8.1 Introduction Interactions within a package system refer to the exchange of mass and energy between the packaged food, the packaging material and the external environment. Food-packaging interactions can be defined as an interplay between food, packaging, and the environment, which produces an effect on the food, and/or package (Hotchkiss, 1997). Mass transfer processes in packaging systems are normally referred to as permeation, migration and absorption (Fig. 8.1). Permeation is the process resulting from two basic mechanisms: diffusion of molecules across the package wall, and absorption/desorption from/into the internal/external atmospheres. Migration is the release of compounds from the plastic packaging material into the product (Hernandez and Gavara, 1999). The migration of compounds from polymer packaging materials to foods was the first type of interaction to be investigated due to the concern that human health might be endangered by the leaching of residues from the polymerisation (e.g., monomers, oligomers, solvents), additives (e.g., plasticisers, colourants, UV-stabilisers, antioxidants) and printing inks. Later, absorption or scalping of components originally contained in the product by the packaging material attracted attention. Product components may penetrate the structure of the packaging material, causing loss of aroma, or changing barrier and/or mechanical properties, resulting in a reduced perception of quality (Johansson, 1993). The fundamental driving force in the transfer of components through a package system is the tendency to equilibrate the chemical potential (Hernandez and Gavara, 1999). Mass transport through polymeric materials can be described as a multistep process. First, molecules collide with the polymer surface. Then they 8 Packaging-flavour interactions J. P. H. Linssen, R. W. G. van Willige and M. Dekker, Wageningen University, The Netherlands adsorb and dissolve into the polymer mass. In the polymer film, the molecules ‘hop’ or diffuse randomly as their own kinetic energy keeps them moving from vacancy to vacancy as the polymer chains move. The movement of the molecules depends on the availability of vacancies or ‘holes’ in the polymer film. These ‘holes’ are formed as large chain segments of the polymer slide over each other due to thermal agitation. The random diffusion yields a net movement from the side of the polymer film that is in contact with a high concentration or partial pressure of permeant to the side that is in contact with a low concentration of permeant. The last step involves desorption and evaporation of the molecules from the surface of the film on the downstream side (Singh and Heldman, 1993). Absorption involves the first two steps of this process, i.e. adsorption and diffusion, whereas permeation involves all three steps (Delassus, 1997). 8.2 Factors affecting flavour absorption As polymer packaging is more and more widely used for direct contact with foods, product compatibility with the packaging material must be considered. Flavour scalping, or the absorption of flavour compounds, is one of the most important compatibility problems. The problem of aroma absorption by plastic packages has been recognised for many years (Johansson, 1993). Several research groups throughout the world investigated flavour absorption phenomena extensively. It is a complex field, and several factors have been Fig. 8.1 Possible interactions between foodstuff, polymer film and the environment, together with the adverse consequences (Nielsen and Ja¨gerstad, 1994) Packaging-flavour interactions 145 proven to have important effects on the extent of absorption of different flavour compounds by various packaging materials (Nielsen and Ja¨gerstad, 1994). An understanding of absorption between flavour compounds and polymeric packaging materials requires knowledge of the chemical and physical structures of both the flavour compound and the polymer. The properties of a plastic packaging material are the foremost important parameters that control the amount of flavour absorption. The properties of a polymer result from its chemical nature, morphology, formulation (compounding with additives), processing, and even storage and conditions of use. Important parameters derived from the chemical structure, such as glass transition temperature, crystallinity and free volume that have an effect on flavour absorption are essentially determined upon the selection of a particular polymer. 8.2.1 Glass transition temperature (Tg) Figure 8.2 shows the behaviour of one of the many properties of an amorphous and semicrystalline polymer: the modulus of elasticity. There are two sharp breaks indicating phase transitions. At low temperatures the polymer is rigid and brittle: it forms a ‘glass’. At the glass transition temperature Tg the modulus of elasticity drops dramatically. Many of the properties of the polymer change a little at this temperature. Above Tg the polymer becomes soft and elastic; it forms a ‘rubber’. At high temperatures, the polymer may melt, to form a viscous liquid (Wesselingh and Krishna, 2000). The polymers that we know as glassy polymers, such as the polyesters polyethylene terephthalate (PET), polycarbonate (PC) and polyethylene nafthalate (PEN), have a Tg above ambient temperature. At room temperature, glassy polymers will have very stiff chains and very low diffusion coefficients for flavour molecules at low concentrations. Rubbery polymers, such as the polyolefins polyethylene (PE) and polypropylene (PP), have a Tg below ambient temperature. Rubbery polymers have high diffusion coefficients for flavour compounds and steady-state permeation is established quickly in such structures (Giacin and Hernandez, 1997). Stiff-chained polymers that have a high glass transition temperature generally have low permeability, unless they also have a high free volume (Miller and Krochta, 1997). 8.2.2 Free volume The free volume of a polymer is the molecular ‘void’ volume that is trapped in the solid state. The permeating molecule finds an easy path in these voids. Generally, a polymer with poor symmetry in the structure, or bulky side chains, will have a high free volume and a high permeability (Salame, 1989). 8.2.3 Crystallinity The importance of crystallinity to absorption has been recognised for many years. All polymers are at least partly amorphous; in the amorphous regions the 146 Novel food packaging techniques polymer chains show little ordering. However, polymers often contain substantial ‘crystalline’ parts, where the polymer chains are more or less aligned. The crystalline areas are typically a tenth denser than the amorphous parts; for many permeants they are practically impermeable. So, diffusion occurs mainly in the amorphous regions in a polymer, where small vibrational movements occur along the polymer chains. These micro Brownian motions can result in ‘hole’ formation as parts of the polymer chains move away from each other. It is through such ‘holes’ that permeant molecules can diffuse through a polymer (Johansson, 1993; Wesselingh and Krishna, 2000). Therefore, the higher degree of crystallinity in a polymer, the lower the absorption. 8.2.4 Concentration and mixtures of flavour compounds There are relatively few reports relating flavour absorption to the relative concentrations of the sorbants in a liquid or vapour. Mohney et al. (1988) reported that low sorbant concentrations will affect the polymer only to a very limited extent and the amount of absorbed compounds will be directly proportional to the concentration of the sorbants. At higher concentrations, however, the absorption of compounds into a polymer material may alter the polymer matrix by swelling (Charara et al, 1992; Sadler and Braddock, 1990). Consequently, to avoid overestimation of the amounts of absorbed compounds or swelling of the polymer, it is advisable to use a mixture of compounds in the concentration range that can be expected to be found in a food application (Johansson and Leufve′n, 1997). However, to generate reliable and reproducable analytical data, experimental procedures are usually carried out with enhanced Fig. 8.2 Modulus of elasticity against temperature, showing the glass transition and melting temperatures (Wesselingh and Krishna, 2000). Packaging-flavour interactions 147 concentrations. Interactions between different flavour compounds may also affect the absorption of low molecular weight compounds into polymer food packaging materials (Delassus et al, 1988; Kwapong and Hotchkiss, 1987; Letinski and Halek, 1992). Some flavour compounds exhibit a lower absorption rate in mixtures compared to systems containing the individual flavour compounds. This may be due to a competition for free sites in the polymer and/or alteration of the partitioning between the solution and the polymer due to an altered solubility of the compounds in the solution. Therefore, the use of single compound model solutions may cause an overestimation of the amount absorbed in an actual food packaging application (Johansson and Leufve′n, 1997). 