(I'll). .v ‘u u!!.llh.!’ I ‘l I'FIIIYI o - sail J M p, I” I I ._ I‘Z-Z'.’ ‘ IWII’IWIIW’ L- i??? - This is to certify that the thesis entitled THE AROMA PERMEABILITY AND SOLUBILITY OF TWO CEREAL LINER MATERIALS AND THEIR RELATIONSHIP TO PRODUCT QUALITY presented by SUSAN MARIE BRAUN MOHNEY has been accepted towards fulfillment of the requirements for M.S. degreein PACKAGING Drs. Jack R.Giacin 8 Bruce Harte Major professor Date October I7, 1986 0-7639 MS U is an Waive Action/Equal Opportunity Institution RETURNING MATERIALS: IV1531_] Place in book drop to remove this checkout from M your record. FINES will be charged if book is returned after the date stamped below. s— 82 M~.,....~.-~u FEBWNZEUI R33 0 8 lg? THE AROMA PERMEABILITY AND SOLUBILITY OF TWO CEREAL LINER MATERIALS AND THEIR RELATIONSHIP TO PRODUCT QUALITY BY Susan Marie Braun Mohney A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE School of Packaging 1986 Hzfl-szz, Copyright by Susan Marie Braun Mohney 1986 ABSTRACT THE AROMA PERMEABILITY AND SOLUBILITY OF TWO CEREAL PACKAGE LINER MATERIALS AND THEIR RELATIONSHIP TO PRODUCT QUALITY by Susan Marie Braun Mohney Permeability and Solubility studies were performed on two typical cereal package liners using d-limonene vapor as the volatile test penetrant. Permeability coefficient (P) values were found to be concentration dependent and the relationship between P and vapor concentration can be represented by an exponential equation. The glassine based structure exhibited between three to four orders of magnitude reduction in relative permeability as compared to the polyolefin structure, over the entire vapor concentration range studied. Solubility coefficient values were determined by an independent procedure, involving gravimetric analysis. Additionally, a qualitative relationship was established between the limonene headspace concentration and sensory evaluation for a packaged fruit-flavored cereal product. Results from these studies indicate that the loss of volatile aroma moities may be due to sorptive and/or permeation mediated processes and both mechanisms must be considered in packaging a product where quality is. associated with the retention of aroma compounds. DEDICATION This thesis is dedicated to my family in appreciation and thanks for their assistance and guidance through all my academic endeavors. Also, to my husband for his patience and support throughout this work. ii ACKNOWLEDGEMENTS The author wishes to express her appreciation to Dr. Bruce Harte and Dr. Jack Giacin for their combine guidance and support while serving as co-advisors. Appreciation is also expressed to Dr. Charles Stine for serving on the guidance committee. Special thanks is extended to Ruben Hernandez, Steven Tan, and Larry Baner for their patience and technological support throughout this research study. iii LIST OF TABLES LIST OF FIGURES INTRODUCTION TABLE OF CONTENTS LITERATURE REVIEW The Use of Barrier Materials Mass Transfer Parameters d-Limonene: Characteristics Methods of Analysis Aroma Odor: Characteristics & Molecular Structure Sensory Analysis: Selecting Panelist MATERIALS AND METHODS Materials Permeation Measurements Sorption Measurements Analytical Sensory Analysis RESULTS AND DISCUSSION Sorption Studies Permeation Studies Sensory Analysis SUMMARY AND RECOMMENDATIONS EXPERIMENTAL ERROR ANALYSIS APPENDIX List of Appendix Appendix A: Appendix B: Appendix C: Appendix D: BIBLIOGRAPHY Permeation Data Sorption Data Standard Calibration (Data and Curves) Sensory Evaluation Data iv Page vii 11 12 19 20 22 29 29 3O 4O 45 46 48 48 64 73 77 81 84 84 86 101 105 112 117 Table 10 11 12 13 14 LIST OF TABLES Solubility Data for HDPE Base Structure As A Function of Limonene Vapor Concentration Solubility Data for Glassine Base Structure As A Function of Limonene Vapor Concentration Permeability Data for HDPE Base Structure As A Function of Limonene Vapor Concentration Permeability Data for Glassine Base Structure As A Function of Limonene Vapor Concentration Permeation Data for HDPE Based Structure (Appendix A): Transmission Rate Profile Permeability (slope) Permeability Coefficient Apparant Lag Time Diffusion Coefficient Driving Force Vapor Concentration Permeation Data for Glassine Based Structure (Appendix A): Transmission Rate Profile Permeability (slope) Permeability Coefficient Apparant Lag Time Diffusion Coefficient Driving Force Vapor Concentration Sorption Data (Appendix B): Page 49 50 65 66 86 86 88 9O 92 94 95 95 97 98 99 100 101 Table Page 15 Solubility Data for HDPE Based Structure 101 16 Solubility Data for Glassine Based Structure 103 Standard Calibration (Appendix C): 105 17 d-Limonene Standard Calibration Data 106 18 d-Limonene Standard Calibration Data 108 19 d-Limonene Standard Calibration Data 110 Sensory Evaluation for Glassine Based Structure Only (Appendix D): 112 20 Summarized Data 112 21 Data Following 3 Months of Storage 114 22 Data Following 6 Months of Storage 115 23 Data Following 9 Months of Storage 116 vi FIGURE 10 11 12 13 14 15 LIST OF FIGURES Schematic of Quasi-Isostatic Permeation Test Apparatus Photograph: Apparatus Quasi-Isostatic Permeation Test Schematic of Permeant Cell System Photograph: Photograph: An Assembled Cell System Comparison of Cell Volumes Utilized for the Center Chamber Permeability Cell Schematic Diagram of the Sorption Apparatus Photograph: Photograph: Absorption at 1.5 ppm Absorption at 3.9 ppm Absorption at 4.1 ppm Absorption at 6.3 ppm Absorption at 6.6 ppm Absorption at 7.2 ppm Absorption Sorption Apparatus (Overall View) Sorption Apparatus (Close-up View) of d-Limonene in HDPE Base Structure and 20.5°c of d-Limonene and 20.5°c of d-Limonene and 20.5°c of d-Limonene and 20.5°c of d-Limonene and 20.5°c of d-Limonene and 20.5°c of d-Limonene Structure at 2.5 ppm and vii in in in in in in 20. HDPE Base Structure HDPE Base Structure HDPE Base Structure HDPE Base Structure HDPE Base Structure Glassine Base 5°C PAGE 32 34 35 37 37 42 44 44 51 53 54 55 56 57 59 Figure Page 16 d-Limonene Vapor Solubility in HDPE Base and Glassine Base Structures 60 17 d-Limonene Vapor Solubility Coefficient of HDPE Base and Glassine Base Structures 63 18 The Effect of Vapor Concentration on the Diffusion of d-Limonene Through HDPE Base Structure At 23°C 67 19 The Effect of Vapor Concentration on the Diffusion of d-Limonene Through Glassine Base Structure At 23°C 70 20 The Effect of d-Limonene Vapor Concentration on the Log P For HDPE Base and Glassine Base Structures 71 21 Bar Graph Data Relating Quantitative Headspace Analysis With Qualitative Sensory Analysis of a Fruit-Flavored Cereal Product 75 22 d-Limonene Standard Calibration Curve 107 23 d-Limonene Standard Calibration Curve 109 24 d-Limonene Standard Calibration Curve 111 25 Aroma Evaluation Form 113 viii INTRODUCTION Volatile, low molecular weight organic compounds, represent important constituents of foodstuffs because of their influence on the characteristic odor and taste properties of a food. In general, protection against the loss of volatile aroma constituents from foodstuffs may be achieved through the selection and use of appropriate packaging materials. The intensity of the aroma of a packaged foodstuff thus depends, at least in part, upon: (i) the vapor pressure of the individual components of the total aroma: (ii) the interaction of these volatile organic moieties with non-volatile food components; and (iii) the aroma barrier characteristics of the package. The solubility and transport properties of the aroma moieties-package system, is of major concern in the. selection and use of plastic packaging materials for food packaging. Because of this inherent concern with plastic packaging materials and their wide utilization in food packaging, this work deals specifically with the inter- relationship between mass transport and the shelf life of a product whose quality is associated with the retention of volatile aroma moieties. The phenomenom of permeability can be considered a function of penetrant diffusion (D) and penetrant solubility (S) in the polymer matrix. The diffusion coefficient (D) is the rate which a penetrant molecule moves through the film and the solubility coefficient (S) describes the number of penetrant molecules permeating the barrier. For fixed or non-interactive gases the permeability coefficient (P) is related to the two fundamental mass transfer parameters by (Crank & Park, 1968) 3': D x S (1) However, unlike the transport properties of non- interacting penetrants, for permeability involving organic penetrants and multilayer laminate structures, this simple relationship is not always applicable. For organic penetrants, the penetrant/barrier system can exhibit non- ideal diffusion and solubility properties due to the swelling of the polymer matrix by the sorbed organic penetrant (Crank, 1975: Berens, 1977: Bagley and Long, 1958: and Fujita, 1961). Thus, for such cases, diffusion, solubility and permeability data determined experimentally are necessary to completely describe the mass transfer and sorption behavior between organic vapors (i.e. aroma compounds) and a multilayer polymeric barrier structure. Knowing the solubility of essential flavor ingredients in polymer structures typically used for food packaging is of paramount importance in avoiding the effect of "flavor scalping" or loss due to sorption. For example, d-limonene, a common flavor component present in citrus foods, has a relatively high solubility in polyolefins (DeLassus, 1985). Since aroma compounds such as d-limonene are normally present in low concentrations in a foodstuff, there is an increase in the potential for "lose" of the aroma constituents due to their absorption by the packaging materials (i.e. solubilization). DeLassus (1985) has alluded briefly to this phenomenom in his discussion of barrier layer location in multilayer structures. Further, knowledge of the diffusion coefficient (D) and permeability coefficient (P) can provide information with regard to the time required to attain a steady rate of transmission and the steady-state permeation rate, for a specific penetrant/polymer system under some end-use application. The studies reported here were undertaken to determine under well defined experimental conditions, the relationship of mass transport to the storage stability of a packaged fruit flavored cereal product, whose quality is associated with the retention of volatile aroma constituents. It is proposed that a selected volatile organic compound, or selected compounds contributing to product aroma be identified and the permeability and solubility of these moieties through the barrier films be determined. Such a constituent (or constituents) should be selected, based upon their contributions, to perceived product quality. d- Limonene, a compound present in the products aroma profile, was selected as the probe aroma compound to be monitored, because of its recognizable sensory characteristics and its ease of analysis. Seperate studies were performed involving permeation and sorption measurements of d-limonene vapor to evaluate the barrier properties of two typical cereal package liners, namely a high density polyethylene/sealantl laminate structure and a wax/barrier/glassine/barrier/wax structure. The specific objectives included developing an accurate and reproducable method of measuring the permeability of d- limonene vapor through the selected packaging films and evaluating the effect of d-limonene partial pressure on the permeation rate. The solubility of d-limonene vapor in the respective film samples was also determined using a gravimetric technique, at a series of penetrant concentration levels. Sensory analysis was carried out to evaluate the effectiveness of the glassine based structure in providing the necessary protection from loss of aroma. A qualitative relationship between d-limonene headspace concentration and perceived product quality was developed by packaging cereal product of varying d-limonene concentration in the glassine based structure. After allowing the system to equilibrate, the packaged product was evaluated by sensory analysis and 1 Sealant is a polyblend of EVA, Surlyn and Polybutalene. the amount of d-limonene in the headspace quantified. In summary, the major objectives of this study were to: (1) (2) (3) (4) (5) (5) (7) (8) Evaluate the quality of a fruit-flavored cereal product undergoing loss via aroma permeation and/or sorption, Determine the aroma barrier properties of selected packaging liner materials to d-limonene vapor, Determine the effectiveness of the package liner materials for use under the specified d-limonene concentration levels and conditions of storage, Evaluate the effect of permeant concentration on the permeation constant of the respective film structures (concentration dependency), Determine the solubility of d-limonene in the respective film structures, Compare the barrier properties of the HDPE based structure and the glassine based structure as a function of d-limonene vapor concentration, To establish a qualitative relationship between d-limonene headspace concentration and sensory evaluation, and Indicate the mechanism responsible for the loss of volatile aroma constituents. LITERATURE REVIEW THE USE OF BARRIER MATERIALS Preservation of food from environmental hazards is of major concern