers 1 and 2 ................... 116 Consistency of toluene vapor driving force in permeation test chambers 3 and 4 ................... 117 Total toluene uptake by chocolate for different package/product systems (Experimental data) ........ 124 Comparison of estimated and experimental data for toluene uptake by OPP/product system at a = 0.0015...129 Comparison of estimated and experimental data for toluene uptake by HDPE/product system at a = 0.0015..130 Comparison of estimated and experimental data for toluene uptake by Glassine/product system at a = 0.0015 ......................................... 131 33. 34. 35. 36. Comparison of estimated and experimental data for toluene uptake by Saran OPP/product system at a = 0.0015 ...................................... Calibration Curve of ethyl acetate by GC-FID ....... Calibration Curve of toluene by GC-FID ............. Calibration curve of toluene by GC/MS - thermal desorption unit (SIM Mbde) ......................... xii 143 144 INTRODUCTION Since 1970, the use of flexible packaging materials has continued to grow, and has found wide use in food, snack food, confectionery product and beverage packaging (Minifie, 1970, and Martin, 1987). Selection of the appropriate packaging material for a specific food packaging application is essential in order to prevent the development of off-odor or flavor by the product, as a result of permeability of low molecular mass organic compounds from the storage environment through the package to the product, or the migration of low molecular mass organic compounds regularly contained by the package itself into the product. With respect to migration, package printing with organic solvent based inks can result in the development of off-odor or off- taste in packaged products such as snack foods, confectionery products, and baked foods (Kontominas, 1985; Kontominas and Voudouris, 1982; and Wilks and Gilbert, 1968). Another example of the development of off-odor or flavor by a product is the uptake of diesel odor or soap aromas by a packaged food product, as a result of either improper storage or inappropriate packaging material selection (Franz, 1993 and Hotchkiss, 1995). 2 The flavor or aroma of a food product is comprised of a particular and often delicate balance of organic compounds. With the extensive use of flexible films for food packaging, consumers have become more aware of the occurrence of product off-flavor and off-odor. Consumer perception of inferior product quality, as a result of the development of off-odor and flavor, has resulted in the food industry becoming more sensitive to off-flavor complaints (Whitfield, 1986). The basic mechanism of transmission of vapors and gases through polymeric films is either by flow through macro— voids or by permeation through micro—voids. Flow through macro—voids may be due to pinholes, cracks, or a faulty package seal. Permeation is a complex process which includes dissolution of the permeant at the high permeant concentration contact surface, diffusion through the polymer film bulk phase, and desorption or evaporation of the permeant from the downstream surface of the film. In practice, knowledge of the permeability of water vapor, carbon dioxide, oxygen and nitrogen has been utilized to design packaging systems to prevent or retard product deterioration during product transportation and storage. The rate of permeation is affected by several parameters, such as the nature of the permeant molecule, to include molecular structure and polarity; as well as the nature of the barrier material to include percent 3 crystallinity, glass transition temperature, level of plasticizer added, and polymer morphology. The barrier properties of polymer films can also be affected by environmental parameters such as the temperature, relative humidity, and the driving force or partial pressure gradient of the permeant (Lebovits, 1966, and Kontominas, 1985). Recent developments have made available high barrier flexible packaging materials. For example, the present combination of aluminum foil in laminate structures, surface deposition of silicon oxide and aluminum to polymer films, and the incorporation of high barrier polymers like ethylene-vinyl alcohol copolymers (EVAL) in multilayer laminations have resulted in the availability of high barrier flexible structures exhibiting barrier characteristics similar to glass and metal (Schaper, 1989). In addition to the barrier characteristics of the packaging material, the packaging seal characteristics, as previous indicated, are also very important. Traditionally, a heat sealing technique was used to form and seal the package structure. However, at present there is a trend to use a cold seal closure in fabrication of simple wrappers and bags (Stone, 1976 and Jones, 1985). For a cold seal closure, a self adhesive coating is applied to the polymer films. Advantages of a cold seal closure include an increase in the speed of wrapping and a decrease in the total cost of sealing. Therefore, the development of high barrier polymer 4 films, together with the application of a pattern cohesive to allow sealing by a cold seal closure is very important to the food industry, for protecting packaged products frdm the external package environment. Toluene, methyl ethyl ketone, and ethyl acetate are among the most common organic solvents found as residuals from laminating and printing processes (Kontominas, 1985). Mueller (1990) also showed that printed materials, such as newspapers, are a major source of toluene contamination of food. Based on these findings, the organic permeants evaluated in the present study included toluene and ethyl acetate. For the respective permeant, the temperature and concentration dependency of the permeability process was determined. For seal integrity studies, toluene was selected, based on the fact that it was representative of a residual solvent, could serve as a model compound for diesel fuel volatiles and it had a known sensory threshold level in chocolate. The specific objectives of the study include: 1. Determine the temperature and concentration dependence of the permeability of toluene and ethyl acetate through a series of commodity packaging structures used for packaging confectionery products. 2. Utilize data on the temperature and concentration dependence of the permeability of toluene through the 5 respective test structures to estimate the permeance of the barrier films at permeant concentration levels below which it would be impractical to measure experimentally. Develop a continuous flow system to allow determination of the rate of toluene uptake by a packaged confectionery product at conditions of constant temperature and toluene vapor activity. Determine the extent of toluene vapor uptake by a confectionery product packaged in a series of commodity packaging structures sealed with a cold seal closure and compare the predicted and experimentally determined levels of toluene uptake. LITERATURE REVIEW Permeation Nbchanism.£or the Transfer of Organic vapor Through Polymeric Packaging materials The deterioration of food flavor during storage may be due in part to the loss of volatile components from the packaged food product or to the uptake of contaminants permeating from the external environment. Therefore, measurement of the rate of permeation of various organic permeants through polymeric packaging materials is of significant importance. Permeability is the phenomenon of transmission of a gas or vapor through a polymer. This process is characterized by three basic steps which are summarized below: 1. Adsorption - the dissolution of a gas or vapor into the polymer matrix at the high penetrant concentration surface. 2. Diffusion - the movement of a gas or vapor penetrant through the polymer bulk phase, along a concentration gradient. 3. Desorption - the desorption or evaporation of the penetrant from the low concentration surface of the polymer membrane. 7 Mass transfer in polymeric packaging materials can be referred to as either permeability, sorption or migration. When small molecules permeate through a polymeric membrane, the rate of permeation is governed by the physical and chemical nature of the penetrant and the polymer membrane, as well as external factors which include: temperature, vapor pressure, and relative humidity. In general, the permeability of a gas or vapor through polymer films is a function of two fundamental parameters: the diffusion coefficient (D), and the solubility coefficient (S). Functionally, the diffusion coefficient is a kinetic term and describes how rapidly permeant molecules are advancing and the time required to reach steady state and the solubility coefficient is a thermodynamic term, describing how many permeant molecules are sorbed by the barrier structure. The diffusion process is the result of polymer molecules having a random kinetic agitation or heat motion. Above the glass transition temperature (TB), polymer chain segments have vibration, rotation, and translation properties which continually create temporary - "holes" or voids within the polymer matrix. The "holes" allow penetrant molecules to pass through the matrix under a concentration gradient. For polymer membranes belOW'T' the rate of g! diffusion will be a function of the size and frequency of preexisting "holes" or the void volume between polymer chains. Whereas, the solubility involves the affinity of the 8 permeant for the polymer (Baner, 1986 and Imbalzane et al., 1991). Permeation is a measure of the steady-state transfer rate of the permeant, which is usually described by the permeability coefficient, P. The permeability constant can be expressed in terms of two fundamental parameters: the diffusion coefficient (D) and the solubility coefficient (S). Their relationship is described by Equation 1. P=DxS (1) Where the diffusion coefficient (D) is a measure of the rate at which penetrant molecules are advancing through the polymer, and the solubility coefficient (S) describes the amount of permeant contained or dissolved in the polymer matrix or slab at equilibrium conditions (Crank, 1968). The mass transfer of organic volatiles is a major concern associated with polymeric packaging films. First, there is the potential loss of aroma/flavor compounds due to their permeation through the film, or to their sorption by the polymeric packaging material. Further reason for concern involves the possibility of contamination of the food due to the permeation of organic vapor from the external environment through the packaging film (Mbhney et al., 1988; Tou et al., 1990; and Franz, 1993) and sorption by the product. 9 In food product/package systems, understanding the sorption mechanism is necessary to assure quality control and prediction of product quality change. In addition, measurement of the loss of product volatiles by permeation is also associated with product quality change or shelf life. Sorption can be described as a function of sorbate concentration by a sorption equilibrium isotherm, that can be represented by Henry’s Law or other mathematical models. Specifically, Henry's Law describes the relationship between the vapor pressure of the organic penetrate and the mole fraction of penetrant sorbed by the polymer membrane at equilibrium. The partition coefficient K is defined as the ratio of the equilibrium concentration of a component in a fluid phase (e.g. food phase), and the equilibrium concentration of the component in a contracting polymeric material. For a specific value of concentration, the partition coefficient K provides a practical way to calculate the change in a component concentration, either in the food or packaging material, from the time that the food product and packaging material are contracted, up to the moment they reach equilibrium, provided the initial concentrations are known. Since sorption and migration are essentially the same mass transport phenomenon, migration can also be described by a partition and diffusion coefficient. The diffusion coefficient determines the dynamics of the sorption process. 10 The larger the value of D, the shorter the time to reach equilibrium. (Giacin, 1995). None-interaction molecules like oxygen, nitrogen, and carbon dioxide will have little or no effect on polymer morphology when sorbed into the polymer matrix. However, organic vapors are comparable in size or larger than the polymer chain segments associated with chain segmental mobility at temperatures above TE. These organic penetrants diffuse by a complicated mechanism which is dependant upon the motions of both the polymer and diffusant molecule. Organic molecules may also have a higher solubility in the polymer, which can result in significant swelling of the polymer matrix by the sorbed organic molecules. Organic molecules sorbed by the polymer can act like a plasticizer, lowering the glass transition temperature (TE) and increasing the polymer’s segmental motions at any temperature. This will results in further plasticization and swelling (Meares, 1965 and Marner, 1986). W The rate of transfer of a diffusing substance can be expressed mathematically by Fick's first and second laws of diffusion, as described in Equations 2 and 3. The solubility can be described by Henry's law (Crank, 1975). do F = - D -——-—— (2) 11 Q. Q. 0 Q. N 0 dc —--—— = ( D -————— ) = D —————- (3) dt dx dx dxz where: F = flux (the rate of transfer of the diffusing substance per unit area) = diffusion coefficient — concentration of diffusing substance = direction of diffusion O x C) U l = time. When the permeant molecules enter into the polymer, a boundary layer is created. The boundary layer phase of the polymer is swollen and the diffusion coefficient of the permeant in this phase is not equivalent to that of the non- swollen phase. The diffusion coefficient in the swollen phase is larger than in the unswollen phase. Also, Fick's laws of diffusion do not apply because the boundary phase is changing with time (Crosby, 1981 and Maekawa, 1994). Under low vapor concentration levels, the vapor pressure of the penetrant is proportional to its concentration in the polymer. According to Henry's law (Equation 4), the rate of transfer of a diffusing substance can be expressed in terms of the permeant partial pressure between surface x = 0 to x = L by the relationship: C = S x Ap (4) 12 where: C = concentration of the penetrant in the polymer phase S = solubility coefficient Ap = partial pressure of the penetrant in the gas phase. At steady state, the diffusion flux (F) (Equation 5) of a permeant in a polymer can be defined as the amount of penetrant which passes through a unit surface area per unit time. Q F = ———————— (5) A x t where: Q = quantity of diffusing substance transferring through the film A = area t = time. Also the diffusion flux (F) (Equation 6) of the permeant can be integrated through the total thickness of the polymer between the two concentrations, assuming that the diffusion coefficient (D) is constant and independent of concentration. L where: c“ (a = concentration of gas or vapor on the surface x = 0 and x = L, respectively (c1 > c2) L = film thickness. n: I!) 13 From Equations 4, 5 and 6 one can then calculate the amount of penetrant: D ( c1 — c2 ).x A x t Q = (7) L D x S ( 91"55 ) x A,x t Q = (8) and by using the definition of permeability: P = D x S one can then solve for the permeability constant using Equation 9: Q x L p = (9) A x t x ( p1-;§ ) W Generally, there are three methods for measuring gas and vapor permeability of packaging materials: (i) the absolute pressure method; (ii) the isostatic method; and (iii) the quasi-isostatic method. Basically, each test method is designed to measure the permeability by mounting the film in a hermetic environment between two test cell chambers. A fixed and constant gas or vapor concentration is introduced into the high concentration cell chamber. The 14 permeant then diffuses through the barrier membrane and is desorbed from the low concentration film surface into the low concentration cell chamber, where it is then quantified using some appropriate method of detection. The absolute pressure method involves use of no gas or vapor other than the permeant/vapor in question. There is a total pressure gradient between the two chambers which provides the driving force for the mass transfer process (ASTMID3985-81). Mere recently, the isostatic and quasi—isostatic procedures have become two widely employed methods for measuring permeability. Since organic vapor and aroma permeability measurements are very complicated procedures, there are no standard technique for analysis. Also, there is a very limited number of commercial instruments available for measuring organic vapor permeability. To date, most research in this area has been conducted using a permeation system of the investigator's own design (Blakesley, 1974; Niebergall et al., 1979; Kontominas, 1985; Barner, 1986; Hernandez et al., 1986; Hatzidimitriu et al., 1987; Liu et al., 1991; and Franz, 1993). Most of these have been designed to measure the transmission rate of organic vapor through barrier membranes by maintaining a constant partial pressure differential across the test membrane. 15 I i h In this procedure, the total pressure in both chambers of the permeation cell is kept constant by maintaining the same atmospheric pressure in both chambers of the test cell. The permeability measurement involves maintaining a constant permeant partial pressure, or a concentration gradient between the two cell chambers. The permeant which has permeated through the film and into the lower concentration chamber is then conveyed by a carrier gas to a detector where is quantified. Figure 1 presents a typical transmission rate profile curve obtained using the isostatic procedure. The permeability measurement involves determining the change in the ratio of [(AM/At)t/(AM/At)a] as a function of time. Equation 10, which was derived by Pasternak et al. (1970), describes the transmission rate profile curve (Hernandez et al., 1986): (NM/At) —L2 t )"2 exp(——) (10) 4 L2 = ( )(--—- (AM/ At) . i H 4Dt 4m: where: (Am/At)t and (AM/At)°== the transmission rate of the penetrant at time (t) and at steady state (m) t = time L = film thickness D = diffusion coefficient. 16 A value of (IF/4Dt) can then be calculated for each value of [(ANVAt)t/(AM/At).]. A straight line can be obtained by plotting (4Dt/LF) as a function of time. From the slope of the straight line, the diffusion coefficient (D) was obtained from Equation 11. (slope) x L2 D = (11) 4 Ziegel et a1. (1969) derived Equation 12, from a different expression for [(ANVAt)t/(&M/At)o] (Hernandez et al., 1986) to solve for D. L2 D = (12) 7.199 x t0_l5 where: tms = the time required to reach a rate of transmission (AM/At)t'value equal to half the steady state (AMJAt)¢'value. Finally, the permeability coefficient (P) can be calculated from data obtained using the isostatic method by substitution into Equation 13: a x G x f x L P = (13) A x Ap where: a = calibration factor converting detector response to units of mass of permeant/unit of volume G = response units from detector output at steady state 17 f = flow rate of sweep gas conveying penetrant to detector A = area of the film exposed to permeant in the permeation cell L = film thickness Ap = partial pressure gradient. u i-i i This method involves measurement of accumulated permeated gas or vapor as a function of time. In this method, one of the two chambers between which the film is placed is completely closed. The gas or vapor being tested is allowed to flow through the high concentration cell chamber at atmospheric pressure. At the same time, the gas permeates into the lower concentration cell chamber and aliquots are withdraw from the low concentration cell chamber for analysis at predetermined time intervals. Using this method, the quantify of permeant accumulated increases as a function of time. When the relationship between quantity accumulated per unit of time is constant, steady-state rate of diffusion has been reached. By extrapolating the linear segment of the curve to the x-axis, the lag time (9) can also be calculated. It is the intersection of the projection of the steady state portion of the transmission curve (Hernandez, et al., 1986 and Wangwiwatsilp, 1993). Figure 2 illustrates a typical transmission rate profile curve for the quasi-isostatic test 18 procedure. The diffusion coefficient is expressed by Equation 14: L2 D = -—-——— (14) 66 where: e = lag time L = film thickness and the permeability coefficient at steady state can be described by Equation 15. slope x L P = (15) A x Ap where: slope straight line portion of the transmission rate curve (quantity/time) film thickness I." II A = surface area Ap = partial pressure. Taints and Off-flavors in Foods and Confectionery Products As a result of increased competition in the market place and stronger consumer demands for high product quality, the food industry has strived to minimize adverse consumer reactions. The unpleasant nature of food contaminants or taints can cause severe problems for retailer, producer, ingredient supplier, farmer, and equipment supplier. 19 According to the International Standardization Organization (ISO, 1992), the standard definition of a taint is a taste or odor foreign to the product. In addition, ISO also defines an off-flavor as a typical flavor usually associated with deterioration. An alternative means of defining taints and off—flavor has been described by Kilcast (1993): Taints: unpleasant odors or flavors imparted to food through external sources. Off-flavors: unpleasant odors or flavors imparted to food through internal deteriorative change. Perception of a taint in a food product is dependent on the volatile chemicals, chemical structure, concentration, the type of food, and the sensitivity of the consumer (Kilcast, 1993). W The packaging requirements for chocolate confectionery products are similar to those of other food products (Martin, 1987): Protection: The product should be protected by the package from the time of manufacture until consumption is completed. Keep product flavor inside the package, while water vapor, oxygen and 20 undesirable environmental odors need to be excluded. Function: The package should be functional and facilitate the manufacture, shipping, storage, display and use of the product. Sales Appeal: The package must attract the shopper, appeal to emotions, inform and trigger the sale. There are several common polymeric films and flexible materials used as confectionery packaging structures. These include cellulose, high density polyethylene (HDPE), polypropylene (PP), polyvinylidene chloride (PVDC), polyethylene terephthalate (PET), aluminum foil, and paper/glassine. In addition, a number of newer materials with very good barrier properties are suitable for confectionery packaging. r f int Basically, taints in foodstuff are the result of poor handling, transportation or storage of the product, or exposure of the product to an alien compound. Air acts as a good carrier for a broad range of compounds which can be deposited on, the surface of a sensitive food product. For example, a cake product or biscuits can absorb volatile compounds from newly-painted walls or new flooring in factories. Food may also pick up 21 pesticide, diesel or petrol fumes flavor, which can cause tainting (Saxby, 1982). Packaging can serve as yet another source of taints in food. Packaging materials in direct contact with food may result in the transfer or migration of volatiles into the food product. During polymer manufacture, residual volatiles such as monomer can become incorporated into the resin beads, which when converted into film, can affect food flavor. Other factors that can affect package flavor contribution during film processing include: sealing temperature, sterilization, residual solvent from printing inks (eg., ethyl acetate, hexane, toluene, and heptane), types of adhesives, coatings and coextrusion processes (Wilkes and Gilbert, 1972; Halek and Levinson, 1988; and Thompson et al., 1994). Residual solvent can migrate into a packaged food by either direct contact or via head space inside the package. Thrgghglg fgr vglggilg gggpggggg Generally, threshold can be defined as the minimum concentration of a stimulus which can be detected (absolute threshold), discriminated (just-noticeable-difference) or recognized (recognition threshold) by 50% of a specific population (Pangborn, 1981). ISO 5492 (ISO, 1992) has presented a more detailed definition for threshold levels (kilcast, 1993): 22 Detection threshold: Nfinimum value of a sensory stimulus needed to give rise to sensation. Recognition threshold: Nunimum value of a sensory stimulus permitting identification of the sensation perceived. Difference threshold: Value of the smallest perceptible difference in the physical intensity of a stimulus. Terminal threshold: Nflnimum value of an intense sensory stimulus, above which no difference in intensity can be perceived. Several researchers have studied the perception threshold for organic vapors in air and in foods. The American Conference of Governmental Industrial Hygienist (ACGIH) publishes an annual listing of threshold limit values (TLV). Amoore and Hautala (1983) summarized water odor threshold levels and TLV for 214 organic vapors. In their paper, they reported the TLV value for toluene is 100 ppm (v/v) (ACGIH, 1982), the air-dilution threshold for toluene is 0.042 ppm (w/v). When a consumer is exposed to a complex flavor profile, the threshold value for a specific organic flavor compound will increase. Granzer et al. (1986) reported the threshold of toluene is 1 - 20 ppm (mg/kg) in water, 50 — 500 ppm (mg/kg) in butter cookies, and 100 - 500 ppm (mg/kg) in 23 coconut oil. The amount that is sorbed by the food depends on the partitioning of the migrant between the packaging material and the contained food product (Gilbert, 1979 and Halek et al., 1988). Several factors have been found to affect the partitioning from the packaging materials into the closed contents. (Koros and Hopfenberg, 1979; Adamson, 1982). In general, direct migration of substances from packaging materials occurs more with fatty foods. Fats and oils can penetrate some packaging surfaces and actually promote residual solvent migration into the food product. Studies conducted by Halek and Hatzidimitriu (1988) have suggested that the partioning of the solvents varies according to the target materials, and the order of affinity for organic solvents is as follows: soybean oil > chocolate liquor > cookie. This order is related to the fat content percentage in the target materials. Ruter (1992) also investigated the partition coefficient (Km!) of typical printing solvents in food stuffs at 23°C. The results of this study are summarized in Table 1. The residual solvents from laminate films or printing inks can cause off-odor and off-taste of the packaged product. Rfiter (1992) investigated and determined the partition coefficients and the sensory threshold values of commonly used solvents in different foodstuffs. The sensory thresholds in chocolate at 23°C are shown in Table 2. 24 Table 1. The partition coefficient (K91) of organic sorbates at 23°C (Rfiter, 1992) Partition coefficient (Km!) Sorbate Potato Coffee Yogurt Chocolate chips Acetone 53.0 52.0 8.6 100 Cyclohexane 18.0 33.0 600.0 53 Ethyl 18.0 46.0 36.0 60 acetate MEKm 22.0 38.0 11.0 66 Toluene 3.6 9.4 57.0 11 ‘U. MEK: Methyl ethyl ketone 25 Table 2. Sorbate sensory threshold in chocolate (Rfiter, 1992) Smell (odor) Taste (flavor) Sorbate , , . . . . stimulus recognition stimulus recognition threshold threshold threshold threshold .Acetone 10 - 20 20 - 100 20 - 50 50 — 100 Cyclo- 50 - 100 100 - 500 500 - 1000 1000 - 2000 hexane Ethyl 10 - 50 20 — 100 20 ~ 50 50 - 100 acetate MEK‘” 20 - 50 50 - 100 50 - 100 100 - 500 Toluene 5 - 15 20 - 25 5 - 15 20 - 25 .H. MEK: Methyl ethyl ketone 26 NMeller et al. (1990) evaluated the food sold at service stations, which may have been contaminated with petroleumrbased constituents. In this study, the investigators found that the confectionery products sold in service stations and placed near newspapers showed the highest concentration of toluene contamination (216 ppb, ug/kg). Products sold in supermarkets which were displayed near printed materials such as newspaper and magazine, showed higher toluene concentration level than the products displayed in the absence of printed materials. Eweller et al. (1990) concluded that the toluene contamination can depend on the presence of printed materials in the vicinity of the tainted food product. Techniques of Analysis of Flavors The aroma profile of a food product is a complex mixture of volatile organic chemicals. Aroma virtually never stands alone, but is combined with other sensory attributes. Difficulty in the isolation and analysis of flavor or aroma compounds is due to the presence of such compounds in very low concentrations. These trace aroma compounds are generally in a complex food system which contains thermally labile constituents (eg. sugar, protein, lipid and vitamins), food emulsifiers, aqueous and fat soluble components and other volatile components. The basic principle used in the isolation of aroma compounds from the 27 major food components is either by means of the compounds volatility or solubility properties. Most flavor components are soluble in organic solvents, while most food components are soluble in water. The presence of a high level of lipids (which are soluble in organic solvents) in a food product may cause the isolation to be more complicated (Reineccius, 1993). Normally, there are three methods that are commonly use for flavor isolation: distillation extraction, direct headspace analysis, and a dynamic purge and trap procedure with a porous polymer absorbent (dynamic headspace) (Alberola and Izquierdo, 1978; Reineccius, 1984 and Reineccius, 1993). Di il 1 n Ex r i n h One of the simplest and most efficient approaches for aroma isolation is direct solvent extraction. However, one major limitation of this method is that it is useful only on foods or samples that do not contain any lipids. Also, the biases imposed on the aroma profile by solvent extraction which are related to the relative solubility of various aroma constituents in the organic/aqueous phases (Reineccius, 1993). One of the most commonly used methods for separating volatile compounds from a food or beverage is steam distillation followed by extraction with an organic solvent. 28 The distillation method can be defined broadly, to include: high vacuum molecular distillation, steam distillation or simple heating of the food product. Each method is based on volatility. It is the technique which allows a larger concentration of solvent to evaporate. The most commonly used extraction method involves some modification of the Linkens—Nikerson extraction apparatus. This apparatus involves aroma volatiles isolation by distillation and extraction procedures carried out simultaneously under vacuum, which allows for more efficient extraction of the distillate and the use of a minimal volume of solvent. Reduced pressure Operation results in minimal thermally induced artifact formation (Alberola and Izquierdo, 1978). However, the loss of low-boiling volatiles and the trapping of solvent impurities may occur during concentration of the sample volatiles. Thus, care must be exercised in the use of this apparatus (Parliment, 1987). Wicca Headspace sampling is the simplest, fastest and one of the most precise techniques for analysis of volatile compounds. This technique is based on an equilibrium distribution of the volatiles between the sample and the gas phase in a closed and thermosetted vial (Kolb, 1984). The analysis of headspace vapors above food stuffs by gas chromatography (CC) or gas chromatography/mass spectroscopy 29 (GC/MB) has been widely applied in flavor chemistry. Wyllie (1978) defined headspace as the gaseous mixture surrounding a sample within a closed system at equilibrium. This method was recommended by the working group on Methods of Analysis of the International Organization of the Flavor Industry (IOFI), because this direct injection method is very suitable for routine analysis of the quality of orange juice volatile fraction (Alberola and Izquierdo, 1978). Unfortunately, this method can not provide adequate sensitivity for trace compound analysis, and is limited to volatile partial pressure levels in the headspace equal to concentrations above 10‘7 g/L (Schaefer, 1981). Use of inlet splitless injection when using GC can enhance these detection limits. Direct headspace sampling has been widely used for rapid analysis and major component analysis. The cryogenic trapping technique offers some advantages and disadvantages. The cryogenic trap collects headspace vapors irrespective of compound polarity and boiling point. However, water is typically the most abundant volatile in food products and further procedures may be required to remove the trapped water vapor (Reineccius, 1984). r r in - h l i n hni Many "headspace" determinations involve the flowing of a non-condensible gas over the sample to sweep the volatiles 30 into a trapping device. This technique is also commonly referred to as a dynamic headspace method or the dynamic purge and trap method. Normally, the dynamic headspace technique can provide a sensitivity 10 times greater than the sensitivity obtained using the static headspace technique. The sample is continuously purged with an inert gas such as nitrogen or helium, which strips volatiles aroma constituents from the sample, and conveys the aroma volatiles to a trapping system where they are sorbed. The aroma constituents can be trapped by Tenax (porous polymer), charcoal, or other suitable trapping adsorbents. Tenax-GC (poly 2-6 diphenylepara-phenylene oxide) is widely used for aroma trapping. Tenax-GC has a low affinity for polar compounds (water vapor) and high affinity for non- polar compounds (volatile and semi-volatile compounds). Investigators initially used Tenax as the trapping medium for isolating and concentrating volatile organics in food, environmental and biochemical samples (Bellar and Lichtenberg, 1974 and Bertsch et al., 1974). Activated carbon traps have a strong affinity and high capacity for most aroma constituents. Generally, adsorbents begin to efficiently trap most classes of volatile organic flavor compounds when their carbon skeletons are above C-3. The pore diameter and particle size distribution of activated carbon profoundly affects the breakthrough volumes of analities. Smaller particles and pores are more efficient 31 for trapping highly volatile, low molecular weight hydrocarbons such as C2 - C4 species. Silica gel has been used to trap low molecular weight polar compounds. Often, the ideal adsorbent trap may be a mixed bed design with a number of adsorbents in combination (Supelco, 1990). Porous polymer and carbon-type adsorbents have a low affinity for water vapor, which will elude the pores of the adsorbent and inhibit the diffusion of volatile organic compounds into the porous polymer resin, thereby decreasing trapping efficiency. This problem can be eliminated