8.2.5 Polarity The polarities of a flavour compound and polymer film are an important factor in the absorption process. The absorption behaviour of different classes of flavour compounds depends to a great extent on their polarity. Different plastic materials have different polarities; hence their affinities toward flavour compounds may differ from each other (Gremli, 1996). Flavour compounds are absorbed more easily in a polymeric film if their polarities are similar (Quezada Gallo et al., 1999). Polyolefins are highly lipophilic and may be inconvenient for packaging products with non-polar substances such as fats, oils, aromas etc., since they can be absorbed and retained by the package (Hernandez- Mun?oz et al., 2001). The polyesters, however, are more polar than the polyolefins and will therefore show less affinity for non-polar substances. 8.2.6 Molecular size and structure The size of the penetrant molecule is another factor. Smaller molecules are absorbed more rapidly and in higher quantities than larger molecules. Very large molecules plasticise the polymer, causing increased absorption into the newly available absorption sites (Landois-Garza and Hotchkiss, 1987). Generally, the absorption of a series of compounds with the same functional group increases with an increasing number of carbon atoms in the molecular chain, up to a certain limit. Shimoda et al. (1987) reported that absorption of aldehydes, alcohols and methyl esters increased with increasing molecular weight up to about ten carbon atoms. For even larger molecules the effect of molecular size overcomes the effect of the increased solubility of the compounds in the polymer, and the solubility coefficient decreased. Linssen et al. (1991a) reported that compounds with eight or more carbon atoms were absorbed from yoghurt drinks by HDPE, while shorter molecules remained in the product. They also observed that highly branched molecules were absorbed to a greater extent than linear molecules. 148 Novel food packaging techniques 8.2.7 Temperature Temperature is probably the most important environmental variable affecting transport processes. The permeability of gases and liquids in polymers increases with increasing temperature according to the Arrhenius relationship. Possible reasons for increased flavour absorption at higher temperatures are (Gremli, 1996): ? increased mobility of the flavour molecules ? change in polymer configuration, such as swelling or decrease of crystallinity ? change in the volatile solubility in the aqueous phase. 8.2.8 Relative humidity For some polymers, exposure to moisture has a strong influence on their barrier properties. The presence of water vapour often accelerates the diffusion of gases and vapours in polymers with an affinity for water. The water diffuses into the film and acts like a plasticiser. Generally, the plasticising effect of water on a hydrophilic film, such as ethylene-vinyl alcohol (EVOH) and most polyamides, would increase the permeability by increasing the diffusivity because of the higher mobility acquired by the polymer network (Johansson, 1993). Absorbed water does not affect the permeabilities of polyolefins and a few polymers, such as PET and amorphous nylon, show a slight decrease in oxygen permeability with increasing humidity. Since humidity is inescapable in many packaging situations, this effect cannot be overlooked. The humidity in the environment is often above 50%RH, and the humidity inside a food package can be nearly 100%RH (Delassus, 1997). 8.3 The role of the food matrix The quality and the shelf-life of the packaged food depend strongly on physical and chemical properties of the polymeric film and the interactions between food components and package during storage. Several investigations have shown that considerable amounts of aroma compounds can be absorbed by plastic packaging materials, which can cause loss of aroma intensity or an unbalanced flavour profile (Hotchkiss, 1997; Arora et al., 1991; Lebosse′ et al., 1997; Linssen et al., 1991a; Nielsen et al., 1992; Paik, 1992). The composition of a food matrix is of great importance (besides other factors, see Fig. 8.3) in determining the amount of flavour absorption by plastic packaging materials. There is only limited information available in literature about the influence of the food matrix on flavour absorption by polymers. Linssen et al. (1991b) and Yamada et al. (1992) showed that the presence of juice pulp in orange juice decreased absorption of volatile compounds into polymeric packaging materials. They suggested that pulp particles hold flavour compounds (e.g., limonene) in equilibrium with the Packaging-flavour interactions 149 watery phase, which could be responsible for the decrease of absorption of these compounds by the plastics. Fukamachi et al. (1996) studied the absorption behaviour of flavour compounds from an ethanolic solution as a model of alcoholic beverages. The absorption of a mixture of homologous volatile compounds (esters, aldehydes and alcohols with carbon chain length 4-12) into LDPE film first increased with a maximal absorption at 5–10% (v/v) aqueous ethanol and then decreased remarkably with increasing ethanol concentration. EVOH film showed similar absorption behaviour, with maximal absorption at 10–20% (v/v) aqueous ethanol. Nielsen et al. (1992) investigated the effects of olive oil on flavour absorption into LDPE. Olive oil and, thereby, the flavours dissolved in the oil, were absorbed in large amounts by the plastic. The partition coefficients for alcohols and short-chained esters in an oil/polymer system were higher than in a water/polymer system, while the partition coefficients for aldehydes and long- chained esters were lower in an oil/polymer system than in a water/polymer system. Not only the type of plastic used is of importance for the uptake of aroma compounds, but also possible interactions between flavour and food components. Flavour components may be dissolved, adsorbed, bound, entrapped, encapsulated or retarded in their diffusion through the matrix by food components. The relative importance of each of these mechanisms varies with the properties of the flavour chemical (functional groups, molecular size, shape, volatility, etc.) and the physical and chemical properties of the components in the food (Kinsella, 1989; Le Thanh et al. 1992). Knowledge of the binding behaviour of flavour components to non-volatile food components and their partitioning between different phases (food component/water and water/polymer) is of great importance in estimating the rate and amount of absorption by polymers. Because many food products are emulsions of fat and water, such as milk and milk products, the fat content is an important variable in the food matrix. Fat/oil content is often reduced in order to Fig. 8.3 Factors influencing flavour absorption by plastic polymers (Van Willige, 2002c). 150 Novel food packaging techniques decrease caloric intake to make food healthier. Removal or reduction of lipids can lead to an imbalanced flavour, often with a much higher intensity than the original full fat food (Widder and Fischer, 1996; Ingham et al., 1996). De Roos (1997) reported that in products containing aqueous and lipid phases, a flavour compound is distributed over three phases: fat (or oil), water, and air. Flavour release from the oil/fat phase of a food proceeded at a lower rate than from the aqueous phase. This was attributed, first to the higher resistance to mass transfer in fat and oil than in water and, second to the fact that in oil/water emulsions flavour compounds had initially to be released from the fat into the aqueous phase before they could be released from the aqueous phase to the headspace. Kinsella (1989) reported that several mechanisms might be involved in the interaction of flavour compounds with food components. In lipid systems, solubilisation and rates of partitioning control the rates of release. Polysaccharides can interact with flavour compounds mostly by non-specific adsorption and formation of inclusion compounds. In protein systems, adsorption, specific binding, entrapment, encapsulation and covalent binding may account for the retention of flavours. Oil and fatty acids can also be absorbed by polymers (Arora and Halek, 1994; Riquet et al., 1998) resulting in increased oxygen permeability (Johansson and Leufve′n, 1994) and delamination of laminated packaging material (Olaffson and Hildingson, 1995; Olaffson et al., 1995) However, the availability of data about the influence of oil on the absorption of flavour compounds by plastic packaging materials is limited. Nielsen et al. (1992) found that some apple aroma compounds added to and stored in pure olive oil were lost to a greater extent to LDPE than from an aqueous