PFAS adsorption to air-water interfaces: Effects of velocity and PFAS concentration

Understanding how per- and poly-fluoroalkyl substances (PFAS) are transported in soil and groundwater is critical to site characterization, monitoring, risk assessment, and remediation planning.  This includes an understanding of PFAS retention and release associated with adsorption to air-water interfaces, which is particularly important for transport through the vadose zone.  Many previous laboratory studies have focused on individual PFAS with experiments conducted over a narrow range of concentrations and pore-water velocities.  However, the effects of PFAS mixtures, concentrations and velocities are important to extend our understanding, and the application of numerical models, to realistic site conditions.

In this study, a series of laboratory experiments was conducted using one-dimensional sand-packed columns (40 cm × 5 cm dia.).  Trapped air bubbles were emplaced in the sand (quasi-saturated conditions) by sequential drainage and imbibition.  Similar to fluctuations in the water table, this emplacement technique was used to create immobile air-water interfaces that are uniformly distributed throughout the column and are readily accessible to flowing water.  Each experiment included separate injections of non-reactive tracer (NaCl) and PFAS solutions through both water-saturated and quasi-saturated columns.  A clean, low organic carbon sand was used to eliminate solid-phase sorption (verified through comparison of non-reactive tracer and PFAS breakthrough in the water-saturated columns) and to isolate the effect of air-water interfaces.  Experiments were conducted using single-component solutions of PFOS over a range of concentrations (2 to 1000 μg/L) and pore-water velocities (0.8 to 2.6 cm/day).  Experiments were also conducted using diluted aqueous film-forming foam (AFFF) solutions containing PFOS. 

The results showed that PFOS breakthrough was significantly delayed in the presence of trapped air bubbles, and that breakthrough varied considerably with concentration and velocity.  Greater retardation occurred generally at lower PFAS concentrations and slower velocities.  However, the change in retardation due to a change in velocity was sensitive to concentration, with greater changes occurring for lower concentrations.  PFOS breakthrough was also affected by the presence of other PFAS and surface active components in AFFF, with PFOS in the diluted AFFF arriving earlier than expected for an equivalent concentration of PFOS alone.  Mixture effects were also observed in the breakthrough of other PFAS in AFFF, particularly concentration overshoot (effluent concentration temporarily greater than the influent concentration) of some less surface active PFAS.  The results highlight the need for more comprehensive models of PFAS transport that incorporate non-ideal and competitive behaviour to capture processes occurring in complex field scenarios.