Project

Migration of linear and cyclic oligomers from (bio)polyesters in food contact

Updates
0 new
0
Recommendations
0 new
0
Followers
0 new
4
Reads
0 new
62

Project log

Marie Kubicova
added a research item
Polyesters labelled as bio-based or compostable are increasingly common among the 'bioplastics' in use as food contact materials (FCM). The knowledge of material composition is mandatory to predict potential leachable oligomers as well as to partly evaluate the correctness of the label 'bioplastic', which is used for promotional purposes. The composition of (bio)polyesters can be determined by alkaline hydrolysis of the entire material and subsequent analysis of the monomers via high-performance liquid chromatography with diode array detection and GC-MS detection. Thirty-three frequently used monomers (polycarboxylic acids, hydroxy carboxylic acids, polyols) including highly polar monomers such as lactic acid were analysed with detection limits below 10 g/kg of the material. Lactic acid enantiomer elucidation was performed using an enzyme assay. The content of non-hydrolysable residue was determined gravimetrically after hydrolysis, and the inorganic residue after washing. The composition of 12 polyesters mostly in food contact, labelled as bio-based or compostable and sampled from the market was elucidated recovering 92-101% of the total mass by summing up the determined monomers and non-polyester contents. Seven different monomers were detected in the 12 samples (up to four different monomers per sample), lactic acid being the most common (9 samples) with contents ranging from a minor component (about 11 mol%) up to the only monomer found in the material. The ratio of D-to L-lactic acid ranged from 0.3:99.7 to 4.7:95.3 (w/w). The non-hydrolysable (in)organic residue was quantified in amounts of up to 390 g/kg. Overall, the presented analytical protocol is a fundamental tool helping both to verify the appropriateness of labelling as biopolyesters as well as to predict potential leachables such as oligomers during an FCM risk assessment.
Martin Eckardt
added 3 research items
In terms of risk assessment especially for known and unknown substances migrating from food contact materials, quantification without corresponding reference substances currently poses a challenge. In the present study, the opportunity of a universal response quantification approach was evaluated by using a corona charged aerosol detector (CAD) for liquid chromatography combined with inverse gradient compensation. Characteristics of CAD detection in dependence of substance properties were analyzed with 46 randomly chosen reference substances. An almost equal CAD response (±20%) was achieved for non-volatile substances with a molecular weight of minimum 400 g/mol and a vapor pressure of maximum 10−8 Torr. We empirically defined an analytical parameter, Q50/35, the quotient of CAD peak areas at CAD evaporator temperatures of 50 °C and 35 °C, to predict the adequacy of the CAD universal response approach for quantification of known and unknown analyte substances. Exemplarily, we applied the CAD universal quantification approach for the determination of extractable oligomers below 1000 g/mol from a variety of food contact polycondensate plastic materials (e.g. polyesters like polyethylene terephthalate, polybutylene terephthalate, Tritan copolyester, polyamides 6, 6.6 and 6 T/6I and polyarylsulfones polyphenylsulfone and polyethersulfone). Quantitative results for in total 44 oligomers out of 11 materials were compared with established material-specific quantification methods using extracted oligomer mixtures as well as individual oligomers isolated from the mixtures. CAD-based quantification results were generally in accordance to published quantification approaches for polyamide oligomers and oligomers from polyarylsulfones. For oligomers extracted and isolated from polyester materials a slight underestimation was determined by CAD universal response approach. In terms of detection limits and accuracy, the universal CAD approach exhibits no advantages compared to established UV-methods, to date.
Coatings for cans or closures are essential to protect the metal from corrosion and the food from migration of hazardous metal ions. Since coatings are no inert materials, they can release substances of potential health concern into food. In the present study, a comprehensive analysis is presented for a complex two-layered polyester–phenol-coating commercially used for metal closures of complementary infant food in sterilised glass jars. Focussed on the identity and migration of cyclic polyester oligomers as a kind of predictable non-intentionally added substances, polyester resin raw materials (n = 3) as well as individual coating layers (n = 3) were characterised by several analytical strategies (size exclusion chromatography, high-performance liquid chromatography mass spectrometry, diode array detection, charged aerosol detection, monomer determination after alkaline hydrolysis, overall migrate). The main polyester monomers were terephthalic acid, isophthalic acid, trimellitic acid, ethylene glycol, diethylene glycol, neopentylglycol, 2-methyl-1,3-propanediol, 1,4-butanediol and tricyclodecanedimethanol. The coatings were extracted with solvents acetonitrile and ethanol (24 h, 60°C), food simulants 50% ethanol, 20% ethanol and water (1 h, 121°C) as well as homemade and commercial baby food (1 h, 121°C). The released total polyester content determined by alkaline hydrolysis ranged from 288 µg/dm² (water, 1 h, 121°C) to 6154 µg/dm² (acetonitrile, 24 h, 60°C). However, individual cyclic oligomers, mainly dimers, were released from the coating to up to about 140 µg/dm². Migration into infant food was best represented by the food simulants water (up to 1% fat) and 20% ethanol (up to 5% fat). Cyclic polyester oligomers are classified as Cramer III substances by the threshold of toxicological concern concept associated to an exposure threshold of 1.5 µg/kg body weight per day. Exposure to cyclic polyester oligomers might be a potential concern for highly exposed infants.
Linear and cyclic oligomers are unavoidable non-intentionally added substances (NIAS) present in food contact materials made from common polyesters such as polyethylene terephthalate (PET), polybutylene terephthalate (PBT) and polyester coatings. Although polyester oligomers can migrate into fats in significant amounts in high-temperature processes such as baking or frying, little is known about their toxicological properties and their behaviour in the human gastrointestinal tract. In the present study, first indications of a possible digestibility of polyester oligomers formed from the commonly used aromatic dicarboxylic acid terephthalic acid (TPA) are provided by in vitro experiments. Three cyclic polyester oligomers originated from PET (trimer) and PBT (dimer and trimer) were extracted from the raw materials, isolated and subjected to a simulated intestinal digestion. A fast cleavage (≥75% of the initial amount) of all three cyclic oligomers into their linear counterparts was detected already within the first hour of in vitro intestinal incubation. Subsequent hydrolysis to shorter chained linear oligomers was determined especially for the PET cyclic trimer. Degradation down to the monomer TPA was not observed. In terms of risk assessment and prioritisation for non-evaluated NIAS, the threshold of toxicological concern (TTC) concept is an appropriate tool. While cyclic polyester oligomers based on TPA are assigned to the TTC Cramer class III (high potential concern, exposure threshold 1.5 µg/kg body weight per day), the corresponding linear oligomers are expected to be of a lower probable toxicological concern (Cramer class I, 30 µg/kg body weight per day). A cleavage of cyclic polyester oligomers under human intestinal conditions, which was assessed to be likely by the provided in vitro experiments, could consequently affect the risk assessment on polyester oligomers.