A.G.J. Voragen

Influence of flavour absorption on oxygen permeation through LDPE, PP, PC and PET plastics food packaging

The effect of flavour absorption on the oxygen permeability of low-density polyethylene (LDPE), polypropylene (PP), polycarbonate (PC) and polyethylene terephthalate (PET) was studied using an isostatic continuous flow system. Polymer samples were exposed to a model solution containing limonene, hexyl acetate, nonanone and decanal at 40°C. After exposure, one part of each sample was analysed for absorbed flavour compounds using a Large Volume Injection GC Ultrasonic ‘in vial’ extraction method, and from the other part, oxygen permeability was measured in a permeation cell at 25°C. After 8h of exposure, LDPE and PP samples showed a significant linear (R2 = 0.82 and 0.99) increase in oxygen permeability of 21 and 130%, respectively. Owing to swelling of the polymer samples resulting from flavour absorption, the structure of the polymeric network changed (i.e. opened) and consequently increased oxygen permeability. The oxygen permeability of exposed PC showed a significant linear (R2 = 0.78) decrease of 11% after 21 days. PC obviously did not swell like LDPE or PP. Therefore, it was suggested that absorbed flavour compounds occupied or blocked ‘microcavities’ through which normally oxygen is transported. Absorption of flavour compounds by PET did not affect the oxygen permeability of PET significantly.

Introducing porous graphitized carbon liquid chromatography with evaporative light scattering and mass spectrometry detection into cell wall oligosaccharide analysis.

Separation and characterization of complex mixtures of oligosaccharides is quite difficult and, depending on elution conditions, structural information is often lost. Therefore, the use of a porous-graphitized-carbon (PGC)-HPLC-ELSD-MS(n)-method as analytical tool for the analysis of oligosaccharides derived from plant cell wall polysaccharides has been investigated. It is demonstrated that PGC-HPLC can be widely used for neutral and acidic oligosaccharides derived from cell wall polysaccharides. Furthermore, it is a non-modifying technique that enables the characterization of cell wall oligosaccharides carrying, e.g. acetyl groups and methylesters. Neutral oligosaccharides are separated based on their size as well as on their type of linkage and resulting 3D-structure. Series of the planar beta-(1,4)-xylo- and beta-(1,4)-gluco-oligosaccharides are retained much more by the PGC material than the series of beta-(1,4)-galacto-, beta-(1,4)-manno- and alpha-(1,4)-gluco-oligosaccharides. Charged oligomers such as alpha-(1,4)-galacturonic acid oligosaccharides are strongly retained and are eluted only after addition of trifluoroacetic acid depending on their net charge. Online-MS-coupling using a 1:1 splitter enables quantitative detection of ELSD as well as simple identification of many oligosaccharides, even when separation of oligosaccharides within a complex mixture is not complete. Consequently, PGC-HPLC-separation in combination with MS-detection gives a powerful tool to identify a wide range of neutral and acidic oligosaccharides derived from various cell wall polysaccharides.

Influence of food matrix on absorption of flavour compounds by linear low-density polyethylene: proteins and carbohydrates

The influence of oil and food components in real food products on the absorption of four flavour compounds (limonene, decanal, linalool and ethyl 2-methyl butyrate) into linear low-density polyethylene (LLDPE) was studied using a large volume injection GC in vial extraction method. Model food systems and real food products investigated included oil/water emulsions, oil/casein models, oil/pectin models, skim milk and whole milk. A small amount of oil (50 g l-1) had a major influence on the amount of flavour absorption. Because of solubilization of the more apolar flavour compounds limonene, decanal and linalool into the oily phase, only the remaining flavour compounds in the aqueous phase were available for absorption by LLDPE. After 14 days of exposure, absorption of limonene and decanal decreased by 97°and that of linalool by 86ÐDue to a salting out effect, absorption of the less apolar ethyl 2-methylbutyrate (E2MB) first increased with increasing oil concentration, but decreased at higher oil concentrations (>2.5 g l-1). Oil/casein and oil/pectin models showed that the more apolar flavour compounds were mainly dissolved in the oily phase and that the compounds present in the aqueous phase could interact with casein or pectin. Oil influenced the level of flavour absorption by LLDPE to a much greater extent than pectin or casein. However, the low amount of fat (1.11 g l-1) in skim milk had no influence on the absorption of flavour compounds. Only the proteins in skim milk (especially casein) decreased the absorption of limonene and decanal, because the fat was probably entrapped. Whole milk, which contained a higher concentration of (free) fat, suppressed the absorption of all flavour compounds by LLDPE to the same extent as was found for the oil model solutions. In general, absorption results from skim milk and whole milk were in good agreement with the results of the investigated model solutions containing individual food components.

