Epsilon Archive for Student Projects

Abstract

Arsenic (As) is a toxic metalloid that can lead to serious health effects. Based on promising lab results, microscale and nanoscale zero valent iron (mZVI and nZVI) were injected into an anoxic As-contaminated site in Sweden as a remediation effort. However, monitoring showed a limited effect on groundwater As concentrations. For this study, we took groundwater and soil samples, along with x-ray absorption spectroscopy (XAS) measurements, to determine why the pilot test was mostly unsuccessful. Aggregation upon injection was the main cause of poor performance for mZVI: there were concentrated high levels of solid-phase Fe and visible black clumps in soil cores. As a result, mZVI had no impact on groundwater conditions. On the other hand, nZVI showed an impact on groundwater conditions at one sample point. At this point, the nZVI had a better distribution than mZVI, but was limited to upper layers of the aquifer, and did not have a large-scale impact. Additionally, after 2 years in the field it had not fully corroded—up to 38% of Fe(0) remained in some areas. Groundwater conditions were more reducing than expected, leading to As(III), the more toxic and mobile species of As, dominating over As(V) in the groundwater at all sites, further limiting As immobilization. Immobilization mechanisms were very different compared to previous lab studies, with sulfur (S) playing a bigger role than expected, as monothioarsenate, orpiment, realgar, and arsenopyrite were present. Sulfidized mZVI (S-mZVI) has been shown to be more mobile than unmodified ZVI, so follow-up short-term column experiments were conducted at varying flow speeds with sediment material from this field site to determine feasibility for future implementation. Overall, recovery rates (of <2% for all flow rates) were not promising. Subsequent modeling with HYDRUS-1D, however, showed evidence of heavy blocking (when the binding of particles to a collector limits the binding of subsequent particles), a property which could be beneficial for field distribution of S-mZVI. This study shows the limitations of transferring successful lab tests to the field. It also indicates that ZVI, regardless of size, is not an effective remediation strategy for As at this site, where heterogeneous ZVI distribution and very reducing conditions are strong limiting factors. S-mZVI, however, could be a promising remediation strategy in the field, despite recovery rates in the lab.