by introducing a "bone dry" make-up gas line into the purge gas outlet directly before the adsorbent trap. The dry make—up gas will serve to lower the relative humidity of the purge gas (Hartman et al., 1993). The volatile organic moieties concentrated in the adsorbent trap are subsequently released by thermal desorption and transferred into a gas Chromatograph. This technique was termed the purge and trap/thermal desorption (PT&T-TD) procedure, and has been accepted as a means of isolating flavor chemicals and enriching headspace flavors prior to CC and GC/MS analysis (Boyko et al., 1978 and Wyllie et al., 1978). Investigators at Supelco, Inc. (Bellefonte, PA) evaluated carbotrap adsorbents and found that they can sorb a more wide range of airbone organic compounds than Tenax-GC or AD-2 resins. In addition, they found that carbotrap 32 adsorbents have no surface ions or active functional groups. Therefore, the entire surface is available for interaction that depends solely on dispersion (London) forces. In comparison, the other widely used adsorbents, Tenax-GC and Amberlite AD—2 resins, have localized surface charges for specific adsorbent/adsorbate interactions. Furthermore, carbotrap adsorbents are hydrophobic in nature, allowing their performance to be unaffected by humidity (Anonymous, 1986). Identification and Quantification of Trace Volatile Compounds by Gas Chromatography/mass Spectrometry Gas Chromatography/Mass Spectrometry (GC/MB) combines analytical specificity and sensitivity together for analysis of a diverse range of organic compounds, and can confirm the presence of or identify the structure of a suspected contaminant compound. The gas Chromatograph is used to separate the particular species, while the mass spectrometer is used to determine the identity of the compounds. The detector of the Mass Spectrometer can be operated in three modes: (a) total ion monitoring - used for quantification (b) total mass spectrogram - used for structure elucidation of an unknown compound (c) selected ion monitoring (SIM) - used for quantification of a selected ion 33 characteristic of the volatile There are three major functional parts of the mass spectrometer: (1) the ion source; (2) the mass analyzer; and (3) the output system. Sample molecules are ionized by the ion source, using a variety of techniques. The ions are separated and displayed as a function of their mass and relative abundance. This constitutes a pattern or profile called the mass spectrum, which usually is uniquely diagnostic of a particular molecule (Merritt and Robertson, 1982). The most commonly used ion source is the electron impact (EI) type, where the sample molecules are bombarded by an energetic electron beam. Normally, the energy of the electron beam is 70eV which is above the ionization potential for organic compounds. After ionization occurs, positive singly charged ions are produced predominantly, with a small co-existing population of double positively charged and single negatively charged ions. (Gilbert, 1984) For an organic molecule x, the production of a molecular ionlx”'occurs as follows: X+e -e (X)°++2e' The collection of ions of differing mass to charge 34 ratio (m/z) produced in the ion source of the mass spectrometer is then accelerated away as a beam by a series of focusing slits operated at different voltages. This beam then passes to the next stage of the mass spectrometer. This involves the separation of these ions and their collection. Two fundamentally different instrumental methods of separation of ions are used and these form the basis of the two major types of mass spectrometers. One is by deflection through a magnetic field (magnetic sector instruments), the other is by using a collection of four rods at radiofrequency and direct current voltages (quadrupole instruments). The quadrupole type instruments have the capacity of rapid scanning and additionally, a simplicity of operation for obtaining selected ions by rapid switching between various rod voltages. Normally, the total mass spectrogram is utilized as a confirmation tool. Each specific organic compound has a unique mass spectrum. By comparing the mass spectrogram of an unknown with standard spectra, the structure of the unknown sample can be identified. If other organic compounds are eluting from the column at the same retention time, this will be evident by comparing the mass spectrogram. By focusing on the relative abundance of a limited number of ions monitored simultaneously, rather than operating the mass spectrometer in the full scanning mode, the sensitivity increase dramatically. This is called the 35 selected ion monitoring (SIM) mode, which can be defined as "the dedicated use of a mass spectrometer to acquire and record ion currents only at certain selected mass-to-charge (m/z) values" (Gilbert, 1987). In the roles of confirmation and quantification, the selected ion monitoring (SIM) mode is most commonly employed. Quadrupole instruments are typified by good calibration stability, but SIM is normally operated at constant magnetic field with switches in the accelerating voltage to bring the desired ions into focus. In the selected ion monitoring mode of operation, specificity and quantitative analysis is obtained by monitoring one or more ion fragments in combination with the gas chromatographic retention time. Sometimes, it is necessary to monitor more than one ion fragment in each standard to establish an area ratio, and any charge in the area ratio. By using SIM the particular choice of either one ion or a number of ions in the spectrum is governed by two principles. First, to achieve maximum sensitivity, ions of a high relative abundance are desirable, as the absolute sensitivity attainable by SIM is frequently better than is likely to be necessary for a particular assay. Second, as a general rule specificity tends to improve with increasing m/z value and the ideal choice is of a high m/z, preferably of greater than m/z 200 (Gilbert, 1987). The migration of residual monomer, such as vinyl chloride, vinylidene chloride, acrylonitrile and styrene 36 from food packaging to a contained product has been a major concern as a source of food contaminant. By using the GC/MS techniques, identification and quantification of monomers have been accomplished (William and.NHles, 1975; and Lierop and Stek, 1976). The SIM mode of the mass spectrometer also provides a useful technique to identify specific contamination compounds from a complex flavor profile (Gilbert and Startin, 1981; and Gilbert and Startin, 1982). This particular application of the mass spectrometer has been used for extensive surveys of a wide range of retail foods for styrene monomer contaminates. Low resolution SIM, in the majority of cases, has proved to be quite adequate in determining the presence of specific volatiles in food products. Characteristics of Packaging Seal In selecting an adhesive system used to provide a package seal, two major factors should be considered. One is the chemical nature of the adhesive components which acts as the binder. Second is to consider the manner in which the bond is formed. In practice, adhesives are widely used in forming multilayer structures involving the combination of polymeric films or other non-polymeric polymers, as well as to seal fabricated package systems (Lazarus, 1990). In general, the primary bonding associated with adhesives include three types: (1) covalent bonds; (2) ionic 37 bonds; and (3) metallic bonding. Three general methods can be described for adhesive bond formation (Lazarus, 1990) and are summarized below: (1) Loss of carrier - This bond is formed when the carrier (water or solvent) is lost through adsorption into the substrate or by evaporation. Water—based adhesives are preferred from an environmental point of view, but its removal requires more energy. For most cases, solvent based adhesive systems are more common. However, if solvent based adhesives are employed, there is a need to consider residual solvent remaining in the packaging material or sealing area following package fabrication. (2) Loss of heat - The application of a molten adhesive, such as a hot melt, which solidifies on cooling. Such an adhesive type has application with both porous and impermeable substrates. However, this molten adhesive can not be applied to an heat-sensitive substrate. (3) Chemical reaction - A solution of the adhesive system is applied to the substrate and then 38 reacts chemically to yield a solid coating on the substrate surface. This method is commonly used for structural adhesives, where high bond strengths are needed. E£A&_§§§1 By their very nature, hot melt adhesives are thermoplastic and largely amorphous, softening gradually over a wide temperature range. This can give rise to potential problems, unless the thermal properties of the adhesive system are known and fully understood by the user. Adhesive performance can be best understood by discussing the solid to liquid change in relation to adhesive application, and the liquid to solid change in relation to the formation of a good bond (Kettleborough, 1990). The speed of bond formation on modern high-speed machinery depends on several factors such as the specific adhesive system, temperature, porosity and surface characteristics of substrates, as well as application conditions, and the time and extent of compression of the surfaces. If all factors are carefully considered, the relative high price of 100% solids hot melts may be offset by various economic factors, including operational costs and performance (Skeist and Miron, 1990) Booth (1990) defined the process of a heat seal as "to \“ 39 reactivate a layer of thermoplastic material or adhesive on an adhered substrate by heat, enabling it to form a bond with another adhered substrate, usually assisted by pressure". Heat seal adhesives are based on several different chemical compositions. Those based on ethylene — vinyl acetate (EVA) copolymers represent the majority of the market. The EVA resin emulsions have a high degree of versatility and potential to tailor the formulation of the adhesive system for specific end use applications. The EVA polymer systems offer excellent machine operation, setting and adhesion characteristics, coupled with good low temperature flexibility. as well as very good compatibility with other formulating constituents used in the manufacture of hot melts or heat seal adhesives. Other types of polymers employed as hot melt adhesives include polyethylene, polyamides, polyurethanes, amorphous polypropylene, and thermoplastic copolymers (Coggin, 1990). Hot melts are well known for their versatility and functionality. There are also certain limitations in their application which include: (i) a limited heat resistance due to the thermoplastic nature of the adhesive system. It should be noted that most hot melts used for packaging have little mechanical strength at 60%:; (ii) susceptibility to solvent and plasticizer attack; (iii) bond failure under high stress or load; and (iv) a clean substrate surface is a 40 requirement for effective bonding. Hot melts based upon ethylene-vinyl acetate (EVA) copolymers are dominant in this field, with 50-55% of the market volume. Polyolefins, such as LDPE and amorphous PP account for 30-40% of the market volume, with the remaining 5-15% balance made of "specialty" hot melts (Kuzma, 1990). QQIQ_£§AI Cold seal adhesives were first introduced in the 1960's for confectionery packaging. Since then, their usage has increased significantly, and a wide range of substrates have been evaluated (Stone, 1976 and Anonymous, 1980). The term "Cold seal" is used to describe the creation of an effective seal without the use of heat. This technique requires two surfaces of cold seal adhesives being bonded together under pressure alone. Booth (1990) defined cold seal as "the ability of a non-tacky adhesive film to bond to itself under pressure at room temperature". In the field of packaging, cold seal closures have been widely use for food products as well as for other heat sensitive products. Early applications of this sealing technique were hindered. However, by the need to use organic solvents, which were not suitable for food contact, in applying the cold seal adhesive pattern. Now, with the development of new cold seal materials, which are water- based, rubber latex stabilized with alkali (Jones, 1985), 41 the use of cold seal adhesives has found considerable utility in the packaging of confectionery products. Cold seal adhesives are normally used on flexible film webs to replace a laminated or extruded polyethylene layer as the sealing medium. This improves the output and economy in the flexible packaging industry. Cold seal adhesives are based on a mixture of natural and synthetic rubber latexes, combined with a number of lesser ingredients, such as antifoams, antifungicides, and filling agents. The natural latex confers the characteristics of self adhesion. Synthetic latexes are used to control the variable cohesive properties, to improve stability, to increase resistance to oxidation, or to improve adhesion to difficult substrates, and may also be used to reduce costs (Anonymous, 1980). The application of cold seal adhesives in the confectionery industry is recognized as revolutionary with respect to optimization of machine efficiency and reduction of running costs (Layfield, 1990). According to Levine (1976) and Jones (1985), the most important factors contributing to the dramatic growth in the use of cold seal coated materials were: (1) Elimination of heat damage to heat sensitive products, such as chocolate. There is no heat required in the bonding process, only contact pressure required. 42 (2) Less distortion and shrinkage of polypropylene packaging films. (3) Water based adhesive formulation eliminate solvent related concerns. (4) Elimination of costly temperature control equipment on the packaging line and lower power consumption. (5) Easier machine maintenance, ease of clean-up and non- sticking of the coating to the product. (6) A faster process than heat sealing, packaging equipment can run at a faster line speed. Compared to heat sealing techniques, machine speeds can be doubled with the use of a cold seal. Heat sealing is a more difficult procedure due to the short heater dwell time and the excessively high jaw temperature needed to compensate for this. Cold seals are normally based on the type of natural rubber latex and are widely categorized into two types which can be described by the nature of the dried film (Layfield,1990): "Hard" cold seals are essentially non pressure- sensitive and usually exhibit low bond strength. The nature of the dried film is less prone to blocking and stringing. These would be more likely to give good flow properties on a packing line, as a result of the low coefficient of surface friction. 43 "Soft" cold seals show a greater degree of pressure sensitivity and will seal more readily. This characteristics lead to much higher seal strengths and in some cases to cause film tear. It has a higher probability of blocking and stringing. The softer varieties of cold seals are usually more tolerant to surface contamination. However, they exhibit a slight reduction in seal strength. In recent years, the trend with respect to cold seals has been towards maximizing bond strength and obtaining 100% fiber tear on single web films, especially low density pearlized polypropylene. Some fragile products need an easy- open, peelable bond, this requirement necessitates a seal strength in the region of 100g/25mm (Layfield, 1990). In addition, to the well-established confectionery and biscuit markets, cold seal adhesives may find use in other applications, such as medical packaging, tea packaging or frozen food packaging. Also development of adhesive technology to maintain performance with new films and inks/lacquers will be critical to the industry (Stone, 1976 and Layfield, 1990). MATERIALS AND METHODS Part I: Permeability of Organic velatiles Through Packaging Film Roll Stock Packaging materials: Six commercial barrier film structures were evaluated in the present studies. All six packaging films were coated with a 1/2" wide pattern cold seal (PCS) on the side of film. The six polymer structures evaluated included: O Printed Oriented Polypropylene (OPP) - 1.7 mil O High Density Polyethylene (HDPE)- 1.7 mil O White Opaque Glassine (Glassine) - 1.7 mil O Saran Coated Oriented Polypropylene (Saran OPP) - 1.5 mil O Acrylic Coated Oriented Polypropylene (Acrylic OPP) - 1.7 mil O Metallized Polyester/Polypropylene (Met PET/OPP) - 1.5 mil All film samples were approximately 1.5 to 1.7 mil in thickness. Ethyl acetate and toluene were evaluated as the organic penetrants. Acetonitrile and methanol were used as solvents to prepare standard solution used for developing the respective calibration curves for quantification. 