Back Propagation Artificial Neural Network Modeling and Migration Analysis of Siloxane D5 Migration from Selected Food Contact Materials

Abstract

Abstract The detection and quantification of environmental pollution compounds migration in food contact silicone rubber materials remains a prospective issue to be solved in the consideration of toxicology and safety assessment. In this study, an artificial neural network (ANN) model was established to predict migration property of non-target compound decamethylcyclopentasiloxane (D5) molecule in food contact silicone rubber. The average prediction accuracy of the model was 99.8%. The analysis of ANN indicates that high temperature condition accelerates the migration of D5 from silicone rubber into two typical food simulants, namely H2O and acetic acid. The migration of D5 is more apparent when the silicone rubber is in contact with acetic acid. The combination of experiment and simulation analysis of D5 migration indicates that high temperature and acetic acid food simulant environment threaten the safety of food contact silicone rubber. These fundamental studies can provide a comprehensive understanding of the migration of cyclic organosiloxane oligomer from silicone rubber and guidance for the safety evaluation and early warning mechanisms.