solution, probably due to differences in polarity of the aromas, polymer and solutions. Thus, oil/fat has a major influence on flavour compounds (perception, intensity, volatility, etc.) and on the properties of packaging material. Van Willige et al. (2000a, b) did a more detailed study on the influence of the composition of the matrix on food products. These authors used a model system, consisting of limonene, decanal, linalol and ethyl 2-methylbutyrate to study flavour scalping in LLDPE from different models representing differences in food matrices. The proteins, -lactoglobuline ( -lg) and caseine were able to suppress absorption of decanal and limonene, because -lg interacted irreversibly with decanal and caseine was capable of binding limonene and decanal by hydrophobic and covalent interactions. Dufour and Haertle′ (1990) and Charles et al. (1996) reported that -lg does not bind terpenes as limonene and linalol. The behaviour of ethyl 2-methylbutyrate could not be fully explained and needs further investigation. The presence of carbohydrates also affected the absorption of flavour compounds by LLDPE. Absorption rates of limonene and to a lesser extent of decanal were decreased in the presence of pectine and carboxymethylcellulose. Increasing viscosity slowed down diffusion of flavour compounds from the matrix to LLDPE. Roberts et al. (1996) also reported that thickened solutions of similar viscosity did not show the same flavour release. Their results showed an Packaging-flavour interactions 151 influence of both viscosity and binding interactions with the thickener on the release of flavour. Binding interactions with carbohydrate-based thickeners are often due to adsorption, entrapment in microregions, complexation, encapsulation and hydrogen bonding between appropriate functional groups (Kinsella, 1989; Damodaran, 1996). The presence of disaccharides, lactose and saccharose was able to bind water and cause a salting out effect of the lesser polar flavour compounds, linalol and ethyl 2-methylbutyrate, resulting in an increased absorption in the polymer. Also Godshall (1997) reported that disaccharides can lower the amount of bulk water due to hydration, which increases the effective concentration of flavour compounds and therefore can enhance their absorption into polymers. The main effect of the influence of the food matrix on flavour scalping, however, is the presence of oil or fat. Even a small amount of oil (50g/l) had a major effect on the amount of flavour absorption. Absorption of limonene and decanal is reduced to approximately 5%. A quantity of oil as low as 2 g/l results in a decrease of about 50% of absorption, meaning that the presence of oil very strongly influences the level of absorption of flavour compounds in polymeric packaging material (Fig. 8.4). The composition of the food matrix plays a major role in the absorption of flavour compounds by LLDPE. Several studies have already revealed that flavour compounds interact with oil, carbohydrates and proteins, but the Fig. 8.4 Influence of oil on the relative absorption of limonene, decanal, linalool and ethyl-2-methylbutyrate (E2MB) by LLDPE after one day of exposure at 4o (Van Willige et al., 2000a) 152 Novel food packaging techniques influence on flavour absorption by plastic packaging materials in different food matrices has been unclear for a long time. Van Willige et al. (2000a,b) showed that food components can affect the quantity of absorbed flavour compounds by LLDPE in the following order: oil or fat >> polysaccharides and proteins > disaccharides. Because of the lipophilic character of many flavour compounds, food products with a high oil or fat content will lose less flavour by absorption into LLDPE packaging than food products containing no or a small quantity of oil. 8.4 The role of differing packaging materials An important requirement in selecting food-packaging systems is the barrier properties of the packaging material. Barrier properties include permeability of gases (such as O 2 , CO 2 , N 2 and ethylene), water vapour, aroma compounds and light. These are vital factors for maintaining the quality of foods. A good barrier to moisture and oxygen keeps a product crisp and fresh, and reduces oxidation of food constituents. Plastics are widely used for food packaging due to their flexibility, variability in size and shape, thermal stability, and barrier properties. PE and PP have been used for many years because of their good heat sealability, low costs and low water vapour permeability. However, poor gas permeability makes laminating of PE with aluminium foil and paper necessary. During the last decades, PET and, to a lesser extent, PC have found increased use for food packaging. PET has good mechanical properties, excellent transparency and relatively low permeability to gases. PC is tough, stiff, hard and transparent, but has poor gas permeability properties and is still quite expensive. Unlike glass, plastics are not inert allowing mass transport of compounds such as water, gases, flavours, monomers and fatty acids between a food product, package and the environment due to permeation, migration and absorption. The quality and shelf-life of plastic-packaged food depend strongly on physical and chemical properties of the polymeric film and the interactions between food components and package during storage. Several investigations showed that considerable amounts of aroma compounds can be absorbed by plastic packaging materials, resulting in loss of aroma intensity or an unbalanced flavour profile (Van Willige et al., 2000a,b; Arora et al., 1991; Lebosse′ et al., 1997; Linssen et al., 1991b; Nielsen et al., 1992; Paik, 1992) Absorption may also indirectly affect the food quality by causing delamination of multilayer packages (Olafsson and Hildingsson, 1995; Olafsson et al., 1995) or by altering the barrier and mechanical properties of plastic packaging materials (Tawfik et al., 1998). Oxygen permeability through the packaging is an important factor for the shelf-life of many packed foods. Little information is available in literature about the influence of absorbed compounds on the oxygen permeability of packaging materials. Hirose et al. (1988) reported that the oxygen permeability of LDPE and two types of ionomer increased due to the presence of absorbed d-limonene. Johansson and Leufve′n (1994) studied the effect of rapeseed oil on the oxygen barrier properties Packaging-flavour interactions 153 of different polymer packaging materials. They found that amorphous PET remained an excellent oxygen barrier even after storage in rapeseed oil for 40 days. The polyolefins (PP and high-density PE) showed an increased oxygen transmission rate (OTR) after being in contact with rapeseed oil for 40 days. This was attributed to swelling of the polymer matrix. However, the increase in OTR was not proportional to the amount of absorbed oil. Sadler and Braddock (1990) showed that the oxygen permeability of LDPE was proportional to the mass of absorbed limonene. In another paper, they concluded that oxygen permeability of LDPE and the diffusion coefficients of citrus flavour volatiles in LDPE were related to the solubility of these compounds in the polymer (Sadler and Braddock, 1991). The increased oxygen permeability of LDPE could only be explained by absorption. Attachment of volatile molecules at the polymer surface (adsorption) might hinder oxygen permeation, which would lower the oxygen permeation, or leave it unchanged. An increased oxygen permeability of LDPE indicated that absorption of volatiles must be responsible for structural changes in the polymer. Flavour absorption can have a major influence on the oxygen permeability of plastic packaging materials, and consequently on the shelf-life of a food product, making it necessary to investigate this important aspect more thoroughly. Van Willige et al., (2002b) investigated the influence of oxygen permeation on absorption of several flavour compounds (limonene, decanal, hexyl acetate and 2-nonanone) into LDPE, PP, PC and PET packaging materials. They measured the oxygen permeability of the exposed polymer specimens with a set- up based on the isostatic continuous flow technique. In the isostatic method the pressure differential across the test film remains constant during the total permeation process. Whereas the high-pressure side (oxygen chamber) remains constant at a certain value, the low-pressure side (nitrogen chamber) is maintained by sweeping the permeated molecules with a continuous flow of carrier gas (Hernandez and Gavara, 1999). Figure 8.5 gives a good picture of the influence of the total amount of flavour absorption on the oxygen permeability of all four investigated polymers. PP and LDPE showed an increase in oxygen permeability after absorption of flavour compounds. This increase in oxygen permeability indicated that molecular changes occurred in the polymer