Recovery of Rare Earth Elements from Acid Mine Drainage with Supported Liquid Membranes: Impacts of Feedstock Composition for Extraction Performance.

Acid mine drainage (AMD) from inactive coal mines can be enriched in rare earth elements (REEs) and has gained much attention as an alternative source for these technology-critical metals. However, AMD is a relatively low-grade REE resource in which the abundance of impurities and the composition variability of the feedstock create major uncertainties for the performance of REE extraction technologies. This study sought to identify AMD feedstock variables that influence the extraction efficiency of REEs by supported liquid membranes (SLMs). SLM separation is a process involving a hydrophobic membrane embedded with an extracting solvent that facilitates the selective extraction of REE ions. The major aims were to (1) assess the effectiveness of SLM-based REE separation from several AMD samples representing a spectrum of aqueous composition, (2) determine the effects of AMD storage and holding time on extraction performance, and (3) assess the impact of AMD pretreatment (e.g., filtration and pH adjustment) on REE recovery. The results showed that relative extraction fluxes of REE correlated with AMD characteristics such as pH and major ions such as Fe, Ca, and Mn. The purity of the acid strippant product, expressed as the REE dry weight content, depended on the initial REE concentrations in the AMD source rather than the flux of individual REEs across the membrane. For AMD samples stored for 3 months prior to extraction, REE recovery by SLM separations was substantially decreased if oxidation of Fe(II) to Fe(III) was observed during sample storage. Pretreatment of AMD feedstocks by pH adjustment did not substantially improve the separation performance. Overall, this study establishes primary water quality parameters of AMD that influence the SLM separation flux and product purity. Such insights contribute to a mechanistic understanding of critical metals extractions by SLM for complex and nontraditional feedstocks such as AMD wastes.

Distribution and Homogenization of Multiple Mercury Species Inputs to Freshwater Wetland Mesocosms.

Mercury (Hg)-impaired aquatic ecosystems often receive multiple inputs of different Hg species with varying potentials for transformation and bioaccumulation. Over time, these distinct input pools of Hg homogenize in their relative distributions and bioaccumulation potentials as a result of biogeochemical processes and other aging processes within the ecosystem. This study sought to evaluate the relative time scale for homogenization of multiple Hg inputs to wetlands, information that is relevant for ecosystem management strategies that consider Hg source apportionment. We performed experiments in simulated freshwater wetland mesocosms that were dosed with four isotopically labeled mercury forms: two dissolved forms (Hg2+ and Hg-humic acid) and two particulate forms (nano-HgS and Hg adsorbed to FeS). Over the course of one year, we monitored the four Hg isotope endmembers for their relative distribution between surface water, sediment, and fish in the mesocosms, partitioning between soluble and particulate forms, and conversion to methylated mercury (MeHg). We also evaluated the reactivity and mobility of Hg through sequential selective extractions of sediment and the uptake flux of aqueous Hg in a diffusive gradient in thin-film (DGT) passive samplers. We observed that the four isotope spikes were relatively similar in surface water concentration (ca. 3000 ng/L) immediately after spike addition. At 1-3 months after dosing, Hg concentrations were 1-50 ng/L and were greater for the initially dissolved isotope endmembers than the initially particulate endmembers. In contrast, the Hg isotope endmembers in surface sediments were similar in relative concentration within 2 months after spike addition. However, the uptake fluxes of Hg in DGT samplers, deployed in both the water column and surface sediment, were generally greater for initially dissolved Hg endmembers and lower for initially particulate endmembers. At one year postdosing, the DGT-uptake fluxes were converging toward similar values between the Hg isotope endmembers. However, the relative distribution of isotope endmembers was still significantly different in both the water column and sediment (p < 0.01 according to one-way ANOVA analysis). In contrast, selective sequential extractions resulted in a homogeneous distribution, with >90% of each endmember extracted in the KOH fraction, suggesting that Hg species were associated with sediment organic matter. For MeHg concentrations in surface sediment and fish, the relative contributions from each endmember were significantly different at all sampling time points. Altogether, these results provide insights into the time scales of distribution for different Hg species that enter a wetland ecosystem. While these inputs attain homogeneity in concentration in primary storage compartments (i.e., sediments) within weeks after addition, these input pools remain differentiated for more than one year in terms of reactivity for passive samplers, MeHg concentration, and bioaccumulation.