Migration of bisphenol A from can coatings—effects of damage, storage conditions and heating

Bisphenol A (BPA) is an important monomer used in the manufacture of epoxy resins for internal food can linings. Experiments were conducted to investigate the effects of different storage conditions and can damage on the migration of BPA to foods. These experiments were conducted in a systematic fashion by filling empty epoxyphenolic coated cans with four foods: soup, minced beef, evaporated milk and carrots and a food simulant (10% ethanol). Filled cans of each food type or simulant were then sealed and processed using appropriate conditions, before storage at three different temperatures: 5°C, 20°C and 40°C. For each of the storage regimes, 50% of the cans were dented to establish if this would lead to increased BPA migration. Cans were removed from these stocks at intervals of 1, 3 and 9 months storage at 5°C and 20°C or 10 days, 1 and 3 months at 40°C. Some initial problems of heterogeneity between samples was overcome by determining the amount of BPA in food as well as in the can lining. It was found that 80–100% of the total BPA present in the coating had migrated to foods directly after can processing by pilot plant filling with food or simulant, sealing and sterilization. This level was not changed by extended storage (up to 9 months) or can damage, indicating most migration was occurring during the can processing step. There was no noticeable difference, in this respect, between the different foods or the food simulant. Analysis of control samples (foods fortified with ∼0.1 mg kg−1 BPA and contained in Schott bottles) showed that BPA was stable under both processing and storage. Experiments were also conducted to investigate the potential effects, on the migration of BPA from can coatings, of cooking or heating foods in the can prior to consumption. Food cans were purchased and the food either cooked or heated in the can. BPA was analysed prior to and after the heating/cooking process. It was concluded from the results that there were no appreciable differences in the BPA level before and after cooking or heating.