On February 28, 2013, the peer-reviewed journal Analytica Chimica Acta published the review article “The challenge of identifying non-intentionally added substances from food packaging materials: A review” online (Nerin et al., 2013). The Spanish researchers from University of Zaragoza scrutinize analytic procedures for identifying, detecting and assessing contents of volatile and non-volatile non-intentionally added substances (NIAS) in food contact materials (FCMs) and in food simulants. NIAS originate mainly from three sources: (1) interaction between constituents of the packaging material, (2) degradation processes and (3) impurities of the raw materials. The ever more refined analytical procedures at hand yield a more detailed picture of components found in FCMs and often also enable their quantitative evaluation. As the researchers point out, most of the NIASs might not cause health concerns as their concentration is assumed to be minute. Still, the time consuming and costly identification and evaluation of substances present in a material migrating at potentially hazardous levels is legally required (EC 10/2011, Art. 19), even though, as of yet, no guidance for the analysis has been set forth. According to Nerin and colleagues the evaluation of NIAS is a delicate task due to the following reasons: Firstly, packaging materials are complex in structure with up to seven layers of different materials. Each layer might further be laminated, glued, coated and printed with a further set of substances. The fact that manufacturers of packaging material tend to keep the most compositions of materials confidential and reveal only main components aggravates this circumstance further. In addition, the packaging producers themselves remain mostly ignorant of substances contained in products supplied to them by other producers throughout the supply chain. Secondly, constituents might be tailor-made in order to yield certain properties. The authors name among others UV absorbers, plasticizers and varnishes as additional additives. Not only is their composition usually not disclosed, but neither is there a dedicated EU regulation in place for these materials. Thirdly, Nerin and colleagues state that accurate NIAS detection and evaluation is made difficult because they are present at very low concentrations. Proper analysis requires sophisticated analytic tools relying on gas and liquid chromatographic separation and mass spectrometry (MS) detectors coupled to computer based databases. To facilitate this analysis, the researchers suggest that a protocol for NIAS identification should be put in place. Furthermore, they advocate for the traceability of packing materials and the regulation of the description of NIAS. Reviewing the literature the authors focus mainly on substances with a molecular weight below 1000 Da, since heavier chemicals are considered substantially hindered in migration, and are therefore not expected to become part of the packaged foodstuff. For the several sources of NIASs as degradation processes, additive degradation, impurities, neoformed compounds and contaminants, the authors present a comprehensive list of which a selection is shown below (Table 1).
|Impurities||N2-dodecanoyll-arginine (LAS)||Ethyl lauroyl-arginates (LAE) impurity|
|1-hexanol-2-ethyl||acrylic adhesives in multi-layer packaging|
|cyclic adipate 1,4,7-trioxacyclotridecane-8,13-dione||polyurethane (PU) adhesives|
|Neoformed compounds||Primary aromatic amines (PAAs)||polyurethane (PU) adhesives|
|BADGE·HCl, BADGE·2HCl||bisphenol A diglycidyl ether (BADGE)|
|chlorohydrins||epoxidized soybean oil (ESBO) in PVC|
|Oxidation products||2,4-ditertbutylphenol (2,4-DTB)||Irgafos 1010 and Irgafos 168, both antioxidantants|
|alkylphenols||tris(nonylphenyl)phosphite (TNPP) (anti-oxidant in poly-|
(vinyl chloride) (PVC), polyolefins and acrylics);
polyethoxylated nonylphenols (APEOs) (surfactants from PET bottle production, adhesives and polymeric dispersions)
|Thermal degradation products||formaldehyde||polyethylene terephthalate (PET)|
|(Z)-9-octadecenamide||Chimasorb 944 (applied in LDPE polymers)|
|2,4-diterbutyl-6-nitro-phenol and 2-cyclohexene-1-dione|
|3,5-dimethyl or o-methyloxime|
|4b-8-dimethyl-2-isopropylphenanthrene||abietic acid (tackifier in hot melt adhesives)|
|dehydroabietic acid methyl ester|
The very first step in NIAS risk assessment is the unambiguous identification of migrating compounds. As composition analysis of FCMs is a challenging endeavor Nerin et al. stress the importance to introduce a protocol covering as many NIASs as possible present in FCM samples. Special attention should be paid to compounds most likely migrating into food stuff. NIAS analysis is carried out in accordance with EU regulation (EU/10/2011) both in actual samples of packaging, where the NIAS concentration is higher, as well as in food simulants. Four classes of analytical tools are available for polymer FCMs: (1) direct thermal desorption couple with GC–MS analysis; (2) direct MS analysis as atmospheric solids analysis probe (ASAP), direct analysis in real time (DART) and desorption electrospray ionization (DESI); (3) total polymer dissolution analysis via GC–MS or LC–MS; and (4) solvent extraction analysis via GC–MS or LC–MS. These approaches are best suited for confirmation of the presence of known NIAS, but are less useful for identification of novel NIAS, as MS fragment determination is difficult without a priori knowledge of present compounds.