in food packaging. Oxygen, moisture, light and organic flavor compounds are to name but a few of the variables important in maintaining the quality of a product. Adverse conditions such as temperature, humidity, contamination, and adsorption of off odors may reduce the quality of even the finest products. There is a definite movement towards the use of barrier plastics for food packaging applications. A recent survey (Hamilton, 1985) predicts a 72% annual growth rate over the next 12 years in the use of barrier films. Such growth is expected to occur based on: - "New packaging manufacturing processes that will allow barrier plastics to be commericialized," - "New barrier and adhesive resins for use in sophisticated coextrusions", and - "Increasing consumer acceptance of plastics for food packaging." In the past, food packaging has been concerned primarily with problems associated with the transmission of oxygen, carbon dioxide and water vapor through the polymer structures (Manathunya, 1976: Gyeszi, 1971). Standard test methods were available to determine the transmission rates for these permeants (ASTM E96-66, ASTM D3985-81). More recently, the transport of organic vapors through polymeric packaging materials has become of increasing importance and has been the subject of several recent investigations (Gilbert, et al., 1983: DeLassus, 1985: Hamilton, 1985; Baner, et al., 1986). Food packagers rely on the package to protect food from potential changes in sensory qualities (Hamilton, 1985). Undesirable odors are likely to occur in storage and handling and can be transmitted to other goods. Since with many foodstuffs, there is no natural protection against the loss of aroma nor the take-up of foreign odors from the environment, this is generally achieved by the use of appropriate packaging materials. MASS TRANSFER PARAMETERS The rate at which a gas, vapor, or liquid will pass through a film sample is dependent on several factors (Giacin, 1983; Gilbert, et al., 1983). These factors may be controlled by the properties of the material (degree of crosslinking, cystallinity, and size of lattice area for molecular passage), the properties of the gas, vapor, or liquid (degree of volatility, the size and shape of the permeating molecule), and/or the degree to which interaction occurs between the permeant and the film sample forming the barrier. Environmental conditions such as temperature and relative humidity affect the rate of permeation as well. Other parameters to consider are the concentration gradient, thickness of laminate structure as well as the gauge of the individual laminate plies and the surface area exposed. Migration, absorption/adsorption, wicking/delamination, package catalyzed reaction, swelling and/or delamination, are among the list of potential compatability problems facing a product and its primary package (Giacin, 1986). Only recently has there been experimental evidence illustrating the importance of both permeability and solubility in understanding the behavior between penetrant/polymer interaction (Baner, 1986: DeLassus, 1985: Hernandez, 1984). Crank 8 Park (1968) describe the Permeability Coefficient (P) as the product of Solubility and Diffusion (P's D * S); where: Permeability (P) is a measure of the ease with which a gas or vapor can pass through a polymeric structure, as a function of the end use application, Solubility (S) is that which adheres to the matrix of a polymer film structure causing changes to occur in the physical as well as mechanical properties (i.e. swelling, bowing, molecular separation), and Diffusion (D) refers to the rate at which a permeant is transmitted from a volatile (or high) vapor phase surface area, through to a non-volatile (or low) surface area [path length (cm2)/time (sec)]. In other words, the rate which a molecule travels in the process of diffusion. For gases, such as oxygen, a linear relationship exists between penetrant concentration and the permeability coefficient (independent of concentration). Results obtained from tests performed at high or low levels of penetrant concentration allow for extrapolation in estimating the results for other concentrations. This is not necessarily true for aroma permeability. As pointed out by Hamilton (1985) "Many equate aroma protection to oxygen permeability but this relationship is at best tenuous". Zobel (1982) reported that "in many organic vapour/packaging film combinations, the permeability coefficient is strongly dependent upon concentration. This effect occurs because the vapour interacts with and swells the polymer, increasing the permeation rate. When exposed to certain saturated vapours, this effect can be so extreme as to cause distortion of the film, resulting in very high permeation rates limited principally by the rate at which the vapour is removed from the surface of the film.” Zobel (1982) also noted that "much of the published work has involved the use of saturated solvent vapours. 10 Whilst these data are useful in estimating how well a packaging material will withstand accidental high-level contamination, it is not valid to use such data to estimate permeation rates at the very much lower levels of vapour encountered in typical retailing situations, whether from foreign contaminating odours or from the flavour components within the package." For cereal product containing volatile aroma constituents, vapor is absorbed onto the surface layer of the film and will either pass through to the other side or adhere to the matrix of the polymer film. That portion trapped within the matrix of the film may cause the film to swell. Permeability data in conjunction with solubility data are therefore integral parts of any study involving the storage stability of a product whose quality is related to the retention of aroma constituents. Berens (1978) presented a detailed review of a sorption method for measuring the sorption and diffusion of small molecules in polymers. In the case of polyolefin structures (i.e. High Density Polyethylene, a non-polar polymer) swelling was observed when subjected to a non-polar substance such as d-limonene (DeLassus, 1985). When swelling takes place, diffusion, permeability and solubility change. 11 d-Limonene d-Limonene, a hydrocarbon which occurs in essential oils, has recognizable odor properties (lemon) which can be monitored by sensory and gas chromatographic (G.C.) methods. Hydrocarbons contain only the elements hydrogen and carbon (Hart & Schuetz, 1978). d-Limonene, an unsaturated hydrocarbon contains double bonds. When two double bonds are present, the compounds are called alkadienes, or more commonly, just dienes. d-Limonene is characterized (Weast, et al., 1985 - 1986) by the following: Structural Formula: 2>—<:;>—- Molecular structure = C10 H16 Molecular Weight = 136.24 Density 3 0.842 g/ml Boiling Point = 178°C Melting Point = -74.3°C Molar Density - 6.17 x 10-3 gmole/ml at 25°C Soluble in: Ethyl Alcohol, Diethyl Ether Vapor Mole fraction in equilibrium with pure liquid = 1.9 x 10-3 at 25°C Among the essential oils, d-limonene constitutes over 90% of orange oil (fruit). "Essential oils are mixtures of various volatile organic substances along with some non-volatile waxy materials." The term "oil" does not 12 refer to any chemical characteristic but rather implies these substances are insoluble in water and soluble in non- polarsolvents" (Amerine, et al., 1965). METHODS OF ANALYSIS The loss of odor/flavoring components were first studied in the forties and fifties through means of direct gravimetrical analysis for products packaged in plastic lined containers (Becker, et al., 1983). Research and technology have developed more effective methods to analyze the chemical nature of various aromas within the last two decades. Although aroma/flavor research has been under investigation for many years it has not been until recently that methods have been developed to quantitatively measure barrier properties of polymeric films to a variety of organic penetrants (Zobel, 1982: Gilbert, 1983; Murray & Dorschner, 1983: Murray, 1985). The majority of published studies have used the method developed by Barrer (1939). In studies involving an organic substance, a constant saturation vapor pressure is exposed to one side of a test film, while a vacuum is maintained on the other side so as to generate permeation. This is refered to as the Absolute Pressure Method. Here the Absolute Pressure Method is one of three quantitative procedures used to obtain diffusion data of organic vapors in semi-permeable membranes. The other two 13 procedures include the Quasi-Isostatic Method and the Isostatic Method (Giacin, 1983). The Absolute Pressure Method incorporates an apparatus where no gas other than the permeant in question is present. A pressure differential, provides the driving force for permeation between two chambers. Here, the permeation rate is determined from the change in pressure on the volume of the low pressure chamber of the cell. Zobel (1982) developed an isostatic method for measuring the permeability of packaging films to organic vapors at low penetrant concentrations and described a modification of the procedure which incorporated an adsorption/desorption cycle (Zobel, 1984, 1985). In the Isostatic Method, both chambers of the permeation cell are maintained at atmospheric pressure. The gas which permeates through the film and into the lower chamber of the cell system is swept to a detector system by use of an inert carrier gas. Quantitative measurement, by means of gas chromatographic (G.C.) analysis, are then employed to determine the presence of a partial pressure differential or a concentration gradient between the two cell chambers. Hernandez (1984) and Baner, et a1. (1984, 1986) have employed both isostatic and quasi-isostatic test methods in evaluating the diffusion of organic penetrants through barrier films. The Quasi-Isostatic Method is a modification of the 14 Isostatic Method. Murray & Dorschner (1983) developed an accumulation method for determining the permeability of organic vapors through barrier films and in a more recent publication, Murray (1985) expanded on this procedure and reported a number of examples for which the test apparatus was employed to determine the "relative permeation rates" of organic vapors through barrier structures. Three related procedures which are based on the Quasi- Isostatic method are (Giacin, 1983): (1) (2) In this process, the system is maintained at atmospheric pressure. One chamber of the test cell allows for a continuous flow of permeant vapor while the other portion(s) of the cell system contains the gas or vapor which has permeated through the test material. Samples are withdrawn from each of the concentration chambers of the cell system at predetermined time intervals and injected into the G.C. for quantitative measure. From the concentration data, the film permeability can then be determined. In this technique the penetrant gas or vapor thoroughly purges the upper chamber of the cell which is then closed. G.C. analysis is used to monitor the gas concentration in both the upper and lower chambers initially, and throughout the permeation run at predetermined time intervals. 15 Permeability can be determined from the difference in pressure between the cell chambers or from concentration data. (3) In this method the lower chamber of the permeation cell is filled with a liquid above which is maintained permeant vapor at a constant partial pressure. The vapor permeating into the upper chamber is then monitored as a function of time until the permeation rate reaches a steady state. In most cases, the concentration of permeant in the upper chamber is determined by G.C.. Odor/flavor quantification is most commonly obtained through means of Flame Ionization Detector (FID) systems, such as that used in G.C. analysis. "Ionization detectors operate on the principle that the electrical conductivity of a gas is directly proportional to the concentration of charged particles within the gas." (Giacin, 1983) "The effluent (flowing out or forth) gas from the column is mixed with hydrogen and burned in air or oxygen." "The FID responds to virtually all organic compounds. The lack of response to air and water make an FID especially suitable to headspace analysis of aqueous samples." Gas Chromatography(GC), Gas Chromatography/Mass Spectrometry(GCMS) and High Pressure Liquid Chromatography (HPLC) methods are frequently utilized in mass transport studies. Separation and identification of several hundred 16 to thousands of components naturally occuring in aromas may be accomplished by such methods. It is possible to correlate the quality of a food product based upon permeation and sorption values with that relating to sensory evaluation. Recently, DeLassus (1985) described the transport of d- 1imonene vapor through a series of polymer films typically used in food packaging. Permeability (P) was determined from experimental data, diffusivity (D) was calculated from a derived equation, and the solubility coefficient (S) was calculated from the equation P'- D * S using measured values of P and D. In this study, limonene was used as the penetrant vapor, with argon as the carrier gas. Measurement was made through means of a Photoionization Detector. Five films, of different structures, were