Cheng-Shiuan Lee

Per- and polyfluoroalkyl substances (PFASs) commonly occur as complex mixtures, yet the relationships among PFAS structure, cytotoxicity, and DNA strand-break responses remain incompletely resolved. We evaluated 11 PFASs selected to span environmentally prevalent and regulatory-relevant perfluorocarboxylic acids (PFCAs), perfluorosulfonic acids (PFSAs), and fluorotelomer sulfonates (FTSs) in HepG2 cells using MTT cell-viability and alkaline comet assays. Individual PFAS showed a wide range of cytotoxic potency after 24 h, with IC50 ranging from 5700 μM for perfluorobutanoic acid (PFBA) to 0.3 μM for 8:2 FTS. Cytotoxic potency followed the order 8:2 FTS > PFOS > PFNA > PFOA > PFHxS > 6:2 FTS > HFPO-DA > PFHxA > PFBS > 4:2 FTS > PFBA. A PFAS mixture test was conducted to investigate the effects of PFAS type and concentration. Short-chain PFAS pairs exhibited predominantly antagonistic interactions, whereas 4:2 FTS + 8:2 FTS showed the most consistent synergistic pattern. DNA strand-break responses did not parallel cytotoxic potency: PFOA produced the largest tail moment, whereas 8:2 FTS was the most cytotoxic PFAS. LC–MS/MS exposure verification quantified PFAS concentration in the medium and identified condition-dependent losses for PFOS and 8:2 FTS. These findings provide screening-level evidence that PFAS hazard ranking depends on both chemical structure and end point, while the high administered concentrations and in vitro exposure limitations preclude direct inference to environmental risk.

Per- and polyfluoroalkyl substances (PFAS) have garnered considerable attention over the last decade due to their persistence in the environment and potential negative health effects on living organisms, including humans. However, our understanding of their occurrence, transport, and fate in the marine environment remains limited. For example, the interaction of PFAS with marine phytoplankton, which serve as entry points for several contaminants into the marine food web, has not yet been thoroughly examined. In this study, we overcame the challenging issues related to PFAS sorption onto filter membranes and successfully investigated algal uptake of PFAS by various marine diatoms (i.e., Thalassiosira pseudonana, Thalassiosira weissflogii, Phaeodactylum tricornutum, and Chaetoceros muelleri) under environmentally realistic conditions. Volume concentration factors (VCFs) of PFAS by T. pseudonana showed a positive correlation with PFAS carbon chain length, and the presence of a sulfonic group enhanced uptake. For perfluorocarboxylic acids (PFCAs), logVCFs ranged from 0.17 (C4-PFCA) to 5.65 (C14-PFCA), and, on average, VCF increased by 4-fold per carbon added from C6 to C14. The uptake of PFAS isomers by algae was also examined. Linear PFAS showed a higher affinity for algal cells compared to branched PFAS (e.g., logVCFL-PFOS = 3.40, logVCFbr-PFOS = 3.20). Emerging PFAS such as HFPO-DA (GenX) exhibited very little accumulation in diatoms (logVCF = 0.39). We observed that temperature (18°C vs. 4°C) and light (light vs. dark) had no significant effect on PFAS uptake. Interestingly, reducing salinity appears to slightly increase algal PFAS uptake. Among the four diatom species tested in this study, T. weissflogii exhibited the lowest PFAS accumulation, and the degree of PFAS uptake was proportional to the surface-to-volume ratio of algal cells. The cellular distribution of PFAS in diatoms was also investigated. C8 to C10 PFCAs were primarily associated with the cytoplasmic fraction, whereas C11 to C14 PFCAs were mostly associated with the frustule. Our findings represent an important systematic study of PFAS uptake by marine phytoplankton and of algal PFAS accumulation in response to environmental factors, which will benefit modeling of emerging contaminants in the marine food web/environment.

