RESUMO
In this study, an optimized random forest (RF) model was employed to better understand the soil-water partitioning behavior of per- and polyfluoroalkyl substances (PFASs). The model demonstrated strong predictive performance, achieving an R2 of 0.93 and an RMSE of 0.86. Moreover, it required only 11 easily obtainable features, with molecular weight and soil pH being the predominant factors. Using three-dimensional interaction analyses identified specific conditions associated with varying soil-water partitioning coefficients (Kd). Results showed that soils with high organic carbon (OC) content, cation exchange capacity (CEC), and lower soil pH, especially when combined with PFASs of higher molecular weight, were linked to higher Kd values, indicating stronger adsorption. Conversely, low Kd values (< 2.8 L/kg) typically observed in soils with higher pH (8.0), but lower CEC (8 cmol+/kg), lesser OC content (1 %), and lighter molecular weight (380 g/mol), suggested weaker adsorption capacities and a heightened potential for environmental migration. Furthermore, the model was used to predict Kd values for 142 novel PFASs in diverse soil conditions. Our research provides essential insights into the factors governing PFASs partitioning in soil and highlights the significant role of machine learning models in enhancing the understanding of environmental distribution and migration of PFASs.
RESUMO
Stormwater management practices (SMPs) rely on infiltration and adsorption capabilities of soil and vegetative cover to mitigate the harmful impacts of contaminants in stormwater runoff, including potentially toxic elements (PTEs). Under chemical equilibrium conditions, the soil-water distribution coefficient (Kd) quantifies the relationship between the solid and aqueous phase PTE concentrations, and thus the PTE removal efficiency and mobility through the SMP soil layers during the infiltration process. The SMP loading ratio (LR), the ratio of the drainage area to the SMP infiltration area, combined with runoff concentration determines SMP mass loading and is also expected to impact PTE transport. In this study, a simulation model was developed to investigate PTE breakthrough and build-up in SMP media, considering the impacts of Kd and LR. Eight PTEs were simulated (Cl-, Cr, Fe, Zn, Cu, As, Cd, and Pb), and Cl- was the only PTE that showed high mobility and reached the groundwater table (e.g., ~ 1 year for breakthrough). Conversely, other PTEs were effectively immobilized in the top ~60 cm of soil for a simulated lifespan of 20 years. Soil and porewater contaminant indices, as indicators of SMP lifespan, were estimated based on the ratio of PTE porewater and soil concentrations after 20 years to published standards, suggesting the following order of environmental significance (most concern to least): Cl- > Cr > As > Pb > Fe > Cu > Cd > Zn. After 20 years of simulated use, only Cl- pore water concentrations at the groundwater table exceeded regulatory values, with porewater contamination index values of 4 to 7.5. Chloride also exceeded the surficial media soil contamination index, as did As and Cr, though these exceedences were largely associated with media background concentrations. Generally, higher LR and Kd contributed to higher accumulation of PTEs in top layers; however, simulations showed that the combination of low LR and high Kd may result in lower PTE accumulation in the media, such that the PTE concentration in soil may decrease in deeper layers. In these scenarios, a notable fraction of PTE load was adsorbed on top layers and considerably lower PTE concentrations reached the lower layers. Sensitivity analysis revealed that dispersion, infiltration rate, and kinetically-limited sorption did not impact the PTE accumulation and mobility to a practical extent. The results from this simulation may be adapted to various environmental conditions to enhance the design and maintenance of SMPs.