evaluated. The data showed that as the test temperature increased, the permeability and diffusivity values rose, while a decrease in the solubility coefficient was observed. The permeability (P) of a SaranR film was found to be up to four orders of magnitude lower than that determined for a polyolefin structure. However, little difference was noted with regard to the solubility coefficient values. Theoretical analyses (or modeling) were applied to two product/package systems: Package System (1), a rigid 6 oz. polypropylene (PP) container and Package System (2), a thin- 17 walled LDPE package. Orange juice was contained in each of these containers. A summary of DeLassus's findings are described below. Pkg. System (1): The initial quantity of d-limonene in the container (i.e. 0.09 gm) was determined to be well above its saturation in either water or sugar water therefore the contents would provide a concentration pressure gradient of d-limonene. Information on the vapor pressure of d-limonene at two specific temperatures of interest (25°C and -5°C) was obtained from the literature. Permeability of d-limonene and the solubility coefficient values were estimated from theoretical analysis to determine the steady state rate of diffusion at each temperature. In calculating the potential loss of d-limonene due to sorption, the values obtained were higher than the initial amount of d-limonene in the product. Application of the expression, (Where: D = diffusivity: L = film thickness: and t1/2 = the time required to reach half of the steady state rate], indicated that the time to reach a steady state rate of diffusion was very long, being on the order of 103 to 105 days. Therefore, DeLassus concluded that for the rigid 18 Polypropylene container, all important losses of d-limonene were sorptive. As previously noted, the solubility values were essentially the same for both polymers. The permeability of the SaranR polymer is much lower than the PP, therefore virtually all of the d-limonene pressure drop will be across the barrier layer, SaranR. When adding a layer of SaranR to the inside wall of the PP container, the permeability barrier is significantly improved. Thus, the PP layer becomes temporarily isolated from sorption, reducing the sorption capacity of the container. Pkg System (3): DeLassus's results suggest that for the thin-walled LDPE package, a very significant loss due to permeation would occur within one day. Losses due to sorption however, would not be as severe when compared to the permeation barrier requirements. The author proposed that the permeation rate of d-limonene through the LDPE package could be reduced by as much as 10'6 g/day by adding a barrier layer of SaranR to the LDPE structure. Placement of the barrier material in the overall structure is not as critical a factor in this package system. Presumably, the sorption loss in the LDPE layer would occur quickly as compared to several months for loss to be detected in the barrier layer. Applying the barrier film to the skin layer between the LDPE and the contents would reduce sorption in 19 LDPE to approximately zero. Conclusions: "A barrier layer would help both the rigid container and the thin-walled container. Sorption is controlled in the rigid container. Permeation is controlled in the thin-walled container." If the quality of a product, based on its aroma characteristics can not be maintained, "it makes little difference if a molecule is lost by permeation to the enviornment or by sorption in the package wall". AROMA Defining odor is a difficult task. If one were to establish that odor is "that which can be smelled" then the question remains "by whom?". Since it is impossible to deine odor in physical terms it has been suggested that it be defined in terms of "phenomenolgy"2. Sagarin (1954) proposed a phenomenological definition: "Odor is the property of a substance that is perceived, in the human and higher vertebrates, by inhalation in the nasal or oral cavity; that makes an impression upon the olfactory area of the body: and that, during and as a result of such inhalation, is distinct from seeing, hearing, tasting, or feeling, and does not cause or result in choking, 2The scientific investigation or description of phenomena. 20 irritation, cooling, warmth, drying, wetting or other functions foreign to the olfactory area. A phenomenological definition places the responsibility on the individual." Odor can also be defined in "physiological"3 terms: "sensations perceived from responses of the olfactory nerve or first cranial nerve" (Amerine, et al., 1965). The physiological significance of an odorous substance is that it can easily stimulate appetite. Because odors may attract or repel consumers, the food industry is cognizant of its importance. The task of producing, increasing, or maintaining aroma is not the responsibility of the food technologists alone but also includes those involved in the preparation, processing, packaging and storage of foodstuffs. Odor: Characteristics 5 Molecular structure The detection of odor occurs when molecules of volatile substances reach the olfactory receptors at the top of the naval cavity. All known odorous substances are gases having a high vapor pressure and boiling below 300°C (Amerine, et al. 1965). Odor is perceived from molecules having neither too low nor too high a molecular weight. Odor intensity increases with increasing molecular weight, however, "the characteristic odor of a chemical compound (which is 3pertaining to the functions of living organisms. 21 specific for that compound) decreases with molecular weight in a homogeneous series" (Amerine, et a1, 1965). Compounds having a molecular weight greater than 300 are characteristically odorless, according to Stoll (1957). "This is particularly due to the low volatility of such compounds and partially to differences in structure" (Amerine, et al., 1965). Research on aromas is difficult due to associated complexities. An enormous variety of organic materials exist each having its own structural significance (aliphatic, aromatic, saturated and unsaturated hydrocarbons, etc...). Molecules of quite dissimilar structure may have similar odor properties. Amoore (1952) stipulated "that the odorous properties of any compound depend on its volatility and on the size, shape, and electronic status of its molecule." "Double bond and ring structures are associated with odor." "The quality and intensity of odor are influenced by the position of the double bond in the molecule, the distribution of electrons, resonance or induction of the molecule (particularly in the 5- or 6- membered rings), and the kind of group adjacent to the osmophore. [Osmophore refers to a chemical entity which confers odor on an otherwise odorless compound. Strong osmophores include phosphorus, arsenic, sulfur, selenium, chlorine, and bromine (Amerine, et al., 1965).] In general, 22 molecules with greater adsorption capacity are more odorous" (Amerine, et al., 1965). Rigid molecules of specific shape have been found to be more effective olfactory stimuli than flexible molecules. Odor may be attributed to the internal attractive forces of the compound and by the size and shape of the molecule. Odor may also be influenced by the polarity and form of the molecules (DeLassus, 1985; Amerine, et al., 1965). Cyclic or polycyclic compounds (rigid structures) are more odorous than aliphatic compounds (less rigid structures). "When the stereoisomerism is the result of a ring, the kind of odor and their intensity vary. Optical isomers generally have very similar odors. The odor of cis and trans - isomers is very distinct but their intensities are about the same (Amerine, et al., 1965). When performing research where odor is of prime importance, a volatile organic compound (or compounds which contribute to product aroma) should be identified prior to the initiation of work. Such a consitiuent (or constituents) should be selected based on contribution to perceived product quality. In monitoring the behavioral characteristics of odorous substances, a sample must be isolated under controlled conditions. 23 Sensory Analysis (selecting panelist) Observation based on past research shows that the sense of smell is more highly developed than the sense of taste (Parker & Stabler's 1913). The human sense of smell is the best way by which to initially detect the quality of a food product, but unfortunately odors cannot be measured quantitatively by the nose. In conducting experiments it is difficult for normal individuals to indicate differences associated with odorous substances. For sensory evaluation to be significant, a panel should consist of either six highly trained judges or fifty untrained judges. Semi-trained judges may also be used if at least ten to fifteen individuals are able to differentiate and identify a set of samples (Filadelfi, 1985; Hamilton, 1985). Although not always possible to find, highly trained experts are more useful in evaluating quality. Metzner (1943) noted that a connoisseur is not necessarily more sensitive to stimuli but instead attributes his or her skill to knowledge of what signs to look for and how to interpret them. If threshold tests are selected then judges must have the ability to detect specific properties (semi-trained or trained individuals). Laboratory panels are generally used as qualified judges in studies involving human perception of food attributes. Threshold refers to the "value which gives 24 the limiting concentration at which an odor is still just perceived" (Amerine, et al., 1965). Consumers are seldom trained and presumably react similar to an untrained laboratory panel. Consumers, normally unfamiliar with taste tests, are generally influenced by appearance. Many times color changes are accompanied by undesirable change in odor. Therefore, objectives should be established (initially) and well understood by all subjects involved in sensory evaluation. Food attributes, such as color and texture should not distract panelists when performing odor evaluations on products. Many precautions should be followed when doing taste panel work. Panel members should avoid such things as coffee, mints, smoking, perfume, and any other substances which may interfer with the odor of interest, at least 30 minutes prior to the test (Filadelfi, 1985). Hammer (1951) found that sensitivity increases throughout the day if no lunch is acquired. Furchtgott and Friedman (1960) found that a mild degree of hunger lowers the olfactory thresholds - but only slightly and not in all individuals. In physiological terms, research indicates that appetite is affected by odor. After meals, Mancioli (1921) observed a decrease in olfactory insensitivity which was attributed to excessive stimulation of the olfactory region (during eating). Olfactory sensitivity may be due to 25 alcohol, sugar, and amphetamine (10 mg) (Amerine, et al., 1965). By including tannic, tartaric, or acetic acid with a meal preserves after eating acuity. Kuehner (1954) pointed out that extreme variations in sensitivity may occur in an individuals response and found it necessary to "standardize" a subject from day to day. With some compounds the age of a panelist may effect the olfactory threshold (Fortunato, 1958: Fortuneto and Niccolini, 1958). The ability to smell is limited by a fatigue factor causing interference in the detection of similar odors, but rarely affects the detection of dissimilar odors. Interference may be caused by several factors (Amerine, et al., 1965) including: fatigue, obstructed nasal passages, partial anosmia (loss of the sense of smell), brain lessions, or injured olfactory. In addition, it is possible for one odor to overwhelm another. Fatigue tends to set-in and remain more so in sniff studies than in sight, sound, or taste evaluations (Amerine, et al., 1965). In certain cases, loss of the sense of smell may be beneficial providing it's a different odor than that being evaluated. Sensory rooms should be ventilated, have odor free air, good lighting, and partitioning between individuals. Odors move downwind, thus in certain situations, some odors could be detected from a distance causing erroneous conclusions. 