Migration modeling to estimate exposure to chemicals in food packaging for application in highthroughput risk-based screening and Life Cycle Assessment

Specialty software and simplified models are often used to estimate "worst-case" migration of potentially toxic chemicals from packaging into food. Current approaches, however, cannot efficiently and accurately provide estimates of migration for emerging applications, e.g. in Life Cycle Assessment and risk prioritization and screening. To fulfill the need for a migration model flexibly suitable for such tools, we develop an accurate and rapid (high-throughput) approach. The developed model estimates the fraction of an organic chemical migrating from polymeric packaging into food for user-defined scenarios and requires limited parameters (i.e. physicochemical properties). Several hundred step-wise simulations optimized the coefficients of the model to cover a wide-range of scenarios (e.g. packaging thickness, food etc.). The developed model, implemented in a disseminatable spreadsheet, nearly instantaneously estimates migration from packaging into food for user-defined scenarios, and has improved performance over common model simplifications. The common practice of setting the package-food partition coefficient = 1 for specific "worst-case" scenarios is insufficient to predict the equilibrium concentration in food for diverse scenarios. Therefore a partition coefficient model, as a function of a chemical’s octanol-water partition coefficient and a food’s ethanol-equivalency, was also developed. When using measured diffusion coefficients the model accurately predicted (R2 = 0.9, SE = 0.5) hundreds of empirical datapoints for various scenarios. Diffusion coefficient modeling, which determines the speed of chemical transfer from package to food, was found as a major contributor to uncertainty and decreased model performance (R2 = 0.5, SE = 1). In all, this study provides a migration modeling approach that rapidly estimates the fraction migrated for emerging screening and prioritization approaches. To estimate exposure, chemical concentrations in packaging are essential.