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

The influence of flavour absorption by low-density polyethylene (LDPE), polycarbonate (PC) and polyethylene terephthalate (PET) on taste perception of a model solution containing seven flavour compounds and orange juice in glass bottles was studied with and without pieces of the respective plastic films after dark storage at 20°C. Owing to absorption, the amount of flavour compounds in the model solution exposed to LDPE decreased substantially. From the model flavour solution valencene was almost completely absorbed by LDPE, followed to a lesser extent by decanal, hexyl acetate, octanal and nonanone. Less flavour compounds were absorbed from the model solution by PC and PET. In contrast to LDPE, valencene was absorbed in the lowest amounts and decanal in the highest. Limonene was readily absorbed from orange juice 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 the flavour content between controls and polymer-treated samples differed substantially, the loss of flavour compounds due to absorption by LDPE, PC and PET did not influence taste perception of a model solution and orange juice significantly up to 29 days of dark storage at 20°C as determined by triangular taste panel tests.

Influence of food matrix on absorption of flavour compounds by linear low-density polyethylene: oil and real food products

The influence of oil and food components in real food products on the absorption of four flavour compounds (limonene, decanal, linalool and ethyl 2-methyl butyrate) into linear low-density polyethylene (LLDPE) was studied using a large volume injection GC in vial extraction method. Model food systems and real food products investigated included oil/water emulsions, oil/casein models, oil/pectin models, skim milk and whole milk. A small amount of oil (50 g l-1) had a major influence on the amount of flavour absorption. Because of solubilization of the more apolar flavour compounds limonene, decanal and linalool into the oily phase, only the remaining flavour compounds in the aqueous phase were available for absorption by LLDPE. After 14 days of exposure, absorption of limonene and decanal decreased by 97°and that of linalool by 86ÐDue to a salting out effect, absorption of the less apolar ethyl 2-methylbutyrate (E2MB) first increased with increasing oil concentration, but decreased at higher oil concentrations (>2.5 g l-1). Oil/casein and oil/pectin models showed that the more apolar flavour compounds were mainly dissolved in the oily phase and that the compounds present in the aqueous phase could interact with casein or pectin. Oil influenced the level of flavour absorption by LLDPE to a much greater extent than pectin or casein. However, the low amount of fat (1.11 g l-1) in skim milk had no influence on the absorption of flavour compounds. Only the proteins in skim milk (especially casein) decreased the absorption of limonene and decanal, because the fat was probably entrapped. Whole milk, which contained a higher concentration of (free) fat, suppressed the absorption of all flavour compounds by LLDPE to the same extent as was found for the oil model solutions. In general, absorption results from skim milk and whole milk were in good agreement with the results of the investigated model solutions containing individual food components.

Modelling the effect of oil/fat content in food systems on flavour absorption by LLDPE.

One of the phenomena in food-packaging interactions is flavour absorption. Absorption of flavour compounds from food products into food-packaging materials can result in loss of flavour compounds or an unbalance in the flavour profile changing a products quality. The food matrix influences the amounts of absorbed flavour compounds; the presence of oil or fat especially determines the ability to absorb flavour compounds from the food to the package. On the other hand, the polarity of the flavour compound itself is a characteristic that also influences the level of absorption into synthetic polymers. A model based on the effect of the polarity (logP) of flavour compounds and on their partitioning coefficients between the food (matrix) and the packaging material is described. The model can be used for predicting absorption of flavour compounds from foods into LLDPE. However, an attempt to apply the proposed model on real foods shows serious limitations of the model for (very) low fat products. Predictive values deviate from the measured values, probably due to other interaction phenomena, e.g. with proteins. Predictive and measured values from a product with a substantial amount of fat match much better, suggesting that the model is valid for products having a substantial amount of (free) fat.