Arsenic Removal Is Mediated by Biofilms in Non-Aerated and Aerated Wetland Mesocosms

Planted and unplanted recirculating constructed wetland mesocosms maintained under non-aerated (reducing) and aerated (oxidizing) conditions were compared in their ability to remove 1 mg/L As(V) from simulated domestic wastewater over a 33-week period. Each week, one day prior to the addition of arsenic, non-aerated mesocosms removed 70% of the nitrate and 95% of the organic carbon likely through denitrification while aerated mesocosms removed organic carbon only (97%). Upon arsenic addition and treatment for the remainder of the week, non-aerated mesocosms removed more arsenic (72%) than aerated mesocosms (43%). For the first time in the literature, we estimated the proportion of arsenic found in biofilms detached from gravel and incorporated this aspect into mass balance calculations. We discovered that biofilms from non-aerated mesocosms contained a higher proportion of arsenic (28%) than biofilms from aerated mesocosms (5%). Plants from both types of mesocosms removed a negligible amount of arsenic (<0.41%). Through X-ray fluorescence (XRF) analysis, it was discovered that biofilms from non-aerated mesocosms had the highest iron concentrations (35,000 mg/kg), and this iron was identified as pyrite (2–15 wt%) through X-ray diffraction (XRD) analysis. The iron concentration of aerated biofilms was no different from that found in clean gravels (17,500 mg/kg), and no iron minerals were detectable in these samples using XRD analysis. Although some of the arsenic removal in both types of wetlands was thought to be attributable to adsorption onto substrates and biofilms, as well as uptake into biofilms, removal of arsenic in the non-aerated mesocosms was likely enhanced by the presence of pyrite because pyrite can adsorb arsenic directly or indirectly through the formation of iron oxides that can adsorb or coprecipitate with the iron oxides.