Polyethylene terephthalate polyesters (PETs), formed of terephthalic acid and ethylene glycol units, are some of the most commonly found polymers in plastics in our daily life and environment. PET is used as a generic term for plastics containing a majority of polyethylene terephthalate polyesters but also encompasses blends with terephthalate polyesters with different diols. Everyday-life PET-based plastics are becoming more complex through the use of virgin PET in blends with recycled PET plastics. Furthermore, the molecular complexity of PET-based plastics is increased by weathering or microorganisms' degradation. So, the identification of the precise PET-based plastics' composition and their structural modifications is still very challenging. They are principally identified by optical spectroscopy and Py-GC-MS. Unfortunately, these techniques are not efficient for identifying and quantifying PET at the molecular level, and they provide very little structural information on their modifications. In this work, PET was depolymerized under soft conditions and analyzed by complementary mass spectrometry approaches. After presolubilization with 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP), a direct transamidation using N,N-dimethyl-1,3-propanediamine (DMAPA) at lower temperatures was performed to break ester bonds and improve MS ionization, allowing the detection of acids and aromatic di- and poly acids. The resulting products were analyzed by ultrahigh-resolution mass spectrometry using MALDI ionization on a 9.4 T FTICR MS and ESI ionization on an Orbitrap MS hyphenated to liquid chromatography. Daily-life PET samples from different countries, brands, and contents were studied, and their depolymerized products were identified and quantified. A specific fingerprint for each sample was attributed. Dimers, oligomers, scission products, oxygenated species, cross-linking structures, and additives were successfully identified with high accuracy and resolution. This strategy can be applied to other polyesters and copolyesters.
Almasri et al. (Tue,) studied this question.