44 Ethyl Acetate (CH3COOC2H5): J. T. Baker Chemical Co. Analytical Reagent grade Molecular weight Density at 25°C, g/ml Boiling range Purity Toluene (CbHSCH3) : J. T. Baker Chemical Co. Analytical Reagent grade 45 (Phillipsburg, NJ) 88.11 0.894 77.2 d: 0.5°C 99.9% (Phillipsburg, NJ) Melecular weight 92.14 Density at 25°C, g/ml 0.893 Boiling range 110.4 i 0.2%: Purity 99.9% Acetonitrile (CH3CN): EM Industries, Inc. (Gibbstown, NJ) HPLC grade Molecular weight 41.05 Density at 25°C, g/ml 0.78 Boiling range 82.0 i 0.1°C Purity 99.8% Methanol (CH3OH): Sigma Chemical Co. (St. Louis, MO) HPLC grade Molecular weight 32.04 Density at 25°C, g/ml 0.791 Boiling range 64.7 i 0.1°C Purity 99.9% Permeation test: Permeability studies were carried out by an isostatic procedure utilizing the MAS 2 (MAS Technologies, Inc., 000” Permeability Test System Zumbrota, MN). The system allows for the continuous collection and measurement of the 46 permeation rate of an organic vapor or gas through a polymer membrane, from the initial time zero to steady state conditions. The system incorporates a flame ionization detector (FID), precise temperature and flow rate control and is interfaced to an IBM 486SX computer system. The computer controls many of the parameters that are established during set-up. It will activate and deactivate the gas flow direction, cell opening/closing, as well as display data on the screen, while recording all pertinent instrument parameters. Finally, all permeation data can be stored in the computer hard drive or on a floppy disc, and LOTUS 1-2-3 used to recall the permeation data to calculate and print out the respective mass transfer parameters and the transmission rate profile curve. For each test film, a sample approximately 6" x 6%" was cut, mounted on a paperboard film holder with tape, and the mounted film sample placed in the permeability cell. The film was mounted with the pattern cold seal facing away from the high permeant concentration stream. The permeability studies were carried out at three temperatures, and for each temperature, three vapor activity levels were evaluated. A constant concentration of permeant vapor for the "high concentration" cell chamber was produced by bubbling nitrogen through the liquid permeant. The liquid permeant is contained in a vapor generator consisting of a Kontes Fritted.Nfidget Bubbler, $24/40, 25 ml (Fisher Scientific, 47 Pittsburgh, PA). The organic vapor stream was then mixed with another source of pure nitrogen. Rotameters were used to provide an indication of settings required for the desired vapor activities. The gas flows to the rotameters were regulated by NUpro "M? series needle valves. The vapor generator system was maintained at a constant temperature of 23 1 1°C. Before actual testing was conducted, flowmeter settings were determined to provide vapor activity (a) values of approximately 0.067, 0.21, and 0.41, respectively. Because of the wide variation in the saturation vapor pressure of toluene and ethyl acetate at the respective temperatures of test, the partial pressure gradient for the permeability experiments was expressed as vapor activity. This allowed for barrier performance to be compared at a standard driving force for the two permeants evaluated. Vapor activity values were calculated by dividing the experimentally determined vapor pressure by the saturated vapor pressure for the respective permeants (Equation 16). P Vapor activity (a) = (16) P 0 where: P = partial vapor pressure P°== saturated vapor pressure All vapor pressure values were determined at 23 i 1°C. A sampling port was installed between the dispensing 48 manifold and the test cell to provide an accurate measure of the permeant vapor concentration or activity. To determine the specific vapor concentration, a 100 pl sample was removed from the sampling port with a Hamilton 500 pL, 1750SN gas tight syringe (Hamilton Co., Reno, Nevada), and the sample injected directly into the gas chromatograph. Analysis of permeant concentration was carried out by a gas Chromatograph (Model 5890A, Hewlett Packard, Avondale, PA) equipped with flame ionization detector and interfaced to a Hewlett-Packard integrator (Model 3395, Avondale, PA). A setting of 1 minute purge "ON" was utilized for all analyses. The gas chromatography conditions are presented as following: GC condition (Model 5890): Column: Supelcowax 10 (Supelco Inc., Bellefonte, PA) Fused silica capillary column Polar bounded stationary phase 60 m length x 0.32 mm I.D. Carrier gas: Helium at 1.5 ml/min Range: 2 Attenuation: 0 Zero: dependant on the GC signal after ignition Temperature cyele fer ethyl ecegete: Injection temperature 220 %: Initial temperature 40 %: Initial time 1 min Temperature rate 5 °C/min Final temperature 200 °C Final time 10 min Detector temperature 250 %: 49 Temperetere eyele fer telgene: Injection temperature 220 %: Initial temperature 60 %: Initial time 1 min Temperature rate 57°C/min Final temperature 200 °C Final time 10 min Detector temperature 250 %: Integrator setting (Model 3395): Zero: 0, 0.021 Attenuation: 6 Chart speed: 0.5 cm/min Peak width: 0.04 Threshold: 0 Area reject: 0 Retention time for ethyl acetate: 5.4 minutes Retention time for toluene: 7.85 minutes In both cases, a standard curve of response vs permeant quantity was constructed from standard solution of known concentration. Calibration standard solutions were prepared by dissolution of known quantities of ethyl acetate in acetonitrile and toluene in methanol. A 1 pl sample was removed from standard solution and injected into HP 5890 Gas Chromatograph. The standard calibration curves for ethyl acetate and toluene are shown in Appendix I. Once the operational parameters of gas flow rates, temperature and signal base line have attained steady state, the MAS 2000 Permeability test system was calibrated as follows: A 100 uL sample of permeant vapor of known concentration was removed from the sampling port interfaced 50 to the dispensing manifold of the vapor generator system and the gaseous sample conveyed directly to the flame ionization detector of the MAS 2000 test system via the injection port incorporated on the instrument chassis. The detector response (picoamps) is then related to the quantity of permeant injected to determine a calibration factor. All instrument operational parameters were as recommended by the manufacture in the operating manual. After calibration for the permeant, by striking the "Test" key, the internal valve will activate and the organic vapor will enter the test cell and start the permeation test of the film. The specific temperatures selected for the permeability studies were based on the structures of the respective barrier films. The temperatures selected for the respective test films are presented below: Ethyl agegate Toluene OPP : 30°C, 40°C, 50°C 30°C, 40°C, 50°C HDPE: 30°C, 40°C, 50°C 3 0°C , 40°C, 50°C Glassine: 23°C, 30°C, 40°C 30°C , 40°C, 50°C Saran OPP: 40°C, 50°C, 60°C 40°C, 50°C, 60°C Acrylic OPP: 50°C, 60°C, 70°C 50°C, 60°C, 70°C Met PET/OPP: 50°C, 60°C, 70°C 50°C, 60°C, 70°C 51 Part II: Permeability of Toluene Through a Packaged Confectionery Product with A Cold Seal Closure Samples of a chocolate type confectionery product packaged in the respective test package structures were provided by a commercial supplier. For this study, toluene was selected as the organic permeant and only one vapor activity was evaluated for the packaged product permeation studies. The organic vapor activity selected was based on the lower bound activity level which could be adjusted with the vapor generator system designed. The toluene level in the four test chambers was determined every third day to monitor the consistency of the vapor generator system. To determine the specific toluene vapor concentration, a 100 pl sample was removed from the sampling port with a Hamilton 500 pL, 17508N gas tight syringe (Hamilton Co., Reno, Nevada), and injected into the gas chromatograph. The analytical procedure for toluene analysis was identical to that previously described (see page 48). All confectionery products (coated with milk chocolate) were wrapped individually with the six barrier membranes which were evaluated in Part I of the study. Sealing of the respective barrier structures was carried out with the same pattern cold seal adhesive and sealing conditions. The confectionery products were manufactured in the same dimension of 11 cm x 3 cm x 2 cm (length x width x height). As an initial postulate, it was assumed that the 52 package system was hermetically sealed, and the permeated toluene vapor was sorbed totally by the chocolate layer. Permeation chamber design: A.multi-chamber test system was designed and fabricated to allow exposure of the packaged product to a fixed and constant toluene vapor concentration at constant temperature. The test chambers consisted of four (4) stainless steel desiccant chambers of dimension, 30.5 cm x 28 cm x 31 cm (length x width x height). Each chamber was equipped with four open metal shelfs to allow multiple sample exposure under identical conditions. Each test chamber was also equipped with a gas inlet and outlet port, and a gas sampling port to allow continuous monitoring of the toluene vapor concentration over the course of the study. A schematic diagram of the test system utilized is presented in Figure 1. As shown, the test system was designed to evaluate multiple product/package samples concurrently. All studies were carried out at 23 1 1°C. A constant concentration of toluene vapor to be flowed continually through the respective exposure test chambers was produced by bubbling nitrogen through the liquid permeant. This is carried out by assembling a vapor generator consisting of a gas washing bottle, with a fitted dispersion tube, containing the organic liquid. To obtain a lower vapor concentration, the permeant vapor stream is 53 .uosooum pmmmxomm uo auaaflommEumm on» mcfinfleuouoo MOM Seaman Down 00 Emummflo oflumEooom .H ousmam 2mm; 25.3% H... 1290:0080“; n0> .8820 05:36:00 ”00 105.529 ”m o>_m> 0600: ”>0 56.302 ”mm ... . .l ‘ 0> a m 00 mo 00 . mo ‘ _ . a _ _ >0 >0 >0 >0 > m m m 0 O O O O 4 O _ _ _ _ m V 502:2 54 mixed with another stream of pure carrier gas (nitrogen). Flow meter settings were determined to provide the desired vapor activity value of approximately a = 0.003. As shown in Figure 1, flow meters were used to provide a continuous indication that a constant rate of flow of permeant vapor was maintained to the individual test chambers. Gas flow was regulated with NU PRO needle valves, Type B-25G. The vapor generator system was kept at a constant temperature of 23 1 1°C. The penetrant vapor concentrations are expressed throughout in ppm (mass/volume) vapor in nitrogen, where 1 ppm equals 1 pg toluene per cm; of gas mixture at 1 atm.(23%3)..A photograph of the vapor generator system and test chambers is shown in Figure 2. Storage condition: All studies were carried out at 23 1 1°C. As a function of time, a series of individual packaged confectionery products were removed from the test chambers and the product analyzed for the quantity of volatile organic permeant (toluene) sorbed by the chocolate. The sampling frequency varied according to the barrier properties of the packaging structures. The six barrier membranes can be separated into two groups. Samples from group I, which included HDPE, Glassine, and OPP based structure, were evaluated weekly for the first month, then evaluated every two weeks for a two month period. Samples 55 .0056000 oommxomm H00 mnooemno umwu huflefiocwEme pom Emumhm ecumuocom Homw> uo zmmumoDonm .N whomfim 56 from group II, which included Saran OPP, acrylic OPP, and the Metallized PET/OPP based structures, were evaluated after 11 and 19 days, then evaluated every third week for a period of three months. Toluene quantification apparatus for trapping of volatiles: A dynamic gas purge and trap system was designed for dynamic headspace sampling of the dry product. Six 250 ml Erlenmeyer flasks were modified to 29/42 standard taper male joints to which the dispersion tube assembly of a gas washing bottle could be fitted (Stopper assemblies for Corning 31770 gas washing bottles, Fisher Scientific, Pittsburgh, PA). Modification of the dispersion tube assembly of the gas washing bottles and the erlenmeyer flasks were performed by the Chemistry Department Glass Blowing Shop at MBU. A schematic diagram of the dynamic purge and trap system is shown in Figure 3. As shown, a cylinder of compressed nitrogen was interfaced to a dispersing manifold which consisted of three flow meters and needle valves, all connections were through 1/8" O.D. copper tubing and swagelok fitting. Flow meters were used to provide a continuous indication that a constant rate of flow of nitrogen was maintained to the individual purge and trap cells. Gas flow was regulated with NU PRO needle valve, type B-25G. The trapping system was designed to ensure that the .Emum>m mono pom omwom 02856 no Emummflo oaumaonom .m whom?” 57 509:2 339.3%: N... 81.08.29 ”I £3 19m; nm>> 30605061 55:09 ”mm on. 00 on. m m E O O O mm 58 sample was continuously flushed with nitrogen gas and the desorbed volatiles conveyed to the trapping tube attached. The sorption trap was connected to the exit port of the dispersion head via swagelok adopters. The dispersion head exit port of 8 mm O.D. glass tubing was connected by a 5/16" swagelok nut and a series of reducing adopters to a 1/4" male swagelok fitting. The sorption trap was mounted to the dispersion head with a 1/4" thumb wheel swagelok fitting (Supelco Inc., Bellefonte, PA) for easy removal and can figuring to both glass and metal desorption traps. By changing the ferrule materials to teflon or graphite, the thumb-wheel swagelok fitting can be employed to fit both metal (Scientific Instrument Services, NJ) or glass (Supelco Inc, Bellefonte, PA) thermal desorption traps. Figures 4 and 5 show the trapping cell and the whole purge and trap system. The glass lined stainless steel GLT thermal desorption tubes (4 mm I.D. x 100 mm, threaded on both ends) used in the present study were prepacked by Scientific Instrument Services, Inc. (Ringoes, NJ). The trapping tubes were packed with three layers of adsorbent resins (Supelco, PA) - 20/40 Carbon C, 20/40 Carbon B, and 60-80 Carbosieve S-III. Before each trapping procedure, the sorption traps were conditioned overnight at 330°C with a nitrogen flow rate of 20 ml/min. 59 Figure 4. The apparatus of trapping cell. 60 .Eoumhm ammo pom wmuom ueEmc%o mo smoumoDonm .m ousmflm 61 Flavor Analysis Procedure: r r r To establish optimum procedure conditions for the analysis of sorbed toluene in the product samples, several temperature/time combinations were evaluated in a series of preliminary runs. A water bath temperature of 40°C and a purge time of 20 hours provided chromatograms which showed good peak resolution. These conditions were therefore selected for analysis of the sorbed toluene. To determine if a water bath temperature of 40°C and a purge time of 20 hours would provide quantitative desorption of sorbed toluene from the confectionery product samples, the procedure was run twice on the same sample, which was known to contain sorbed toluene. A comparison of the chromatograms obtained from the first and second runs showed no detectable levels of toluene present when the purge and trap procedure was repeated. It was therefore assumed that with an water bath temperature of 40°C and a purge time of 20 hours, more than 90% of any sorbed toluene would be desorbed from the sample during the initial heating period. Individual packaged confectionery products were removed from the permeation test chambers as a function of storage time. The package was opened, the product removed and a sample (11 cm) cut from the end of the chocolate layer. Approximate 0.3 grams of the chocolate layer weighed 62 accurately were placed in the trapping flask. The desorption cells were assembled and the traps connected. The desorption cells were then placed in a water bath maintained at 40%:. Nitrogen gas was continually flowed through the desorption cells at a flow rate of approximately 25 cubic centimeters per minute for 20 hours. The water bath temperature was maintained at 40°C throughout the purge period. hr r h - r ri 1 i The trapped volatile compounds were directly desorbed and injected into a gas chromatograph — mass spectrometer (GC/MS) system using a Short Path Thermal Desorption unit (Scientific Instrument Services, Ringoes, NJ). Before carrying out the desorption procedure, the sorption trap was screwed on a 30 mm long side hole syringe needle (Scientific Instrument Services, Ringoes, NJ), and attached to the thermal desorption unit. The complete injection and desorption mechanism is shown in Figure 6. The trapped volatile compounds were heated and desorbed from the carbon adsorbent by the desorption heater block into the injection port of the gas chromatograph (GC) (Figure 7). The sorption trap was heated at 290°C with helium gas flushing (10 ml/min) for 6 minutes. The GC/MS analysis was simultaneously started with the beginning of the desorption process. The volatile compounds were separated and identified using a HP 5890 gas 63 .Abz .mmomcflm .moofi>umm ucmEdLumcH encasemeomv can: conuduommo fineness 20mm uuoam one H00 Emacmoooa cofiumuomoo one ouoHQEoo mo ofiumEonom .m ousmfim gramme a Leo: 8Lo_gseu =S_Lou_=_ u=_Loo_=_ =o_0_me; was; I 4 V . ...... 