Having identified a new migrating NIAS, a risk assessment has to be carried out, which should include the elucidation of its origin. It is, however, not always possible to explain the toxicity of a material by the identified substances alone. Nerin et al. give an example of an in vitro study analyzing the genotoxicity of recycled and virgin paper. The overall toxicity of the material could not solely be explained by the known substances found to be migrating. A further study referred to showed that only 2% of the toxicity was explained by the starting material (BADGE, BPA-BADGE 2H2O and TMA) for epoxy coatings. Hence, mostly reaction products are to be held responsible for the overall toxicity.
A further obstacle is that toxicity data is not available for all chemical substances present in FCMs. The authors refer to the Threshold of Toxicological Concern (TTC) concept as a potential solution. TTC is an approach estimating toxicity based on chemical structures (FPF has reported extensively). Importantly, Nerin et al. also point to the limitation of the approach naming some high potency carcinogens unsuitable for this approach: e.g. aflatoxin-like compounds, azoxy compounds, nitroso compounds, 2,3,7,8-dibenzoparadioxin and dioxin like compounds and steroids. The same holds true for substances accumulating in the body including polyhalogenated dibenzodioxins, -dibenzofurans, and –biphenyls, heavy metals, and certain proteins. The evaluation of NIAS migration and their toxicity is both time intensive and difficult. Nerin et al. also point out that this setting presupposes equal responses in detection and that substances present at very low doses do not raise concern. Likewise, it cannot replace real individual identification. Nonetheless, the authors consider it a good starting point, which is constantly being refined by advancing analytic tools and helpful for the prioritization of different NIAS.
In conclusion, the researches lay out the difficulty and challenges of identifying and quantifying NIAS, which are related to factors as a compound’s physical and chemical properties, its concentration in the sample, the availability of a pure standard and the possibility to separate it from the sample. Further complications arise from complex mixtures and isomers with similar properties that hinder unequivocal analytical confirmation. The authors expect that standardized treatment procedures, separation and concentration methods along with high resolution detection tools, software and database applications will eventually lead to successful NIAS identification. Currently, volatile compounds can be detected more readily due to the availability of mass spectra libraries in contrast to non-volatile compounds, where no such libraries are available. Here, the experimental set-up is crucial to successful identification. Nonetheless, Nerin et al. are confident that also these drawbacks will eventually be resolved. More sophisticated tools are being developed, which integrate more refined detection with software and database applications. Apart from the necessity to resolve technical issues, the authors state that the analytic and identification processes should be regulated. The authors propose that the origins of packaging material should be made traceable, which would allow for a more rapid exclusion of unwanted and hazardous compounds from FCMs.
Nerin C. et al. (2013). “The challenge of identifying non-intentionally added substances from food packaging materials: A review” Analytica Chimica Acta (published online February 28, 2013).
Bradley, E.L., and Coulier, L. (2007). “An investigation into the reaction and breakdown products from starting substances used to produce food contact plastics.” Central Science Laboratory York.