Predicting the relationship between PFAS component signatures in water and non-water phases through mathematical transformation: Application to machine learning classification

An artificial intelligence platform for automated PFAS subgroup classification: A discovery tool for PFAS screening

Since structural analyses and toxicity assessments have not been able to keep up with the discovery of unknown per- and polyfluoroalkyl substances (PFAS), there is an urgent need for effective categorization and grouping of PFAS. In this study, we presented PFAS-Atlas, an artificial intelligence-based platform containing a rule-based automatic classification system and a machine learning-based grouping model. Compared with previously developed classification software, the platform's classification system follows the latest Organization for Economic Co-operation and Development (OECD) definition of PFAS and reduces the number of uncategorized PFAS. In addition, the platform incorporates deep unsupervised learning models to visualize the chemical space of PFAS by clustering similar structures and linking related classes. Through real-world use cases, we demonstrate that PFAS-Atlas can rapidly screen for relationships between chemical structure and persistence, bioaccumulation, or toxicity data for PFAS. The platform can also guide the planning of the PFAS testing strategy by showing which PFAS classes urgently require further attention. Ultimately, the release of PFAS-Atlas will benefit both the PFAS research and regulation communities.

A fundamental model for calculating interfacial adsorption of complex ionic and nonionic PFAS mixtures in the presence of mixed salts

Per- and polyfluoroalkyl substances (PFAS) are emerging contaminants that have been used extensively as firefighting agents and in a wide range of commercial applications around the world. As many of the most-common PFAS components are surfactants, they readily accumulate at interfaces, a process that can govern their environmental fate. There are thousands of PFAS compounds, and they have nearly always been used as mixtures, so it is common to find many different PFAS components present together in the environment. Furthermore, the interfacial behavior of ionic PFAS can be strongly influenced by the presence of salts, with adsorption dependent on both the composition and concentration of salts present. Any predictions of PFAS interfacial behavior made without considering both the mixed nature of PFAS present, as well as the composition of the salts present, have the potential to be off by orders of magnitude. To date, models capable of making predictions of PFAS interfacial adsorption when both mixed PFAS and mixed salts are present have not been presented. The work described here addresses this need by extending a mass-action model developed previously by the authors to allow predictions in cases where complex combinations of mixed PFAS and mixed salts are present. Predictions of PFAS interfacial affinity for a range of PFAS mixture conditions and ionic strengths are verified using experimentally-measured surface tension data. The new model provides physically-realistic prediction of interfacial adsorption of a wide range of PFAS mixtures over a wide range of salt concentrations and compositions. The model is capable of predicting interfacial adsorption of ionic/nonionic PFAS mixtures in the presence of salts, and can also make predictions when there is competitive adsorption between different PFAS components, a common case in PFAS source zones where high concentrations of multiple components are present and in foam fractionation reactors.

A Review on Per- and Polyfluoroalkyl Substances in Pregnant Women: Maternal Exposure, Placental Transfer, and Relevant Model Simulation

Per- and polyfluoroalkyl substances (PFASs) are important and ubiquitous environmental contaminants worldwide. These novel contaminants can enter human bodies via various pathways, subsequently posing risks to the ecosystem and human health. The exposure of pregnant women to PFASs might pose risks to the health of mothers and the growth and development of fetuses. However, little information is available about the placental transfer of PFASs from mothers to fetuses and the related mechanisms through model simulation. In the present study, based upon a review of previously published literature, we initially summarized the exposure pathways of PFASs in pregnant women, factors affecting the efficiency of placental transfer, and mechanisms associated with placental transfer; outlined simulation analysis approaches using molecular docking and machine learning to reveal the mechanisms of placental transfer; and finally highlighted future research emphases that need to be focused on. Consequently, it was notable that the binding of PFASs to proteins during placental transfer could be simulated by molecular docking and that the placental transfer efficiency of PFASs could also be predicted by machine learning. Therefore, future research on the maternal-fetal transfer mechanisms of PFASs with the benefit of simulation analysis approaches is warranted to provide a scientific basis for the health effects of PFASs on newborns.

