Machine Learning-Based Models with High Accuracy and Broad Applicability Domains for Screening PMT/vPvM Substances.

Persistent, mobile, and toxic (PMT) substances and very persistent and very mobile (vPvM) substances can transport over long distances from various sources, increasing the public health risk. A rapid and high-throughput screening of PMT/vPvM substances is thus warranted to the risk prevention and mitigation measures. Herein, we construct a machine learning-based screening system integrated with five models for high-throughput classification of PMT/vPvM substances. The models are constructed with 44 971 substances by conventional learning, deep learning, and ensemble learning algorithms, among which, LightGBM and XGBoost outperform other algorithms with metrics exceeding 0.900. Good model interpretability is achieved through the number of free halogen atoms (fr_halogen) and the logarithm of partition coefficient (MolLogP) as the two most critical molecular descriptors representing the persistence and mobility of substances, respectively. Our screening system exhibits a great generalization capability with area under the receiver operating characteristic curve (AUROC) above 0.951 and is successfully applied to the persistent organic pollutants (POPs), prioritized PMT/vPvM substances, and pesticides. The screening system constructed in this study can serve as an efficient and reliable tool for high-throughput risk assessment and the prioritization of managing emerging contaminants.

Freshwater ecotoxicity characterization factors for PMT/vPvM substances.

This study addresses the gap in freshwater ecotoxicological characterization factors (CFs) for Persistent, Mobile, and Toxic (PMT) and Very Persistent and Very Mobile (vPvM) substances. These CFs are vital for integrating the ecotoxicity impacts of these chemicals into life cycle assessments. Our goals are twofold: first, to calculate experimental freshwater CFs for PMT/vPvM substances listed by the German Environment Agency (UBA); second, to compare these CFs with those from the USEtox database. The expanded UBA list includes 343 PMT/vPvM substances, each representing a unique chemical structure, and linked to 474 REACH-registered substances. This study successfully computed CFs for 244 substances, with 107 overlapping the USEtox database and 137 being new. However, ecotoxicity data limitations prevented CF determination for 97 substances. This research enhances our understanding of freshwater CFs for PMT/vPvM substances, covering 72% of UBA's 343 PMT/vPvM substances. Data scarcity remains a significant challenge, which invariably impedes CF calculations. Notably, the disparities observed between CF values in the USEtox database and those derived in this research largely stem from variations in ecotoxicity data. Consequently, this research underscores the dynamic nature of CFs for substances, emphasizing the need for regular updates to ensure their accuracy and relevance.

Probing the chemical complexity of tires: Identification of potential tire-borne water contaminants with high-resolution mass spectrometry.

Tire wear particles are not only the type of polymer particles most prevalent in the environment but also act as source of various organic micropollutants, many of which are likely still unknown. We extracted particles prepared from nine tires in artificial freshwater (28 d) with the goal to characterize leachables (max intensity >105 in artificial freshwater), which are tire-borne water contaminants. A subsequent extraction of these particles with acetone (3 h) was used to assess the long-term leaching potential. A suspect and nontarget screening in aliquots of each extract led to the detection of 214 organic substances of which 145 were classified as leachables. The intrinsic polarity of some leachables (mean log D (pH 7.4) 3.9), which facilitates an increased aquatic mobility, highlights their potential as environmental water contaminants. With N,N'-diphenylguanidine (DPG) and benzothiazole, two of the ten unequivocally identified leachables, are classified as potential persistent, mobile and toxic substance by the German Environment Agency. Of the identified chemicals DPG showed the highest intensities in aqueous extracts and N-(1,3-dimethylbutyl)-N'-phenyl-p-phenylenediamine (6-PPD), the precursor of 6-PPD-quinone, in acetone extracts. A comparison between the 69 detected suspects and 174 high-intensity signals (>106) detected in the nontarget screening led to an overlap of only 29 features. A detailed investigation of the remaining high-intensity suspects revealed the presence of 13 proposed DPG reaction products, further highlighting the chemical complexity of tires. Consequently, we conclude that there are many, often still unrecognized chemicals entering the aquatic environment through leaching from tire wear particles.