Two recent articles, both published in the peer-reviewed journal Food Additives & Contaminants: Part A, report on the determination of key parameters for modeling of contaminant migration from low density polyethylene (LDPE) into food simulants or foods.

In an article published on February 19, 2016, a group of scientists led by Joaquium Maia and Perfecto Paseiro from the Department of Analytical Chemistry, Nutrition and Bromatology at the University of Santiago de Compostela, illustrate the data collection process for the FACET project (Flavourings, Additives and food Contact materials Exposure  Tool) (FPF reported). This is done on the example of benzophenone (CAS 119-61-9) migration from immersion-contaminated LDPE film. Migration into solid (cooked ham and turkey ham), semi-solid (Gouda cheese, chocolate spread, and pâté spread), and liquid (tomato sauce, orange juice, and red wine) foods was assessed at three different temperatures, 20°C, 40°C, and 60°C, for storage up to 15 days. The amount of benzophenone migrating into food was calculated as the difference in its initial and final concentration in the LDPE film for each time period. From the collected data, the scientists determined diffusion coefficients (DP, within the polymer, and DF, within the food), and partition coefficient between polymer and food (KP/F).

Diffusion coefficients increased with increasing storage temperature; the partition coefficient was not affected significantly. DFs for solid and semi-solid foods were similar; in the liquid foods, 70-700 times increase was observed, indicating a much faster diffusion rate. Nearly 100% of benzophenone migrated into high-fat foods, while 70-80% migrated into foods with lower fat content. Interestingly, the experimentally measured KP/F in several cases deviated from those predicted in the applicable worst-case scenario. Namely, for the high-fat foods, it was up to an order of magnitude lower than 1, which is a worst-case value suggested for migrants that are “very soluble in the food”. For the aqueous foods, maximum KP/F value observed was still around 20 times lower than 1000, which is a worst-case value suggested for migrants that are “sparingly soluble in the food”. Since in both cases these deviations would lead to an underestimation of the real amounts of chemical migrating into the foods, the authors suggest that a revision of worst-case scenario KP/F values may be needed.

While the above-described article looked at benzophenone migration from LDPE into real foods, another study evaluated its migration into food simulants (20% ethanol, 50% ethanol, and olive oil) at the same three temperatures. This article by Rafael Paseiro-Cerrato, formerly from the Department of Analytical Chemistry, Nutrition and Bromatology at the University of Santiago de Compostela, and now at the U. S. Food and Drug Administration, and colleagues from the Fraunhofer Institute for Process Engineering and Packaging (IVV), Freising, Germany, was published on March 21, 2016. Apart from benzophenone, partition coefficients KP/F were determined for six other model migrants, covering a wide range of polarity: Styrene (CAS 100-42-5), 1,2,3-Propanetricarboxylic acid, 2-(acetyloxy)-, tributyl ester (ATBC, CAS 77-90-7), 1-octene (CAS 111-66-0), trans,trans-1,4-Diphenol-1,3-butadiene (DPBD, CAS 538-81-8), Bis(2-ethylhexyl) adipate (DEHA, CAS 103-23-1), and 2,2’-(2,5-Thienediyl)bis[5-(2-methyl-2-propanyl)-1,3-benzoxazole] (Uvitex OB, CAS 7128-64-5).

For the selected model compounds and food simulants, a correlation was observed between logKP/F and logPO/W (a measure of partition between two immiscible phases, octanol and water). The authors demonstrate that the derived regression lines can be used to select appropriate simulants for a particular food type, and to estimate worst case migration using the logPO/W value of a potential migrant. Comparison of partition coefficient values obtained for benzophenone in food simulants and real foods demonstrates that the most reliable prediction for migration into high-fat foods is provided by olive oil, for both alcoholic and non-alcoholic liquid foods by 20% ethanol, and for solid foods with relatively low fat content, like ham, by 50% ethanol. Migration into the dry food simulant Tenax was not assessed.

Read more

JRC (January 5, 2017). “Improving the prediction of exposure to chemicals from food packaging.


Maia, J., et al. (2016). “The determination of key diffusion and partition parameters and their use in migration modelling of benzophenone from low density polyethylene (LDPE) into different foodstuffs.Food Additives & Contaminants: Part A (published online February 19, 2016)

Paseiro-Cerrato, R., et al. (2016). “A study of the partition coefficients K p/f of seven model migrants from LDPE polymer in contact with food simulants.Food Additives & contaminants: Part A (published online March 21, 2016)