1..., O «A. \_ 0‘) 64 ,————— Carrier Gas Flow / Desorption Tube ‘ Desorption Tube /Heater Blocks GC Oven Needle (Transfer Line) 60 Injection Port Cryo Trap Sample ——-> on Column Capillary Temperature Column Program for GO Oven Analysis To MS or CO Detector A Figure 7. Short Path Thermal Desorption Unit operation with the gas chromatograph. ‘ 65 chromatograph (Hewlett Packard, Avondale, PA) interfaced to a HP 5970 mass selective detector (MSD) mass spectrometer. This GC/MS system was equipped with a HP 59970 Chemstation Data System. The GC/MS analysis was performed at the Mass Spectrometry Laboratory in the Department of Biochemistry at Nfichigan State University. A SPB-5 non-polar fused silica capillary column (60 m x 0.32 mm I.D., 1 pm film thickness, Supelco Inc., Bellefonte, PA) was used to separate the desorbed compounds. The GC temperature program was set as following: the initial temperature was 35%: for 6 minutes, and the temperature was increased at 5°C/min to 80°C then increased at 15°C/min to a final temperature of 250°C maintained for 10 minutes. Helium was the carrier gas with a flow rate of 10 ml/min. The injection port temperature of the GC and the transfer line temperature between the GC and the MS were set at 250%:. The mass spectrometer was operated in the electron impact mode with an electron energy of 70eV and an ion source temperature of 250°C. The separated volatile compounds were then introduced into the mass spectrometer directly from the capillary column. A selective ion mode (SIM) with 4.7 cycles per second, and three ion masses (m/z) 65, 91, and 92 were selected. The total ion chromatograph was used for quantification. However, by using the SIM mode of operation, the total ion spectrum screened out or eliminates all 66 eluting compounds except for volatile compounds containing the three selected ion masses (m/z = 65, 91 and 92). For identification of the toluene, elution time is compared with the elution time of a toluene standard solution (8.8 minutes), or by comparison of the sample mass spectrum with the standard toluene mass spectrum (the ratio of ion mass 91/92 is about 1.75/1). Toluene quantification was carried out by integrating the peak area under the total ion chromatogram (i.e. area of m/z = 65, 91 and 92 ion peaks). The standard calibration curve for toluene was constructed using thermal desorption - GC/MS system, operated under the same conditions as previously described. Toluene standard solutions of 4, 10, 20, and 40 ppm (wt/v) were prepared by using methanol as the solvent. A 1.2 pl sample of the respective standard solutions was injected into the sorption trap, then the trap was heated and the toluene vapor directly injected into the GC/MS. The calibration curve of toluene vapor by the thermal desorption - GC/MS procedure is presented on Appendix II. The standard toluene mass spectrum is showing on Appendix III. The total ion chromatograms obtained by GC/MS analysis were recorded, stored and analyzed by using a HP Chemstation data system. 67 Part III: Leak Testing The package leak testing was performed by a Nikka Packaging Leak Tester (Mbdel PLT - 3501) (Nikka Densok U.S.A., Inc., Lakewood, CO). The leak tester was connected to a test sample cell of 14.5 x 4.4 x 2.5 cm?. This test cell was designed to handle the packaged product system evaluated in the present study. This leak tester is based on measuring the different vacuum pressure change between the test cell and the packaged product itself. The operational parameters were set up as the instrument manufacture recommended. The parameters set up to be used in this study are listed below: BAT 1 = 99.9 STAB = 1.5 seconds Measure = 2 seconds A Comp = 2000 mm-Hg) HI = 999 min-H20 L0 = 0 mm-HZO where: STAB = stabilized time after the test cell was pulled out air Measure = actual measuring time for observing the vacuum pressure change A Comp = actual vacuum pressure to apply on the packaged product HI = the highest measuring pressure to be set up as reject during packaging line L0 = the lowest measuring pressure 68 Before initiating the package system permeability studies, 20 package/product samples for each packaging material were randomly selected and their leak test values measured. Leak test values for packaged samples removed from the permeation test chambers were determined directly upon their removal and prior to opening. RESULTS AND DISCUSSION Part I: Permeation Study of Packaging Films Eff f V r A ivi T r r n E h l W Permeance values determined at ethyl acetate vapor activity levels of a = 0.095, 0.21 and 0.41 for the respective barrier structures are summarized in Tables 3 - 7. For the metallized polyethylene terephthalate/OPP laminate, no measurable rate of diffusion was detected, following continuous testing for 44 hours at 70°C and a vapor activity of a = 0.41. To better illustrate the effect of vapor activity on the transmission rates for the respective test films, the results are presented graphically in Figures 8 - 11, where log permeance (P) is plotted as a function of vapor activity, at constant temperature. From Figures 8 - 11, it becomes evident that at the vapor activity levels evaluated, that the effect of penetrant concentration on the permeance values is minimized. From a least squares fit, the following expressions were derived to describe the correlation between the permeance constant and penetrant concentration. For the oriented polypropylene structure, the 69 70 Table 3. Permeance constant of ethyl acetate through printed oriented polypropylene (OPP) as a function of vapor activity and temperature”’. Kg Vapor Temperature P( ) x 1042 Activity“) (9C) n@.sec.Pa log(P) 30 0.62 -12.2134 0-095 40 1.72 -11.7696 50 4.37 -11.3610 30 0.72 -12.1442 0.21 40 1.94 -11.7134 50 4.42 -11.3546 30 0.91 -12.0409 0.41 40 2.08 -11.6831 50 4.77 -11.3214 “ ‘”. The results reported are the average of duplicate analyses. (0. Vapor activity values were determined at ambient temperature (249C). 71 Table 4. Permeance constant of ethyl acetate through high density polyethylene (HDPE) as a function of vapor activity and temperature". K9 Vapor Temperature P( ) x 10’11 Activity‘z’ (°C) m2 . sec . Pa log (P) Ella—=1 = SI 30 1.02 -10.9918 0.095 40 1.72 -10.7668 50 2.66 -10.5760 30 ‘ 2.36 -10.6268 0.21 40 2.80 -10.5531 50 3.37 -10.4720 30 2.89 -10.5386 0.41 40 3.76 -10.4256 50 4.81 -10.3185 ‘”. The results reported are the average of duplicate analyses. (3. Vapor activity values were determined at ambient temperature (24%3. Table 5. Permeance constant of ethyl acetate through Glassine as a function of vapor activity and temperature“). K9 Vapor Temperature P( ) x 10'12 Activity‘z’ (°C) . sec . Pa log(P) I. 23 1.17 -11 0.095 30 1.96 -11 40 4.23 ~11 23 2.28 -11 0.21 30 2.95 -11 40 5.20 -11 ‘”. The results reported are the average of duplicate analyses. ‘3. Vapor activity values were determined at ambient temperature (249C). 73 Table 6. Permeance constant of ethyl acetate through Saran coated oriented polypropylene as a function of vapor activity and temperature“’. K9 Vapor Temperature P( ) x 1043 Activity‘z’ (°C) m2 . sec . Pa log (P) L n 40 0.99 -13.0033 0.095 50 5.01 -12.3006 60 15.07 -11.8220 40 1.61 -12.7950 0.21 50 5.23 -12.2846 60 14.90 -11.8271 40 2.83 -12.5481 0.41 50 7.74 -12.1121 ' 60 26.48 -11.5772 n (n. The results reported are the average of duplicate analyses. (0. Vapor activity values were determined at ambient temperature (249C). 74 Table 7. Permeance constant of ethyl acetate through Acrylic coated oriented polypropylene as a function of vapor activity and temperature”). Kg Vapor Temperature P( ) x 10'12 Activity‘z’ (°C) m2 . sec . Pa log (P) 50 0.25 -12.6033 0.095 60 1.03 -11.9856 70 2.65 -11.5769 50 . 0.27 -12.5628 0.21 60 1.16 -11.9354 70 3.60 -11.4434 50 0.34 -12.4728 0.41 60 1.20 -11.9226 70 3.29 -11.4824 ‘”. The results reported are the average of duplicate analyses. (0. Vapor activity values were determined at ambient temperature (249C). 32.30323 3223 no.5... no 35252. of :o 32:: 33> 323m :56 .6 Soto of. .o 2.3:. 33:: 33> 75 ov.o and 0N6 o_..o _ _ _ 1 H _ Om.N_... Ell mm NF- 1 0| i cod—3 w (.m S ) aili‘inlillilf 3:. x. m Tul w. l 3:. a 3 1 . it. u i mmé... .d r W 0.3 o . 000* I TI 00 2.. 0.3 a e mud? 76 237213321 2133.. .12.. 3 3113:1121 2: 3 53:: 33> 323m .33 3 323 2:. .a 6:31“. 3.33.4 33> ovd 00.0 0N6 o...o _ 1 T t 1 1 Om._.—.. 6.8 o ‘ l 3:. I 0.3. a r m a ) 0.3 . . Si. N (M a 2.2. w s I . S . . l 3. 2. u a ..I NC . U l 3.2. 00.2.. 77 n—n—O bounce 00.0w *0 oocuthon 0.: co >:>_uoa hoad> o~u~ooa :guo —o gowto 313314 33> 0v.0 00.0 0N0 01.0 1 1 1 1 1 1 00.3.. 00.01.. 00.9. 00.01.. 00.0w. 00.1.... 00.1.... 2: .2 2.11111 (aa°009'z.wlbx) (aloe: 78 n:10 .1230 0:32 .0 3.32521 2: .3 1:1an 33> 223m 12$ 1o 82? of. .2 2:01“. 1:13.: 33> 0¢.o om.o 0N0 o...o 1 1 1 1 1 1 ON.N_.. mw mw\\\\\\\\\\\@\\\\\I 3.9. I o 1 W; 1C I: OWN... l\ 1 Mn. N I 9...... w. a f . I 00.2.. .0 d I U 0.8 a I 2.11. 0.3 a - om._._.. 79 relationship between the permeance value (P) and ethyl acetate vapor activity (a) was found to be: log(p),3o.c, = -12.263 + 0.5442(a) (R2 = 1.00) 109(9),“..0 = -11.734 + 0.2609(a) (R2 = 0.90) log(p),50.c, = -11.377 + 0.1302(a) (R2 = 0.95) For the high density polyethylene structure, the relationship between the permeance value (P) and ethyl acetate vapor activity (a) was found to be: 109(9),,0.” = -11.035 + 1.3275(a) (R2 = 0.78) 109(9),,mc, = —10.828 + 1.0333(a) (R2 = 0.91) log(P)(50.c) = -10.649 + 0.8118(a) (R2 = 1.00) For the Saran coated oriented polypropylene structure, the relationship between the permeance value (P) and ethyl acetate vapor activity (a) was found to be: log(P)“0.,c) = -13.1209 + 1.4214(a) (R2 = 0.99) log(P)(50.c) = -12.3821 + 0.6279(a) (R2 = 0.92) log(P)(6o.C) = -11.9399 + 0.8300(a) (R2 = 0.86) For the Acrylic coated oriented polypropylene structure, the relationship between the permeance value (P) and ethyl acetate vapor activity (a) was found to be: -12.6460 + 0.4182(a) (R2 = 1.00) 109 (P) (50°c) log(P)(6o.c) —11.9919 + 0.1847(a) (R2 0.86) From the permeance values summarized in Tables 3 to 7, it becomes evident that the permeance (P) is highly 80 dependent on the temperature, at a constant permeant concentration level. This is illustrated in Figures 12 - 16, where permeance is plotted as a function of temperature [T'1 (°K)] for studies carried out on oriented polypropylene, high density polyethylene, glassine, Saran coated oriented polypropylene, and Acrylic coated polypropylene, respectively. As can be seen, the temperature dependency of the transport process associated with the respective barrier membranes, over the temperature range studied, follows well the Arrhenius relationship. From the slopes of the Arrhenius plots, the activation energy for the permeation process (Ep) was determined for the respective film samples, as a function of vapor activity. The determined activation energy values are summarized in Table 8. From the general Arrhenius equation, Ep P = P; exp (————) (17) RT Ep 1 log(P) = ( - —————— )*( ) + log(Po) (18) 2.3 R T Ep Slope = —————— (19) 2.3 R the following expressions were derived by using a least squares fit and applying experimental data. 81 22222.8 03.31:. 1: 2203 .23 3 3.32521 2: 1o 3112:1212. 2322120... .01 2.3101 c: 1.: 0000.0 0000.0 1.0000 _ r F 1 OM.N_.. E T I 0120... | 0 1 m... H )I 00.3. \I ”I I a ... w m I 2.11. 1... g .I O m. 3.0 a a o . . d 3.0 .. a I cm 2 My 000.0 I a @ r 00._._.. 82 000150200 1:12.00 .1010 E 2230 72:0 00 3:000:00 2: .0 300210000 01220qu... .0— 20201 31.1 :2 200.0 2.00.0 58.0 1 1 1 OO._._.. 3... . a o - 3.0 u 0 I .I 00.0... I 11.8.0 . .1 0 m 8.2. a 2.2. m w. 8.2. 1,. 9 O 8.2. .a d ray 1 3.2. . 8.2. 83 211320 111 3330 12:0 .0 0230:1100 01: .0 0212:1002. 013200111: .1: 2001.1 31.1 :1 «000.0 0000.0 0000.0 «000.0 1 1 1 1 1 1 1 OO.N—.. .1 00.3. I 00.3. 05.3. 00.3. 00.3. 00.3. (9d‘009'3.ull°)l) (d)5°I as . 1. 1. 80.0 .. 1. a 00.3. 00.3. 84 0.10 02000 110100 0.2000 12.0 00 0000001100 01: .0 00110000000 01301001110... .03 01001.1 31.1 :1 «80.0 1000.0 800.0 1 _ 1 COIVFu I 00.2. - m w 00.2. H ..um a 21.2. W a. ... 00.2. u no W 8.11. 00.3. 855 000 00.000 0:30.. :1 0.0.000 ...—1.0 .0 000005100 011. .0 0000000000 0.0.0.0050... .0. c... .... 1000.0 0000.0 0000.0 1 1 0:... .1 00.03. I. 00.01.. I 0...“... .| 00.3. I 05.. ....0 . 11 o / 10.0 . 11 0 // 00.... 000.0 . 11 0 00.3. 0501.1 (Bd'Ofl'ZJflIM) (0)60) 86 Table 8. Activation energy values for the permeation of ethyl acetate through polymer membranes. E(kcal/mole) Polymer Mbmbranes a = 0.095 a = 0.21 a = 0.41 OPP 17.7 14.6 14.5 HDPE 9.3 4.4 4.9 Glassine 14 .6 9 .6 N/A‘“ Saran OPP 24.9 23.0 24.5 Acrylic OPP 25.9 27.9 25.0 Metallized PET/OPP N/A‘“ N/A‘” N/A‘” ‘”. N/A denotes not available 87 P r n log(P)(a=0.095) = 1.8571 - 4262.0 x (1/T) (R2 = 1.00) log(P)(a=o.21) = 0.8967 - 3948.2 x (l/T) (122 = 1.00) log(P)(a=0M) = - 0.1696 - 3597.6 x (l/T) (R2 = 1.00) High Density Polyethylene: log(P)(a=o_o95) = - 4.1255 - 2079.0 x (l/T) (R2 = 1.00) log(P)(a=0_21) = - 8.0745 - 7739.0 x (l/T) (R2 = 1.00) log(p),a=0_,1, -1- - 6.9068 - 1100.2 x (l/T) (R2 = 1.00) Glassine: log(P),a=o_095) = - 1.2647 - 3161.4 x (l/T) (R2 = 1.00) log(P),a=0-21) = - 4.6129 - 2088.9 x (l/T) (R2 = 0.99) ar n e orien ed Pol ro lene: log(P)(a=0.095) = 5.9344 - 5906.3 x (1/T) (R2 = 0.99) 109(P),a=0_21, = 2.7010 - 4839.8 x (l/T) (R2 = 1.00) log(P)(a=0.,.1) = 2.9711 - 4854.9 x (l/T) (R?- = 1.00) Aprylic coated oriented Polypropylene: log(P)(a=0_095) = 3.3406 - 5132.0 x (l/T) (R2 = 0.99) log(P),a=0-21) = 4.8100 - 5596.9 x (l/T) (R2 = 0.99) log(p)(a=o_,1, = 2.8966 — 4952.0 x (l/T) (R2 = 1.00) if f V r A ivi T r ur on T 1 10100011112 Determined permeance values for toluene activity levels of a = 0.067, 0.22 and 0.44 for the respective barrier structures are summarized in Tables 9 — 13. There was no 88 Table 9. Permeance constant of toluene through printed oriented polypropylene (OPP) as a function of vapor activity and temperature“’. L =' K9 Vapor Temperature P( ) x 10'11 Activity”) (°C) In2 . sec . Pa log (P) 30 0.49 -11.3078 w 0-067 40 1.14 -10.9445 50 1.77 -10.7512 30 4 0.75 -11.1242 0.22 40 1.44 -10.8408 50 2.67 -10.5739 30 1.37 -10.8648 0.44 40 2.04 -10.6905 50 3.05 -10.5157 ‘”. The results reported are the average of duplicate analyses. (3. Vapor activity values were determined at ambient temperature (24°C) . 89 Table 10. Permeance constant of toluene through High Density Polyethylene (HDPE) as a function of vapor activity and temperature“). K9 Vapor Temperature P( ) x 10‘11 Activity“) (PC) n@.sec.Pa log(P) E F 30 2.14 -10.6704 0-067 40 3.36 -10.4742 50 5.65 -10.2479 30 . 3.99 -10.3989 0.22 40 5.49 -10.2603 50 8.00 -10.0970 30 7.69 -10.1144 0.44 40 8.56 -10.0675 50 9.20 -10.0362 ‘”. The results reported are the average of duplicate analyses. ‘3. Vapor activity values were determined at ambient temperature (249C). 90 Table 11. Permeance constant of toluene through Glassine as a function of vapor activity and temperature“). Kg Vapor Temperature P( ) x 10'12 Activity“) (°C) In2 . sec . Pa log (P) E LI 30 2.83 -11.5488 0-067 40 3.86 -11.4134 50 5.08 -11.2943 30 3.50 -11.4557 0.22 40 5.27 -11.2785 50 6.60 -11.1802 30 4.09 -11.3882 0.44 40 5.52 -11.2581 50 6.80 -11.1677 “ ‘”. The results reported are the average of duplicate analyses. 3. Vapor activity values were determined at ambient temperature (24°C) . Table 12. Permeance constant of toluene through Saran coated oriented polypropylene (Saran OPP) as a function of vapor activity and temperature“’. K9 Vapor Temperature ) x 10'13 Activity“) (°C) . sec . Pa log (P) F: 40 0.97 -13.0121 0-067 50 5.85 -12.2329 n 60 21.90 -11.6601 40 2.55 -12.5928 0.22 50 9.73 -12.0121 60 25.00 -11.6026 40 3.12 -12.5053 0.44 50 12.80 -11.8943 60 28.90 -11.5393 I __ ‘”. The results reported are the average of duplicate analyses. (0. Vapor activity values were determined at ambient temperature (24°C) . Table 13. Permeance constant of toluene through Acrylic coated polypropylene as a function of vapor activity and temperature“’. ‘____‘-—-—_—T Vapor Temperature ) x 1043 Activity“) (°C) .sec .Pa log (P) 50 ‘3’ - 0 .067 60 (3) _ 7o ‘3’ - 50 (3) _ 0 .22 60 ,3, __ 7O (3) _ 50 .37 -10.8648 0.44 60 .04 -10.6905 70 .05 ~10.5157 ‘”. The results reported are the average of duplicate analyses. ‘3. Vapor activity values were determined at ambient temperature (24%3 (n. Without detectable response after 44 hours test. 93 measurable rate of diffusion for the polyethylene terephthalate/OPP laminate structure following continuous testing for 44 hours at 70°C and a vapor activity of 0.44. Figures 17 - 20 illustrate the effect of toluene vapor activity on the transmission rates for the respective test films, where log permeance (P) is plotted as a function of vapor activity, at constant temperature. The following expressions were derived to describe the correlation between the permeance constant and penetrant concentration, using a least squares fit equation. For the oriented polypropylene (OPP) structure, the relationship between the permeance value (P) and toluene vapor activity (a) was found to be: log(P)(3o.c) = -11.3866 + 1.1871(a) (R2 = 1.00) 109(9),,0.c, = -1o.9903 + 0.6811(a) (R2 = 1.00) log(P)(50.c) = -10.7604 + 0.6057(a) (R2 = 0.86) For the high density polyethylene (HDPE) structure, the relationship between the permeance value (P) and toluene vapor activity (a) was found to be: log(P)(30.c) = -10.7524 + 1.4768(a) (R2 = 0.99) log(P)“o.c) = -10.5279 + 1.0754(a) (R2 = 0.98) 109(9),”..0 = -10.2596 + 0.5472(a) (R2 = 0.89) For the Glassine structure, the relationship between the permeance value (P) and toluene vapor activity (a) was found to be: 94: 0110;020:011 00.110110 00.11110 .0 00000111100 0... .10 1:13.00 1000.. 011020. .0 .00..0 011.1. .2 01:01.1 >._>_.0< .000> 00.0 00.0 03.0 00.0 1 1 1 1 1 1 1 0¢._._.. 00.3. 00.3. 03.3. 00.3. 00.03. 00.0... 00.0... 00.03. 00.0... 00.0... (30'003'Z.Wl5)|) (d)5°| IIITIITTTIITT‘IIIIT 95 1011011. 011211110021 E0000 1.0111 .0 0011001500 01: 00 >._>_.00 10001, 00020. .0 .00..0 01:. .0. 0:101". 2.3.0.01 .000> 10.0 2.0 11.0 10.0 r 1 1 1 1 1 1 1 ON.O_.. 0.00 o w . . 0.00 1. I 00 2 I a O 0.00 . I 00.2. w 0I 1.... a I 00.2. 0 I N 1I 00.2. N. I Y r F 00.2. m .l d 0 (Br , 1. I 2.2- 00.03- 596 00100010 .0 001.001.1100 01.. .10 11.13.00 .0001. 01100.0. .0 .00..0 01... .0. 0:10.". 0.1.11.3. 33> 1.0.0 .00 :0 00.0 1 1 1 1 1 1 _ 1 °C.—.ru 00.... m m 00.... m w. z 00.... ... 0 0 ”d 00.... u 0...... $97 0.101.110.0201. 00.0013 00.000 110.00 .0 00000111101. 01.. .10 1:13.00 .0001. 011030. .0 .00..0 01.... .00 0:10... >._>_.0< .000> .00 Z0 ...0 3.0 1 1 1 1 1 1 . OO.V—.