Hydrophobic ion pairing (HIP) was used as a novel pretreatment approach to enhance the removal of short-chain per- and polyfluoroalkyl substances (PFAS) by granular activated carbon (GAC). The formation of ion pairs between PFAS and cationic surfactants was tested through controlled batch experiments with varying types and dose of HIP (e.g., cetyltrimethylammonium chloride (CTAC)) and surface tension measurements. Interactions between PFAS and CTAC were observed to be disturbed in high ionic strength solution, likely due to competitive electrostatic interactions from the presence of inorganic ions. Addition of CTAC (2 μM) prior to GAC treatment in batch adsorption experiments greatly improved the sorption of short-chain PFAS (C₇) by 23-96%. The changes in sorption capacities were hypothesized to result from the altered physicochemical properties of the formed ion pair complex, thereby affecting its uptake on GAC. Both the linear free energy relationship (LFER) calculation and molecular dynamic (MD) simulation suggested that short-chain PFAS sorption was improved via enhanced hydrophobicity of formed ion pairs with a low self-assembly tendency, while long-chain PFAS sorption was inhibited due to the formation of large complexes with a high self-assembly tendency. Rapid small-scale column tests were performed for tap water spiked with an equal mass of perfluorobutanoate (PFBA), perfluorobutanesulfonate (PFBS), and perfluorooctanesulfonate (PFOS). The bed volume for 50% breakthrough was improved by 350% and 230% for PFBA and PFBS, respectively, in the presence of CTAC and PFOS.

Per- and polyfluoroalkyl substances (PFAS) enter the marine food web, accumulate in organisms, and potentially have adverse effects on predators and consumers of seafood. However, evaluations of PFAS in meso-to-apex predators, like sharks, are scarce. This study investigated PFAS occurrence in five shark species from two marine ecosystems with contrasting relative human population densities, the New York Bight (NYB) and the coastal waters of The Bahamas archipelago. The total detected PFAS (∑PFAS) concentrations in muscle tissue ranged from 1.10 to 58.5 ng g^-1 wet weight, and perfluorocarboxylic acids (PFCAs) were dominant. Fewer PFAS were detected in Caribbean reef sharks (Carcharhinus perezi) from The Bahamas, and concentrations of those detected were, on average, ∼79% lower than in the NYB sharks. In the NYB, ∑PFAS concentrations followed: common thresher (Alopias vulpinus) > shortfin mako (Isurus oxyrinchus) > sandbar (Carcharhinus plumbeus) > smooth dogfish (Mustelus canis). PFAS precursors/intermediates, such as 2H,2H,3H,3H-perfluorodecanoic acid and perfluorooctanesulfonamide, were only detected in the NYB sharks, suggesting higher ambient concentrations and diversity of PFAS sources in this region. Ultralong-chain PFAS (C ≥ 10) were positively correlated with nitrogen isotope values (δ¹⁵N) and total mercury in some species. Our results provide some of the first baseline information on PFAS concentrations in shark species from the northwest Atlantic Ocean, and correlations between PFAS, stable isotopes, and mercury further contextualize the drivers of PFAS occurrence.