26 The period set aside for sniffing should also be controlled. The time of day, number of samples per test period, frequency of tests per day (and/or per week), and time allotment per sample, are all factors to consider (Filadelfi, 1985). Interference may arise if care is not exercised in selecting the appropriate medium for test specimens. The material selected must maintain and not react with the quality of a particular product (Filadelfi, 1985: Harte, 1985). The container opening should be wide enough to allow a subject to sniff an adequate quantity of the odor upon first opening the sample. Thus the type and ease of opening of a container is also of great importance. An adequate and simple technique for odor (threshold studies) determination is through sniffing. This technique is popular, inexpensive, and in most cases easy to perform with large numbers of subjects. Means of expressing the intensity of the stimulus vary. Analysis involving new or improved packaged products may be evaluated through means of scoring tests. Large consumer groups are necessary to determine consumer reaction to new or improved packaged products. This type of test is also valuable for quality control, storage stability, screening of intensity levels, and measuring panel characteristics (reproducibility). In detecting a difference between two items, a paired- 27 stimuli procedure is followed (Harte, 1982). A similar study, involving three samples, is known as a triangle test. Here one of the three samples is identified as being different. A test more effectively applied by untrained consumers refers to Hedonic scaling (Amerine, et al., 1965). Judges express the degree of liking by selecting a point on a scale ranging from extreme disapproval to extreme approval. Values obtained are treated by rank analysis or analysis of variance. Sensory panels may base their judgements on an acceptance or preference basis (Amerine, et al., 1965). In accepting a product, the panelist shows a willingness to use or eat a product. Where as, preference relates to a greater degree of acceptance of one product over another. Sensory stimuli makes it possible to recognize certain foods and choose food in accordance to our needs. It also initiates appropriate responses in establishing digestion, promotes satiety, and makes it possible to anticipate the pleasure of eating (Amerine, et al., 1965). Odor is often the key to consumer acceptance of foods (Hamilton, 1985). Detecting a difference in taste is seemingly an easier task than with odor. Measurement of the odor threshold, which directly relates to loss of quality can be a difficult and lengthy task (depending on the product and type of organic compound). 28 Odor quality is retained more precisely than odor intensity (Amerine, et al., 1965). Moncrieff (1957) termed odor-intensity as "the ratio of the olfactory threshold determined after sniffing the undiluted substance, to the threshold determined after sniffing the diluent". In other words, the more diute a sample, the less smell detected, and the lower the odor-intensity [full strength/diluent]. The human sense of smell is the best means by which to initially detect the quality of a food product, but unfortunately these odors cannot be measured quantitatively by the nose. Therefore, sensory tests are correlated with chemical measurement in the determination of a products quality and/or its shelf life (DeLassus, 1985: Wyatt, 1985: Zobel, 1985: Amerine, 1965). MATERIALS AND METHODS MATERIALS Analytical grade d-limonene, obtained from Eastman Kodak Company (Rochester, N.Y. 14650), was used throughout this study. The films which were evaluated included two typical cereal package liners, namely a High Density Polyethylene/Sealant laminate and a Wax/PVOH4/Glassine/PVOH/Wax structure. Throughout the remainder of this thesis, the d-limonene will be refered to simply as limonene, the High Density Polyethylene/Sealant Laminate liner will be referred to as the HDPE base structure and the Wax/PVOH/Glassine/PVOH/Wax liner will be referred to as the glassine base structure. The composition of the two film structures are as follows: * HDPE BASE STRUCTURE (1.9 mil Thickness) POUND/3000 ft2 Total weight 33.5 HDPE 27.9. Sealant ' 5.6 * Obtained from supplier of film. 4 PVOH refers to Polyvinyl alcohol. 29 3O * GLASSINE BASE STRUCTURE (2.2 mil Thickness) £0UNDS£3099 {:2 Total weight 38.3 Wax 9.0 PVOH (polyvivyl alcohol) Coating 2.4 Glassine 26.9 * Obtained from supplier of film. Film samples were stored in a desiccator over CaSO4 desiccant (0% relative humidity) at ambient temperature (23 C) prior to testing. The product chosen for evaluation involved a fruit flavored cereal, whose quality is associated with the retention of volatile aroma constituents. The product was purchased, as fresh as possible, from Michigan State University stores. EXPERIMENTAL METHODS Permeation Measurements The transmission rate, as determined by a Quasi-Isostatic method, is defined as that quantity of vapor passing through a unit area of the parallel surfaces of a plastic film per unit time, under specified conditions of test. This procedure is refered to as "Quasi-Isostatic" because the test compartments are maintained at an 31 essentially constant total pressure of 1 atmoshpere. It is, however, an accumulation procedure where permeant collects, as a function of time. A schematic diagram of the permeation test apparatus is presented in Figure 1. A constant concentration of permeant vapor is produced by bubbling nitrogen gas through the liquid permeant. This is carried out by assembling a vapor generator consisting of a gas washing bottle, with a fritted dispersion tube, containing the organic liquid. The apparatus was designed to have the capability of controlling penetrant concentration through a broad range of levels. The effect of limonene partial pressure on the permeation rate and permeability coefficient was determined over a concentration range of 0.4 to 3.6 ppm (HDPE based) and from 1.5 to 4.8 ppm (Glassine based) for the respective film 'structures. Vapor concentration (ppm) in nitrogen is expressed throughout on a weight per volume basis. .The permeability studies were carried out at temperatures ranging from 21.1°c (70°F) to 26.7°C (80°F) and approximately 0% relative humidity. To obtain a low vapor concentration, the permeant vapor stream is mixed with another stream of pure carrier gas (nitrogen). Before being directed to the permeation cell, the vapor stream was passed through a glass reservoir as a means of dampening perturbations. The vapor generator system was mounted in a constant temperature water bath, 32 Z n In N Am 1 Tank L ......... J B = Water Bath 81 Glass Mlxlng Devlce 2 Glass Vapor Generator R. = Regulator V = Needle Valve R C F T = Rotameter = Cells (Double Chamber) = To Waste and Gas Flow Bubble Meter = Three Way Valve Figure l. SCHEMATIC OF OUASl-ISOSTATIC PERMEATION TEST APPARATUS 33 maintained at 1°C above ambient temperature so as to avoid condensation after the permeant vapor passed through the glass reservoir. As shown, flow meters were used to provide a continuous indication that a constant rate of flow was maintained. A micro-flow meter was utilized for tests performed at the low permeant concentration levels. For a better illustration of the permeation apparatus, see Figure 2. Care is taken to ensure that the permeation cell is free of limonene vapor prior to the initiation of each run. This is achieved by mounting a sheet of foil in place of a film sample and allowing the system to set for a period of 2 to 3 days, under closed conditions. After which time, the headspace of each chamber is measured for trace amounts of residual limonene which may have leached off from the side walls of the cell. If any limonene is detected, the cell is disassembled and baked in a 43°C (110°F) oven for 3 to 4 days and then re-evaluated. This procedure is performed following each permeation test and repeated until the system is clean. The permeability of the respective films was determined under identical conditions, so as to compare their relative barrier properties. Duplicate runs on the same film type are carried out simultaneously in specially designed permeability cells. Figure 3 provides a detailed view of the Permeant Cell System. Each permeability cell, 34 . . . . _ . 2:3: .- E .r. r .2: ., : Hmwfi COHfimmfiHOm OfluvMUmOmHIHmMDO .N ounwflm 35 Lego! e325 30.... .50 use See? 0... I u .30 22¢ I 0: ..e0 h3:06 I 00 :00 :04 I 04 3.0a 353.com I o 2...) e500... I > 33530: I a some) asses—sen. I >1 .2326 .33 5325.. .n .53 36 constructed of stainless steel or aluminum, is comprised of two cell chambers and a hollow center ring. Both cell chambers and the center ring are equipped with an inlet and outlet valve and a sampling port. An assembled cell is shown in Figure 4. Unless otherwise stated, the low concentration cell chambers each have a volume of 50 cc, and the volume of the center cavity is approximately 90 cc. Tests were also performed utilizing permeability cells having a smaller volume for the center cavity (50 cc)(see Appendix A). In this case, the low concentration cell chamber volumes remained at 50 cc. A comparison of these volumes is shown in Figure 5 for the center chamber of the two permeability cells utilized. In operation, test films are mounted in the permeability cell so that the center ring effectively isolates the right and left cell chambers. Hermetic isolation of the chambers from each other and from the atmosphere is achieved by compression of overlapping Viton "0" rings (from Detroit Ball Bearing Company) on the film sample. Viton is a fluorocarbon elastomer which is resistant to attack and swelling by most organic vapors. For the permeability cell with the lower center cavity volume, isolation of the cell chambers from each other and from the atmosphere was achieved through compression of the film against a smooth metal face which resulted in a metal/film/metal seal. 37 Figure 4. An Assembled Cell System Figure 5. Comparison of Cell Volumes 38 In use, the films to be tested are mounted in the permeability cell and the cell assembled. Unless otherwise stated, a constant concentration of permeant vapor is then flowed continuously through the high concentration (center) chamber of the permeability cell at a flow rate in the range of 10 to 40 cc/min. Initial studies with the HDPE based structure revealed a problem related to maintenance of a constant limonene vapor concentration within the center cavity of the permeability cell, particularly in the early stages of the diffusion process. Adjustment of either the center cell cavity size or the rate of permeant flow, however, eliminated this problem and allowed for a more representative collection of data. By increasing the flow rate of the penetrant vapor through the center chamber, rapid displacement of the void volume was successfully achieved. This in effect, decreased the time necessary to establish a steady concentration prior to actual permeation through the test material. Therefore, in two cases of limonene vapor concentration, a higher flow rate ranging from 130 to 140 cc/min was employed (see Appendix A for further details). As shown in Figure 1, to perform multiple runs concurrently, a series of four (4) cells can be attached to a dispensing manifold which allows delivery of a constant concentration of permeant vapor to each cell. This allows the permeability of up to eight film samples to be determined concurrently, under identical conditions. 39 The increase in penetrant level in the low concentration cell chambers is determined by gas chromatography analysis with flame ionization detection. At predetermined time intervals, an aliquot (500 ul) of headspace is removed from the low concentration cell chambers with a gas tight syringe (Hamilton no. 1750, side port type) and injected directly into the gas chromatograph for quantitation. A constant total pressure of one atmosphere is maintained in both the upper and lower cell chambers by replacing the sample volume removed with an equal volume of pure nitrogen. Samples are removed a number of times over the period of test and an array of time vs. area response values recorded. To evaluate the concentration dependency of the diffusion process, permeation runs were carried out at several penetrant concentration levels. The transmission rate (P) and permeability constant (P) values were determined from the resultant transmission data. To determine the diffusivity and permeability values, the increase in penetrant quantity in the lower concentration cell chambers was plotted as a function of time and the resultant transmission profile related to the permeability of the film sample. The lag time (9) value is obtained experimentally as the intercept on the time axis of the steady rise portion of the penetrant-time plot and the apparent lag time diffusion coefficient (Dlag) determined 4O (Barrer, 1939) by: Dlag = 12/66 (3) where: Dlag = apparent lag time diffusion coefficient (cmZ/sec) l = total film thickness (cm) 9 = lag time (sec) The lag time diffusion coefficient for laminate structures is considered an apparent diffusion coefficient value, being a composite of the diffusivity properties of the respective individual laminate layers. Due to the complex nature of the glassine base liner material, an apparent diffusion coefficient was not determined for this structure. The time interval during which the permeability data was evaluated to obtain a steady state rate of transmission was determined by graphical analysis of the time versus area response values. In all cases, the data was evaluated statistically by linear regression analysis to obtain the best straight line fit. Sorption Measurements Sorption measurements were carried out on a Cahn-RG Electrobalance by the continuous flow method (Cahn Instruments Inc., Cerritos, CA). The electrobalance and sample tube were maintained at a constant temperature of 21.5 + 0.5°C. A schematic diagram of the test apparatus is 41 shown in Figure 6. As shown, the polymer film sample is suspended directly from one of the arms of the electrobalance and a constant concentration of penetrant vapor is flowed continually through the sample tube (hang-down tube), such that the polymer sample is totally surrounded by the vapor. A constant concentration of penetrant vapor is produced by employing a vapor generator system, similar to that described above. The level of limonene concentration found in the glass hang down tube is dependent on the amount of nitrogen gas flowing over the liquid phase (limonene) and on the amount of nitrogen gas selected for mixing. The equilibrium sorption and solubility coefficient values of limonene in the films were evaluated within a vapor concentration range of 0.3 to 7.0 ppm for the HDPE based structure and between 1.5 and 7.3 ppm for the Glassine based structure. The test system, as designed, allows for the continuous collection of sorption data from the initial time (t = 0) to the time the system has equilibrated (steady state conditions), as a function of penetrant concentration. The penetrant vapor surrounding the test film, is quantitatively measured by a G.C. procedure. Using a gas tight 500 pl syringe, a portion of the vapor within the sample holding facility is removed and injected into the G.C. 42 .32 509:2 I p 00:23 305020 500.53 .255 It 03? >02, 095.12. 