Is it possible to describe the behaviour of per- and polyfluoroalkyl substances (PFAS) and biodegradable precursors using the MACRO model? – Approach and preliminary results.

Abstract

<p>Per- and polyfluoroalkyl substances (PFAS) are anthropogenic substances, which moved to the scientific focus due to their ubiquity in the environment. Several thousands of individual PFAS are known, which differ in chemical structure as well as in their chemical and physical properties. Due to the huge number of substances, the assessment of their environmental behaviour is challenging. Still, more information about these substances, which are partly already declared PBT (persistent, bioaccumulative and toxic), is needed for risk assessments and remediation.</p><p>The overall aim of this study was to describe the environmental behaviour in soil during leaching of ten different PFAAs (perfluoroalkyl acids), a group of persistent PFAS, and two diPAPs (polyfluoroalkyl phosphate diesters) with physical equations and parameters to gain knowledge about substance-related characteristics. For this purpose, we simulated the leaching of all surveyed PFAAs and of the biodegradable precursors 6:2 diPAP and 8:2 diPAP, which can transform into PFAAs. Soil and percolation data received from two experimental studies, a laboratory soil column study and a field lysimeter study, were used to evaluate the leaching and transformation behaviour using the MACRO model. In both studies the behaviour of diPAPs were simulated, which enables a comparison between natural and unnatural conditions. Modelling results from the laboratory study, in which climate impacts are limited, were used as input for the field study with natural climatic conditions. Parameters used for calibration were substance-related such as the adsorption distribution coefficient (K<sub>D</sub>). The amount of non-recovered PFAS, which is potentially related to the formation of non-extractable residues (NERs), was described using a first-order degradation equation. The evaluation of simulations was done using the goodness-of-fit function KGE (Kling-Gupta Efficiency) and a fitting score comparing simulated and observed soil and percolation data. Optimisation was done using the caRamel evolutionary algorithm with multi-objectives within the GNU R environment and up to 15,000 runs per substance, which resulted in a pareto front. Results of parameter values were then used to describe their leaching behaviour as transformation products (PFAAs) of diPAPs in two steps: (1) optimisation of leaching and transformation of diPAPs with soil data, (2) simulation of leaching of transformation products using results of PFAA simulations. The preliminary modeling results are promising for simulating the behavior of PFAAs as well as their precursors with MACRO.</p>