network. Several researchers reported that swelling of a polymer by a permeant (i.e. plasticising) greatly increased the diffusivity. During the absorption process molecules are absorbed in the free volume (‘holes’) which is always present in the amorphous regions. Diffusion and a slow relaxation of the polymer, reducing the intercatenary forces and even promoting polymer swelling control the rate of absorption. This further enhances the rate of diffusion, which further influences the relaxation. As a result, the permeation of one component affects the permeation of another component, i.e. the plasticising effect within the polymer matrix becomes apparent (Halek, 1988; Hernandez-Mun?oz et al., 1999). Absorbed water has a similar effect on the permeability of some hydrophilic polymers, such as ethylene vinyl alcohol (EVOH) and most polyamides. Water 154 Novel food packaging techniques molecules absorbed at high relative humidities are believed to combine with hydroxyl groups in the polymer matrix and weaken the existing hydrogen bounds between polymer molecules. As a result, the interchain distances increase and thus free volume, facilitating the diffusion of oxygen and perhaps other gases. The presence of water in the hydrophilic polymer matrix not only influences how a permeant is sorbed and diffused, it also leads to depression of the glass transition temperature (Tg) of the polymer due to the plasticising effect of water. When the Tg drops below storage temperature, a substantial increase in oxygen permeability is expected (Zhang et al., 1999; Delassus et al., 1988). Krizan et al. (1990) reported that free volume in a polymer is the dominant factor in determining the permeation properties. A plot of the log of the oxygen permeability coefficients versus the reciprocal of the specific free volume showed a good linear relationship. Also Sadler and Braddock (1990) reported that the oxygen permeability was proportional to the mass of absorbed limonene. The specific molecular composition of a flavour compound seems to play a more important role than the mass of absorbed flavour compounds. Each individual absorbed flavour compound caused swelling of PP; i.e. increased the specific free volume. Rubbery polymers (LDPE and PP) have very short relaxation times and respond very rapidly to stresses that tend to change their physical conditions. Glassy polymers (PC and PET) have very long relaxation times. Penetrant (molecules) can therefore potentially be present in ‘holes’ or irregular cavities with very different intrinsic diffusional mobilities (Stern and Trohalaki, 1990). Hernandez-Mun?oz et al. (1999) reported that there are two possible effects of absorbed flavour compounds on oxygen mass transport: (i) flavour compounds Fig. 8.5 Influence of total flavour absorption on oxygen permeability of PP, LDPE, PC and PET at 25oC (Van Willige et al., 2002b) Packaging-flavour interactions 155 and oxygen compete for the same sites, reducing the solubility of oxygen since many sites are already occupied and (ii) the flavour compounds swell the polymer, opening the structure and increasing polymer free volume, i.e. oxygen transport. The presence of holes is assumed for rubbery polymers as well as for glassy polymers. ‘Hole filling’ is suggested as an important sorption mode above as well as below Tg, with one crucial difference between the sorption mechanisms in the rubbery and glassy regions; hole saturation does not occur in the rubbery state because new holes are formed to replace those filled with penetrant molecules (Stern and Trohalaki, 1990). Landois-Garza and Hotchkiss (1988) reported that the presence of water molecules in the polymer matrix occupied ‘holes’ that otherwise would be available for the diffusion of permeant molecules, effectively increased the length of the viable diffusion paths, and diminished the permeant diffusivity. The linear decrease of the oxygen diffusivity of PC due to flavour absorption suggests that ‘hole filling’, resulting in an increased oxygen diffusion pathway, was also found in this study. However, the oxygen permeability of PET, which is also in its glassy state at 25oC, was not significantly affected by absorption of flavour compounds. Because of the low oxygen permeability of PET, which was close to the detection limit of the oxygen analyser, a significant effect of flavour absorption on oxygen permeability cannot be ruled out. A more sensitive oxygen analyser or a smaller permeation cell should be used in order to investigate the influence of absorption of flavour absorption on the oxygen permeability of PET. Van Willige et al. (2002b) concluded that flavour absorption increased oxygen permeability of PP and LDPE by 130% and 21% after 8 hours of exposure to various flavour compounds. Because of the higher oxygen permeability a reduction in the shelf-life of oxygen sensitive products, which are packed in LDPE or PP and contain the tested flavour compounds (such as orange juice and apple juice) can be expected. Furthermore, flavour absorption has probably a positive effect on the shelf-life of oxygen-sensitive products packed in PC, because of the reduction in oxygen permeability of 11% after 21 days of exposure to various flavour compounds. Oxygen permeability of PET was not influenced by the presence of flavour compounds, meaning that PET remained a good oxygen barrier. One should realise that the concentrations of flavour compounds in real food products are usually substantially lower, with the exception of limonene, than the concentrations used in this study. Therefore, the observed effects may be less or even not significant in foods and beverages. 8.5 Flavour modification and sensory quality One of the main aspects of flavour scalping is how this phenomenon is able to affect the quality of foods. In this field a lot of research has been carried out on fruit juices. During the last decades the quality of juices, aseptically packed in laminated cartons, has been investigated extensively. Loss of organoleptic 156 Novel food packaging techniques characteristics during storage has been commonly observed (Marshall et al., 1985, Moshonas and Shaw 1989a, b). LDPE laminated carton packs, such as Tetra Brik and Combibloc , are commonly used for packaging aseptically filled juices. LDPE is able to absorb considerable amounts of flavour compounds (Arora et al., 1991, Nielsen et al., 1992, Van Willige et al., 2002a). Therefore, food industries often correct this absorptive effect by adding excess flavour compounds to the food for keeping taste and flavour acceptable for consumers until the end of the product’s shelf-life (Lebosse′ et al., 1997). Although instrumental analysis indicates that considerable amounts of flavour compounds are absorbed into polymeric packaging and that the loss of flavours may be high enough to affect the sensory quality of a packaged food, only few authors have conducted sensory tests to go along with the analytical results (Du¨rr et al., 1981, Kwapong and Hotchkiss, 1987, Mannheim et al., 1987, Moshonas and Shaw, 1989b, Sharma et al., 1990). Du¨rr et al. (1981) reported that absorption of d-limonene up to 40% did not affect the sensory quality of orange juice during three months storage at 20oC. d-Limonene was suggested scarcely to contribute to the flavour of orange juice. Moreover, they even considered limonene absorption as an advantage, since limonene is known as a precursor to such off-flavour compounds as -terpeniol. They also reported that the storage temperature was the main quality parameter for the shelf-life of orange juice. Kwapong and Hotchkiss (1987) found that assessors were able to detect significant differences in odour profile due to absorption of citrus essential oils from aqueous model solutions by LDPE strips. Moshonas and Shaw (1989a) reported significant reduced flavour scores using a sensory panel in a commercial aseptically packaged orange juice stored for six weeks at 21oC and 26oC. The detected flavour changes were caused by the combined loss of limonene due to absorption together with the increase of potential off-flavour compounds. Mannheim et al., (1987) found that the product shelf-life of orange and grapefruit juices was significantly shorter in LDPE laminated cartons than in glass jars. A loss of ascorbic acid and an increase in brown colour was observed and a 40% decrease of limonene was found; other volatiles were not assayed. After ten weeks of storage at 25oC they revealed a difference in taste. Sharma et al. (1990) reported that PE and PP contact did not cause perceptible changes in sensory quality of fruit squash (orange and lemon) and beverages (mango, orange and blue grapes). Pieper et al., (1992) stored orange juice in glass bottles and in LDPE laminated cardboard packages at 4oC for 24 weeks. Absorption of d-limonene up to 50% and small amounts of aldehydes and alcohols by the packaging materials did not affect the