The application of electron beam (e-beam) technology for water treatment has been proposed to be a faster and safer approach to decomposing persistent contaminants in water, because of its ability to rapidly generate high amounts of both oxidizing and reducing reactive species without the addition of chemicals. In this study, we utilized a lab-scale 9 MeV e-beam accelerator to investigate the feasibility of treating 1,4-dioxane in various water matrices in batches with low sample volumes (−1), without the need for any sample modification or pH adjustment. Low dissolved oxygen in the solution enhanced the degradation efficiency by 21–23% when treating 1000 mg L⁻¹ of 1,4-dioxane, presumably because of the increased H˙ and O˙⁻ that can react with 1,4-dioxane. Although the degrading intermediates were not fully mineralized to carbon dioxide at the tested doses, the detected intermediates, such as aldehydes and organic acids, were not as persistent as 1,4-dioxane and were more vulnerable to conventional treatment methods and natural attenuation. The slowest reaction rate constant was observed when treating wastewater samples (k = 0.13–0.62 kGy⁻¹), followed by contaminated groundwater (k = 0.16–1.4 kGy⁻¹), suggesting other organics and ions could scavenge the generated reactive species. The electrical energy per order (EE/O) ranged from 0.39 (DI water) to 6.3 (wastewater) kWh m⁻³ per order, depending on the initial concentration of 1,4-dioxane and water matrix. The EE/O values were comparable to other traditional advanced oxidation processes (AOPs) to treat 1,4-dioxane, suggesting the feasibility of utilizing e-beam to treat contaminated waters. The organic carbon content of the sample positively correlated (R² >0.9) with the EE/O values and thus can be utilized to predict e-beam treatment performance. Our results show that e-beam radiolysis is a promising technology to treat 1,4-dioxane and could potentially outperform other AOPs and 1,4-dioxane disposal methods (e.g., incineration) in terms of energy consumption and treatment time, leaving no trace of oxidants.

Although most manufacturers stopped using long-chain per- and polyfluoroalkyl substances (PFASs), including perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS), short-chain PFASs are still widely employed. Short-chain PFASs are less known in terms of toxicity and have different adsorption behavior from long-chain PFASs. Previous studies have shown electrostatic interaction with the adsorbent to be the dominant mechanism for the removal of short-chain PFASs. In this study, we designed a high charge density cationic quaternized nanocellulose (QNC) to enhance the removal of both short- and long-chain PFASs from contaminated water. Systematic batch adsorption tests were conducted using the QNC adsorbent to compare its efficiency against PFASs with varying chain lengths and functional groups. From the kinetic study, PFBA (perfluorobutanoic acid), PFBS (perfluorobutanesulfonic acid) and PFOS showed rapid adsorption rates, which reached near equilibrium values (>95% of removal) between 1 min to 15 min, while PFOA required a relatively longer equilibration time of 2 h (it obtained 90% of removal within 15 min). According to the isotherm results, the maximum adsorption capacity (Q_m) of the QNC adsorbent exhibited the following trend: PFOS (Q_m = 559 mg g⁻¹ or 1.12 mmol g⁻¹) > PFOA (Q_m = 405 mg g⁻¹ or 0.98 mmol g⁻¹) > PFBS (Q_m = 319 mg g⁻¹ or 1.06 mmol g⁻¹) > PFBA (Q_m = 121 mg g⁻¹ or 0.57 mmol g⁻¹). This adsorption order generally matches the hydrophobicity trend among four PFASs associated with both PFAS chain length and functional group. In competitive studies, pre-adsorbed short-chain PFASs were quickly desorbed by long-chain PFASs, suggesting that the hydrophobicity of the molecule played an important role in the adsorption process on to QNC. Finally, the developed QNC adsorbent was tested to treat PFAS-contaminated groundwater, which showed excellent removal efficiency (>95%) for long-chain PFASs (C7–C9) even at a low adsorbent dose of 32 mg L⁻¹. However, short-chain PFASs (i.e., PFBA and perfluoropentanoic acid (PFPeA)) were poorly removed by the QNC adsorbent (0% and 10% removal, respectively) due to competing constituents in the groundwater matrix. This was further confirmed by controlled experiments that revealed a drop in the performance of QNC to remove short-chain PFASs at elevated ionic strength (NaCl), but not for long-chain PFASs, likely due to charge neutralization of the anionic functional group of PFASs by inorganic cations. Overall, the QNC adsorbent featured improved PFAS adsorption capacity at almost two-fold of PFAS removal by granular activated carbons, especially for short-chain PFASs. We believe, QNC can complement the use of common treatment methods such as activated carbon or ionic exchange resin to remove a wide range of PFAS pollutants, heading towards the complete remediation of PFAS contamination.