- 0.00 0 - 0.00 a I 00.2. I 0.0.. a - m 00.2. .0 .m flu 00.2. w 1... n I 00.2. m I ..11 a I 00.... ( 00.3. 98 109(9),”..0 = -11.5665 + 0.4219(a) (R2 = 0.96) 109(9),,0.“ = -11.4121 + 0.3937(a) (R2 = 0.77) log(P)(50.c) = -11.2915 + 0.3197(a) (R2 = 0.74) For the Saran coated oriented polypropylene structure (Saran OPP), the relationship between the permeance value (P) and toluene vapor activity (a) was found to be: log(P),,o.c, = —13.0164 + 1.2914(a) (R2 = 0.80) log(P)(50.c) = —12.2601 + 0.8819(a) (R2 = 0.93) log(P)(6o.c) = -11.6785 + 0.3214(a) (R2 = 0.99) From Tables 9 to 13 it can be seen that the permeance (P) is highly dependent on the temperature, at a constant permeant concentration level. Figures 21 to 25 illustrate this graphically, where log permeance is plotted as a function of temperature [T‘1 (°K)]. The following expressions were derived using a least squares fit and applying experimental data. ri n l r lene: log(P)(a=0.067) = - 2.0956 - 2783.0 x (l/T) (R2 = 0.97) log(P)(a=o_22) = - 2.0415 - 2751.5 x (l/T) (R2 = 1.00) log(P)(a=o_“) = - 5.1047 - 1745.5 x (l/T) (R?- = 1.00) Hi n i l h len : log(P),“ofim = - 3.7042 - 2112.5 x (l/T) (R2 = 1.00) log(P)(a=o_22) = - 5.4217 - 1509.5 x (l/T) (R2 = 1.00) log(p),a=o_m = - 8.8215 - 391.0 x (1/T) (R2 = 0.99) 99 0.10.20.01.30 00.11013 :1 00030. .0 0011001500 0;. .0 00110000000 0.0.0.0050. ..0 0:101“. 31.. :1 0000.0 0000.0 0000.0 0000.0 .0000 1 1 L 1 1 1 1 0.32.. % ¢¢.0 I 0 .l 00.... 00.0 I 500.0 I 00.... 0.... 00.... 00.0.. 00.0.. 00.0.. 00.0.. 00.0.. (Bd'oos'z.w/0)I) (8)60: 100 000.110.0201. 11.101100 1.0.1. .1. 0000.0. .0 00.100.11.01. 01.: .0 00000110000 0.0.0.0050. .00 0:12.. c... :1 0000.0 0000.0 0000.0 0000.0 .0000 F 1 1 1 1 1 1 0 08.07 3.0 .. 1. o 0 . . 00.0.. 0 W002 m .000 1. I 00.0.. m 00.2. m m- 002. 1a 8 u 00 2. .d h 2.2. 00.2. l()1 011100010 11. 0110010. .0 001100111100 01.. .0 0000000000 0.0.0.0050... .00 0:10... c... 1.1. 0000.0 0000.0 0000.0 0000.0 1000.0 1 1 1 . _ 1 1 1 °O.F—.n 3.0 . 0 o I 00.0 n 0 a . . I. .000 . 0 00 1. m U1 0...... n N w. 00.... 1a 9 9 0 “d 00.... u 0..... 102 011030010301. 00.110110 00.000 00.00 11. 01100.0. .0 00110001101. 01.. .0 0000000000 010.010.1010.. .00 011101.. 101.. .1. 0000.0 0000.0 1000.0 1000.0 0000.0 1 1 1 1 P 1 1 °°.V—.I 3.0 . 11 o T 00.0 . 1. .1 I 00.0.. oI 100.0 . 1. 0 - m 00.0.. ( m 00.0.. w 1... . . 8 00 0. u 0. 01...... ( 00.... 103 0110.20.02: 00.110110 00.000 0:50.. 11. 011030. .0 00110001101. 01.. .0 0000000000 010.010.1010.. C... :1 .000.0 1000.0 0000.0 0000.0 0000.0 1 L 1 L 1 1 1 OO.¢F. I 00.0.. m I I 00.0.. a I 00.0.. II 00.0... /@| vv.0I¢ @ II 001.... 00.... .00 01001.1 (vd'ou'z.wlbx) (8)60I 104 21225122: log(p),a=0_067) = - 7.3468 - 1272.5 x (l/T) (R2 = 1.00) log(P),a=0_22, = - 6.8968 — 1377.5 x (l/T) (R2 = 0.97) log(P)(a=o.“) = - 7.7433 - 1102.5 x (l/T) (R2 = 0.99) r n ri n l r l n log(P)(a=o_067) = 8.6543 - 6760.0 x (1/T) (R2 = 0.99) log(P)(a=o.22) = 3.2789 - 4951.0 x (1/T) (R2 = 0.99) log(P)(a=o_“) = 2.9934 - 4830.0 x (l/T) (R2 = 0.99) Aopylic ooeteo oriented polypropylene: log(P)(a=0M, = 8.1508 - 6884.0 x (l/T) (R2 = 0.99) These results are similar to the results of the ethyl acetate permeation studies, as the temperature dependency of the transport process associated with the respective barrier membranes, over the temperature range studied, follows well the Arrhenius relationship. The activation energy values for the permeation process (Ep), determined from.the slopes of the Arrhenius plots for the respective film samples, are summarized in Table 14. The Arrhenius expression used to describe the permeance for both ethyl acetate and toluene as a function of temperature is typically applied over a temperature range above or below the glass transition temperature (T3) of the polymer, but not within a temperature range which includes T'..A straight line extrapolation typically cannot be made through.T; and a graphical analysis is expected to show a 105 Table 14. Activation energy values for the permeation of toluene through polymer membranes. E(kcal/mole) Polymer Mbmbranes a = 0.067 a = 0.22 a = 0.44 OPP 12.63 12.53 7.79 HDPE 9.45 6.90 1.90 Glassine 5.78 6.21 5.02 Saran OPP 34.00 34.10 35.41 Acrylic OPP N/A‘“ N/A‘” 30 .25 Metallized PET/OPP N/A‘” N/A‘” N/A‘“ ‘”. N/A denotes not available. 106 change in slope at the T; (DeLassus, et al., 1988). This was assumed not to be a concern for the respective polymer structures evaluated in the present study, since the temperatures evaluated in this steady were well above the glass transition temperatures of polyethylene (T; = -120%D, polypropylene (T9 = -19°C) and Saran (T9 = -19°C), while the T9 value of PET (T9 = 81°C) is above the maximum temperature evaluated in the present study. Results involving studies designed to evaluate the activation energies of the respective test structures for the permeability of ethyl acetate and toluene (see Tables 8 and 14) showed that for both oriented polypropylene and polyethylene activation energy values decreased with an increase in vapor activity, while the activation energy values for the Saran and Acrylic coated polypropylene structures were not affected by permeant concentration levels. While not fully understand, the concentration dependency of the activation energy for the polyolefins may be due in part to penetrant-induced relaxation effects occurring within the polymer matrix. Such relaxation effects would be most favorable above the glass transition temperature (TE) of the polymer membrane, which is the case for PE and OPP. The absorption of organic vapors can result in polymer swelling and thus change the conformation of the polymer chains. These conformational changes are not 107 instantaneous, but are controlled by the retardation times of polymer chains. If these times are long, stresses may be set up which relax slowly. Thus, the absorption and diffusion of organic vapors can be accompanied by concentration as well as time-dependent processes within the polymer bulk phase, which are slower than the micro-Brownian motion of polymer chain segments which promote diffusion (Mears, 1965). There is precedence in the literature in support of such long time period relaxation effects occurring in polymer films above their glass transition temperature (Berens, 1977, and Blackadder and Keniry, 1973). Thus, there may be concentration dependent relaxation effects occurring during the diffusion of ethyl acetate and toluene through the polyolefin films investigated. Such relaxation processes which occur over a longer time-scale than diffusion may be related to a structural reordering of the free volume elements in the polymer. Thus, providing additional sites of appropriate size and frequency of formation, which promote diffusion and may account for the observed decrease in activation energy levels with increased vapor activity. From this study, the metallized PET/OPP based structure was found to have the best organic vapor barrier properties among the six polymer membranes evaluated, when tested in film fonm. The order of barrier properties for the six polymer membranes evaluated, in order of decreasing barrier 108 performance is as follows: Metallized PET/OPP, Acrylic OPP, Saran OPP, OPP, Glassine, and HDPE. 1 n f P l V r A ivi L v 1 l t r i 11 Knowledge of the temperature and concentration dependence of the mass transfer process provided a means of estimating the permeance of the barrier membranes, at toluene concentration levels below which it would be impractical to measure experimentally. As discussed above, by application of the Arrhenius Equation, expressions describing the relationship between permeance and temperature were derived for the respective toluene vapor activity levels evaluated. Based on the linear expressions derived, permeance values at ambient temperature (249C) were estimated for each individual vapor activity. Table 15 summarizes the calculated permeance values at 249C, for the respective packaging structures evaluated, at three vapor activity levels. Permeance values varied as a function of the vapor activity levels. The concentration dependency of the permeance values for the respective test structures at 24%L is presented graphically in Figure 26, where the permeance values are plotted as a function of vapor activity. The relationship between the permeance constants and vapor activity for the respective films evaluated at 249C can be 109 Table 15. Estimated permeance values of test packaging structures for toluene at 249C (based on Arrehenius Equation). Vapor Permeance (kg/n@.sec.Pa) activity OPP HDPE Glassine Saran OPP Acrylic OPP 0.067 3.42x10'12 1.52x10'11 2.35x1042 7.71x10'15 N/A”’ 0.22 4.96x10'12 3.12x10'11 2.93x10'12 3.99x1044 N/A“> 0.44 1.05x10'11 7.27x10'11 3.52x10'12 5.66x104‘ 5.40x1046 ‘”. N/A denotes not available. 110 0.00 .0 0010.001.» 10111301. .0010.._0 10. 0011001501. 00 1:12.00 .0001. 011030. .0 .00..0 01:. .00 0:10... >._>_.0< .000> 00.0 00.0 3.0 no.0 r 1 1 1 1 1 1. °+m°.° 43’ .> <1 11.00.. .1. . 0 ...000 m u 2.00.0 u 0 9 11.00.. .m m 1100.0 0. .100 11.1.0 0 . 1.... 00100010 0 I 3.000 M 0.10... a - . . .8 4 .1110 0 II 1. 00. ...... 3.00.0 111 described by the following expressions derived from a least squares fit. OPP P = 1.5885 x 10'12 + 1.9415 x 10‘”(a) R2 = 0.96 HDPE P = 1.7591 x 10'12 + 1.5656 x 10"°(a) R2 = 0.98 Glassine P = 2.1809 x 10'12 + 3.1049 x 10"2 2323 3 3.3.3200 Km 2:2“. 3:3 as: O m .3520 1?- whonE¢;OIfll ooog. Foogu «cog. moogu voogu moogu ooogu noogu ooogu aoogu 0280 human JOdBA 117 e 23 m «33:33 32 32353.. E 323 9:33 33> 3.322 3 3333.30 .3 352". 3:3 2.: r _ _ _ _ _ _ _ r _ _ a t. _ _ _ t oon Foogv «oogv Moog" voogv mcogu ooogu scary ooogv «cog. 028° human JOdBA v 33528 1?- a 33:3;OI—T ‘I‘I‘WI'IW nf i n Pr u n f Ti 24° Based on estimated permeance values for the respective barrier structures, at a vapor activity level of a = 0.0015, and the following assumptions: 0 Hermetic seal integrity 0 Total sorption of permeated toluene by the chocolate layer 0 Vapor sorption is a function of permeation of toluene through the barrier structure 0 Sensory threshold levels of toluene in chocolate taken from Rfiter (1992) The time required for the packaged product to exceed the sensory threshold level, at the conditions of test (i.e. 24°C, a = 0.0015), for the respective package structures was estimated. A detailed example of the calculation is illustrated in Appendix V. This estimation is based on the assumption that mass transfer only results from permeability and not flow through microvoids within the seal area. The time required to reach or exceed the reported toluene threshold level in chocolate (20 ppm, wt/wt) (Rfiter, 1992) for a chocolate product packaged in the respective test structures, was estimated as follows: (i) Determine the quantity (Q) of permeant (i.e. toluene) to be absorbed by the product, equivalent to the literature threshold concentration level. This is based on the assumed threshold level of 20 ppm (wt/wt) and (ii) 119 the mass of the product. The product mass was taken to be only the chocolate coating, which was approximately 28 gram. Based on the mathematical expressions describing the relationship between permeance and vapor activity, permeance values corresponding to a vapor activity of a = 0.0015 were derived. Substitute the maximum allowed permeant quantity (Q), the vapor activity level expressed as a partial pressure, and the corresponding permeance values into the following expression and solve for time (t): Q 1 t = -—-————— X (20) A x AP P where: t = time to sorb allowed permeant quantity (Q) A = package surface area (0.0122 n?) AP = vapor pressure or vapor concentration P = Permeance constant (iii) Based on the assumptions described above, the time required to reach or exceed the purported sensory threshold level of toluene in chocolate was estimated for the respective package structures at a toluene vapor activity of a = 0.0015. The predicted shelf life values are summarized in Table 18. For comparison, predicted shelf life values for toluene vapor activity levels of 0.003, 0.005 and 0.01 are also summarized in Table 18. 120 Table 18. Estimated uptake time to exceed toluene sensory threshold levels in a packaged chocolate product“) Time (days) Toluene vapor OPP HDPE Glassine Saran OPP activity 0.0015 67.3 54.6 49.8 26568.4 (P = 4.9 Pa) 0.003 32.7 23.2 24.7 12381.7 (P = 9.8 Pa) 0.005 19.4 12.9 14.9 7190.9 (p = 16.3 Pa) 0.01 9.1 4.9 7.4 3153.0 (P = 32.5 Pa) ‘”. Assumed toluene threshold level in chocolate is 20 ppm (wt/wt) from Rfiter (1992). The calculated toluene uptake levels, based on the estimated permeance constant, and the experimentally determined levels of sorbed toluene, quantified by the dynamic purge/trap - thermal desorption GC-MB procedure, are summarized in Tables 19 and 20. To better illustrate the relative uptake of toluene vapor by the respective product/package systems, the results are presented graphically in Figure 29, where total toluene uptake is plotted as a function of storage time for the individual package systems. As shown, all six package/product systems had detectable levels of toluene in the chocolate product. The statistical analysis for toluene uptake by chocolate was based on a completely randomized design (CRD). The toluene uptake values for the respective package/product systems were analyzed using a two—way analysis of variance (ANOVA) (Table 21). The MSTAT (ver 4.0, 1987) microcomputer statistical program was used for calculation. Significant differences between mean squares were evaluated at the 95% confidence level (P s 0.05). The toluene uptake values for the six package/product systems were compared, with the oriented polypropylene (OPP) package/product system being assigned as the control group. The Dunnett's t-value was calculated for the respective package/product systems and the t-values compared with two- 122 Table 19. Total toluene uptake by chocolate for OPP, HDPE and Glassine based package structures as a function of time (Experimental and calculated data) Total Quantity (g) x 10* Time (days) OPP HDPE Glassine Exp“’ Cal“) Exp Cal Exp Cal 6 0 4 0 5 3.2 0 6 4 8 0 7 14 0.8 1.2 1.8 1.4 7.6 1.6 20 1.3 1.7 1.6 2.1 12.0 2.3 27 2.3 2.3 4.3 2.8 6.5 3.0 43 3.3 3.6 4.9 4.4 5.4 4.8 54 4.5 4.5 5.3 5.5 5.6 6.1 62 1.9 5.2 6.1 6.4 3.2 7.0 76 5.9 6.3 11.9 7.8 9.2 8.6 83 7.3 6.9 10.1 8.5 16.3 9.4 ‘”. Experimental data are the average of duplicate samples. ‘3. Calculated data based on the estimated permeance constant. 123 Table 20. Total toluene uptake by chocolate for Saran coated OPP, Acrylic coated OPP and Mbtallized PET/OPP based package structures as a function of time (Experimental and calculated data) Total Quantity (g) x 10* Time (days) Saran OPP Acrylic OPP MEt PET/OPP Exp“) Calm Exp Cal(3) Exp CalC” 11 0.5 0.002 0.2 - 0.3 - 19 1.4 0.004 0.6 - 0.4 - 35 2.7 0.007 0.2 - 0.4 - 56 2.7 0.012 0.9 - 2.9 - 79 3.1 0.017 3.5 — 2.2 - ‘”. Experimental data are the average of duplicate samples. (3. Calculated data based on the estimated permeance constant. ‘”. No estimated permeance constant available for calculation. 124 330 323.533me 2333 33333332. 233:3 3_ 33.323 .3 8.33: 2332 .2: .am 332“. A333 25... co cm on om om oe on cm or 0 93:2 :2 A/ mac 33m Allfllu to 23.2 A\\ 93 A.) $2. A\\ wk $23.. ..:+ I to 3526... r to 533.-...- T 0. 2.33.0 ..... o ..... I \ man—Ii!) I 2:235 A\ “Eels! o+moo.o Tmomd Tweed «.momfi mica... «.mmmé «.mom; «.mmné m.moo.m (0) exudn ouonlo; |azol 125 Table 21. Analysis of variance (ANOVA) for toluene uptake for product/package systems Source df Sum of Square MEan Square F value”) Total 47 512.36 Film 5 178.04 35.61 10.13* Time 3 155.00 51.67 14.70* Film x Time 15 94.94 6.33 1.80 Error 24 84.37 3.52 a) . - . - . _ * = Significant d1 ference at p s .05 126 sided Dunnett's t-tabulated values (Gill, 1987). Table 22 shows the comparison of the various package/product systems with the control group. The detailed statistical analysis is presented in Appendix VI. A statistically significant difference between the control system (i.e. OPP/product) and the comparison product/package system at the 95% confidence level is indicated for cases where the Dunnett's calculated t-values exceeds the Dunnett’s table t-value. Based on the statistical analysis, there was no statistically significant difference between the performance of the control package system and the comparison package system with respect to toluene uptake over the storage time, except for the Glassine/package system. Based on statistical analysis of the toluene permeance values for (i) Oriented Polypropylene; (ii) High Density Polyethylene; (iii) Glassine; and (iv) Saran OPP, determined on flat sheet stock, a statistically significant difference between the toluene barrier properties of the respective film structures was found. These findings indicate that the barrier performance of the film structures showed significant differences. However, as indicated above, the films barrier characteristics are negated following their fabrication to a package structure. This suggests that toluene uptake by the packaged product is the result of toluene ingression through the cold seal area as the predominant mechanism and not the Control: 1 2 Table 22. 127 OPP/product HDPE/product Glassine/product Saran OPP/product Acrylic OPP/product MEt PET/product Statistical comparison of toluene uptake for six package/product systems Comparison group Dunnett's t-value“’ (calculate) V8 V8 VS VS VS Control 2 Control 3 Control 1 Control 1 Control 1. .142 .445* .290 .975 826 ‘”. Dunnett’s t-table value: t00£5mfitw%.