sensory quality of orange juice significantly. A reason could be the low storage temperature. Sadler et al., 1995 reported that no evidence was found that flavour absorption directly altered sensory characteristics of orange juice through general or selective absorption of volatile compounds by LDPE, PET and EVOH after three weeks of storage at 4.5oC. Marin et al. (1992) exposed orange juice to LDPE and an ionomer (i.e. Surlyn). The polymers absorbed Packaging-flavour interactions 157 more than 70% of the limonene content in 24 hours at 25oC. However, results from gaschromatography-olfactometry (GCO) analysis indicated that limonene possessed only trace odour activity. Furthermore, the polymers did not alter the odour-active components present in orange juice substantially. Van Willige et al. (2003) investigated three types of packaging materials (LDPE, PET and PC) on flavour scalping in a model system (consisting of the compounds octanal, decanal, ethylbutyrate and 2-nonanone) and a reconstituted orange juice. Sensory evaluation was carried out by a sensory panel of 27 assessors. From the model system valencene was almost completely absorbed by LDPE, followed to a lesser extent by decanal, hexylacetate, octanal and nonanone. Much less flavour compounds were absorbed by PC and PET. In contrast to LDPE valencene was absorbed to the lowest extent and decanal to the highest. From the orange juice limonene was readily absorbed by LDPE, while myrcene, valencene, pinene and decanal were absorbed in smaller quantities. Only three flavour compounds were absorbed from orange juice by PC and PET in very small amounts: limonene, myrcene and decanal. Although instrumental analysis showed a substantial decrease of the flavour content between control and polymer treated sample, meaning that the polymers absorbed a substantial amount of the flavour compounds, the sensory panel was not able to detect any significant differences caused by flavour scalping. Van Willige (2002b) stated that flavour scalping is not the main reason for a possible change of flavour perception during storage of food products into polymer packaging materials. It is more likely that other mechanisms play a more important role, such as chemical degradation, resulting in a development of off-flavour compounds. Sizer et al. (1988) stated that storage temperature remains the single most important factor in delaying flavour loss and achieving satisfactory shelf-life and quality. From all the published data, Gremli (1996) stated that there is ample evidence that flavour compounds migrate from beverages and foods into plastic packaging materials. However, investigations about the relevance of the loss of flavour compounds for the sensory quality of a product are insufficient and sometimes contradictory because flavour alteration depends on many parameters, such as storage temperature and type of packaging material. Therefore, investigations regarding the effect of flavour absorption on sensory quality of a product should be carried out at ambient temperature (i.e. usual storage conditions of aseptic packs), because the rate and amount of flavour absorption by packaging materials increases with increasing temperature (Van Willige et al., 2002a). Moreover, it is important that the polymer treated and untreated (=control) samples are comparable. That means similarly packed and using a sensory procedure that evaluates complete packaging systems with similar properties, e.g. oxygen permeability (i.e. glass-glass, and not glass- laminated carton). 158 Novel food packaging techniques 8.6 Case study: packaging and lipid oxidation Lipid oxidation is an important chemical process in food products containing (poly)unsaturated fats. It is a process that leads eventually to the formation of volatile off-flavours by further reaction of the formed hydroperoxides in secondary lipid oxidation reactions (Nawar, 1996). Dekker et al. (2002) studied the primary oxidation process in a defined model system for a packed food using sunflower oil. The package consisted of a PE/EVOH/PE tub, with a top-film of either LLDPE or PE/EVOH/PE. At regular time intervals measurements were made of: oxygen concentrations in headspace and oil (sensor technology), peroxide value of the oil and the volume of the headspace. Storage of the packages was in the dark at three different temperatures (25, 37 and 50oC). The headspace was either air or nitrogen, and for some experiments oxygen scavengers (Ageless) were used glued to the top-film. The degree of lipid oxidation was studied as a function of temperature, top-film material, initial headspace composition and the presence or absence of an oxygen scavenger. In Fig. 8.6 the effect of temperature is clearly visible, at 50oC the peroxide values obtained are about tenfold the values at 37oC. The effect of initial headspace composition and scavenger is as expected, although the effects of these packaging concepts is limited under the present conditions, especially due to the low barrier properties of the used LLDPE film. When the high barrier EVOH film is used, oxidation stops after all oxygen from the package headspace is consumed. For the description of product quality the level of primary oxidation is only an indicator value. The real problem is the subsequent secondary oxidation process that leads to the formation of off-flavours. In section 8.7.2. the process of lipid oxidation in different packaging concepts is translated into a mathematical model. Fig. 8.6 Peroxide value for different temperatures and packaging concepts for LLDPE top-film. Packaging-flavour interactions 159 8.7 Modelling flavour absorption 8.7.1 Modelling of flavour scalping Knowledge of solubility and binding behaviour of flavour compounds to non- volatile food components and their partitioning behaviour between different phases (component/water, component/oil or component/oil/water on one side and water/polymer, oil/polymer or water/oil/polymer on the other side) is of great importance in estimating the rate and amount of absorption from real food products by polymers. Enormous amounts of different flavour compounds are used in foods. It is impossible to study them all. A determination of the relationship between flavour compounds and polymeric packaging materials for predicting flavour absorption would save research time for the packaging industry. Prediction of flavour absorption in relation to the packed food and the packaging material would be a valuable tool in product development. It can help the food industry in choosing packaging material or in determining product formulation. In the literature little information is available on the prediction of flavour absorption. Attempts were made by using several theories. Tigani and Paik (1993) used the dielectric constants of polymers and flavour compounds to predict flavour absorption. They concluded that the dielectric constant might not encompass all of the factors for an accurate prediction of absorption by polymers. Paik and Tigani (1993) examined the application of Hildebrand’s regular solution theory for predicting the equilibrium absorption of flavour compounds by polymer packaging materials. However, they found a poor correlation between the regular solution theory and flavour absorption values, indicating that the entropy contribution could not be assumed negligible. Paik and Writer (1995) applied the Flory-Huggins equation for prediction of flavour absorption. The Flory-Huggins theory is based on the entropy contribution due to molecular size and shape differences of molecules. They showed that the Flory-Huggins equation gave much better estimations of flavour absorption than the regular solution equation. However, the Flory-Huggins equation still did not adequately predict flavour absorption but it can provide a qualitative prediction of flavour absorption, which can be useful for selection and design of packaging materials. Finally, Li and Paik (1996) tried to estimate flavour absorption by the Universal Functional-group Activity Coefficient (UNIFAC) group contribution model. This is based on a semi-empirical model for liquid mixtures called UNIversal QUasi-chemical ACtivity (UNIQUAC). Comparison between the experimental and calculated data indicated that the UNIFAC model was much more accurate in absorption prediction than the regular solution theory and the Flory-Huggins equation. Flavour absorption by a solid (amorphous) polymer is a meta-equilibrium state that often requires a long time to reach equilibrium. The equilibrium distribution of flavour compounds will depend on their partitioning behaviour of compounds between different phases in the system: polymeric packaging and food matrix. The properties of the package, such as polarity and crystallinity, as 160 Novel food packaging techniques well as the composition of the food matrix (presence of oil, proteins, carbohydrates) are extremely important