030) 0.30th econ salabsas ButteuLm Bomb: .85 aEmuum rose 233 26: mood. 5.: msanmLm :5 6.23.50 too o_aEcmlm HEELS BSQEOULU .08.:momnom 509.22 E 0335 omega ..oac> aebeEobom .. m 2006.3 .0 5:90:00 .53 .303 am m2e on» ma 05He>on .oanofl x omn\~aoIoV .mIOH x Homn> osocoafia Ema « noshaom v\ecocoaaa mAOV .muoa x uoahaom m\osocofifia mAnV .>\u3 .ammAsV "we commoumxo mugs: m CH N.mHH o b h m ¢.Nw v m v.5 H.Nm N.v n b h.mH o N o h n.0H m H m b N.N n o AmVAavusmeomeuumoo Aeckovuseeosuumoo Amvmneausflensaom Aaesoeuanusmosoo comeMuwa wufiawnsHom Baaunaaasvm uoms> ososofiaa SOwHMHuGOUGOU HOQM> QEOCOEHA MO BOflHOr—Pm é m4 wusuosuum mmem use: you sumo suaaensHom H OHQMB 50 .mcsu manuaamou o3» unwed us no omnuo>n on» ma osae>acv .nroa x Momm> ococoawa Sun a uoahaom o\ocosoaaa maov .nuoa x noshaom.m\ocmcoawa many .>\u3 .Emmanv "mm commoumxo hues: w.m w.ow m.b n.~ e.vH v.w v.H ¢.m m.¢ m.a m.< m.~ o.a m.a m.a Leekovusmeoauuwoo AeVIQVsuHHHnsHom Aevsoeusuusmosoo auflawnsaom asfiunfiaflsum uome> ococosflq soflucuucoosoo uomm> ecosoawq mo coauOCSh d ma musuosuum ommm osflmmnaw you even huwafinsHom N OHQMB 51 b e EXPERIMENTAL DATA 1'2 —-CALCULATED FROM EQUATION 5 1.0 - 0.8 - Mt M . 0.6 - 0.4 s 0.2 e o 1 A J J l 0 100 200 800 Time” (secondsv’) Figure 9. ABSORPTION OF D-LIMONENE IN HDPE BASE STRUCTURE AT 1 .5 ppm AND 20.5 C 52 concentration of 1.5 ppm (wt/v). Superimposed in Figure 9 is the sorption profile obtained by solution of the sorption vs. time data fitting equation (Equation 5). 8 -D *Trz * t 1 -9D *772 * t -- = 1 - -- [exp( ----------- ) + ' 3XP( ------------ )] (5) M... 7T2 12 9 12 Equation 5 was derived from the equation, Mt 9° 8 [-(2n + 1)27T2 Dt] -- -.- 1 - Z -------- §--2 exp [ --------- 5 ------ J (5) M” n=0 (2n + 1) 7r [ l I by taking the first two terms. As shown, good agreement was obtained between the experimental data and that calculated from the theoretical expressions. The general shape of the sorption curve supports the assumption that for the HDPE based structure the diffusion process followed apparent Fickian behavior (Fujita, 1961; Crank and Park, 1969) at the limonene vapor concentration level of 1.5 ppm. That is, the diffusion coefficient is not time dependent (Crank, 1975). Sorption profile curves were also determined for the HDPE based structure, at d-limonene concentration levels of 3.9 ppm, 4.1 ppm, 6.3 ppm, 6.6 ppm and 7.2 ppm. These curves are shown in Figures 10 through 14 respectively. For each limonene vapor concentration studied, the experimental and calculated results, following the same procedure as above, are presented. As shown, at the higher d-limonene vapor 53 1.4 H EXPERIMENTAL DATA -' -— CALCULATED FROM EQUATION 5 1.2 .. 1 ’— __ w‘“ .8 '- M Moe 6 i- .4 - .2 I" 0 1 1 O ‘IOO 200 300 Time/2 (seconds‘l’) Figure 10. ABSORPTION OF DLIMONENE IN HDPE BASE STRUCTURE AT 3.9 ppm AND 20.5c C 54 1.4 0 EXPERIMENTAL DATA 1 2 r—CALCULATED FROM EQUATION 5 “O .f 8 _ V M" .6 r- M I 7 4 _ d 2 .- J o J 1 O 100 ZOO 300 TimeV2 (secondsv’) Figure 11. ABSORPTION OF D-LIMONENE IN HDPE BASE STRUCTURE AT 4.1 ppm AND 20.5 C 55 1.4 O EXPERIMENTAL DATA 1,2 .- -—CALCULATED FROM EQUATION 5 O 1 1 O 1 CO 200 300 . T Turnev2 (seconds/1') Figure 12. ABSORPTION OF D-LIMONENE IN HDPE BASE STRUCTURE AT 6.3 ppm AND 20.5 C 1.4 56 1.2- . EXPERIMENTAL DATA 1 » l -—-CALCULATED FROM EQUATION 5 I 1 DO 200 Time” (secondsV’) , 300 400 Figure 13. ABSORPTION OF D-LIMONENE IN HDPE BASE STRUCTURE AT 6.6 ppm AND 20.5° C 57 1 .4 0 EXPERIMENTAL DATA - - CALCULATED FROM EQUATION 5 1.2 - l l J 0 100 200 300 400 450 J Time"’ (secondsV’) Figure 14. ABSORPTION OF D°LIMONENE IN HDPE BASE STRUCTURE AT 7.2 ppm AND 20.5‘ C 58 concentrations, there is apparently non-Fickian relaxation controlled sorptibn which results in additional sorption of the penetrant. A similar sorption process has been described by Berens (1977) for a vinyl chloride/polyvinyl chloride system and by other investigators (Bagley and Long, 1958: Fujita, 1961). The Mt/M vs. tl/2 curve for sorption of limonene by the glassine based structure, at a concentration level of 2.5 ppm, is shown in Figure 15. The data shows what appears to be a smaller relative amount of rapid sorption, followed by a slower approach to apparent equilibrium. In contrast to the apparent Fickian behavior of the HDPE based structure, the sorption of limonene by the glassine based structure appears to be more complex and showed a time dependency. While this phenomenon is not totally understood, it can be attributed to the rapid uptake of limonene vapor by the external wax layer followed by a slower approach to steady state due to diffusion through the PVOH barrier layer. The results of the equilibrium vapor pressure measurements on the HDPE structure and the glassine structure are shown in Figure 16, where limonene solubility is plotted against vapor concentration for the respective films. Equilibrium solubility was determined by: weight of penetrant uptake at equilibrium Cs ' """""""""""""""""""""""""" (7) weight of polymer 59 0 EXPERIMENTAL DATA —SMOOTH FIT CURVE 1.2 r M- _L J . 200 300 J 0 100 1 I Time-7' (second57 ) Figure I5. ABSORPTION 0F 0- LIMONENE m GLASSINE BASE STRUCTURE AT 2.5 ppm AND 20.5'C n NENE - Cs,SOLUBILITY < gglPOLMcIMER no’) .s 3‘ r _s N d O 60 0 HDPE BASE STRUCTURE 0 GLASSINE BASE STRUCTURE "" SMOOTH FIT CURVE 0 2 4 6 LIMCNENE VAPOR CONCENTRATON (ppm) Figure I6. LIMONENE VAPOR SOLUBILITY IN HDPE BASE AND GLASSINE BASE STRUCTURES 61 As shown, over the lower portion of the concentration range studied, the data follow Henry's Law: es s k cv, (8) where C8 is the equilibrium solubility in g limonene/g polymer, Cv is the vapor concentration in ppm (wt/v) and k is a constant. According to Henry's Law the amount of limonene sorbed by the film structure should be directly proportional to the level of limonene vapor surrounding the film sample, at equilibrium. However, at vapor concentrations above 5.5 ppm deviation from Henry's Law is observed. As shown, the limonene solubility in the glassine structure is substantially lower than in the HDPE structure, at the same limonene vapor concentration and temperature. Equilibrium distribution of limonene vapor between the fruit flavored cereal, and the packaging liner systems would result in a much lower limonene concentration in the glassine structure than in the HDPE structure. This can be of paramount importance in avoiding the effect of "flavor scalping" or loss due to sorption, for a product whose quality is associated with the retention of volatile aroma constituents. Solubility coefficient values were determined from the equilibrium sorption data, where: 62 weight of penetrant uptake at equilibrium 5 a ---------------------------------------------- (9) weight of polymer * vapor concentration in ppm The results for the respective films are shown in Figure 17, where the S values are plotted against the relative limonene vapor concentrations. The solubility coefficient values over a limonene vapor concentration range of 0.3 to 5.0 ppm were essentially constant for the HDPE based structure with g limonene s = 7.7 x 10'3 ------------------------------------- . (10) g HDPE structure * ppm limonene vapor Above a limonene vapor concentration of 5.0 ppm, the solubility coefficient appears to be concentration dependent and must be determined for the specific concentration of interest. Similarly, the solubility coefficient (S) determined over the concentration range of 1.5 to 5.5 ppm for the glassine structure can be taken as essentially constant and estimated by linear regression analysis to give 9 limonene s = 1.5 x 10"3 -------------------------------------- . (11) g Glassine struct * ppm limonene vapor The data suggest that at the higher concentration levels, swelling may be induced, thus changing the structural characteristics of the liner materials. In the process of swelling, polymer relaxation occurs forcing the molecular bonds to expand. This increases the molecular space into which penetrant molecules may absorb. In general, the further apart the molecular bonds, the higher the penetrant solubility and the lower the barrier properties. S, SOLUBILITY COEFFICIENT x106) LIMONENE 63 18, . HDPE BASE STRUCTURE 0 GLASSINE BASE STRUCTURE 15. —SMO0TH FIT CURVE “I 14» 12> 10r 8>,,______¢_ o 5. 4r- 2*' e O 90 2 4 6 LIMONENE VAPOR CONCENTRATION (ppm) Figure I7. LIMONENE VAPOR SOLUBILITY COEFFICIENT 0F HDPE BASE AND GLASSINE STRUCTURES 64 PERMEATION STUDIES The results of the studies on the diffusion of limonene vapor through the HDPE and the glassine structures, as a function of penetrant concentration, are summarized in Tables 3 and 4. The data show, as anticipated, a concentration dependent permeability constant (P) for the respective film samples, with P increasing with an increase in penetrant concentration. The concentration dependent permeability constant suggests penetrant/polymer interaction, or swelling of the polymer matrix by the sorbed limonene vapor, which can result in alteration of polymer chain conformational mobility and thus penetrant diffusivity. If there was no interaction between the penetrant and film, the permeability constant should remain constant over the entire range of penetrant concentrations investigated. Representative transmission rate profile curves of the HDPE based structure, for the respective limonene penetrant levels investigated, are presented graphically in Figure 18, where the total quantity of limonene permeated is plotted as a function of run time. Figure 18 illustrates the effect of penetrant concentration on transmission rates and lag time values for the limonene/HDPE system. The transmission profile curves of the test film show an induction period, followed by a non-steady state rate of diffusion, after which a constant transmission rate is observed. 65 .cfls\oo and was Robsmno Haoo coaunuucmucoo no“: nmsounu ouch aonAev .»e u mean you .oos x Ixxev .sufiaansesa> «0 usmsoeuumoo maeum> no ucoaoauuooo oveuo>n on» ma osHe>AnV .mcsu on» no omusoo may ocflusu we I\+ mo coflunfl>mc oasuo>n an o>nv A>\p3v neweuucoocoo Roan> ecosoawqanv m.w me man o.n b.m ms bag H.m H.h Hm Hmn w.~ ~.o moa mma m.H N.m mma moa m.H Imvo.s Amvna .mveea Amvo.a m.v and Nb «.0 oaroa x oom\meo Amoussaav and a mum « ma A>\u3 .ammv ucofiowmuooo Anvmfiaa ousuonuum a m coflusuusoosoo soflmsuufio own usofiofiuumoo Anvuomc> ococoaflq AcVAbvucoumaad “ovanvhuflawnwosuom coflucuucoosoo uoan> ocososfiq no scauocsm < m< ousuosuum omnm man: you name aufiaannosbom m OHQMB 66 .se.