= 2.70 = significant difference'at'p k 0.05 'k 128 primary result of diffusion through the package structure. This hypothesis is supported by the fact that the metallized PET/OPP based packaging system showed comparable levels of toluene uptake by the product, as compared with the polypropylene package system, while no measurable rate of permeation was observed for the metallized PET/OPP flat sheet structure. Based on the assumptions described in the Section "Predicted Uptake Levels of Toluene Vapor by the Packaged Confectionery Product as a Function of Time (24PC)", the calculated values of total quantity of toluene uptake for the respective product package Systems at 249C and a = 0.0015 were obtained and compared to the experimentally determined values. The results are summarized in Table 19 and 20, respectively. For better illustration, the experimental and calculated toluene uptake values obtained for the OPP, HDPE, Glassine and Saran/OPP package systems are presented graphically in Figures 30 - 33, respectively. As shown, for the OPP, HDPE and Glassine package systems, the calculated toluene uptake values were similar to those determined experimentally, while for the Saran/OPP package system, the calculated (estimated) toluene uptake values were significantly higher than the values determined experimentally. With respect to the Saran/OPP package system and the lack of agreement between the estimated and (experimentally determined toluene uptake values; like any 129 8.43 3.8.3 . a s as»: 32.2.. ":3 of 3 8.33: 2322 3a 332, 3:35.333 23 33:53 3 .333 :30 .3 332“. 3:3 .2: cm on on om om ow on on o_. 323.5336 3 332330;- I T o+moo.o v.wom.~ Tweed v.wom.h m.woo._. mimmé «flow... «.mmsé m.moo.m (a) oxudn ouonloz mm; 130 3.3: 28.3 .. a 3 as»: 333:: .3: of .3 3.33: 3.323 3— 3...: 323.5333 23 33.533 3 .333 :30 .2." 332". 7:3 .2: o+moo.o «.mom.~ I. Tmood w W Ymomfi m. ..n: 3.38.. m 0 3.333 m .. n I 3.33 a 323.5336 M r m... 33303019. I aims; ( m.moo.m 131 83$ 98... . a s 5223 33222323 of 3 333: 2333 .2 332, 3325333 23 33523 3 33:32.30 .mm 23:. 7.33 2.: om cm on on em cc on on or o ___rt___t_t_____.Jv)o+moo.o \Q\ .I #33.“ I. w 4.33.3 w Y3: m n 9 3.33 w b 3.333 n I .u. I 3.33 m n 3325336 I I My: “33333019- _I «.mmué ( 3.38.3 132 Aooemv 336 n u E 539$ .0233an :Saw 2: .3 3.22. 2332 .3 32: 3225338 2; 3.2573 3 53.3800 .3 2:2”. 3:3 2.: 9&8... Ian.“ I.— 0 I 3.8.... m I 33.4. m n I 9 I 38.. w I b I 33; m I w I 33.. N 3225386 a w ) 33:330-? I aims; w 38.“ 133 other predictive model, the validity of the estimation method will only be as accurate as the assumptions made in developing the model. Implicit in the estimation method used to calculate toluene uptake levels is the assumption that the package system is hermetically sealed and that diffusion through the package structure, as described by Fick's First and Second Laws, is the principle mechanism resulting in toluene ingression. Since the calculated toluene uptake values for the Saran/OPP package system were significantly higher than the uptake values determined experimentally, the assumption that vapor sorption is a function of permeation of toluene through the barrier Structure is not valid. Thus, as indicated above, a mechanism involving flow or transfer through microvoids in the cold seal closure area is the predominant mechanism of toluene ingression and not the result of diffusion through the package structure. Part III. Seal Integrity Evaluation of the Cold Seal Closure for Packaged Confectionery Systems The seal integrity tests were performed with a Nikka Packaging Leak Tester (PLT-3501) (Nikka Densok U.S.A., Inc., Lakewood, CO). The average leak test values obtained for the respective packages provided are summarized in Table 23. In performing the leak tests, the Nikka Packaging Leak Tester Model PLT-3501 was found to have limitations when evaluating package systems, such as the Glassine structure, which had 134 Table 23. The average values for leak testing by using Nikka Packaging Leak Tester (PLT-3501) Film Structures Leak test“) Ummng) OPP 25.4 t 6.2 HDPE _ 32.4 i 6.1 Glassine“) --- Saran OPP 25.0 1 6.2 Acrylic OPP 21.9 i 3.3 Metallized PET/OPP 24.3 i 3.2 ‘”. Average for each product/packaging system, based on 20 replication testing samples. (a. Glassine packages exhibited very poor seal properties as well as poor mechanical characteristic. The values of glassine package system were found over the limitation of leak tester machine. 135 serious leakage or seal integrity problems. For package systems with poor seal quality, which could be seen by visual inspection, the leak tester output gave values indicative of a well sealed package system. Therefore, prior to preforming a seal integrity test on the package system, all packages were examined visually for the presence of major defects in the seal area. In this study, it was found that a large number of the Glassine packages had serious visually observed structural defects in both the seal area and the package itself, although the test output gave values indicative of a well sealed package. Even when an effort was made to leak test Glassine packages which showed no visual indication of structural defects, it was evident that under the applied vacuum, there was no pressure differential created between the package internal and external environments, which is the principle upon while the test is based. Because of this finding, the leak test values for the glassine package are not presented in Table 23. The results obtained from evaluating the integrity of the glassine packages were consistent with the finding of the toluene uptake studies, in that the glassine package system showed the highest level of toluene ingression and was the package system showing the poorest seal quality. The various other package structures, showed no visible evidence of structural or seal defects and the leak test 136 values for the OPP, HDPE, Saran OPP, Acrylic OPP and Metallized PET/OPP package systems were assumed to be reliable and indicative of the seal integrity of the respective packages. As shown in Table 23, the leak test values determined for OPP, Saran OPP, Acrylic OPP and MEtallized PET/OPP package systems were similar, while the value for the HDPE package system was higher. Since the leak test values are inversely related to seal quality, the higher value obtained for the HDPE package system is representative of a package with a lower seal integrity, as compared to the other package systems evaluated. The results obtained from the package integrity studies for the OPP, Saran OPP, Acrylic OPP, Metallized PET/OPP and HDPE package systems were consistent with the finding of the toluene ingression studies. As described in the Materials and Methods Section, leak test values for packaged samples removed from the permeation test chambers were determined directly upon their removal and prior to opening for product sampling and analysis of sorbed toluene. Summarized in Appendix VII are the seal integrity values obtained for the packaged samples and the corresponding toluene uptake levels as a function of storage time. As shown in Appendix VII, in general, package systems exhibiting leak test values above the mean and standard deviation for the specific package structures showed higher toluene ingression levels, as compared to replicate samples 137 which gave leak test values within the expected range. These finding are consistent with the hypothesis that a mechanism involving flow or transfer through microvoids in the cold seal closure is the predominant mechanism of toluene uptake by the packaged product. SUMMARY AND CONCLUSION For a series of commodity films, permeance values were obtained from permeability studies based on an isostatic procedure utilizing the MAS Technology Mbdel 2000 Permeability Test System. Parameters evaluated included vapor activity and temperature. For each temperature, three vapor activity levels were evaluated. Permeability studies were carried out at three temperatures to allow evaluation of the Arrehenius relationship. In addition, the respective commodity films were used to package a confectionery product via a cold seal closure technology. To evaluate the integrity of the cold seal closure, with respect to organic vapor ingression, package permeability studies were carried out. Toluene was selected as the organic permeant. Package permeability was evaluated based on the level of permeated vapor sorbed by the packaged confectionery product. The results of this study are summarized below: 1. The temperature dependency of the transport process associated with the respective barrier membranes, over the temperature range studied, was found to follow well the Arrehenius relationship. 2. From this study, the metallized PET/OPP based structure 138 139 was found to exhibit the best organic vapor barrier properties among the six polymer membranes evaluated, when tested in film form. Based on statistical analysis, there are statistically significant differences between the toluene barrier properties of the respective polymer structures, in flat sheet form. The barrier properties of the six polymer membranes evaluated, in order of decreasing barrier performance, are as follows: Metallized PET/OPP, Acrylic OPP, Saran OPP, OPP, Glassine, and HDPE. By application of the Arrehenius Equation, expressions describing the relationship between permeance and temperature were derived for the respective toluene vapor activity levels evaluated. Based on the linear expressions derived, permeance values at ambient temperature (24°C) were estimated for the respective vapor activity levels evaluated. Permeance values varied as a function of the vapor activity levels. The relationship between the permeance constants and vapor activity for the respective films evaluated at 249C was described by mathematic expressions derived from a least squares fit. Based on statistical analysis, there was no statistically significant difference between the performance of the control package system and the comparison package systems, with respect to toluene uptake, except for the 140 Glassine/package system. The oriented polypropylene (OPP) package/product system was assigned as the control group. Based on statistical analysis of the toluene permeance values for (i) polypropylene; (ii) polyethylene; (iii) Glassine; and (iv) Saran OPP, determined on flat sheet stock, a statistically significant difference between the toluene barrier properties of the respective film structures was found. These findings indicate that the barrier performance of the film structures showed significant differences, when evaluated as flat sheet stock. This suggests that toluene uptake by the packaged product is the result of toluene ingression through the seal area as the predominant mechanism, and not the primary result of diffusion through the package structure. This hypothesis is supported by the fact that the metallized PET/OPP based packaging system showed comparable levels of toluene uptake by the product, as compared with the polypropylene package system, while no measurable rate of permeation was observed for the metallized PET/OPP flat sheet structure. FUTURE STUDIES Since cold seal closures are widely employed in food packaging applications, evaluation of seal performance or seal integrity for the cold seal closure warrants further investigation. Based on the results of the present study, other issues that have potential for further investigation include: 1. Identity and evaluate alternate adhesive formulations, which have utility as a cold seal pattern adhesive, to be used in formation of a cold seal closure. Also consider increasing the jaw pressure or jaw configuration in effecting the cold seal closure. 2. Future studies may also include seal performance evaluation, with other permeants, such as water vapor, oxygen and additional organic vapors to develop a better understanding of the seals overall vapor barrier performance. Studies may also include determination of seal properties, such as seal strength, or peel strength. 3. This study has focused on a packaged confectionery product and the extent of toluene vapor ingression and 141 142 subsequent sorption by the product. Other product/package systems may also be of importance and warrant investigation. Further, the cold seal integrity, as a function of temperature, may also effect the barrier performance of the package system. This can be very important for some packaged frozen food products. APPENDICES Appendix I 4.00E+4 y - -1.9018E+3 + 1.0025E+12x r‘2 - 0.99 3.5094 ? Calibration Factor - 9.98E-13 '7 3.0094 // o ,/// 3 2.5094 / 3 // .. 2.0094 2 3 1.5094 //@ h / < / LOOE+4 ////// 5.0093 /' 0.0090 A , . I I 0.0090 1.00E-8 2.005-8 3.00E-8 Total Quantity (9) Figure 34. Calibration curve ior Ethyl Acetate by GC - FID Quantity (g) Area Response o o 8.94 x 10’9 5746 17.88 x 10'9 14157 35.76 x 10'9 35158 143 Area Response 6.00E+5 l 5-°°E+5 7 Calibration Factor - 6.805-14 3‘3 4.00E+5 — y - 20.1540 + 1.4713E+13x r‘2 - 0.99 3.0095 1 / [E] 2.0095 A / 1.0095 — // 7 A 0.00E+0 #9 I T I 1 0.0090 1.005-8 2.00E-8 3.00E-8 Total Quantity (0) Figure 35. Calibration curve of Toluene by GC - FID Quantity (9) Area Response 8.67 1: 17.34 )c 34.68 x: 0.0 42712 . 0 124476.5 274730 . 3 502068 . 0 144 Area Response Appendix II 2.00E+7 y - -7.0324E+5 + 1.8566E+14x r'2 - 0.98 Calibration Factor - 5.386E-15 E9] L50E+7 —1 ////////// t00E+7 —‘ ’///////////// 5.0090 — E] d E] 0.00E+0 E? / 1 . 1 I I 1 l 0.00E+O 2.00E-8 4.00E-8 6.00E-8 8.00E-8 Quanthy (g) Figure 36. Calibration curve of Toluene b GCIMS - Thermal desorption unlt (SIM ode) Quantity (9) Area Response 17.34 34.68 X X X X 10'9 10'9 10'9 10'9 0 1051470 2512015 5378503 9167813 145 Appendix III Standard Toluene Mass Spectrum NH: 92 Toluene (:7... Species: Toluene, Formula: C7H8 Contributor: D.HENNEBERG, MAX-PLANCK INSTITUTE, MULHEIM, WEST GERMANY 104 4614 CAS# 108-88-3 EPA# 61278 MW 92 Dual Idx 812 Ten Largest Masses and Abundances 91 92 39 65 51 63 50 27 93 90 999 725 204 136 109 105 74 63 53 51 56 Masses and Abundances 2 5 43 22 63 105 83 1 12 2 44 8 64 23 84 3 13 3 45 44 65 136 85 8 14 5 46 29 66 18 86 10 15 24 50 74 67 1 87 6 25 3 51 109 69 1 88 1 26 30 52 26 70 1 89 40 27 63 53 12 71 1 90 51 28 6 55 4 73 3 91 999 29 3 56 2 74 13 92 725 39 204 57 2 75 8 93 53 4O 23 60 3 76 5 94 1 41 24 61 25 77 14 42 3 62 50 78 1 Graphics, List, Structure, Names, Quit 2146 Appendix IV The analysis of variance (ANOVA) table of permeance for toluene using the MSTAT (ver 4.0, 1987) microcomputer statistical program. Table 24. Analysis of variance for toluene permeance constant Source of Degree of Mean variance freedom (df) Square value“) Polymer 3 10.426 4401.787* membranes (P) Temperature (T) 1 0.917 386.964* Vapor activity (VA) 2 0.330 139.188* P x T 3 0.198 83.498* P x VA 6 0.016 6.778* T x VA 2 0.010 4.148* P x T x VA 6 0.005 2.022 Error 24 0.002 ‘1’. F-table value at P $0.05: F124 = 4.26; F2 2,. = 3.40 133'“ = 3.01; 176'“ = 2.51 * = significant difference at p s 0.05 ' 147 148 Significant differences between film composition, temperature and vapor activity levels were evaluated at the 95% confidence level (P s 0.05). The interaction between the polymer membranes and the temperature or vapor activity levels showed significant differences. The following t-test compares the respective polymer structures with OPP assigned as the control structure. For each of the comparison groups the t-value was calculated from the linear Arrhenius Relationship by substitution into the following expression: For T1F1: 571 = a1 + b1 xL T282: 92 = a2 + b2 xL T3F3: 373 = a3 + b3 XL t; 91 ' 172 ?1 " 172 1,2 = = A sen?1 - 92) [se(y1)2 - se()’>2)2]”2 171 ’ S73 91 - 173 t1: = = ‘ se<5>1 - 93) [se(}71)2 - se(y3)2]1’2 1 (x* =5: )2 var(fiflx) = 02 [ + ] n S XX se = [var (nylx) 1 “2 where: X L l/T value at room temperature (24%3 log(P) value at room temperature for first polymer structure 5 ll se = standard error for 9n Appendix V Estimation of toluene uptake by the packaged confectionery product by a permeation mechanism. Q p = A x t x AP Q = Quantity permeated (kg) in time = t A = Package surface area = 0.0122 n? aP = Vapor pressure at a = 0.003, P = 9.8 Pa For shelf life estimation assume 20 ppm (g/g) as sensory threshold and total uptake is in chocolate layer. For product assumed chocolate layer = 28 g. 20 x 10'6 (g/g) x 28(9) = 560 x 10‘6 (g) = 0.56 x 10* (kg) Time to equal or exceed concentration of 20 ppm (g/g) based on permeability for oriented polypropylene wrapper (assumes only ingression of toluene vapor is by permeation, no macro- channels through seal area, seal is hermetic). Q Q 1 A x t x aP A x P P 149 150 0.56 x 10'6 kg 1 m2.sec.Pa t = x 0.0122 m2 x 9.8 Pa 1.6 x 10'12 kg 2.9274 x 106 sec 1 min 1 hr 1 day t(days) = 2.9274 x 106 sec x—— x——— x 60 sec 60 min 24 hr 34 days To estimate quantity uptake as a function of time for oriented polypropylene at a = 0.003 and 24%: Selected time = 6 days 24 hr 60 min 60 sec 6 days x x x— = 518,400 sec day hr min Q p = A x t x AP 1.65 x 10'12 kg 0 (kg) = x 518400 sec x 0.0122 m2 x 9.8 Pa 2 m .sec.Pa 1.02 x 10'7 kg 1.02 x 10" g Appendix VI The statistical calculation for the comparison of toluene uptake by the packaged confectionery product From the ANOVA Table (Table 21), no significant differences were found for the interaction between film and time. It only shows that there are statistically significant differences for the individual factors, which are film types and time. Therefore, the comparison group does not need to consider each individual sampling time, but rather the average toluene uptake data for each polymer membrane as the comparison group. A For each of the comparison groups, the Dunnett's t- value was calculated by substitution into the following mathematic expression: Y.