factors. Modelling of flavour absorption could be based upon a set of equations describing these equilibrium distributions together with mass balances of the flavour compounds. The final goal is to make predictions of flavour absorption for other food matrices and other compounds based upon their characteristics, such as polarity and molecular weight. In the future, a fitting model could be extended with the dynamics of the absorption phenomena (including mass transfer effects as a consequence of product texture, viscosity, etc.) and also for different packaging materials. In a food-packaging material system flavour molecules will strive for a thermodynamic equilibrium situation in which their chemical activities in all phases of the system will be equal. The time it will take to reach this equilibrium will depend on the composition of the food matrix. It can be assumed that in liquid foods this equilibrium will be reached well before consumption of the product. Experimental data of flavour absorption confirm this (van Willige et al., 2000a, b). In solid or highly viscous food products this equilibrium might take longer, in that case only the outer part of the food will be affected by the flavour absorption effect. Dekker et al. (2003) proposed a formula based on equilibrium in a food-package system. The equilibrium concentration in the packaging material can be calculated from equation 8.1: C P;1 M FP C FP;t 0 M O K O=A K P=A M A K P=A M P 8:1 in which C is concentration (mg/g), M is mass (g) and K is the partitioning coefficient ( ). The indices F, P, O and A refer to food, polymer, oily and aqueous phase, respectively. To make predictions of the extent of flavour absorption, information is required about the value of the partition coefficients for the flavours of interest. The partitioning will depend largely on the nature of the flavour, especially on its polarity. Experimental determination of all partition coefficients is a very laborious task, therefore models describing the relation of the partition coefficients with known quantitative information on the nature of the flavour molecules are valuable. Dekker et al. (2003) made an attempt to do this based upon the log P value of the flavours, which is a good measure of their polarity. These values are reported for many molecules and can also be calculated from their chemical structure. Figure 8.7 shows the relationship between log P values and partitioning coefficients of the flavour compounds limonene, decanal and linalool. Figure 8.7 also shows that a linear relationship between the log P values and the partition coefficients is obtained (R 2 is 0.95 and 0.99 for K O/A and K P/A respectively). In equations 8.2 and 8.3 the relations are given: Log K O=A 0:85 0:88 log P 8:2 Packaging-flavour interactions 161 Log K P=A 5:00 1:60 logP 8:3 With equations 8.2 and 8.3 a prediction can be made of the partition coefficient of other flavour compounds, which have log P values in the range of the studied compounds (log P 3 to 5). In this way the amount of flavour absorption is predicted and this can be used for selection of packaging concepts for giving indications for adjustment of the formulation of the product accordingly. The modelling of the absorption of flavour molecules into LLDPE based on the partitioning behaviour between the different phases in the systems enables the prediction of this phenomenon based on the polarity of the flavour compounds involved. This can limit the amount of work that would be required for experimental determination of the amount of absorption in product development. Future research could focus on the extension of this modelling approach for other polymer packaging materials and other conditions. 8.7.2 Modelling of packaging and lipid oxidation To enable predictions of the degree of lipid oxidation in a given situation mathematical models can be useful. Dekker et al., (2002) developed a model for primary oxidation that included the effect of packaging material and geometry, temperature by using data on diffusion and reaction rates for sunflower oil. To model the oxidation reaction rates in the packed product one has to consider the barrier properties of the packaging material used. This was described by equation 8.4: Fig. 8.7 Relationship between log-P values of the flavour compounds and their partition coefficients K O/A ( a73 ) and K P/A (a76) at 4oC (Van Willige, 2002d). 162 Novel food packaging techniques V h dO 2;headspace dt permeation P A film p O 2 8:4 In which V h is the headspace volume, P the permeation coefficient of the packaging material, A the surface area, and p O2 the oxygen pressure difference between air and the headspace. It was assumed that at the interface between the product and the headspace an equilibrium exists between the oxygen concentrations in both phases, as described by equation 8.5: O 2;headspace O 2;oil j x O H 8:5 In which H is the partitioning coefficient for oxygen. In the product one has to calculate the rates of the primary lipid oxidation reactions. The formation of radicals was assumed to be mainly due to an oxygen induced initiation reaction. This led to the following equations describing the reaction rates in the product phase (8.6–8.9): dO 2 dt reaction k 1 RH O 2 k 2 R O 2 8:6 d R dt k 1 RH O 2 k 2 R O 2 k 3 ROO RH 8:7 d ROO dt k 2 R O 2 k 3 RH ROOH 8:8 d ROOH dt k 3 ROO RH 8:9 Diffusion will take place because of the concentration difference of oxygen and all lipid components involved in the reactions. The diffusion process was described by the second law of Fick: @C i @t D i @ 2 C i @x 2 8:10 In which D i is the diffusion coefficient of component i, and x is the space- coordinate. All reaction, diffusion and permeation rates will depend on temperature. To describe this the Arrhenius equation was used: k n k 0 e Ea R T 8:11 In which k 0 is the pre-exponential factor, E a is the activation energy, R the gas constant, and T the absolute temperature. With similar experiments to the one presented in section 8.6 using additional temperatures, the parameters used in equations 8.4–8.11 were estimated. An example of the description of the measured values of peroxide concentration, oxygen concentration and headspace volume is shown in Fig. 8.8. The parameters obtained enable prediction of the lipid oxidation for different packaging geometry, temperature, and initial headspace composition. It should be realised that primary lipid oxidation is not Packaging-flavour interactions 163 the final quality indicator for edible oils. For the sensory quality of oils, the further (secondary) reactions of the fatty acid hydroperoxides results in the formation of volatile compounds that are responsible for the perceived off- flavour (rancidity) of the product by consumers (see section 8.6.). The predictive modelling approach enables an efficient way of assessing the performance of different packaging concepts for oxidation sensitive products. Before conducting actual shelf-life experiments it is possible to get good estimates of what will happen with the product as a response to a different packaging condition (inclusion of a scavenger, modified atmosphere, or various material properties). Further research has to focus on the modelling of secondary oxidation in relation to sensory quality and the modelling of more complex food products like emulsions or structured foods. 8.8 Packaging–flavour interactions and active packaging How important are packaging–flavour interactions? It depends on the extent they can affect the quality of the packed food. The food matrix is one of the main aspects which determine how important packaging–flavour interactions are. Food products containing fat or oil are able to keep the flavour compounds in the food itself and the loss caused by flavour scalping will be diminished. Some proteins are able to bind some flavour compounds that are no longer available for absorption into a plastic polymer. Aqueous food products have less ability to bind flavour compounds in the food matrix and therefore these foods are more susceptible to losing flavour compounds in the packaging polymer, which can result in quality defects. Many flavour compounds have a lipophilic character and therefore a good affinity to apolar polymers such as PE and PP. It Fig. 8.8 Experimental data and model description of the lipid oxidation of sunflower oil with EVOH top-film. 164 Novel food packaging techniques has been shown that the highest amounts of absorbed flavour compounds are found in these types of polymers. Polyesters, such as PET, PC and PEN, have a more polar character and therefore they show less affinity to the common flavour compounds. This means that polyesters absorb fewer flavour compounds and these polymers are therefore better packaging materials in the context of loss of flavour compounds due to flavour scalping. On the other hand, generally, there is less evidence that flavour scalping