Comprehensive profiles of per- and polyfluoroalkyl substances in Chinese and African municipal wastewater treatment plants: New implications for removal efficiency

Municipal wastewater treatment plants (WWTPs) can reflect the pollution status of per- and polyfluoroalkyl substances (PFASs) pollution. Here, matched influent, effluent, and sludge samples were collected from 58 municipal WWTPs in China, South Sudan, Tanzania, and Kenya. Target and suspect screening of PFASs was performed to explore their pro-files in WWTPs and assess removal efficiency and environmental emissions. In total, 155 and 58 PFASs were identified in WWTPs in China and Africa, respectively; 146 and 126 PFASs were identified in wastewater and sludge, respec-tively. Novel compounds belonging to per- and polyfluoroalkyl ether carboxylic acids (PFECAs) and sulfonic acids (PFESAs), hydrogen-substituted polyfluorocarboxylic acids (H-PFCAs), and perfluoroalkyl sulfonamides (PFSMs) ac-counted for a considerable proportion of total PFASs (sigma PFASs) in Chinese WWTPs and were also widely detected in African samples. In China, estimated national emissions of sigma PFASs in WWTPs exceeded 16.8 tin 2015, with >60 % originating from emerging PFASs. Notably, current treatment processes are not effective at removing PFASs, with 35 of the 54 WWTPs showing emissions higher than mass loads. PFAS removal was also structure dependent. Based on machine learning models, we found that molecular descriptors (e.g., LogP and molecular weight) may affect adsorp-tion behavior by increasing hydrophobicity, while other factors (e.g., polar surface area and molar refractivity) may play critical roles in PFAS removal and provide novel insights into PFAS pollution control. In conclusion, this study comprehensively screened PFASs in municipal WWTPs and determined the drivers affecting PFAS behavior in WWTPs based on machine learning models.

Automated summary

This study collected samples from municipal wastewater treatment plants in China, South Sudan, Tanzania, and Kenya to investigate the presence and removal efficiency of per- and polyfluoroalkyl substances (PFASs) pollution. It was found that PFASs were present in both Chinese and African WWTPs, but the current treatment processes were ineffective at removing them, with 35 out of 54 WWTPs showing higher emissions than mass loads. The study concluded that the behavior of PFASs in WWTPs is influenced by various factors, as determined by machine learning models.