~n u A non .ooH x Ixxev .nusasnanua> no usmsoeunoou mouse>n on» ma 09He>anv .mssu one no chance on» manuso »¢.m I\+ no conumn>oc oonno>c cs o>sm A>\u3v soneuucoosoo some> ococoanqacv hm.H m e om.H m g mN.o m.n no.0 m N no.0 m H Ema « map a ma ousuosnum a m, A>\u3 .Emmv “ovanvusononnnmoo muHHHnnmanom Anvconusuucoocou uoms> osocofinq conusuusoocoo nomn> mcocofinq no cannonsm < m< musuosuum omnm ocnmmsau Ron sumo annannmoeuom e OHQMB ABSOLUTE OUANTITv PERMEATED (groms x104’) 67 700 630* r ‘ 0 0.4 ppm I I 0 1.3 ppm 560: ‘ 1.9 ppm 4 2.6 ppm 490* D 3.6 ppm 420* 350' I ‘ 280: 210: I 140 70* o . O O 0‘ A ’__’- 0 40 80 150 180 200 240 280 320 360 400 Time ( minutes) Figure I8. THE EFFECT OF VAPOR CONCENTRATION ON THE DIFFUSION OF LIMONENE THROUGH HDPE BASE STRUCTURE AT 23°C 68 As permeation begins to occur, the partial pressure or concentration gradient across the high concentration surface and the low concentration surface of the test film changes. As a result of this change in concentration gradient, the permeation rate of limonene, by the test film will be reduced. In an attempt to treat the data in a consistent manner over the entire concentration range investigated, a standard percentage of vapor permeating from the high concentration side to the low concentration side was selected, above which the data was not included in the analysis. For example, permeability measurements were terminated when 12% to 14% of the driving force concentration was found to have permeated through the HDPE structure and into the low concentration chamber of the cell system. From the transmission data, the permeability constant and lag time diffusion coefficient values were obtained by standard methods (Barrer, 1939: Crank 8 Park, 1968)(Appendix A). It is important to point out that the diffusion coefficient values for the HDPE structure were calculated from the transient state region of the permeation curves and may not necessarily represent the actual diffusion coefficient values at steady state. Therefore, whenever diffusion coefficient values are mentioned, the ”apparent" values are presumed. As expected, the lag time values were found to be inversely related to the vapor driving 69 concentrations. Shown in Figure 19, are a series of typical transmission rate profile curves obtained for the glassine structure. The permeability of the glassine based structure was determined in a manner similar to that used for the HDPE structure. In this case however, permeability runs were terminated when 1% to 4% of the driving force concentration diffused through to the low concentration chamber of the cell system. As was observed for the HDPE based structure, the transmission rate of limonene through the glassine based structure was also found to be concentration dependent. Due to the complex nature of the glassine structure, diffusion coefficient values were not determined. It should be pointed out that for the glassine based structure, approximately ten days were required for the transmission rate to attain steady state, over the low limonene concentration range, as compared to a matter of hours for the HDPE based structure, at similar vapor concentration levels and temperature. The relationship between limonene vapor concentration and the permeability coefficient (P) for the HDPE based structure and the glassine based structure is shown in Figure 20, where log P is plotted as a function of penetrant concentration. It appears that, for both polymeric structures, the permeability coefficient increases ABSOLUTE OUANTITv PERMEATED (grams x10") 70 8m 700- 100 o 1.5 ppm 0 2.5 ppm 3.6 ppm A 4.3 ppm 0 4.8 ppm O Time (mmutes x 103 ) 14 Figure IS. THE EFFECT OF VAPOR CONCENTRATION ON THE DIFFUSION OF LIMONENE THROUGH GLASSINE BASE STRUCTURE AT 23' C 71 1000 L F 100:- I: L Q 5 E 10 0 EXPERIMENTAL DATA .5; a I — CALCULATED FROM EQUATIONS 155 I6 .9 ,. I I“ O I e O * E as + 2 E 0‘ '7 9 x 1 __ I: i 0’ I O -’ I 0.1:P P 0.01 1 J 1 1 1 0 1 2 3 4 5 Limonene Vapor Concentration, ppm (wt /v) Figure 20. THE EFFECT OF LIMONENE VAPOR CONCENTRATION ON Log P FOR HDPE BASE AND GLASSINE BASE STRUCTURES 72 exponentially with increased limonene vapor concentration over the entire vapor concentration range evaluated. Equations 15 and 16 were derived by using the ordinary (exponential) least square method and applying experimental data. From the general Exponential expression: P'= a * ebc (12) By taking the log on both sides: lnP= lna + bC (13) leads to the linear equation: y K + bx (14) where: a, b, and K are constants and C is concentration. Equation (12) was used as a model to determine the correlation between permeability coefficients and penetrant concentration. For the HDPE based structure, the relationship between the permeability coefficient (P) and limonene vapor concentration (CV) was found to be: -' -3 0.6951 C PHDPE 3 4.865 X 10 e V (15) For the glassine based structure the relationship between ('5) and (cv) is found to be: EéLASSINE = 3.3146 X 10-7 81.3301 CV (16) The correlation coefficients were 0.99 and 0.98 for the HDPE structure and the glassine structure respectively. As shown in Figure 20, the permeability coefficient (P) of limonene is substantially lower for the glassine structure than the HDPE structure at similar concentration 73 levels and temperature. The glassine structure exhibited 3 to 4 orders of magnitude reduction in relative permeability as compared to the polyolefin structure, over the entire limonene concentration range studied. Since the permeability coefficient is reflective of the steady state transmission rate, permeation losses of limonene vapor from a fruit flavored cereal product could be greatly reduced by packaging in the glassine structure. The physical and chemical properties of a polymeric film structure may influence the mechanism or mechanisms responsible for aroma loss. For the HDPE based structure, the loss of volatile aroma constituents, such as limonene, was found to be the result of both sorptive and diffusion mechanisms, in that the limonene solubility (dissolution) and permeability (diffusivity) were both high. For the glassine based structure, both loss mechanisms were operative, however, the solubility and diffusivity characteristics of this structure were significantly lower than that of the polyolefin structure. It is important to note that for a polymeric food package, such as the cereal liner systems described in this study, the loss of limonene is the result of both sorptive and permeation losses and both mechanisms must be considered with regards to volatile aroma constituents. 74 SENSORY ANALYSIS Sensory evaluation and G.C. Headspace analysis, were used to qualitatively and quantatively measure components related to the products characteristic aroma. The results of the sensory analysis are shown graphically by a histogram (see Figure 21), where the average relative response of the sensory panel is presented as a function of limonene headspace concentration within the package. The relative response was based on a scale of acceptance which ranged from 0 - 10, where a response of 0 represented an undetectable aroma node and a response of 10 was representative of a product of highly acceptable aroma quality. Freshly packaged product, analyzed within five (5) weeks of commercial packaging, gave a limonene headspace concentration of 0.03 ppm (wt/v) and was rated as highly acceptable. This value represents the average of 18 sample containers. After storage for 3 and 6 months (from the date of production) the limonene headspace concentration showed a decrease to 0.02 ppm (31% loss) and 0.013 ppm (55% loss), for the respective storage periods. The change in headspace concentration reflects a loss of aroma caused by a combination of permeation, sorption and suspected leakage from an ineffective sampling septum. Following nine months of storage, a reduction in the limonene concentration in the headspace of approximately 86% (0.004 ppm) was found. 75 RELATIVE RESPONSE SCALE 0.004 0.013 0.01 8 0.029 LIMONENE HEADSPACE CONCENTRATION (ppm) Figure 21. BAR GRAPH DATA RELATING QUANTITATIVE HEADSPACE ANALYSIS WITH QUALITATIVE SENSORY ANALYSIS OF A FRUIT-FLAVORED CEREAL PRODUCT 76 However, this product was still found acceptable, based on sensory evaluation. A perceived loss in relative product quality was noted, however, as the level of measurable limonene in the headspace diminished. A qualitative relationship was established between the limonene within the headspace of a commercially packaged fruit-flavored cereal product and the acceptability level based on sensory evaluation. The results of the sensory studies provide a means of determining a critical aroma node concentration, below which the product is considered unacceptable regardless of the mechanism of aroma loss. SUMMARY AND RECOMMENDATIONS Both qualitative (sensory evaluation) and quantitative (permeation and sorption) measurements, relating to the aroma character of a packaged fruit flavored cereal product, were performed. Aroma is a key factor in the acceptability of many products. For cereal products in general, moisture is of primary concern followed closely by aroma in terms of typical barrier requirements (Hamiliton, 1985). Of practical importance, this study provides information relating to the aroma barrier properties of selected packaging materials to limonene vapor. Assuming the vapor concentration within the package headspace remains constant, potential candidate packages may be evaluated and shelf life estimated. The results of these studies are summarized below: (1) The permeability of.d-limonene vapor through the respective cereal liner structures is concentration dependent, with the transmission through the HDPE based structure being 3 to 4 orders of magnitude greater than the glassine based structure, at the same vapor concentration. (2) The solubility of d-limonene in the respective cereal liner structures follows a Henry's Law 77 78 relationship at low vapor concentration levels (i.e. below 5 ppm), but showed deviation from linearity at higher penetrant levels. This was attributed to non-Fickian related sorption. (3) d-Limonene solubility was found to be substantially lower in the glassine based structure than in the HDPE based structure, at the same vapor concentration and temperature. This relationship can be of major importance in avoiding the effect of "flavor scalping" or loss due to a sorptive mechanism. (4) The loss of volatile aroma moieties such as d- limonene can be the result of sorptive and/or permeation mediated processes and both mechanisms should be considered in understanding the relationship of mass transport to the storage stability of a product, such as the fruit flavored cereal product, where quality is associated with the retention of aroma compounds. Results show glassine to be a superior barrier to the polyolefin structure in both the permeability and solubility studies. Assuming the primary mechanism for quality loss involves aroma retention for the product package system described, the glassine based structure would appear to be the material of choice. The results also could be used to develop a simulation 79 model to predict the shelf life of a packaged product, where quality is related to the retention of volatile aroma constituents within the package headspace. Because of the inherent errors involved in the commonly used accelerated storage test method for estimation of product shelf life, more emphasis is being placed on the development of the simulation modeling approach for product shelf life estimation. Here, the entire package/product system is taken into consideration and a mathematical expression developed that includes a measure of: (1) Product sensitivity, (2) Package effectiveness, and (3) Environmental severity. For the fruit flavored cereal product where quality is related to the retention of an aroma constituent, the parameters to be considered include: (1) A sorption isotherm which describes the relationship between the concentration of limonene in the product and the vapor pressure of limonene above the product (i.e. in the package headspace). (2) The permeability rate and permeability coefficient of limonene through the packaging material. (3) The solubility and solubility coefficient of limonene in the packaging material. (4) Sensory characteristics of the product. 80 The latter three parameters have been investigated in detail in the present study. Determination of the sorption isotherm, coupled with sensory and mass transport data, would provide the necessary elements for development of a simulation model. This would provide an analytical method for shelf life estimation. Such a simulation model would however only be appropriate for a packaged product whose quality is related to retention of volatile aroma moieties. Products undergoing loss by more than one mechanism would require additional information with regard to the nature of the product and the mechanisms of loss. EXPERIMENTAL ERROR ANALYSIS Temperature, relative humidity, and maintainance of the penentrant driving force concentration are among the conditions which must be properly and continually monitored throughout each section of this study (Permeation, Sorption, and Sensory Evaluation). To avoid characteristic changes due to fluctuating relative humidity, the film samples were stored in desiccators prior to each test. Permeation Evaluation It is often difficult to reproduce data relating to a specific penetrant concentration. As time progresses, a syringe tends to loss its reproducability through loose fittings, plugged tip, seals and/or adsorption. Sampling technique is acquired only through means of practice and may substantially effect the data acquired, leading to the possibility of erroneous conclusions. Flowmeters were positioned in stream for each permeability cell. These were provided as a means for controlling the rate of flow (penetrant vapor) and for detecting possible leaks within the permeation apparatus assembly. 81 82 Difficultly can be encountered in placing a film sample within the permeation cell. The permeation cell system is made up of three chambers: the center ring containing the penetrant vapor and two outer chambers containing the permeated vapor, thus two films are simultaneously evaluated. In the process of mounting these film samples, it's possible for the test material to move out of position causing an unacceptable evaluation (i.e. crease, pinhole, non-effective mount/seal . Permeation of a polymer film sample may be due in part to the presence of defecas such as microscopic or macroscopic pinholes or :racks. One of the two films tested in this study was coated with a layer of wax. Depending on the consistency of this layer, duplication of resultant data can be difficult to achieve. Sorption Evaluation Variable results were also found in determining the sorption lag time values concerning the HDPE structure, especially at the low levels of limonene concentration (0.30 ppm). These results were due most likely to the difficulty experienced in mounting a film sample. Limonene is quite soluble in polyolefin (HDPE) structures and a change in weight of the sample due to adsorption is seen instantaneously. Therefore, the time required to mount a film sample becomes a critical factor. 