Patrice Dole

Properties of thermoplastic blends: starch–polycaprolactone

Different compositions of wheat thermoplastic starch (TPS) and polycaprolactone (PCL) are melt blended by extrusion and injected. Different properties are determined: mechanical properties (tensile and impact tests), thermal and thermomechanical properties (DSC and DMTA) and hydrophobicity (contact angle measurement). A large range of blends is analysed with different glycerol (plasticizer):starch contents ratios (0.14:0.54) and various PCL concentrations (up to 40 wt.%). From the behaviour of each polymeric system, it is possible to analyse the relationship properties of each component proportion on the blends. The ageing of the system is studied and shows a structural evolution of the material after injection during several weeks. We have noticed a fairly low compatibility between both polymeric systems. Finally, the addition of PCL to TPS matrix allows to overcome the weakness of pure TPS: low resilience, high moisture sensitivity and high shrinkage, even at low PCL concentration, e.g. 10 wt.%.

Antioxidant properties of lignin in polypropylene

Abstract This work deals with lignin antioxidant properties in polypropylene. We extracted and characterised 15 lignin samples from various botanical origins (wheat straw and wheat bran) using different processes of extraction and using industrial Kraft lignins fractions (Westaco). These lignins were incorporated in polypropylene (1% weight). We underline the effect of lignin solubility in PP on stabilisation behaviour. The solubility in polypropylene is very low; as a consequence, classically important factors such as phenol content and intrinsic reactivity of the molecule are less important than factors making easier solubility. Solubility in PP could not be measured directly; a morphological criterion (average dimension of lignin aggregates in the blend) has been correlated to antioxidant activity and is thought indirectly to reflect solubility behaviour. The following factors tend to improve the compatibility (blend morphology) and consequently the antioxidant activity: • low molecular weight • low total OH content (aliphatic+phenolic) The non-linear relation between Kraft lignin activity and its concentration in the blend has been explained by molecular weight polydispersity, and correlated to activities of Kraft fractions.

Typical diffusion behaviour in packaging polymers - application to functional barriers

When plastics are collected for recycling, possibly contaminated articles might be recycled into food packaging, and thus the contaminants might subsequently migrate into the food. Multilayer functional barriers may be used to delay and to reduce such migration. The contribution of the work reported here is to establish reference values (at 40°C) of diffusion coefficients and of activation energies to predict the functional barrier efficiency of a broad range of polymers (polyolefins, polystyrene, polyamide, PVC, PET, PVDC, [ethylene vinyl alcohol copolymer], polyacrylonitrile and [ethylene vinyl acetate copolymer]). Diffusion coefficients (D) and activation energies (Ea) were measured and were compiled together with literature data. This allowed identification of new trends for the log D = f(molecular weight) relationships. The slopes were a function of the barrier efficiency of the polymer and temperature. The apparent activation energy of diffusion displayed two domains of variation with molecular weight (M). For low M (gases), there was little variation of Ea. Focusing on larger molecules, high barrier polymers displayed a larger dependence of Ea with M. The apparent activation energy decreased with T. These results suggest a discontinuity between rubbery and glassy polymers.

Functional barriers: Properties and evaluation

Functional barriers are multilayer structures deemed to prevent migration of some chemicals released by food-contact materials into food. In the area of plastics packaging, different migration behaviours of mono- and multilayer structures are assessed in terms of lag time and of their influence of the solubility of the migrants in food simulants. Whereas barriers to oxygen or to aromas must prevent the diffusion of these compounds under conditions of use, a functional barrier must also be efficient under processing conditions, to prevent diffusion of substances when the polymer layers are in contact at high (processing) temperatures. Diffusion in melted polymers at high temperatures is much slower for glassy polymers, than in polymers that are rubbery at ambient temperature. To evaluate the behaviour of functional barriers under conditions of use, a set of reference diffusion coefficients in the 40-60°C range were determined for 14 polymers. Conditions for accelerated migration tests are proposed based on worst-case activation energy in the 40-60°C range. For simulation of migration, numerical models are available. The rules derived from the models can be used both by industry (to optimize a material in terms of migration) or by risk assessors. Differences in migration behaviour between mono- and multilayer materials are discussed.

Influence of photocrosslinking on the retrogradation of wheat starch based films

The ageing of wheat starch based materials was followed as a function of time for different systems suitably hydrated to be in the rubbery state. In particular, one aspect of retrogradation, that is to say the development of crystallinity is analysed in order to study the influence of a specific crosslinking treatment of starch films made by UV irradiation (photocrosslinking). Mechanical and thermal properties showed that an important limitation on this particular ageing occurs after photocrosslinking at room temperature. A process of accelerated ageing, monitored by modulated DSC, was also performed at 60 °C in order to increase retrogradation kinetics. The results showed a recrystallisation decrease and it appeared that the presence of the plasticiser enhances retrogradation. For a plasticised sample (17% glycerol), a decrease of 30% of the melting enthalpy was observed compared to 91% for an unplasticised one. A modification in the nature of the melting peaks, as recorded by modulated DSC, allowed us to propose a model for amylopectin retrogradation after crosslinking.