influences the taste and odour of a product. Although flavour compounds can be absorbed in substantial amounts, sensory defects are rarely found. Another factor is the way the polymer properties are affected. There is evidence that oxygen permeation can be enhanced due to absorption of flavour compounds. This means that as a secondary aspect, food quality can be affected due to oxidative chemical reactions, e.g., lipid oxidation can influence the quality of the product. A second parameter of importance is the mechanical properties of a polymer. Some rare research work could be found dealing with the way flavour absorption affects the mechanical strength of a polymer (Tawfik et al., 1998). On the other hand packaging and flavour interactions may act in a possitive way. Attempts to use food packaging and flavour interactions in a positive sense is also mentioned in literature. We are now entering the field of active packaging applications. Polymer packaging material can be used to remove selectively undesirable compounds, causing off flavours, from the packed food. In certain orange varieties a bitter compound, limonin, is developed during the extraction and pasteurisation process of the juice. Inclusion of an absorbent might remove such a compound selectively (Chandler et al., 1968; Chandler and Johnson, 1979). Also an active packaging concept has been described for reducing bitterness of grapefruit juices, caused by the presence of naringin. A thin cellulose acetate layer is applied to the inside of the packaging as an absorbent. Such a layer contains the enzyme naringinase, which hydrolyses naringin to non-bitter compounds (Soares and Hotchkiss, 1998a,b). Other types of compounds responsible for off flavours are amines and aldehydes. Such compounds could also be removed by applying active packaging. Amines are formed from protein breakdown in fish muscle and include strongly alkaline compounds (Rooney, 1995). Vermeiren et al. (1999) reported that a Japanese patent claimed the removal of amines from food by interaction between acids incorporated in the polymer and the off-flavour compounds. ANICO Company Ltd (Japan) introduced another approach to remove amine odours. Bags made from a polymer containing ferrous salt and an organic acid claimed to oxidise the odours as they are absorbed by the polymer (Rooney, 1995). Aldehydes are formed in the lipid autoxidation reaction and they can reduce the quality of food products considerably. Dupont polymers claimed the selective removal of aldehydes from packaging headspaces by using a layer of Bynel IXP 101, which is a HDPE masterbatch (Rooney, 1995). Packaging-flavour interactions 165 In conclusion, although it is very clear that packaging and flavour interactions exist, this phenomenon does not influence the food quality to the extent that it causes insuperable problems in practical situations. Moreover, in some cases packaging and flavour interactions can help to maintain the desired quality of food products. 8.9 References ARORA A P and HALEK G W, 1994, Structure and cohesive energy density of fats and their sorption by polymer films. J Food Sci, 59, 1325–27. ARORA D K, HANSEN A P and ARMAGOST M S, 1991, Sorption of flavor compounds by low density polyethylene film. J Food Sci, 56, 1421–23. CHANDLER B V and JOHNSON R L, 1979, New sorbent gel forms of cellulose esters for debittering citrus juices. J Sci Food Agric, 30, 825–32. CHANDLER B V, KEFFORD J F and ZIEMELIS G, 1968, Removal of limonin from bitter orange juice. J Sci Food Agric, 19, 83–86. CHARARA Z N, WILLIAMS J W, SCHMIDT R H and MARSHALL M R, 1992, Orange flavor absorption into various polymeric packaging materials, J Food Sci, 57, 963–66, 972. CHARLES M, BERNAL B and GUICHARD E, 1996, Interactions of -lactoglobulin with flavour compounds, in Flavour Science, ed. by Taylor A J and Mottram D S. The Royal Society of Chemistry, Cambridge, pp. 433–36. DAMODARAN S, 1996, Amino acids, peptides, and proteins, in Food Chemistry, ed. by Fennema O R. Marcel Dekker, Inc, New York, pp. 321–429. DEKKER M, KRAMER, M, VAN BEEST, M. and LUNING P 2002, Modelling oxidative quality changes in several packaging concepts, in Proceedings Worldpak 2002, Michigan, USA, CRC Press, Boca Raton, Volume 1, 297–303. DEKKER M, VAN WILLIGE R W G, LINSSEN J P H and VORAGEN A G J 2003, Modelling the effect of oil/fat content in food systems on flavour absorption by LLDPE, Food Add Contam, 20, 180–85. DELASSUS P T, 1997, Barrier polymers, in The Wiley Encyclopedia of packaging technology, ed. by Brody A L and Marsh K S. John Wiley & Sons, Inc., New York, pp. 71–77. DELASSUS P T, TOU J C, BABINEE M A, RULF D C, KARP B K and HOWELL B A, 1988, Transport of apple aromas in polymer films, in Food and Packaging Interactions, ed. by Hotchkiss J H. ACS Symposium Series 365, American Chemical Society, Washington, DC, pp. 11–27. DE ROOS K B, 1997, How lipids influence food flavor. Food Technol, 51, 60–62. DUFOUR E and HAERTLE ′ T, 1990, Binding affinities of -ionone and related flavor compounds to -lactoglobulin: Effects of chemical modifications, J Agric Food Chem, 38, 1691–95. DU ¨ RR P, SCHOBINGER U and WALDVOGEL R, 1981, Aroma quality of orange juice after filling and storage in soft packages and glass bottles. Alimenta, 20, 91–3. 166 Novel food packaging techniques FUKAMACHI M, MATSUI T, HWANG Y H, SHIMODA M and OSAJIMA Y, 1996, Sorption behavior of flavor compounds into packaging films from ethanol solution. J Agric Food Chem, 44, 2810–13. GIACIN J R and HERNANDEZ R J, 1997, Permeability of aromas and solvents in polymeric packaging materials, in The Wiley Encyclopedia of packaging technology, ed. by Brody A L and Marsh K S. John Wiley & Sons, Inc., New York, pp. 724–33. GODSHALL M A, 1997, How carbohydrates influence food flavor. Food Technol, 51, 63–67. GREMLI H, 1996, Flavor changes in plastic containers: a literature review. Perfumer & Flavorist, 21, 1–8. HALEK W H, 1988, Relationship between polymer structure and performance in food packaging applications, in Food and Packaging Interactions, ed. by Hotchkiss J H. ACS Symposium Series 365, American Chemical Society, Washington, DC, pp. 195–202 . HERNANDEZ R J and GAVARA R, 1999, Plastics packaging – methods for studying mass transfer interactions. Pira International, Leatherhead, UK, pp. 53. HERNANDEZ-MUN ? OZ P, CATALA ′ R and GAVARA R, 1999, Effects of sorbed oil on food aroma loss through packaging materials. J Agric Food Chem, 47, 4370–74. HERNANDEZ-MUN ? OZ P, CATALA ′ R and GAVARA R, 2001, Food aroma partition between packaging materials and fatty food simulants. Food Addit Contam, 18, 673–82. HIROSE K, HARTE B R, GIACIN J R, MILTZ J and STINE C, 1988, Sorption of d- limonene by sealant films and effect on mechanical properties, in Food and Packaging Interactions, ed. by Hotchkiss J H. ACS Symposium Series 365, American Chemical Society, Washington, DC, pp. 28–41. HOTCHKISS J H, 1997, Food-packaging interactions influencing quality and safety, Food Addit Contam, 14, 601–7. INGHAM K E, TAYLOR A J, CHEVANCE F F V and FARMER L J, 1996, Effect of fat content on volatile release from foods, in Flavour Science, Recent Developments, ed. by Taylor A J and Mottram D S. The Royal Society of Chemistry, Cambridge, pp 386–91. JOHANSSON F, 1993, Polymer packages for food – materials, concepts and interactions, a literature review. SIK – The Swedish Institute for Food and Biotechnology, Go¨teborg, pp 118. JOHANSSON F and LEUFVE ′ N A, 1994, Influence of sorbed vegetable oil and relative humidity on the oxygen transmission rate through various polymer packaging films. Packag Technol Sci, 7, 275–81. JOHANSSON F and LEUFVE ′ N A, 1997, Concentration and interactive effects on the sorption of aroma liquids and vapors into polypropylene. J Food Sci, 62, 355–58. KINSELLA J E, 1989, Flavor perception and binding to food components, in Flavour chemistry of lipid foods, ed. by Min D B and Smouse T H. American Oil Chemists’ Society, Champaign, pp. 376–403. Packaging-flavour interactions 167 KRIZAN T D, COBURN J C and BLATZ P S, 1990, Structure of amorphous polyamides: effect on oxygen permeation properties, in Barrier polymers and structures, ed. by Koros W J. ACS Symposium Series 423, American Chemical Society, Washington, DC, pp 111–25. KWAPONG O Y and HOTCHKISS J H, 1987, Comparative sorption of aroma compounds by polyethylene and ionomer food-contact plastics. J Food Sci, 52, 761–63, 785. LANDOIS-GARZA J and HOTCHKISS J H, 1987, Aroma sorption. Food Eng, 4, 39, 42. LANDOIS-GARZA J and HOTCHKISS J H, 1988, Permeation of high-barrier films by ethyl esters, in Food and Packaging Interactions, ed. by Hotchkiss J H. ACS Symposium Series 365, American Chemical Society, Washington, DC, pp 42–58. LE THANH M, THIBEAUDEAU P, THIBAUT M A and VOILLEY A, 1992, Interaction between volatile and non-volatile compounds in the presence of water. Food Chem, 43, 129–35. LEBOSSE ′ R, DUCRUET V and FEIGENBAUM A, 1997, Interactions