83 At low limonene concentration levels (0.30 ppm), the glassine film structure showed a loss in weight over a period of time. At somewhat higher concentration values, (1.54 ppm) a slight loss again is observed prior to a gain in weight. The loss in weight was apparently a loss in moisture. Therefore, the glassine based structure was placed in a desiccator and allowed to set for a period of at least one week prior to conducting solubility studies. This eliminated the problem. Sensory Analysis Evidence indicates that loss of limonene through the package liner may have been affected by leakage due to sampling. The more headspace samples removed from a particular package system, the more apparent the loss when compared with other packages stored over the same period of time but sampled less frequently. Data collected through quantitative measure shows that approximately 31% of the headspace area response may have been loss due to pinhole leakage. APPENDIX 84 LIST OF APPENDIX: EXPERIMENTAL DATA Appendix Table Title Permeation Study: HDPE based structure: A 5 Transmission Rate Profile A 6 Permeability (slope) A 7 Permeability Coefficient A 8 Apparrent Lag Time Diffusion A 9 Driving Force Concentration Glassine based structure: A 10 Transmission Rate Profile A 11 Permeability (slope) A 12 Permeability Coefficient A 13 Apparrent Lag Time Diffusion A 14 Driving Force Concentration Sorption Study: B 15 HDPE based structure 16 Glassine based structure Standard Calibration: 17 G.C. Model #5830 Data Fig. 22 Calibration Curve 18 G.C. Model #5890 Data 0000 Fig. 23 Calibration Curve Appendix 85 LIST OF APPENDIX cont... l9 G.C. Model #5890 Data Fig. 24 Calibration Curve Sensory Evaluation UUUUU Glassine based structure only: 20 Summarized Table of Data 21 Aroma Evaluation Form 22 Data: 3 months of storage 23 Data: 6 months of storage 24 Data: 9 months of storage APPENDIX: A 86 APPENDIX A Table 5: Data for Transmission Rate Profile Curve HDPE based structure: Limonene Vapor Conc. Time (x-axis) Quantity (y-axis) ppm (w/v) minutes x 10' grams 0.4 0 0.0 (run #36) 135 18.1 146 24.9 153 25.2 174 42.2 189 41.8 209 59.2 229 77.1 246 98.0 268 128.3 299 180.5 324 209.9 1.3 0 0.0 (run #5) 15 19.4 25 17.4 35 9.5 45 10.1 55 9.8 65 16.3 75 17.1 85 14.2 95 18.2 105 26.9 115 38.5 125 53.3 135 82.4 145 118.0 155 156.9 167 199.2 185 341.9 195 417.6 215 616.7 225 711.0 230 881.0 87 Table 5: continued Limonene Vapor Conc. Time (x-axis) Quantity (y-axis) ppm (w/v) minutes x 10 grams 1.9 0 0.0 (run #3) 15 0.0 30 0.0 45 0.0 63 4.5 71 12.4 77 18.0 85 30.5 95 48.6 105 102.8 115 165.8 125 265.1 135 355.7 145 602.0 2.6 0 0.0 (run #2) 28 0.0 45 0.0 65 14.4 75 34.5 85 79.1 95 185.8 105 351.9 115 622.7 125 972.7 3.6 0 0.0 (run #14) 25 0.0 35 0.0 45 10.1 55 27.3 65 83.8 75 245.4 85 632.6 95 1217.2 105 1913.5 115 2701.1 125 3510.2 135 4356.9 145 4673.7 165 6238.2 175 6648.1 185 7550.2 195 7648.7 88 Table 6: Permeation Data For HDPE Based Structure LIMONENE VAPOR RUN # CONCENTRATION % OF LIMONENE SLOPE PERMEATED THRU FILM (g/min) x 10'9 36 0.39 0.93 to 10.73 10.05 ;§ 0.00 to 10.14 10.22 avg. 10.14 42 0.38 0.94 to 7.68 7.21 51 1.23 to 16.46 9.69 avg. 8.45 44 0.38 1.10 to 13.92 11.67 A; 1.12 to 11.88 8.82 avg. 10.25 average 0.38 9.61 46 0.96 1.59 to 13.90 46.34 45 1.95 to 11.28 62.80 average 0.96 54.57 6 1.22 1.49 to 11.12 52.83 _§ 1.21 to 10.25 65. 3 avg. 59.33 20 1.29 0.91 to 3.35 30.02 12 1.32 to 3.74 34.19 avg. 32.11 average 1.26 45.72 4 1.92 1.21 to 4.25 92.19 3 1.10 to 6.47 118.80 average 1.92 105.50 89 Table 6: continued LIMONENE VAPOR RUN # CONCENTRATION % OF LIMONENE SLOPE PERMEATED THRU FILM (g/min) x 10"9 2 2.58 1.58 to 11.54 296.20 ._1 1.03 to 10.35 201.00 avg. 248.60 8 2.53 3.37 to 11.44 437.50 _1 1.50 to 11.21 306.50 avg. 372.00 average 2.56 310.30 18 2.89 0.95 to 11.49 441.60 11 1.54 to 11.96 537. 0 avg. 489.40 12 3.16 1.04 to 11.74 419.00 11 1.02 to 11.72 336.60 avg. 377.80 10 3.21 2.01 to 11.06 482.90 _2 1.86 to 10.30 447.80 avg. 465.40 average 3.09 441.20 14 3.55 1.29 to 10.10 558.90 13 2.87 to 12.29 892.70 average 3.55 725.80 90 Table 7: AVERAGED (a) RUN PERMEABILITY CORRELATION STANDARD COEFFICIENT COEFFICIENT DEVIATION 36 74.22 0.9751 ;5 75.47 0.9860 avg. 74.84 42 54.65 0.9626 A; 73.46 0.9942 avg. 64.05 44 88.45 0.9917 4; 66.85 0.9707 avg. 77.65 average 72.18 11.1075 46 139.02 0.9941 45 188.40 0.9987 average 163.71 34.9160 6 124.71 0.9778 _§ 155.40 0.9811 avg. 140.06 20 67.02 0.9945 12 15123 0.9995 avg. 71.65 average 105.85 41.6076 4 138.29 0.9835 3 178.20 0.9625 average 158.25 28.2242 (9) greme_1_etrueture * day * ppm In2 x 10'4 Permeation Data Continued For HDPE Based Structure COEFFICIENT OF VARIABILITY 15.4 % 21.3 % 39.3 % 17.8 % 91 Table 7: continued AVERAGED (a) RUN PERMEABILITY CORRELATION STANDARD COEFFICIENT COEFFICIENT 2 330.64 0.9891 _1 224.37 0.9351 avg. 227.51 8 498.02 0.9994 _1 348.90 0.9966 avg. 423.46 average 350.48 112.6475 18 440.07 0.9743 11 535.24 0.9835 avg. 487.66 12 381.87 0.9795 11 306.77 0.9809 avg. 344.32 10 433.26 0.9878 _2 401.76 0.9855 avg. 417.51 average 416.50 75.3548 14 453.42 0.9923 13 724.22 0.99999 average 588.82 191.4853 OVERALL COEFFICIENT OF VARIABILITY (a) greme_1_§trueture m2 * day * ppm x 10'4 COEFFICIENT DEVIATION OF VARIABILITY 32.1 % 18.1 % 32.5 % 25.2 % 92 Table 8: Permeation Data Continued For HDPE Based Structure LAG TIME TIME DURING APPARENT DIFFUSION RUN UNSTEADY STATE STEADY STATE cos FICIENT # (min) (hrs) (min) (hrs) (cm /sec) x 10'10 38 132 2.2 192 3.2 4.9385 2; 130 2.2 186 3.1 4.9745 avg. 131 2 2 189 3.2 4.9555 42 113 1.9 166 2.8 5.7229 21 175 2.2 295 4.9 3.8252 avg. 144 2.4 231 3.8 4.7091 44 138 2.3 220 3.7 4.6861 22 123 2.1 220 3.8 5.2578 avg. 131 2.2 225 3.8 4.8788 average 134 2.2 215 3.6 4.8788 48 98 1.8 137 2.3 8.5988 45 88 1.5 84 1.4 7.3487 average 93 1.8 111 1.9 8.9737 8 128 2.1 84 1.4 5.1324 _5 13_Q 2-_-Z 5.6. 1-_4. 4-97 5 avg. 128 2.2 85 1.4 5.0534 20 128 2.1 72 1.2 5.1324 12 120 2.0 19 122 .3890 avg. 123 2.1 71 1.2 5.2807 average 128 2.1 78 1.3 5.1571 4 110 1.8 40 0.7 5.8789 3 100 1.7 45 0.8 8.4888 average 105 1.8 43 0.7 8.1729 93 Table 8: continued LAG TIME TIME DURING APPARENT DIFFUSION RUN UNSTEADY STATE STEADY STATE COE FICIENT # (min) (hrs) (min) (hrs) (cm /sec) x 10-10 2 92 1.5 43 0.7 7.0292 _1 100 1.7 42, 0.8 6.4668 avg. 96 1.6 46 0.8 6.7480 8 85 1.4 34 0.6 7.6080 _1 22 1,5 46 0.8 7.3487 avg. 87 1.5 40 0.7 7.4783 average 91 1.5 43 0.7 7.1132 18 65 1.1 35 0.6 9.9490 11 pg 1,1 22 0.5 9.79 0 avg. 66 1.1 32 0.6 9.8736 12 86 1.4 41 0.7 7.5196 1; 29 1.5 59 0.8 7.1854 avg. 88 1.5 46 0.8 7.3525 10 70 1.2 45 0.8 9.2383 _2 15 1.3 25 0.6 8.6225 avg. 73 1.2 40 0.7 8.9304 average 76 1.3 39 0.7 8.7188 14 72 1.2 33 0.6 8.9817 13 74 1.2 26 0.4 8.7390 average 73 1.2 30 0.5 8.8603 94 Table 9: Permeation Studies For HDPE Based Structure Continued Driving Force Vapor Concentration d-LIMONENE AVERAGED RUN VAPOR AREA STANDARD COEFFICIENT No.8 CONC. RESPONSE DEVIATION OF VARIABILITY 35/36 0.39 *** 1398600 123616 8.8 % 41/42 0.38 *** 1362995 204577 15.0 % 43/44 0.38 *** 1374113 152218 11.0 % average 0.38 1378569 45/46 0.96 *** 3460500 127468 3.7 % 5/6 1.22 * 339256 40263 11.9 % 19220 1.29 * 361440 9337 2.6 % average 1.26 350348 3/4 1.92 * 536960 20023 3.7 % 1/2 2.58 * 720200 50796 7.1 % 218 2.53 * 707800 16880 :2.4 % average 2.56 714000 17/18 2.89 ** 359565 13333 3.7 % 11/12 3.16 * 883886 65476 7.4 % 9/10 3.21 * 896640 62489 7.0 % average 3.09 13/14 3.55 * 992900 45200 4.6 % * Samples # 1 through 24 (except #‘s 17 & 18) were analyized using G.C. Mode 5830 and the limonene standard calibration of 1.79 x 10' grams/area unit. .C. Model 5890 was employed. ** Samples # 17 & 18 were analyzed usingl and a standard calibration of 4.016 x 10 *** Samples # 25 through 46 were analyzed usingl .C. Model 5890 and the Calibration standard of 13.8 x 10 g/a.u. 95 Table 10: Data for Transmission Rate Profile Curve GLASSINE based structure: Limonene Vapor Conc. Time (x-axis) Quantity (y-axis) ppm (w/v) minutes x 10' grams 1.5 0 0.0 (run #20) 75 0.0 96 7.8 128 10.5 218 13.4 367 16.7 794 27.8 941 34.6 1081 37.9 1224 41.3 1372 39.3 1804 49.2 1955 52.0 2127 55.7 2.5 0 0.0 (run #13) 73 0.0 115 10.1 133 13.4 140 7.8 258 . 13.7 401 19.0 419 18.7 587 30.1 704 34.2 840 36.1 984 47.6 1126 53.8 1270 58.9 1410 80.0 96 Table 10: continued Limonene Vapor Conc. Time (x-axis) Quantity (y-axis) ppm (w/v) minutes x 10 grams 3.6 0 0.0 (run #18) 134 13.1 171 12.9 282 21.8 456 44.4 615 81.6 712 106.2 743 107.2 855 144.9 999 166.3 1144 192.2 1292 234.1 1447 257.2 1722 283.7 4.3 0 0.0 (run #10) 113 16.6 119 23.4 128 23.3 143 33.5 158 42.6 164 48.2 261 170.4 267 180.3 282 218.0 288 230.4 306 284.8 544 775.1 550 969.3 4.8 0 0.0 (run #12) 10 6.4 17 6.0 138 187.8 140 200.5 149 232.0 177 320.4 200 407.9 281 699.7 294 731.0 97 Table 11: Permeation Data For Glassine Based Structure LIMONENE VAPOR RUN # CONCENTRATION % or LIMONENE SLOPE PERMEATED THRU FILM (g/min) x 10’11 20 1.54 0.14 to 0.72 2.23 19 " 0.13 to 0.36 0.96 average 1.54 1.60 14 2.51 0.28 to 0.97 8.43 13 2.39 0.00 to 0.66 4.86 average 2.45 6.65 6 3.50 0.19 to 1.67 44.62 5 " 0.22 to 1.53 39. 3 avg. 3.50 41.83 18 3.63 0.25 to 1.46 21.56 16 " 0.28 to 1.69 27.93 25 3.59 0.32 to 1.57 19.73 avg. 3.60 23.07 average 3.55 34.52 8 4.27 0.58 to 2.93 301.40 7 4.26 0.24 to 2.57 271 0 avg. 4.27 286.45 10 4.45 0.19 to 3.52 191.90 9 5.40 0.26 to 4.35 223. 0 avg. 4.43 207.50 4 4.34 0.19 to 3.59 193.70 3 4.35 0.26 to 2.79 lzgégg avg. 4.35 182.10 average 4.35 225.35 12 4.82 0.79 to 3.08 351.10 11 " 0.46 to 2.43 309.20 average 4.82 330.15 Table 12: AVERAGED (a) RUN PERMEABILITY COEFFICIENT 20 4.2 19 1.8 average 3.0 14 9.7 13 5.9 average 7.8 6 37 _§ 12 avg. 34.5 18 17.1 16 22.4 15 15.8 avg. 18.4 average 24.8 8 203 7 184 avg. 194 10 124 2 141 avg. 133 4 129 2 113 avg. 121 average 149.7 12 210 11 185 average 197.3 OVERALL COEFFICIENT OF VARIABILITY (a) W m2 * day * ppm 98 CORRELATION COEFFICIENT 0.9872 0.9709 0.9947 0.9895 0.9808 0.9973 0.9756 0.9808 0.9994 0.9992 0.9911 0.9972 0.9785 0.9898 0.9997 0.9989 x 10"6 Permeation Data Continued For Glassine Base Structure STANDARD COEFFICIENT DEVIATION OF VARIABILITY 1.7 56.1 % 2.8 35.6 % 9.2 37.2 % 36.0 24.0 % 17.7 9.0 % 32.4 % 99 Table 13: Permeation Data Continued For Glassine Based Structure LAG TIME TIME DURING RUN UNSTEADY STATE STEADY STATE fl (min) (days) 1m1n) (days) 20 750 0.5 20520 14.3 19 966 0.7 20306 14.1 average 858 0 6 20413 14.2 14 4014 10086 7.0 13 ---- --- --- 6 2643 1.8 5719 4.0 5 2641 1.8 5719 4.0 avg. 2642 1.8 5719 4.0 18 4559 3.2 12662 8.8 16 2805 2.0 8611 6.0 15 4552 3.2 9905 6.9 avg. 3972 2.8 10393 7.2 average 3440 2.4 8523 5.9 8 1153 0.8 1682 1.2 7 1037 0.7 1682 1.2 avg. 1095 0.8 1682 1.2 10 1580 1.1 3864 2.7 9 14 5 1.0 40 7 222 avg. 1503 1.0 3941 2.7 4 1577 1.1 3690 2.6 3 2196 1.5 2212 22; avg. 1887 1.3 3466 2.4 average 1495 1.0 3030 2.1 12 1379 1.0 1558 1.1 11 1377 1.0 1558 1.1 average 1378 1.0 1558 1.1 100 Table 14: Permeation Studies For Glassine Based Structure Continued Driving Force Vapor Concentration d-LIMONENE AVERAGED RUN VAPOR AREA STANDARD COEFFICIENT No.s CONC. RESPONSE DEVIATION OF VARIABILITY 19 1.54 428807 51729 12.0 % 20 1.54 428807 51729 12.0 % average 1.54 428807 13 2.39 666293 70547 10.6 % 14 42.51 700863 52317 7.5 % average 2.45 683578 5 3.50 976600 61898 6.3 % 6 3.50 976600 61898 6.3 % avg. 3.50 976600 15 3.59 1002225 57707 5.8 % 16 3.59 1003904 63891 6.4 % 18 3.63 1014019 46387 4.6 % a g. 3.60 1006716 average 3.55 994670 7 4.26 1189889 38388 3.0 % 8 4.27 22221§Q 40602 3.4 % avg. 4.27 1190820 9 4.40 1229000 62492 5.1 % 10 4.45 241625 53139 4.3 % avg. 4.43 1235313 3 4.35 1215333 98584 8.1 % 4 4.34 2222222_ 98144 8.1 % avg. 4.35 1214278 average 4.35 1213470 11 4.82 1347571 68274 5.0 % _;2 4.82 1347571 68274 5.0 % average 4.82 1347571 All samples (# 1 through 20) were analyized using G.C. 5830 and the limonene standard calibration of 1.79 x 10 grams/area unit. 22239-1 APPENDIX: B 101 APPENDIX B Table 15: Solubility Data For HDPE Based Structure (1) (2) (3) (4) (5) LIMONENE ROOM EQUILIBRIUM SOLUBILITY DIFFUSION RUN VAPOR TEMP SOLUBILITY COEFFICIENT COEF ICIENT # CONCEN (°F) (mg/mg) (mg/mg ppm) (cm /sec> 39 0.00 71.4 0.00 0.00 --- 40 I! II n --_ 10 0.29 71.4 0.26 8.84 --- 11 0.30 71.7 ---- ---- --- 12 0.29 70.2 0.21 7.34 --- 13 0.29 69.9 0.23 7.87 --- 15 0.30 69.1 0.23 7.64 —-- avg. 0029 7002 0022 7.62 --- 16 1.52 68.5 1.03 6.76 3.63 19 -- 68.8 1.02 8.97 4.59 30 1.48 69.1 1.05 7.07 4.65 avg. 1.50 69.0 1.03 7.60 4.29 37 2.50 68.7 1.78 7.11 4.71 38 2363 67.9 1.95 7.43 5.01 avg. 2.57 67.9 1.87 7.27 4.86 20 4.35 68.4 2.74 6.30 --- 21 3.92 68.2 3.05 7.79 5.61 24 4.60 69.2 3.55 7.71 --- _25 4.09 67.6 3.50 8.56 6.14 avg. 4.24 68.4 3.21 7.59 5.88 33 6.58 69.6 6.44 9.79 --- 34 6.28 68.5 6.04 9.62 8.35 avg. 6.43 68.7 6.24 9.71 8.35 (1) (2) 102 Table 15: continued (3) (4) (5) LIMONENE ROOM EQUILIBRIUM SOLUBILITY DIFFUSION RUN VAPOR TEMP SOLUBILITY COEFFICIENT COEF ICIENT # CONCEN <°F> (mg/mg) (mg/mg ppm) (cm /sec) 5 6.80 71.5 10.38 15.26 12.38 7 7.15 70.9 12.86 17.98 13.07 avg. 6.98 71.2 11.62 16.62 12.73 (1) Concentration express as ppm, vapor in nitrogen on a weight per volume basis. (2) Room Temperature +/- 2 degrees (3) Units expressed as: mg limonene x 10'2 mg HDPE (4) Units expressed as: mg limoneng x 10'3 mg HDPE * ppm (5) Units expressed as: cmz/sec x 10'10 103 Table 16: Solubility Data