Fate and Bioavailability of Per- and Poly-Fluorinated Substances (PFASs) in Soils

Abstract

Per- and poly-fluorinated substances (PFASs) are a group of compounds with similar structures and have a unique set of properties which make them resistant to biodegradation, available to bioaccumulate, and potentially toxic to many organisms in the environment. Yet there is currently limited information on the fate and bioavailability of PFASs in the soil environment. The thesis investigates the sorption and bioavailability of three commonly reported perfluoroalkyl acids (PFAAs) in a wide range of soils with varying properties and determines if sorption coefficients and bioaccumulation can be predicted from soil properties and are affected by residence time in soil. Preliminary studies found that sorption of 14C-perfluoroocatanioc acid (PFOA) was considerable on common laboratory consumables, e.g. centrifuge tubes and syringe filters. In contrast to reports in the literature, sorption losses on polypropylene tubes were found to be significantly higher than on glass tubes. Sorption losses were also significant using syringe filter membranes. This highlights that significant errors can occur, creating analytical bias during routine laboratory procedures. Sorption (Kd) of perfluorooctane sulphonic acid (PFOS), 14C-PFOA and perfluorohexane sulphonic acid (PFHxS) was investigated in a wide range of Australian soils with varying properties. Modelling was conducted to determine if soil properties could predict sorption coefficients using two modelling strategies: multiple linear regression (MLR) using traditional laboratory chemical analyses of soils; and diffuse reflectance Fourier transform mid-infrared spectroscopy coupled to partial least squares regression (DRIFT-PLSR). The sorption coefficients for all three compounds were at the lower end of the ranges previously reported, perhaps due to the low amounts of organic carbon (OC) and alkaline pH of many Australian soils. The retention of PFOA was weak in all soils, but even more so in subsoils, indicating that PFOA could easily be leached through surface- and sub-soil horizons into ground-waters. The retention of PFOS and PFHxS was not affected by soil depth. The sorption of all three PFAAs was influenced by their structural differences and different soil properties. The sorption of PFOA was positively affected by OC and silt-plus-clay content, PFOS sorption was positively correlated to OC content and negatively correlated to pH and PFHxS sorption was positively correlated to OC content, clay content and concentrations of exchangeable cations. DRIFT-PLSR modelling indicated that soils dominated with quartz and pyrophyllite minerals had a low affinity for PFOA sorption. The DRIFT-PLSR modelling of PFOS and PFHxS sorption was unreliable, likely due to the low and narrow range of Kd values found. For PFOA modelling, similar results were found between the MLR and DRIFT-PLSR modelling strategies which suggested that DRIFT-PLSR could be used as a quicker and cheaper technique to predict sorption compared to traditional laboratory analyses of soil coupled with MLR. The aging and bioavailability of the same three PFAAs was investigated in 20 soils varying in physicochemical properties. The soils were spiked with PFOS, PFOA and PFHxS and incubated for up to six-months. A second batch of soils was spiked with the same compounds at the same rate creating ‘aged’ and ‘freshly spiked’ soil treatments. These soils were planted with Phaseolus vulgaris seeds and uptake of the PFAAs into plant shoots was determined. Over the six-month incubation period sampling of porewater was also conducted. Concentrations of the three PFAAs in soil porewater and in plant tissues did not significantly and consistently change (p >0.05) with increasing time of soil contact with the PFAAs. This indicates that significant aging of PFAAs does not occur in soils, a behaviour unlike other persistent organic pollutants (POPs) e.g. polyaromatic hydrocarbons. The sorption affinity of all three PFAAs in these soils was PFHxS<PFOA<PFOS and the inverse relationship was observed for bioaccumulation factors (BAFs) in bean plants, PFHxS>PFOA>PFOS, indicating that PFASs with low sorption have a greater potential to bioaccumulate in terrestrial plants. The lack of any significant effect of time of soil:chemical contact on partitioning or bioavailability to plants indicates that these chemicals are not only resistant to degradation but will remain bioavailable over time, unlike other POPs. The results of this thesis show that while the sorption and bioavailability of PFASs in soils are affected by both soil properties and type of PFAS present, they are not affected by time of residence in soil. These data have the potential to be used by regulators to better manage contaminated soils and the risk assessment of any further contamination from those sites into surface- or ground-waters or the uptake by plants