-Yc I (ZMSE/8)1/2 where: Y} = the mean average for the comparison package/product system, respectively. Y; = the mean average for the control package/product system, respectively. MBE = error mean square 151 Appendix VII Table 25. Leak test values and corresponding toluene uptake levels as a function of time for OPP, HDPE and Glassine based packaging structures (2) OPP HDPE Glassine gfl Toluene uptake“) (gram) 6 84 0.57x10* 42 4.84x104 8.60x10‘ 29 0.23x10‘4 22 1.46x10“ 1.08x10* 14 28 0.82x104 28 1.96x104 14.6x10“ 27 0.78x10‘ 29 1.61x104 0.63x10‘ 20 25 1.60x104 36 ‘1.74x10* 20.9x104 14 1.02x104 69 1.44x104 3.02x10“ 27 38 3.08x10“ 39 2.57x104 2.88x10‘ 29 1.60x10‘ 49 5.99x10‘ 10.1x10* 43 19 0.92x104 90 5.73x104 6.03x104 37 5.70x104 54 3.96x104 4.76x104 54 28 4.07x10* 65 5.74x104 5.54x104 90 4.92x104 53 4.90x10* 5.73x10“ 62 68 0.78x104 36 7.10x104 5.25x10“ 60 2.94x10‘ 35 5.06x10‘ 1.06x104 76 36 7.39x10“ 28 12.4x10* 14.2x10* 33 4.36x10'4 64 11.4x104 4.24x104 83 58 7.20x10* 128 10.6x10* 13.2x104 64 7.32x10‘4 35 __11.1x10* 19.4x10‘ “”1 Unit of leak test is in mm-fig). . Unless otherwise stated values are the average of replicate analysis from individual product/package samples. 152 Table 26. Leak test values and corresponding toluene uptake levels as a function of time for Saran OPP, Acrylic OPP and Metallized PET/OPP based packaging structures Saran OPP Acrylic OPP Met PET/OPP £1123, Leak Toluene Leak Toluene Leak Toluene Best uptake‘z’ asst uptake‘z’ Best uptake‘z’ (gram) (gram) (gram) 11 26 0.62x10" 23 0.13x10" 23 0.21x10" 47 0.44x10" 28 0.17x10" 31 0.36x10" 19 45 1.66x10" 22 0.31x10" 27 0.66x10" 30 1.17x10“ 28 0.93x10": 31 0.05x10" 35 31 1.14x10" 18 , 0.21x10" 22 0.37x10" 50 4.31x10" 22 0.28x10" 32 0.40x10" 56 29 0.52x10" i 27 1.39x10" 33 0.53x10" 35 4.89x10" 20 0.33x10": 52 5.18x10" 79 27 2 .44x10" 35 3 . 67x10" 25 0 . 43x10" L23 3 . 70x10" 24 3 .39x10" 53 4 . 01x10" * . Unit of leak test is in mm—Hg). (3. Unless otherwise stated values are the average of replicate analysis from individual product/package samples. 153 B IBLIOGRAPHY BIBLIOGRAPHY Adam, S. 1984. Analysis of volatile aroma components in fruit and vegetable juices by using gas/liquid extraction, sorptive collection, cryogenic and high- resolution gas chromatography. In Chromatography and Mass Spectrometry in.NUtrition Science and Fbod Safety, A. Frigerio and H. Milon (Ed.) 67-77. Elsevier Science Publishers, Amsterdam, Netherlands. Adamson, A.W. 1982. Physical Chemistry Surfaces, 4th ed. John Wiley and Sons, New York. Alberola, J. and Izquierdo, L.J. 1976. The volatile fraction of orange juice, methods for extraction and study of composition. In Flavor of Foods and Beverages - Headspace Techniques, G. Charalambous (Ed.), 263-304. Academic Press, New York. Amoore, J.E. and Hautala, E. 1983. Odors as an aid to chemical safety: odor thresholds compared with threshold limit values and volatiles for 214 industrial chemicals in air and water dilution. J. Applied Toxicology 3(6):272-290. Anonymous, 1980. Cold seal. Coated Inklings, No.112, March. p.6-8. Anonymous, 1986. CarbotrapTM — an excellent adsorbent for sampling many airborne contaminants, GC Bulletin, 846C. Supleco Inc., PA. Baner, A.L. 1987. The measurement and analysis of the diffusion of toluene in polymeric films. Master Thesis. Nflchigan State University, East Lansing, NH. Baner, A.L.; Hernandez, R.J.; Jayaraman, K. and Giacin, J.R. 1984. Isostatic and quasi-isostatic methods for determining the permeability of organic vapors through barrier membranes. In Current Technologies in Flexible Packaging, M.L. Troedel (Ed.), 49-62. American Society for Testing and Materials, Philadelphia, PA. Bellar, T.A. and Lichtenberg, J.J. 1974. Determining volatile organics at microgram-per-liter levels by gas chromatography. J. Am. Water Works Assoc. 66:739-744. 154 155 Berens, A.R. 1977. Diffusion and relaxation in glassy polymer powders: 1. Fickian diffusion of vinyl chloride in poly(vinylchloride), Polymer 18:697—704. Bertsch, W.; Chang, R.C. and Zlatkis, J. 1974. The determination of organic volatiles in air pollution studies: characterization of profiles. J. Chromatograph Science 12:175-182. Biran, D.; Giacin, J.R.; Hayakawa, K. and Gilbert, S.G. 1979. Vinylchloride sorption by dry casein particles: mechanistic considerations. J. Food Science 44(1): 59-61. Blackadder, D.A. and Keniry, J.S. 1973. Difficulties associated with the measurement of the diffusion coefficient of solvating liquid or vapor in semicrystalline polymer: 1. Permeation methods, J. Appl. Polymer Sci. 17:351-363. Blakesley, C.N. 1974. Kinetic considerations of the permeation of organic vapours through a flexible packaging film. J. Food Technol. vol 9:285-296. Blakesley, C.N. 1975. Organic vapour permeation through packaging films. J. Food Technol. 10:365-367. Boone, J.; Lox, F. and Pottie, S. 1993. Deficiencies of polypropylene in its use as a food-packaging material - a review. Packaging Technology and Science, vol 6: 277-281. Boyko, A.L.; Morgan, M.E. and Libbey, L.ML 1978. Porous polymer trapping for GC/MS analysis of vegetable flavors. In Analysis of Fbods and Beverages - Headspace Techniques, G. Charalamous (Ed.), 57-59. Academic Press, Inc., New York. Briston, J.H. 1970. Plastics in packaging - part III, permeability of plastics. Flavor Industry, vol 11: 779—783. Chen, S.C. 1995. Effect of lipids on flavor development in a mallard reaction model system. Ph.D. dissertation. Nfichigan State University, East Lansing, MI. Crank, J. and Park, G.S. 1968. Diffusion in Polymers. Academic Press, New York. Crank, J. 1975. The Mathematical of Diffusion, 2nd ed., Clarendon Press, Oxford, England. 156 Crosby, N.T. 1981. Food Packaging.Materials. Applied Science Publishers LTD. London. DeLassus, P.T.; Tou, J.C.; Babinec, MLA.; Rulf, D.C.; Karp, B.K.; and Howell, B.A. 1988. Transport of apple aromas in polymer films, In Food and Packaging Interactions, J.H. Hotchkiss (Ed.), 11-27. ACS Symposium Series 365. Denver, CO. Demorest, R.L. 1992. recent developments in testing the permeability of good barriers. J. Plastic Film & Sheeting, vol 8:109-123. Dordick, J.S.; Linhardt, R.J. and Rethwisch, D.G. 1994. Chemical and biochemical catalysis to make swellable polymers. Chemtech, Januaryz33-39. Franz, R. 1993. Permeation of volatile organic compounds across polymer films - part I: development of a sensitive test method suitable for high-barrier packaging films at very low permeant vapor pressures. Packaging Technology and Science, vol 6:91-102. Girard, B. and Nakai, S. 1991. Static headspace gas chromatographic method for volatiles in canned salmon. J. Food Science 56(5):1271-1274. Gavara, R. and Hernandez, R.J. 1993. Consistency test for continuous flow permeability experimental data. J. Plastic Film & Sheeting, vol 9:126—138. Gilbert, S.G. 1979. Modeling the migration of indirect additives to food. Food Technol. 33(4):63. Gilbert, J. 1984. Confirmation and quantification of trace organic food contaminants by mass spectrometry — selected ion monitoring, In Analysis of Fbod Contaminants. 265-319. Elsevier Applied Science, London. Gilbert, J. and Startin, J.R. 1981. Single-ion monitoring of styrene in foods by coupled mass spectrometry—automatic headspace gas chromatography. J of Chromatography 205:434-437. Gilbert, S.G.; Hatzidimitriu; Lai, C. and PAssy, N. 1983. Study of barrier properties of polymeric films to various organic aromatic vapors. In Instrumental Analysis of Foods, Vblume 1, G. Charalambous and G. Inglett (Ed.), 405-411. Academic Press, London. 157 Granzer, V.R.; Koszinowski, J.; Robinson-Mand, L. and Piringer, O. 1986. The influence of rest solvents and monomers acrylate from packages on the sensoria properties of foodstuffs. Verpackungs-Rundschau 37(1):53-58. Halek, G.W. and Hatzidimitriu, E. 1988. Partition coefficients of food package printing ink solvents in soybean oil, chocolate liquor, and a high fat baked product. J. Food Science 53(2):568-570. Halek, G.W. and Levinson, J.J. 1988. Portioning and absolute flavor threshold interactions of aliphatic and aromatic food packaging solvents in high-fat cookies. J. Food Science 54(1):173-176. Halek, G.W. and Levinson, J.J. 1989. Portioning behavior and off- flavor thresholds in cookies from plastic packaging film printing ink compounds. J. Food Science 53(6):1806-1808. Hagenbaugh, J. 1987. Flavor and odor protection of packaging films. The Manufacturing Confectioner, 67(6):111-114. Hartman, T.G.; Lech, J.; Karmas, K.; Salinas, J.; Rosen, R.T. and Ho, C.T. 1993. Flavor characterization using adsorbent trapping-thermal desorption or direct thermal desorption-gas chromatography and gas chromatography- mass spectrometry. In Flavor Measurement, C.T. Ho and C.H. Manley (Ed.), 37-60. Marcel Dekker, Inc. New York. Hatzidimitriu, E.; Gilbert, S.G.; Loukakis, G. and Paik, S. 1987. Permeation studies of model flavor compounds. In Frontiers of Flavor, G. Charalambous (Ed.). 297—305. Elsevier Science Publishers, Amsterdam, Netherlands. Hollifield, H.C.; Brader, C.V. and Dennison, J.L. 1980. Container-derived contamination of maple syrup with methyl methacrylate, toluene, and styrene as determined by headspace gas-liquid chromatography. J. Assoc. Off. Anal. Chem. 63(2):173-177. Hotchkiss, J.H. 1995. Overview on chemical interactions between food and packaging materials. In Food and Packaging Materials - Chemical Interactions. P. Ackermann; M. JAgerstad and T. Ohlsson (Ed.). The Royal Society of Cambridge. Imbalzano, J.F.; Washburn, D.N. and Nehta, PJM. 1991. Permeation and stress cracking of fluoropolymers. In Chemical Engineering. 158 Issenberg, P.; Greenstein, G. and Boékovic, M. 1968. Adsorption of volatile organic compounds in dehydrated food systems. 1. measurement of sorption isotherms at low water activities. J. Food Science, vol 33:621-625. Jones, H.P. 1985. Cold sealing seen as alternative to avoid damage to substrates. Flexo 10(2):21-23. Kilcast, D. 1993. Sensory evaluation of taints and off- flavours. In Fbod Taints and Off-Flavours, MLJ. Saxby (Ed.), 1-34. Blacki Academic and Professional, Glasgow, U.K.. Kolb, B. 1984. Analysis of food contaminants by headspace gas chromatography. In Analysis of Food contaminants, J. Gilbert (Ed.), 117-156. Elsevier Applied Science, London. Kontominas, MLG. 1985. Determination of permeation rates of polymetric films to various organic vapors in relation to retention of flavor quality of packaged foodstuffs. Sciences des Aliments. 5(2):321-329. Kumar, K.R. 1992. Packaging aspects of confectionery items. Indian Food Industry, 11(1):34-39. Levine, M; 1976. Cohesive latex coatings. 1976 TAPPI Annual Meeting at New York, NY. P.83-86. Liu, K.J.; Hernandez, R.J. and Giacin, J. R. 1991. The effect of water activity and penetrant vapor activity on the permeation of toluene vapor through a two-side PVDC coated opaque oriented polypropylene film. J. Plastic Film & Sheeting, vol 7:56-67. Maekawa, S. 1994. Volatilization of vanillin and ortho- vanillin from a corrugated paperboard shipper and sorption by a low density polyethylene primary container. Master Thesis. Nfichigan State University, East Lansing, MI. Martin, R.A. Jr. 1987. Chocolate. In Advances in Fbod Research, Vol 31. C.O. Chichester (Ed.). p.213-343. Academic Press, Inc. San Diego, CA. Nunifie, B.W. 1970. Chocolate, Cbca and Cbnfectionery, 1st Ed. AVI Publishing, New York. NY. NbCarron, R.ML 1972. Package that smells isn't package that sells. Candy and Snack Industry 137(10):54. 159 MCCown, S.M; and Radenheimer, P. 1989. An equilibrium headspace gas chromatographic method for the determination of volatile residues in vegetable oils and fats. LC-GC International 2(12):28-31. Meares, P. 1965. Transient permeation of organic vapors through polymer membranes. J. Applied Polymer Sci. 9:917—932. Merritt, C. JR. and Robertson, D.H. 1982. Part IIb: Techniques of analysis of flavors: gas chromatography and mass spectrometry, In Food Flavours - Part A, Introduction, I.D. Morton and A.J. Macleod (Ed.), 49- 78. Elsevier Scientific Publishing Co., Amsterdam, Netherlands. Nfiltz, J.; Passy, N. and Mannheim, C.H. 1992. Mass transfer from and through packaging materials. Packaging Technology and Science, vol(5): 49-56. Mohney, S.M;; Hernandez, R.J.; Giacin, J.R.; Harte, B.R. and Nultz, J. 1988. Permeability and solubility of d- limonene vapor in cereal package liners. J. Food Sci. 53(1):253-257. Mbntedoro, G.; Bertuccioli and Anichini, F. 1978. Aroma analysis of virgin olive oil by head space (volatiles) and extraction (polyphenols) techniques. In Flavor of Foods and Beverages - Headspace Techniques, G. Charalambous (Ed.), 247-281. Academic Press, New York. leler, U.; Blaas, W.; Mehlitz, I; Vieths, S. and Weber, R. 1990. Benzol und toluol in Lebensmitteln. Deutsche Lebensmittel-Rundschau 86(9):277-290. Nelson, K.L. and Fennema, O.R. 1991. Methylcellulose films to prevent lipid migration in confectionery products. J. Food Science 56(2):504-509. Niebergall, H.; Humeid, A. and Blochel, W. 1979. Aroma permeation of packaging foils and its determination with the aid of a new developed measuring apparatus. Lebensmittel-Wissenschaft und Technologie 12(2):88—94. Niebergall, H.; Humeid, A. and Blbchel, W. 1980. Aroma- Permeation of packaging-foils and its determination with the aid of a new developed measuring-apparatus. Part IV. Lebensmittel-Wissenschaft und Technologie 12(2):88-94. 160 Nijssen, B. 1991. Off-flavors, In volatile Cbmpounds in Foods and Beverages, H. Maarse (Ed.). 698-735. Marcel Dekker, Inc. New York. Pangborn, R.M; 1981. A critical review of threshold insensitive analysis in flavor research. In Flavour ’81. P. Schreier (Ed.), p 3-32. Walter de Gruyter and Co., Berlin. Passy, N. 1983. Off—flavors from packaging materials in food products some case studies. In Instrumental Analysis of Foods, Vblume 1, G. Charalambous and G. Inglett (Ed.), 413- 421. Academic Press, London. Pasternak, R.A.; Schmimscheimer, J.F. and Heller, J. 1970. A dynamic approach to diffusion and permeation measurements.J. Polymer Science, Part A-2, 8:467. Potey, S.D.; Gravier, M. and Benaim, R. 1992. Use of DCI- platine coupled to gas chromatography for measuring the transfer of organic vapours through packaging materials. Pacaging Technology and Science, vol 5: 83-89. Reineccius, G.A. 1984. Determination of flavor components. In Modern Methods of Food Analysis, K.K. Stewart and J.R. Whitaker (Ed.), p 293-316. The AVI Publishing Company, Inc. Westport, Connecticut. Reineccius, G.A. 1993. Biases in analytical flavor profiles introduced by isolation method. In Flavor Measurement, C.T. Ho and C.H. Manley (Ed.). p 61-76. Marcel Dekker, Inc. New York. Rfiter, M; 1992. Effect of solvent residues in package on organoleptic properties of food stuffs. Verpackungs- Rundschau 48(8):53-57. Saleeb, F.Z. and Pickup, J.G. 1978. Flavor interactions with food ingredients from headspace analysis measurements. In Analysis of Foods and Beverages - Headspace Techniques, G. Charalamous (Ed.), 113-129. Academic Press, Inc., New York. Saxby, MLJ. 1993. A survey of chemicals causing taints and off- flavours in foods. In Food Taints and Off- Flavours, M.J. Saxby (Ed.), 35-62. Blacki Academic and Professional, Glasgow, U.K.. 161 Schaefer, J. 1981. Isolation and concentration from the vapour phase. In Handbook der Aroma Forschung, H. Maarse and R. Belz (Ed.), P 44. Akademia-Verlag, Berlin. Schroeder, M.L. 1989. The effect of sorption of flavor volatiles on the adhesive and cohesive bond strength of multilayer laminations. Master Thesis. Nflchigan State University, East Lansing, MI. Schroeder, MLA.; Harte, B.R.; Giacin, J.R. and Hernandez, R.J. 1990. Effect of flavor sorption on adhesive and cohesive bond strength in laminations. J. Plastic Film and Sheeting, 6:232-246. Shaath, N.A. and Griffin, PJM. 1988. Modern analytical techniques in the flavor industry. In Frontiers of Flavor, G. Charalambous (Ed.), 89-107, Elsevier Science Publishers, Amsterdam, Netherlands. Stone, R.K. 1976. The past, present and future of cold seal. 1976 TAPPI Annual Meeting, New York, NY. p.335-338. Tanchotikul, U. and Hsieh, T.C.Y. 1991. Analysis of volatile flavor components in steamed Range clam by dynamic headspace sampling and simultaneous distillation and extraction. J. Food Science 56(2):327-331. Tarkenton, S. 1988. Coated OPP films look good in new study. Snack Worlk, 45(3):46-47. Thomas, A.F.; Willhalm, B. and Flament, I. 1984. Some aspects of GC-MS in the analysis of volatile flavours. In Chromatography and Mass Spectrometry in.NUtrition Science and Food Safety; A. Frigerio and H. Milon (Ed.) 47-65. Elsevier Science Publishers, Amsterdam, Netherlands. Thompson, L.J.; Deniston, D.J. and Hoyer, G.W. 1994. Method for evaluating package-related flavors. Food Technology 48(1):90- 94. Tice, P. 1993. Packaging material as a source of taints, In Food Taints and Off-Flavours, M.J. Saxby (Ed.), 202- 235. Blacki Academic and Professional, Glasgow, U.K.. Tou, J.C.; Rulf, D.C. and DeLassus, P.T. 1990. Mass Spectrometric system for the measurement of aroma/flavor permeation rates across polymer films. Anal. Chem. 62(6):592-597. 162 Wangwiwatsilp, K. 1993. The effect of surface sulfonation on barrier properties of polymer films. Master Thesis. Nflchigan State University, East Lansing, MI. Whitfield, F.B. 1986. Food off-flavours: cause and effect. In Developments in Food Flavours. G.G. Birth and MLG. Lindley (Ed.) 249-273. Elsevier Applied Science. New York, NY. Wyllie, S.G.; Alives, S.; Filsoof, M. and Jennings, W.G., 1978. Headspace sampling: use and abuse. In Analysis of Foods and Beverages - Headspace Techniques, G. Charalambous (Ed.), 1-16. Academic Press, Inc., New York, NY. "IIIIIIIIIIIIIIIIIIIIII“