Brandsch Rainer
a MDCTec Systems GmbH , Augsburg , Germany.
Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2017 Oct;34(10):1743-1766. doi: 10.1080/19440049.2017.1339235. Epub 2017 Jun 29.
Migration modelling provides reliable migration estimates from food-contact materials (FCM) to food or food simulants based on mass-transfer parameters like diffusion and partition coefficients related to individual materials. In most cases, mass-transfer parameters are not readily available from the literature and for this reason are estimated with a given uncertainty. Historically, uncertainty was accounted for by introducing upper limit concepts first, turning out to be of limited applicability due to highly overestimated migration results. Probabilistic migration modelling gives the possibility to consider uncertainty of the mass-transfer parameters as well as other model inputs. With respect to a functional barrier, the most important parameters among others are the diffusion properties of the functional barrier and its thickness. A software tool that accepts distribution as inputs and is capable of applying Monte Carlo methods, i.e., random sampling from the input distributions of the relevant parameters (i.e., diffusion coefficient and layer thickness), predicts migration results with related uncertainty and confidence intervals. The capabilities of probabilistic migration modelling are presented in the view of three case studies (1) sensitivity analysis, (2) functional barrier efficiency and (3) validation by experimental testing. Based on the predicted migration by probabilistic migration modelling and related exposure estimates, safety evaluation of new materials in the context of existing or new packaging concepts is possible. Identifying associated migration risk and potential safety concerns in the early stage of packaging development is possible. Furthermore, dedicated material selection exhibiting required functional barrier efficiency under application conditions becomes feasible. Validation of the migration risk assessment by probabilistic migration modelling through a minimum of dedicated experimental testing is strongly recommended.
迁移模型基于与各种材料相关的扩散和分配系数等传质参数，提供从食品接触材料（FCM）到食品或食品模拟物的可靠迁移估计。在大多数情况下，传质参数无法从文献中轻易获取，因此在估计时存在一定的不确定性。从历史上看，不确定性最初是通过引入上限概念来考虑的，但由于迁移结果被高估，其适用性有限。概率迁移模型能够考虑传质参数以及其他模型输入的不确定性。对于功能性屏障，其中最重要的参数包括功能性屏障的扩散特性及其厚度。一种能够接受分布作为输入并应用蒙特卡罗方法（即从相关参数（即扩散系数和层厚度）的输入分布中随机抽样）的软件工具，可以预测具有相关不确定性和置信区间的迁移结果。从三个案例研究（1）敏感性分析、（2）功能性屏障效率和（3）通过实验测试进行验证的角度展示了概率迁移模型的能力。基于概率迁移模型预测的迁移以及相关的暴露估计，可以对现有或新包装概念背景下的新材料进行安全评估。在包装开发的早期阶段识别相关的迁移风险和潜在的安全问题成为可能。此外，在应用条件下选择具有所需功能性屏障效率的专用材料变得可行。强烈建议通过最少的专用实验测试来验证概率迁移模型的迁移风险评估。
