The frequent detection of the photoinitiators isopropylthioxanthone (ITX) in baby milk, milk products and cloudy juices packaged in beverage cartons has highlighted the necessity to screen foods packaged in paper and board more regularly. UV-filters have been linked to contact sensitization, endocrine activity and cancer (see FPF article on UV filters). In the present study, researchers from the Technical University Dresden (TU Dresden), Germany and food law enforcement authorities (CVUA) of Baden-Wurttemberg, Germany analyzed 310 packaging materials containing dry foods for the presence of 11 photoinitiators and amine synergists that had been previously found in food (Jung et al. 2013). Ultraviolet light (UV) filters like ITX and benzophenone (BP) are used in UV-cured printing inks and varnishes. Certain UV-filters like BP require amine synergists to improve curing. BP has shown positive responses in uterotrophic assays testing for estrogenic activity and is listed as a high production volume chemical by the OECD. Printing inks are currently not specifically regulated under European Law, though substances used in food contact materials shall generally not endanger human health (Article 3, EC 1935/2004) and some printing inks are regulated under the plastics regulation EC 10/2011. In consequence, most printing inks continue to lack toxicological evaluation and specific migration limits (SMLs) in Europe. Yet, the Swiss Ordinance on printing inks contains a positive list of printing inks which also sets regularly updated SMLs.

In their study, Jung and colleagues detected BP in 49% of packaging samples and found 20 food products to violate legally acceptable limits under either the European or the Swiss regulations. Further, the scientists observed packaging surface to food mass ratios to be up to 6 times higher than the default value of 6 dm2/kg specified in EU legislation. This observation indicates that common SMLs for photoinitiators migrating from carton likely underestimate exposure by a factor of 6.

Methods. To analyze substances present in the packaging material, 0.5 dm2 of the packaging material was cut into small pieces and extracted with acetonitrile at 70° C over 24 h. Food stuff was extracted at 70° C over 24 h using 25 or 45 ml in accordance with absorptiveness of the food stuff. Muesli was chosen as a worst case food simulant for the recovery tests. Dry food migration was investigated using Tenax® as a food simulant in accordance with the European Norm Standard 14338. These conditions deviate in duration and temperature from the migration test conditions specified in Regulation EC 10/2011, which will be obligatory by January 1, 2016. Extracts were analyzed using high-performance liquid chromatography with diode array detection (HPLC-DAD). Recovery rates were between 90 and 110%. Limits of detection ranged from 2.8 (BP) to 29 (PBP) µg/dm2 for packaging materials and 2.5 (all other than BP) to 38 (BP) µg/dm2.

Results. BP was detected in 49% of packaging samples. The other photoinitiators and amine synergists were detected in less than 10% of packaging samples. Jung and colleagues found the highest concentrations of BP in the cartonboard packaging of cacao, with BP levels of 2510 µg/dm2. Packages of muesli, Indian pappadums and breakfast cereals also contained high levels of photoinitiators and amine synergists.

The German researchers only analyzed the content of those packages that did not have an obvious barrier material like aluminum foil. They detected photoinitiators and amine synergists in 33 of the 99 food stuffs analyzed. 20 contained photointitiators or amine synergists above the legal limits specified in Article 3 of Regulation EC 1935/2004 and Article 14 of Regulation EC 178/2002. 12 of the 23 food samples in which BP was quantified exceeded the legal limit of 600 µg/kg. 6, 3, 1 and 1 products violated the SML of methylbenzophenone (MBP), methyl-o-benzoylbenzoate (MOBB), ethyl-4-dimethylaminobenzoate (EDAB) and 2-ethylhexyl-4-(dimethylamino)-benzoate (EHDAB) as specified in the Swiss ordinance, respectively. The authors confirmed that polyethylene lining was permeable to BP, and that BP also migrated via the gas phase irrespective of paper or polypropylene barriers.

In order assess migration over time and to avoid cross contamination, eight samples with high levels of photoinitiators were wrapped in aluminum foil until the expiry date. This storage form also simulated the storage of large amounts of packages on pallets. 1-hydroxycyclohexylphenyl-ketone (HCHKP) was found to migrate very rapidly and reached levels of 4260 µg/kg after 3 ½ months. Further, BP and MBP migrated through the inner PE packaging material into the food. Migration of PBP into the food on the other hand remained low, even though high amounts were detected in the packaging. As such, contaminants with small vapor pressure were generally observed to migrate less, but migration was also concentration driven. PE was found to have at best a temporary barrier effect slowing down migration.

To confirm that results were food stuff independent, migration tests were also carried out with Tenax®. For this purpose 3 cartonboard samples were coated with UV varnishes and one with a photoinitiator free dispersion varnish. BP was detected in all materials. Even the dispersion varnish contained low levels of BP, which the authors hypothesize to be due to cross-contamination. BP was detected in the 3 food simulants contained in UV varnished cartonboard. 36% of HCHPK detected in the packaging was transferred into the food simulant Tenax®. The variation measured in migration among different cartonboard samples with UV curing inks is likely due to the level of cross-linking of the polymer network. Cross linking depends amongst other factors on the quality of the light used for the UV curing process.

In conclusion, the authors confirm that many of the cartonboard packaged food stuffs contain photoinitiators and amine synergists when UV curing inks are used. The lack of adequate barriers indicates lack of awareness among food packaging producers. Jung et al. found concentration, vapor pressure and cross linking of the photoinitiators and amine synergists to be linked to transfer rate. Concentration alone did not adequately predict transfer rates. When carrying out migration tests with Tenax®, they advise to use real surface to mass ratios, as otherwise daily intake levels will be significantly underestimated.


Jung et al. (published online September 5, 2013). “Survey on the occurrence of photoinitiators and amine synergists in cartonboard packaging on the German market and their migration into the packaged foodstuff.” Food Additives and Contaminants: Part A.