Modelling aluminium leaching into food from different foodware materials with multi-level factorial design of experiments.

To estimate the contribution of aluminium (Al) leaching from different materials used for food preparation and serving to the dietary Al intake, Al release from foodware typically used in everyday life was investigated using multilevel factorial design (MFD) of experiments. For Al characterisation, sample preparation and an analytical method using inductively coupled plasma atomic emission spectroscopy was developed and validated. Parameter influence (temperature: x₁, contact time: x₂, pH: x₃, salt concentration: x₄, viscosity: x₅), was evaluated with analysis of variance suggesting that the influence of viscosity is not significant compared to the other four studied parameters. Therefore, predictive, exponential quadratic regression models were established with x₁-x₄. Cross-validation and a set of independent experiments in real food products were used to test the prediction force of the different models. They both suggest that the quality of the models established for Al foil, Al plate and ceramic ware is satisfactory, but less good for glassware and stainless steel. Indeed, in the studied conditions, leaching from these latter food wares was often close to or even below the limit of quantification suggesting that the principal sources of Al intake from food contact materials during food processing are utensils made of Al and ceramic ware.