Chemical Composition and Migration Levels of Packaging Adhesives

The project aims were to identify and quantify the major chemical species migrating from different adhesives used in food packaging, with the aim of assessing possible levels of exposure for consumers. Species from the printed surface of packaging could migrate into the adhesive layer, when stored in direct contact, and then migrate into food after the packaging was formed. However the wide range of substances apparently present in coldseal adhesives allowed no more than an initial quantitative study of these.
<P>

Polyurethane (PU) systems are reactive adhesives, i.e. a complex chemical reaction is required for the adhesive bond to form, and thus the nature and extent of potential migrants is time dependent. Study of PU adhesives was difficult because of the speed with which polyurethane industry modified its products to meet customer requirements. At the start of the project the majority of laminates were prepared using solvent-based adhesives but these formulations were rapidly changed to solvent-free systems in which liquid polymeric diphenyl methane diisocyante (MDI) replaced the conventional solvent. Furthermore within each grade of adhesive minor changes in formulation were made for almost every customer. It was not possible to identify all the major migrants from the PU systems using readily-available techniques. <P>Derivatisation HPLC was developed to enhance the detection capability of the polyether polyols. MALDI-MS techniques were used to determine the molecular weight of the principal migrants. These techniques need further refinement for application in this area. Major migrants (10-100 µg/dm2) were identified as: residual polyether polyols; cyclic reaction products formed during preparation of the polyester polyols; and low levels of additives used in the preparation of the adhesive.
<P>
Polyurethane adhesives are generally produced by the one to one reaction of a diisocyanate with a difunctional polyol. The presence of any residual polyether polyol was therefore, not understood since the initial adhesive contains excess diisocyanate. Since the residual polyol has the same mass spectral data as the original material it cannot be an unsaturated system, that is with no hydroxyl functionality. One possible explanation is that more of the diisocyanate than expected reacted with atmospheric moisture.
<P>
The differences in results for the residual –NCO moieties between the HPLC and the Marcli methods were as anticipated since the use of the HPLC method for the determination of residual –NCO species might be expected to produce lower results as only the monomeric species are detected. Further work would be needed on polyurethane systems to corroborate these findings.