Calculating PFAS interfacial adsorption as a function of salt concentration using model parameters determined from chemical structure

Per- and polyfluoroalkyl substances (PFAS) are widely-detected environmental contaminants known to concentrate at surfaces and interfaces. Many of the most commonly-detected PFAS function as ionic surfactants under environmental conditions. The interfacial behaviors of ionic surfactants, including PFAS, are strongly dependent on salt concentration and composition, with interfacial affinity potentially varying by orders of magnitude for the same compound under different conditions. The work described here presents a tool for predicting the salt-dependent adsorption of PFAS compounds based entirely on chemical structure, something of great value for predicting the real-world environmental behavior of many of the large numbers of PFAS compounds for which experimental data are not available. The approach combines two different previously-developed models, one a mass-action model designed to predict the effects of salts on interfacial adsorption of ionic PFAS (the UNSW-OU salt model), and the second a group-contribution model designed to predict interfacial adsorption of PFAS in the absence of salt based on chemical structure. The challenge of combining the two models comes from the fact that both are based on different isotherms. The salt model can produce sigmoidal isotherms under salt-limited conditions (an isotherm shape that is supported by experimental evidence), while the group-contribution model can generate Langmuir parameters from calculations based on chemical structure. Equations were derived to determine salt model isotherm parameters from Langmuir parameters (either from the group-contribution model or experimental sources) by matching surface tension curves in the vicinity of the concentration of highest second derivative. Refined group-contribution model parameters were determined based on data from an additional 40 surface tension curves to allow improved structure-based predictions for important classes of PFAS that were not sufficiently well-represented in the original model. The resulting equations provide a tool allowing quantitative predictions of PFAS behavior under realistic environmental conditions for compounds for which little or no experimental data are available.

Retention of PFOS and PFOA Mixtures by Trapped Gas Bubbles in Porous Media

The transport of per- and polyfluoroalkyl substances (PFAS) in soil and groundwater is important for site investigation, risk characterization, and remediation planning. The adsorption of PFAS at air-water interfaces has been shown to significantly contribute to PFAS retention, with subsequent effects on concentrations and the time scales of transport. In this study, column experiments were conducted to investigate the transport of perfluorooctanesulfonic acid (PFOS), perfluorooctanoic acid (PFOA), and 6:2 fluorotelomer sulfonate (6:2 FTS) individually and in binary mixtures in the presence of a trapped gas phase, using clean sands to isolate adsorption to air-water interfaces. Consistent with previous studies, the transport of PFOS, PFOA, and 6:2 FTS was retarded by adsorption at the air-water interface, with greater retention of PFOS due to its higher affinity for the air-water interface. Chromatographic separation occurred in the experiments using binary mixtures of PFOS and PFOA, with greater retention at lower influent concentrations. The mixture experiments also showed enhanced breakthrough of PFOA in the presence of PFOS, where effluent concentrations of PFOA were temporarily greater than the influent concentration prior to the breakthrough of PFOS. This enhanced breakthrough was attributed to competition between PFOS and PFOA for adsorption to the air-water interface.

Long-distance transport of per- and polyfluoroalkyl substances (PFAS) in a Swedish drinking water aquifer

Use of per- and polyfluoroalkyl substance (PFAS)-containing aqueous film-forming foams (AFFF) at firefighting training sites (FFTS) has been linked to PFAS contamination of drinking water. This study investigated PFAS transport and distribution in an urban groundwater aquifer used for drinking water production that has been affected by PFAS-containing AFFF. Soil, sediment, surface water and drinking water were sampled. In soil (n = 12) at a FFTS with high perfluorooctane sulfonate (PFOS) content (87% of ∑PFAS), the ∑PFAS concentration (n = 26) ranged from below detection limit to 560 ng g^-1 dry weight. In groundwater (n = 28), the ∑PFAS concentration near a military airbase FFTS reached 1000 ng L^-1. Principal component analysis (PCA) identified the military FFTS as the main source of PFAS contamination in drinking water wellfields >10 km down-gradient. Groundwater samples taken close to the military FFTS site showed no ∑PFAS concentration change between 2013 and 2021, while a location further down-gradient showed a transitory 99.6% decrease. Correlation analysis on PFAS composition profile indicated that this decrease was likely caused by dilution from an adjacent conflating aquifer. ∑PFAS concentration reached 15 ng L^-1 (PFOS 47% and PFHxS 41% of ∑PFAS) in surface river water (n = 6) and ranged between 1 ng L^-1 and 8 ng L^-1 (PFHxS 73% and PFBS 17% of ∑PFAS) in drinking water (n = 4). Drinking water had lower PFAS concentrations than the wellfields due to PFAS removal at the water treatment plant. This demonstrates the importance of monitoring PFAS concentrations throughout a groundwater aquifer, to better understand variations in transport from contamination sources and resulting impacts on PFAS concentrations in drinking water extraction areas.