A critical assessment of passive air samplers for per- and polyfluoroalkyl substances

Abstract Since their inclusion in the Stockholm Convention, there has been a need for global monitoring of perfluorooctane sulfonate (PFOS), its salts and perfluorooctanesulfonyl fluoride (PFOSF), along with other non-listed highly fluorinated compounds. Passive air samplers (PAS) are ideal for geographic coverage of atmospheric monitoring. The most common type of PAS, using polyurethane foam (PUF) as a sorbent, was primarily developed for non-polar semivolatile organic compounds (SVOCs) and are not well-validated for polar substances such as the per- and polyfluoroalkyl substances (PFASs), however, they have been used for some PFASs, particularly PFOS. To evaluate their applicability, PAS were deployed for measurement of PFASs in outdoor and indoor air. Outdoors, two types of PAS, one consisting of PUF and one of XAD - 2 resin, were deployed in an 18-week calibration study in parallel with a low-volume active air sampler (LV - AAS) in a suburban area. Indoors, PUF - PAS were similarly deployed over 12 weeks to evaluate their applicability for indoor monitoring. Samples were analysed for perfluoroalkyl carboxylic acids (PFCAs), perfluoroalkyl sulfonates (PFSAs), perfluorooctane sulfonamides (FOSAs), and perfluorooctane sulfonamidoethanols (FOSEs). In outdoor air, 17 out of the 21 PFAS were detected in more than 50% of samples, with a median ∑ 17 PFASs of 18.0 pg m −3 while 20 compounds were detected in indoor air with a median concentration ∑ 20 PFASs of 76.6 pg m −3 using AAS samplers. PFOS was the most common PFAS in the outdoor air while PFBA was most common indoors. Variability between PAS and AAS was observed and comparing gas phase and particle phase separately or in combination did not account for the variation observed. PUF-PAS may still have a valuable use in PFAS monitoring but more work is needed to identify the applicability of passive samplers for ionic PFAS.