PFAS removal from leachate water using electrocoagulation

Abstract

This thesis aimed to elucidate the effects of dissolved organic matter on the applicability of electrocoagulation (EC) with iron electrodes for per- and polyfluorinated alkyl substances (PFAS) removal from landfill leachate. The research methodology consisted of an experimental and a computational part. Galvanostatic EC experiments were conducted in an aerated beaker with 500 mL working volume, using iron electrodes with a surface area of 8 cm2. The pH, voltage and current were measured continuously for 50 minutes. Blank measurements were conducted in a 2 g/L NaCl solution. PFAS removal was tested from 0.25 mmol/L perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS) solutions at current densities of 12.5 and 25 mA/cm2, as well as from solutions with a commercial humic acid (HA) source added at 25 mA/cm2. The removal of HA was also tested separately. Eventually, the removal from real leachate samples was examined. A computer model was constructed in PHREEQC (pH Redox Equilibrium Calculation), a program for modelling chemical processes in water. The model simulates the removal of all tested pollutants based on electrostatic adsorption to a continuously forming surface, associated with the precipitating ferrihydrite. Model rate constants for O2 and CO2 dissolution and Fe(II) oxidation were determined from the results of the blank experiments. Good model fits were obtained for these datasets. HA could be removed completely within 50 minutes at all tested current densities and initial concentrations. PFOA removal was unsuccessful with and without HA as co-solute, with a maximum removal of 17 %. PFOS removal reached 81 % at 25 mA/cm2, but the removal essentially stagnated after 5 minutes treatment time. The presence of HA did not significantly affect the PFOS removal. Instead, HA removal was retarded by the presence of PFOS. For the real leachate samples, no significant removal of PFAS occurred. Conversely, approximately 20 % PFAS removal was observed after 5 minutes treatment of a leachate sample spiked with 0.15 mmol/L PFOA, perfluorobutane sulfonic acid (PFBS) and perfluorobutanoic acid (PFBA). However, this removal reversed during the remaining treatment time. The adsorption equilibrium constants were the most important model parameters. These parameters were determined for HA, PFOA and PFOS based on the experimental data. The model fits were good for HA and PFOA removal, indicating that these were indeed mainly removed by electrostatic adsorption. Contrarily, PFOS removal could not be represented accurately by the current model. Instead of stagnating after five minutes, the simulated removal continued to completion. The poor model fit may indicate that the mechanism of PFOS removal extends beyond strict electrostatic adsorption. Instead, charge neutralization of PFOS molecules causing their aggregation or mass transfer limitations in the vicinity of the electrodes could be involved. Further research is needed to explore these possibilities and determine improved equilibrium constants for the relevant adsorption reactions. In conclusion, this research did not confirm the high PFAS removal efficiency as observed in previous studies. Instead, significant removal of PFOA did not occur and removal of PFOS did not exceed 81 %. Competition effects were not observed in the simultaneous treatment of PFOA and HA. The presence of PFOS impeded the removal of HA, indicating that PFOS was removed preferentially under the current experimental conditions. The established model could simulate most experimental results accurately.