Prediction of worst case migration: presentation of a rigorous methodology.

An improvement of the Piringer model, allowing the prediction of a worst case migration from packaging to food is presented here. The authors are proposing other constants for the calculation of the upperbound value of the diffusion coefficient, using experimental data determined by a film to film method. Considering the plasticizing effects of food simulants, a model involving the variation of the diffusion coefficient versus space and time must be used. Future fields of investigation are discussed: the relationship between diffusion coefficients and the volume of the migrant (instead of molar mass), and the variation of diffusion coefficient activation energy with temperature.

Migration of additives from polymers into food simulants: numerical solution of a mathematical model taking into account food and polymer interactions

The principle of a computing program describing precisely the migration of additives from a polymer into a food simulant is presented. As six parameters are used to fit the simulant sorption and additive extraction kinetics, the parameters have been determined by independent experiments. Owing to the complicated coupling between the liquid and additive diffusion processes, migration kinetics cannot be obtained by a mathematical resolution of kinetic equations, but they must be calculated by numerical analysis. The method is applied to a UV absorber in polypropylene migrating into glyceryl tripelargonate, a pure triglyceride, of which behaviour and average molecular weight are similar to official fatty food simulants. Properly designed experiments validate the model used to fit the migration kinetics. The possibility of erasing any parameters is also discussed.

Integrated approach of migration prediction using numerical modelling associated to experimental determination of key parameters.

The principle of utilizing a computing program describing precisely the migration of additives from a polymer into a food simulant is presented. The model has been validated with a UV absorber in polypropylene migrating into glyceryl tripelargonate, a pure triglyceride whose behaviour and average molecular weight are similar to Myglyol (a synthetic mixture of C8-C12 triglycerides). Six parameters were used to fit the simulant sorption and additive extraction kinetics, and these were determined by independent experiments. The possibility of eliminating any of the parameters is also discussed. This work provides the first consistent set of experimental data that can be used to overestimate the diffusion coefficients of additives both in virgin (without contact with solvent) and in swollen (fat contact) polymer. The influence of mobility increase brought out by temperature or swelling are compared. The effects were more important for high molecular weight compounds.

Modelling of migration from multi-layers and functional barriers: Estimation of parameters

Functional barriers form parts of multi-layer packaging materials, which are deemed to protect the food from migration of a broad range of contaminants, e.g. those associated with reused packaging. Often, neither the presence nor the identity of the contaminants is known, so that safety assessment of the materials has to rely on predictive tools. Several complementary freeware described here allow one to model diffusion in multi-layer films. These tools require the input of parameters that are not easy to determine or predict. Previous work has focused on the prediction of diffusion coefficients at storage temperatures of packaging in contact with food. However, many other kinetic and thermodynamic parameters are needed to describe transport properties during the processing of a material at high temperature and during its shelf-life. All parameters needed for the calculations are discussed. In order to propose default values, the approach consists of (1) reviewing the available literature data, (2) running experiments on polypropylene, polyethylene and poly(ethylene vinyl alcohol) in typical conditions (separately diffusion during processing and migration) and (3) simulating numerical sets for typical situations. Several freeware are proposed to simulate migration from multi-layers and functional barriers using the default parameters.

Acidolysis of a lignin model: Investigation of heterogeneous catalysis using Montmorillonite clay

The use of heterogeneous conditions involving Montmorillonite K10 clay was investigated as a mild alternative to homogeneous acidolysis of a lignin model. Guaiacyl Dehydrogenation Oligomers (DHOs) synthesized by horseradish peroxidase were selected as starting material. Hydrolysis products were analyzed by gel permeation chromatography and by HPLC/mass spectrometry. Several experimental parameters were studied such as catalyst and substrate concentration, as well as reaction solvent composition in order to minimize high molar mass product formation generated by recombination mechanisms. In both catalytic modes, the best hydrolysis conditions were similar in terms of solvent composition with dioxane/water (90/10, v/v) and catalyst H(+) concentration of about 0.01 mol/L. Although the homogeneous catalysis generated only 28% of low molecular weight (LMW) products (monomers and dimers), clay catalysis generated 35%. In light of the qualitative analysis, both catalytic methods gave the same products, thus supporting similar hydrolysis mechanisms.