between reactive aroma compounds from model citrus juice with polypropylene packaging film. J Agric Food Chem, 45, 2836–42. LETINSKI J and HALEK G W, 1992, Interactions of citrus flavor compounds with polypropylene films of varying crystallinities. J Food Sci, 57, 481–4. LI S and PAIK J S, 1996, Flavor sorption estimation by UNIFAC group contribution model. Transactions of the ASAE, 39, 1013–17. LINSSEN J P H, VERHEUL A, ROOZEN J P and POSTHUMUS M A, 1991a, Absorption of flavour compounds by packaging material: Drink yoghurts in Polyethylene bottles. Int Dairy Journal 1, 33–40. LINSSEN J P H, REITSMA J C E and ROOZEN J P 1991b, Influence of pulp particles on limonene absorption into plastic packaging material, in Food packaging: a burden or an achievement? Vol. 2. Symposium Rheims France 5–7 June, pp. 24.18–.20. MANNHEIM C H, MILTZ J and LETZTER A, 1987, Interaction between polyethylene laminated cartons and aseptically packed citrus juices. J Food Sci, 52, 737–40. MARIN A B, ACREE T E, HOTCHKISS J H and NAGY S, 1992, Gas chromatography- olfactometry of orange juice to assess the effects of plastic polymers on aroma character. J Agric Food Chem, 40, 650—54. MARSHALL M R, ADAMS J P and WILLIAMS J W, 1985, Flavor absorption by aseptic packaging materials. Aseptipak 85: Proceedings of the 3rd international conference and exhibition on aseptic packaging (Princeton, NJ: Schotland Business Research, Inc.), 299–312. MILLER K S and KROCHTA J M, 1997, Oxygen and aroma barrier properties of edible films: A review. Trends in Food Sci & Technol, 8, 228–37. MOHNEY S M, HERNANDEZ R J, GIACIN J R, HARTE B R and MILTZ J, 1988, Permeability and solubility of d-limonene vapor in cereal package liners. J Food Sci, 53, 253–57. 168 Novel food packaging techniques MOSHONAS, M. G. and SHAW, P. E., 1989a, Changes in composition of volatile components in aseptically packaged orange juice during storage. J Agric Food Chem, 37, 157–61. MOSHONAS, M G and SHAW P E, 1989b, Flavor evaluation and volatile flavor constituents of stored aseptically packaged orange juice. J Food Sci, 54, 82–5. NAWAR W W, 1996, Lipids, in: Food Chemistry, ed. O R Fennema. M. Dekker, New York. NIELSEN T J and JA ¨ GERSTAD I M, 1994, Flavour scalping by food packaging. Trends in Food Sci & Technol, 5, 353–6. NIELSEN T J, JA ¨ GERSTAD I M and O ¨ STE R E, 1992, Study of factors affecting the absorption of aroma compounds into low-density polyethylene. J Sci Food Agric, 60, 377–81. OLAFSSON G and HILDINGSSON I, 1995, Sorption of fatty acids into ldpe and its effect on adhesion with aluminium foil in laminated packaging material. J Agric Food Chem, 43, 306–12. OLAFSSON G, HILDINGSSON I and BERGENSTAHL B, 1995, Transport of oleic and acetic acids from emulsions into low density polyethylene; Effects on adhesion with aluminum foil in laminated packaging. J Food Sci, 60, 420– 5. PAIK J S, 1992, Comparison of sorption in orange flavor components by packaging films using the headspace technique. J Agric Food Chem, 40, 1822–5. PAIK J S and TIGANI M A, 1993, Application of regular solution theory in predicting equilibrium sorption of flavor compounds by packaging polymers. J Agric Food Chem, 41, 806–8. PAIK J S and WRITER M S, 1995, Prediction of flavor sorption using the Flory- Huggins Equation. J Agric Food Chem, 43, 175–8. PIEPER G, BORGUDD L, ACKERMANN P and FELLERS P, 1992, Absorption of aroma volatiles of orange juice into laminated carton packages did not affect sensory quality. J Food Sci, 57, 1408–11. QUEZADA GALLO J A, DEBEAUFORT F and VOILLEY A, 1999, Interactions between aroma and edible films. 1. Permeability of methylcellulose and low- density polyethylene films to methyl ketones. J Agric Food Chem, 47, 108–13. RIQUET A M, WOLFF N, LAOUBI S, VERGNAUD J M and FEIGENBAUM A, 1998, Food and packaging interactions: determination of the kinetic parameters of olive oil diffusion in polypropylene using concentration profiles. Food Addit Contam, 15, 690–700. ROBERTS D D, ELMORE J S, LANGLEY K R and BAKKER J, 1996, Effects of sucrose, guar gum, and carboxymethylcellulose on the release of volatile flavor compounds under dynamic conditions. J Agric Food Chem, 44, 1321–26. ROONEY M L, 1995, Active packaging in polymer films. in Active Food Packaging, ed. by Rooney M L pp. 74–110. London, Blackie Academic and Professional. Packaging-flavour interactions 169 SADLER G D and BRADDOCK R J, 1990, Oxygen permeability of low density polyethylene as a function of limonene absorption: An approach to modeling flavor scalping. J Food Sci, 55, 587–8. SADLER G D and BRADDOCK R J, 1991, Absorption of citrus flavor volatiles by low density polyethylene. J Food Sci, 56, 35–8. SADLER G, PARISH M, DAVIS J, and VAN CLIEF D, 1995, Flavor-Package Interaction. Fruit Flavors, ed. by R. L. Rouseff and M. M. Leahy (Washington DC: American Chemical Society), pp. 202–10. SALAME M, 1989, The use of barrier polymers in food and beverage packaging, in Plastic film technology, volume one – high barrier plastic films for packaging, ed. by Finlayson K M. Technomic Publishing Company, Inc., Lancaster, Pennsylvania, U.S.A., pp 132–45. SHARMA G K, MADHURA C V, and ARYA S S, 1990, Interaction of plastic films with foods. I. Effect of polypropylene and polyethylene films on fruit squash quality. J Food Sci Technol, 27, 127–32. SHIMODA M, MATSUI T and OSAJIMA Y, 1987, Effects of the number of carbon atoms of flavor compounds on diffusion, permeation and sorption with polyethylene films. J Jpn Soc Nutr Food Sci, 34, 535–9. SINGH R P and HELDMAN D R, 1993, Introduction to food engineering. Academic Press, Inc., New York, pp 499. SIZER C E, WAUGH P L, EDSTAM S, and ACKERMANN P, 1988, Maintaining flavor and nutrient quality of aseptic orange juice. Food Technol, June, 152–9. SOARES N N F and HOTCHKISS J H, 1998a, Bitterness reduction in grape fruit juices through active packaging. Pack. Technol. Sci, 11, 9–18. SOARES N N F and HOTCHISS J H, 1998b, Naringnase immobilisation in packaging films for reducing naringin concentration in grape fruit juice. J Food Sci, 63, 61–5. STERN S A and TROHALAKI S, 1990, Fundamentals of gas diffusion in rubbery and glassy polymers, in Barrier polymers and structures, ed. by Koros W J. ACS Symposium Series 423, American Chemical Society, Washington, DC, pp. 22–59. TAWFIK M S, DEVLIEGHERE F and HUYGHEBAERT A, 1998, Influence of D- limonene absorption on the physical properties of refillable PET. Food Chem, 61, 157–62. TIGANI M A and PAIK J S, 1993, Use of dielectric constants in the prediction of flavour scalping. Packag Technol Sci, 6, 203–9. VAN WILLIGE R W G, LINSSEN J P H and VORAGEN A G J, 2000a, Influence of food matrix on absorption of flavour compounds by linear low-density polyethylene: proteins and carbohydrates. J Sci Food Agric, 80, 1779–89. VAN WILLIGE R W G, LINSSEN J P H and VORAGEN A G J, 2000b, Influence of food matrix on absorption of flavour compounds by linear low-density polyethylene: oil and real food products. J Sci Food Agric, 80, 1790–7. VAN WILLIGE R W G, SCHOOLMEESTER D N, VAN OOIJ A N, LINSSEN J P H and VORAGEN A G J, 2002a, Influence of storage time and temperature on absorption of flavour compounds from solutions by plastic packaging 170 Novel food packaging techniques materials. J Food Sci, 67, 2023–31. VAN WILLIGE, R W G, LINSSEN J P H, MEINDERS M B J, VAN DER STEGE H J and VORAGEN A G J, 2002b, Influence of flavour absorption on oxygen permeation through LDPE, PP, PC and PET plastics packaging materials. Food Addit Contam, 19, 303–13. VAN WILLIGE R W G, 2002c, Effects of flavour absorption of foods and their packaging materials, PhD thesis, Wageningen University, The Netherlands. VAN WILLIGE R W G, LINSSEN J P H, LEGGER-HUYSMAN A and VORAGEN A G J, 2003, Influence of flavour absorption by food packaging materials (low-density polyethylene, polycarbonate and polyethylene terephthalate) on taste perception of a model solution and orange juice, Food Add Contam, 20, 84–91. VERMEIREN L, DEVLIEGHERE F, VAN BEEST M, DE KRUIJF N and DEBEVERE J, 1999, Developments in the active packaging of foods. Trends in Food Sci & Technol., 10, 77–86. WESSELINGH J A and KRISHNA R, 2000, Mass transfer in multicomponent mixtures. Delft. University Press, Delft, NL, pp. 329. WIDDER S and FISCHER N, 1996, Measurement of the influence of food ingredients on flavour release by headspace gas chromatography- olfactometry, in Flavour Science, ed. Taylor A J and Mottram D S. The Royal Society of Chemistry, Cambridge, pp 405–12. YAMADA K, MITA K, YOSHIDA K and ISHITANI T, 1992, A study of the absorption of fruit juice volatiles by the sealant layer in flexible packaging containers (The effect of package on quality of fruit juice, part IV). Packag Technol Sci, 5, 41–7. ZHANG Z, BRITT I J and TUNG M A, 1999, Water absorption in EVOH films and its influence on glass transition temperature. J Polymer Sci: Part B: Polymer Physics, 37, 691–9. Packaging-flavour interactions 171