Methodology / Approach

The basis of this project was a literature review of NTS and SS studies of articles, chemical products, and recycled materials and collection of identified substances.

The literature collected included primary scientific articles and grey literature. The search was performed both via outreach to participant networks for more complete collection of relevant grey literature as well as searches of online databases via Google, Google Scholar, PubMed and Web of Science. Primary search keywords included “non-target screening”, “non-target analysis”, “suspect screening” and the major product/material categories/­subcategories used for the Swedish Chemical Agency’s PRIO database (Appendix , spreadsheet “E”). Relevant references within papers were also used. Due to time constraints, the literature search was not exhaustive and ended upon 76 references which might have resulted in some product categories being underrepresented in the collected literature. Besides that, the PRIO categories did not cover every manmade material.

All references were reviewed and assessed for relevance. Relevance entailed true NTS and/or SS methods used on relevant articles, chemical products, or recycled materials. For the references deemed relevant, substances identified were extracted and entered into a master list. If Schymanski levels were provided, these were also entered. If not, Schymanski levels were assigned as described below. For each study the following information was collected: the study aim, a priori information used for study design, type of screening (NTS, SS, or both), PRIO material category/ies, material subcategory/ies (specific material details), instrumentation used, QA/QC performed (described below), and the file location of tentatively identified substances, if available. The relevant study/­literature list and associated information is provided in Appendix, spreadsheet “A”. For each chemical entry, the following information was collected: chemical name, formula, CAS number, Schymanski level, type of screening (NTS or SS), detection frequency (i.e., the number of product/material samples the substance was found in), sample preparation and chromatographic methods, scanning and ionization modes, PRIO material category, material sub-category (specific material details), study source, and regulatory and hazardous information (described further below). Chemical identifiers (Simplified Molecular Input Line Entry Specification, SMILES) and CAS numbers that were not provided by the literature were collected from the Chemical Identifier Resolver of the U.S. National Cancer institute

https://cactus.nci.nih.gov/chemical/structure

and PubChem database automatically using the Python packages, CirPy and PubChemPy, respectively. Chemicals that still had missing CAS number or SMILES were then manually checked in PubChem and CompTox Chemicals Dashboard

https://comptox.epa.gov/dashboard/

. A mixed product category (i.e., plastic + elastomer) was used in cases where multiple product categories were included in the study, but it was not clear which product type each chemical was found in. An additional ‘Other’ product category was created for product types that didn’t clearly fall under any of the PRIO categories. The following additional information was collected if provided by the study: canonical Smiles, InChIkey, molecular weight, mass-to-charge ratio, retention time, and chemical function. For the final version of the chemical list, only chemicals with a Schymanski level of 1 or 2 were kept.

Schymanski levels if not provided by the study were assigned according to the following criteria:

Quality controls and data processing parameters of each study are described in the QA/QC column in Appendix, spreadsheet “A”. Parameters described include: peak filtering cut off values, usage of procedural blanks, application of internal standards, replicates, precursor mass error, retention time difference, library match score limit, spectral databases used and complementary identification through orthogonal methods such as different ionization techniques and inclusion if ion mobility data. The parameters and set cut off limits included in the QA/QC were used for the assessment and assignment of Schymanski levels to data lacking Schymanski scoring.

The substances given a Schymanski confidence level of 1 or 2 were then screened through a number of regulatory and hazard information databases using the IVL early warning hazard screening tool to determine if these substances are already regulated, prioritized or under assessment within the EU. Those included the following databases from the European Chemical Agency (ECHA) as accessed in November 2024:

as well as: