An article published on May 5, 2016 in the peer-reviewed journal Food Additives & Contaminants: Part A describes a case study for using effect directed analysis (EDA) and accurate mass spectrometry (MS) to identify non-intentionally added substances (NIAS) from paper and board food contact materials. EDA is a method where fractions of a sample containing multiple unknown substances are tested for selected endpoints, usually in vitro. Test results are then used to prioritize the fractions for in-depth MS analysis, tentative identification of unknown substances, and confirmatory testing.

Linda Bengtström and colleagues from the National Food Institute at the Technical University of Denmark (DTU Food) analyzed an extract from a printed pizza box made of recycled corrugated fiber board, and selected two fractions, one acidic and one alkaline, that exhibited a particularly high aryl hydrocarbon receptor (AhR) activity in vitro (tested with AhR-CALUX). Extraction, fractionation and testing experiments are described in a 2014 paper from the same group.

In the two tested fractions, 75 substances were tentatively identified based on MS analysis, and 15 of these compounds were initially selected for further testing in vitro based on structural similarities to substances known to be AhR-active. However, of these 15 substances, in vitro testing and accurate quantitation in the extracts could be performed only for those 7 that were commercially available. Of these 7, the identities could be confirmed for four substances, all found in printing inks, including 2’-(dibenzylamino-)-6’-(diethylamino)-3H-spiro[2-benzofuran-1,9’-xanthen]-3-one (CAS 34372-72-0) and three dyes (Solvent Violet 8, CAS 52080-58-7; Basic Red 1, CAS 989-38-8; Baso Red 546 CAS 509-34-2).

The toxic equivalency factor of the pizza box’ extract, based on the maximum response in the AhR-CALUX, was calculated to be 47.9 ngTCDD /kgpaper. In vitro testing of pure standards of tentatively identified chemicals could not explain a large proportion of AhR activity observed in the total extract. In fact, a weak AhR activity at non-cytotoxic concentrations was observed only for Solvent Violet 8. Similarly, the amount of dioxins and polychlorinated biphenyls present in the extract explained only 3% of the observed AhR activity (1.35 ngTCDD /kgpaper). This indicates that many AhR active substances migrating from recycled board remained unidentified so far. Further investigations are ongoing.

The availability of standards was a significant impediment for quantifying substances and confirming their AhR activity. The scientists plan to work on fine-tuning the fractionation to narrow down the number of substances present in each fraction. Ideally, this would result in one substance per fraction, allowing for toxicity testing and accurate quantitation even when a pure analytical standard is not available commercially. Another improvement of the overall EDA strategy for food contact materials suggested by the authors was to first quantify specific contaminants with known toxicological potency, and proceed to EDA and comprehensive tentative evaluation of unknown compounds only if necessary.

Read more

Philip Lightowlers (2016). “Danish study tracks toxicity in cardboard food packaging.” Chemical Watch (published June 2, 2016).


Bengtström, L. et al. (2016). “Non-targeted screening for contaminants in paper and board food contact materials using effect directed analysis and accurate mass spectrometry.Food Additives & Contaminants: Part A (published online May 5, 2016).

Bengtström, L. et al. (2014). “Fractionation of extracts from paper and board food contact materials for in vitro screening of toxicity.Food Additives & Contaminants: Part A 31:1291-1300.