For Glassine Based Structure (1) (2) (3) (4) (5) LIMONENE ROOM EQUILIBRIUM SOLUBILITY DIFFUSION RUN VAPOR TEMP SOLUBILITY COEFFICIENT COEF ICIENT # CONCEN (°F) (mg/mg) (mg/mg ppm) (cm /sec) 41 0.00 68.6 0.00 0.00 —--- 42 I! n " ---- 14 0.30 69.2 -—-- ---- ---- 27 1.44 69.1 1.76 12.20 2.19 28 1.47 68.7 1.32 9.00 3.09 _29 1.51 68.3 1.32 8.73 4.32 avg. 1.47 68.9 1.47 9.98 3.20 35 2.50 67.8 4.65 18.58 2.39 36 2.49 68.4 4.63 18.59 1.64 avg. 2.50 68.3 4.64 18.59 2.02 23 4.59 69.3 6.56 14.28 2.40 _26 4.61 68.8 6.30 13.66 3.51 avg. 4.60 68.7 6.43 13.97 2.96 31 6.69 68.3 15.37 22.98 4.76 32 6.06 68.1 13.46 22.21 3.79 avg. 6.38 68.1 14.42 22.60 4.28 8 7.40 71.4 58.10 78.51 64.20 9 7.28 70.7 63.11 86.69 22.30 avg. 7.34 71.4 60.61 82.60 68.25 (1) Concentration express as ppm, vapor in nitrogen on a weight per volume basis. (2) (3) (4) (5) 104 Table 16: continued Room Temperature +/- 2 degrees Units expressed as: mg limonene x 10'3 mg glassine Units expressed as: mg limonene x 10'"4 mg glassine * ppm Units expressed as: cmz/sec x 10'10 APPENDIX: C 105 APPENDIX C STANDARD CALIBRATION Basically three standard calibration factors were employed throughout this study, depending on the gas chromatograph (G.C.) model used, as well as test conditions utilized (i.e. oven temp., time factors, cooling time, etc...) Table 17: This standard was used primarily for Permeation run numbers 1 through 24 concerning the HDPE based structure and Permeation run numbers 1 through 20 concerning the Glassine based structure, utilizing the G.C. Model #5830: d-limonene standard calibration factor = 1.79 x 10"12 g/a.u. Table 18: This standard was used only for Permeation run numbers 17 and 18 in regard to the HDPE based film structure, utilizing the G.C. Model #5890: d-limonene standard calibration factor - 4.02 x 10"12 g/a.u. Table 19: The standard listed here was applied to a series of studies which include: (A) Permeation studies: run numbers 25 through 46 concerning the HDPE based structure, (B) All solubility studies for both polymeric films, & (C) All sensory evaluation studies involving quantification of headspace samples in cereal packages fabricated from the glassine based structure. The G.C. Model #5890 was used in obtaining: d-limonene standard calibration factor = 13.8 x 10"14 g/a.u. 106 Table 17: Limonene Standard Calibration Curve Data G.C. Model #5830 INITIAL CALIBRATION: 11/27/84 AREA RESPONSE ABSOLUTE QUANTITY {x 10'9 grams} (x - axis) (v - axis) 0 0.0 30060 60.0 63605 120.0 99705 181.6 132625 242.2 1352500 2421.5 RECALIBRATION: 1/23/86 AREA RESPONSE ABSOLUTE QUANTITY {x 10'9 grams} (x - axis) (v - axis) 0 0.0 2150 4.4 4584 8.8 6905 13.2 10153 17.6 102533 176.0 997650 1760.0 107 no. Xv umzoammm <92 com. oou— coop can 0mm 00% CON 0 hbblrFLFbLPIPLLb PLkFLIPbIbh thFLlL—ybh ...... o. 63w». «Agmflcumjdwmm 4 05324.5 mzwzozjln O O O N O In N N I IWIjTTTTTWj woomm 08mm .30: 0.0 m>m8 zocémzé omézém “13202316 .NN 936: (‘Swm6 .-Ol X) AILLNVFIO almosev 108 Table 18: INITIAL CALIBRATION 4/9/85 AREA RESPONSE ABSOLUTE QUANTITY {x 10"9 grams} (x - axis) (y - axis) 0 0.0 18438 71.8 28617 125.3 47195 180.4 63168 250.5 623680 2505.0 109 cow p. po— 5 mmzoamum 6% com com com com com 0mm oo— h b L )P L— h b h bPth-bbthLrLFnbbPPLLIIPLLb OERm doc: 0.0 misc 29219.15 omézflm mzmzosi: .mN 956: (swmfi ..m X) AlliNVflO almosav 110 Table 19: Limonene Standard Calibration Curve Data G.C. Model #5890 INITIAL CALIBRATION: 8/27/85 AREA RESPONSE ABSOLUTE QUANTITY {x 10'9 grams} (x - axis) (v - ax1s) 0 0 20920 4 53717 8 81261 13 117803 17 1269767 176 RECALIBRATION: 1/11/86 AREA RESPONSE ABSOLUTE QUANTITY {x 10'9 grams} (x - axis) (v - axis) 0 0.0 23757 4.4 62671 8.8 93310 13.2 142003 17.6 1524900 176.0 111 he xv mngmum 3% com? comp 004— coup coop com com com CON —Lbblr)PplPLlPL—LL)Pb|P|PLL+IPLbIPLLLLLLLLbbbLPLLIPLbLPh ....... 6. mx 3 -"ydfléwjdwmm 4 mound} “26:432.. 0 . omddZ/fim mzmzozfiln .\ ommmm d8: do ””1250 29.39.20 @45sz mzmzogjln .vm 8:6,: 0 no N O 0 In 0 IO N Door :3 in 2! O O In F r‘f‘T‘r'fl'j‘T-T’WTT'TTT rT—r r—11'1fi11 1 vfi 1T] 1 TfiWj 7T1 O I!) r\ F O O O N (stJ5 ._01><) AlliNVflO almosav APPENDIX: D 112 APPENDIX D Table 20: Sensory Bar Graph Data Glassine Base Structure Only APPROXIMATE LIMONENE HEADSPACE RELATIVE RESPONSE STORAGE TIME CONCENTRATION SCALE (months) (ppm) x-axis Iqualitv) v-agig 9 0.004 5.2 6 0.013 5.4 3 0.018 8.5 initial 0.029 10.0 113 Figure 25. AROMA EVALUATION OF FRUIT FLAVORED CEREAL Name: Please evaluate the cereal samples in regards to the perceived aroma detected. Then determine, on the given lines, the degree of aroma for each cereal sample. Be sure your mark follows the example as shown. Use only one line per cereal sample. Please feel free to comment on the aroma quality of each sample (i.e. stronger cereal grain aroma than the fruit aroma) in the space provided. Thank you for your participation. Example: 44} N M 7’ S Indicate your response on the following lines: Sample No. No Fruit Moderate Strong Aroma Fruit Aroma Fruit Aroma Sample No. No Fruit Moderate Strong Aroma Fruit Aroma Fruit Aroma Sample No. No Fruit Moderate Strong Aroma Fruit Aroma Fruit Aroma Sample No. Comments Table 21: SAMPLE NUMBER 1031 121 avg; 1041 104 avg? 1051 1052 avg? 1081 12§ avg; 112 l 1122 avg. OVERALL AVERAGE 114 Followed Immediately By A Qualitative Evaluation 6/21/85 thru 9/25/85 LOSS OF LIMONENE __iPPEL_ ___iil__ 41 37 35 36 38 38 SENSORY EVALUATION (scale=10) Sensory Analysis Data; Quantitative Study (Glassine Based Structure Only) 3 MONTHS STORAGE PERIOD: INITIAL FINAL LIMONENE LIMONENE LIMONENE CONCEN ‘aOuOl ‘3032l 108020 61353 11 9 0 68822 109985 65088 0.018 101770 63640 102239 64614 102000 64127 0.018 108190 71225 108200 8907 108195 70066 0.019 99432 62590 99605 64142 99519 63367 0.017 102890 65709 121222 ééééfi 105105 65034 0.018 104961 65536 0.018 Table 22: SAMPLE EHMEEB 1091 1092 avg? 1101 1122 avg? 111 1 1112 avg? 90 1 902 avg. 89 1 89 avg. 88 1 882 avg. OVERALL AVERAGE 115 Sensory Analysis Data: Quantitative Study Followed Immediately By A Qualitative Evaluation (Glassine Based Structure Only) 6 MONTHS STORAGE PERIOD: INITIAL FINAL LIMONENE LIMONENE ‘a.g.) ‘QOual 105890 34816 101789 41997 103835 38377 103690 39989 99451 51911 101566 40950 101630 45020 100640 54439 101135 44730 107200 62874 110599 69985 108875 64379 99643 45690 99735 46499 99689 46090 100260 57011 104870 56494 102565 56753 102944 48546 LIMONENE CONCEN 0.011 0.011 0.012 0.018 0.013 0.016 0.013 6/21/85 thru 12/20/86 LOSS OF LIMONENE __IEEE1_ ___111__ 63 60 56 41 54 45 53 SENSORY EVALUATION (sca1e=10) 116 Table 23: Sensory Analysis Data: Quantitative Study Followed Immediately By A Qualitative Evaluation (Glassine Based Structure Only) 9 MONTHS STORAGE PERIOD: 7/25/85 thru 4/25/86 INITIAL FINAL LIMONENE LOSS OF SENSORY SAMPLE LIMONENE LIMONENE CONCEN LIMONENE EVALUATION NUMBER (a.u.) (a.u.) (ppm) (3) (sca1e210) C1 119460 22201 _92 125890 20312 avg. 122675 21257 0.006 83 5.0 01 110980 14727 _92 116030 13209 avg. 113505 13968 0.004 88 5.0 E1 121380 14903 _E2 1267 0 122§1 avg. 124050 14582 0.004 88 4.1 F1 113180 8577 _E 113909 8295 avg. 113540 8436 0.002 93 8.8 OVERALL AVERAGE 118443 14561 0.004 88 5.7 BI BLI OGRAPHY BIBLIOGRAPHY Abbe, S., Bassett, B.M., Collier, J.C. 1985. Flavor and Odor Considerations With New Types of Food Packages: An Overview. TAPPI Proceedings, Polymers, Laminations and Coating Conference: 445. Chicago, Illinois. September, 1985. Amerine, M.A.: Pangborn, R.M.: and Roessler, E.B. 1965. Principles of Sensory Evaluation of Food. Academic Press Inc., New York and London. ASTM Designation E96-66. 1986. Annual Book of ASTM Standards. Vol. 08.02 Plastics (II): Standard Method of Test for Water Vapor Transmission of Materials in Sheet Form. Philadelphia, PA. ASTM Designation D3985-81. 1986. Annual Book of ASTM Standards. Vol. 08.02 Plastics (II): Standard Method for Oxygen Gas Transmission Rate Through Plastic Film and Sheeting Using a Coulometric Sensor. Philadelphia, PA. Bagley, E. and Long, F.A. 1958. Two-stage Sorption and Desorption of Organic Vapor in Cellulose Acetate. J. Am. Chem. Soc. 77:2172. 117 118 Baner, A.L., Hernandez, R.J., Jayaraman, K. and Giacin, J.R. 1986. Isostatic and Quasi-Isostatic Methods for Determining the Permeability of Organic Vapors Through Barrier Membranes. To be published in current Technologies in Flexible Packaging, ASTM STP 912. Baner, A.L. 1986. Diffusion and Solubility of Toluene in Polymeric Films. Presented at the 13th Annual IAPRI Symposium, Oslo, Norway, May 1986. Barrer, R.M. 1939. Permeation, Diffusion and Solution of Cases in Organic Polymers. Trans. Faraday Soc. 35:628. Becker, K., Koszinowski, J., and Piringer, O. 1983. Permeation of Scenting and Flavoring Substances Through Polyolefins. Berens, A.R. 1977. Diffusion and Relaxation in Glassy Polymer Powders: 1. Fickian Diffusion of Vinyl Chloride in Poly(vinyl chloride). Polymer, 18:697. Berens, A.R. 1978. Analysis of Transport Behavior in Polymer Powders. J. Membrane Science, 3:247. Bischoff, M. and Eyerer, P. 1984. Effect of Orientation on the Sorption of Toluene by High Density Polyethelene. (J. of Membrane Sci., 23:333). Sited in Hernandez, et al., 1986 paper. Crank, J. and Park, 6.8. 1968. Diffusion in Polymers. Academic Press, New York, NY. Crank, J. 1975. The Mathematics of Diffusion, 2nd Edition, Clarendon Press, Oxford, England. 119 Delassus, P.T. 1985. Transport of Unusual Molecules in Polymer Films. TAPPI Proceedings, Polymers, Laminations and Coating Conference:445. Filadelfi, M. 1985. Personal communication. Assistant Professor, Dept. of Food Sci. & Human Nutri., Mich. St. U. Fortunato, V. 1958. Sited in Amerine, et al., 1965. Fortunato, V. and Niccolini, P. 1958. Sited in Amerine, et al., 1965. Fujita, J. 1961. Diffusion in Polymer-Diluent Systems. Fortsch-Hochpolym-Forsch, 3:1. Furchtgott, E. and Friedman, M.P. 1960. The effects of hunger on taste and odor RLs. (J. Comp. and Physiol. Psychol. 53, 576 - 581.) Sited in Amerine, et al., 1965. Giacin, J.R. 1983. Packaging Course (Pkg. 834). Mich. St. U. School of Pkg. Giacin, J.R. 1986. Plastics in Packaging Seminar. Mich. St. U. School of Pkg. June 1986. Gilbert, S.G., Hatzidimitriu, E., Lai, C. and Passy, N. 1983. Studies on Barrier Properties of Polymeric Films to Various Organic Aromatic Vapors. Instrumental Analysis of Food, Vol. 1:405. Gyeszli, I. 1971. Gas and vapor permeability of the double wall compared to single wall plastic packages. Mich. St. U. School of Pkg. thesis. 120 Hamilton, W.E. 1985. Food Packaging -- An Opportunity for Barrier Coextrusions. TAPPI Proceedings, Polymers, Laminations and Coating Conference: 417. Hammer, F.J. 1951. The relation of odor, taste, and flicker-fusion thresholds to food intake. (J. Comp. and Physiol. Psychol. 44, 403 - 411.) Sited in Amerine, et al., 1965. Harte, B.R. 1982. Food Packaging Course. Mich. St. U. School of Pkging. September 1982. Harte, B.R. 1985. Personal Communication. Associate Professor. Mich. St. U. School of Pkging. Hart, H. and Schuetz, R.D. 1978. Organic Chemistry: A Short Course. 5th Edition, Houghton Mifflin Company, U.S.A. Hernandez, R.J., Giacin, J.R., and Banner, A.L. 1986. Measuring the Aroma Barrier Properties of Polymeric Packaging Materials. Pkging Tech. Vol. 16:4 pg. 12 - 15. Hernandez, R.J. 1984. Permeation of Toluene Vapor Through Glassy Poly(ethy1ene) Terephthalate Films. M.S. Thesis, Michigan State University, East Lansing, MI 48824. Kuehner, R.L. 1954. The validity of practical odor measurement methods. (Ann. N.Y. Acad. Sci. 58 (2), 175 - 186.) Sited in Amerine, et al., 1965. 121 Manathunya, V. 1976. Principles of shelf life of Packaged Cereal by an Accelerated Test Technique and a Mathematical Model. Mich. St. U. School of Pkg. thesis. Mancioli, T. 1921. Sited in Amerine, et al., 1965. Mohney, S.M., Hernandez, R.J., Giacin, J.R., and Harte, B.R. 1986. The Aroma Permeability of Packaging Films and It's Relationship to Product Quality. Presented at 13th Annual IAPRI Symposium, Oslo, Norway, May 1986. Murray, L.J. and Dorschner, R.W. 1983. Permeation Speeds Tests, Aids Choice of Exact Material. Package Engineering, Vol. 28, No. 3, pg. 76. Murray, L.J. 1985. An Organic Vapor Permeation Rate Determination for Flexible Packaging Materials. J. Plastic Film and Sheeting, Vol. 1:104. Sagarin, E. 1954. Odor: A Proposal For Some Basic Definitions. (ASTM Spec. Tech. Publ. 164, 3-8.) Sited in Amerine, et al., 1965. Stoll, M. 1957. Facts old and new concerning relationships between molecules structure and odour. (Soc. Chem. Ind. (London) Monograph l, 1 - 12.) Sited in Amerine, et al., 1965. Weast, R.C., Astle, M.J., and Beyer, W.H. 1985 - 1986. CRC Handbook of Chemistry and Physics. 66th Edition, CRC Press, Inc. Boca Ration, Florida. 122 Wyatt, D.M. 1985. Analytical Analysis of Tastes and Odors Impared to Foods by Packaging Materials. TAPPI Proceedings, Polymers, Laminations & Coatings Conference: 325. Zobel, M.G.R. 1982. Measurement of Odour Permeability of Polypropylene Packaging Films at Low Odourant Levels. Polymer Testing, 3:133. Zobel, M.G.R. 1984. Aroma Barrier Properties of Coextruded Films. Proceedings of the 4th Annual International Conference and Exhibition on Coextrusion Markets and Technology, September 19-21, Princeton, NJ. Zobel, M.G.R. 1985. The Odour Permeability of Polypropylene Packaging Film. Polymer Testing, Vol. 5, No. 2:153. "‘)))))))))))))))T