Immobilization of Mercury by Stabilized FeS Nanoparticles and Effects of Oil Dispersant on Sorption/Desorption and Photodegradation of Polycyclic Aromatic Hydrocarbons

Abstract

Iron sulfide (FeS) nanoparticles were prepared with sodium carboxymethyl cellulose (CMC) as a stabilizer, and tested for enhanced removal of mercury (Hg(II)) from water, soil and sediment. The presence of CMC at ≥0.03 wt.% can fully stabilize 0.5 g/L of FeS (i.e. CMC-to-FeS molar ratio ≥0.0006) through concurrent electrostatic and steric stabilization mechanisms. Fourier transform infrared spectroscopy (FTIR) spectra suggested that CMC molecules were attached to the nanoparticles through bidentate bridging and hydrogen bonding. Increasing the CMC-to-FeS molar ratio from 0 to 0.0006 enhanced mercury sorption capacity by 20%; yet, increasing the ratio from 0.0010 to 0.0025 diminished the sorption by 14%. FTIR and X-ray diffractograms (XRD) analyses suggested that precipitation (formation of cinnabar and metacinnabar), ion exchange (formation of Hg0.89Fe0.11S) and surface complexation were important mechanisms for mercury removal. Batch kinetic data revealed that the stabilized nanoparticles facilitated rapid uptake of Hg(II), and the kinetic data can be interpreted with a pseudo-second-order kinetic model. We proposed a dual-mode isotherm model, which considers both precipitation and adsorption mechanisms, to interpret the sorption isotherm data. High mercury uptake was observed over the pH range of 6.5-10.5, whereas at pH<6 significant loss in Hg capacity was observed. High concentrations of Cl- (>106 mg/L) and dissolved organic matter (5 mg/L as TOC) may inhibit Hg uptake, while effect of ionic strength (0-0.2 M) was negligible. When aged for 2.5 years, 14% of sorbed Hg(II) was leached out of the nanoparticles due to pH drop. The leaching, however, can be prevented by maintaining pH above neutral. In situ immobilization of Hg in field-contaminated soil and sediment using soil-deliverable CMC-FeS nanoparticles (CMC-to-FeS molar ratio = 0.0010) was investigated through a series of batch and column experiments. Transmission electron microscopy measurements revealed a particle size of 34.3±8.3 nm (standard deviation), whereas dynamic light scattering gave a hydrodynamic diameter of 222.5±3.2 nm. Batch tests showed that at an FeS-to-Hg molar ratio of 28:1 to 118:1, the nanoparticles reduced water-leachable Hg by 79%-96% and the TCLP (Toxicity Characteristic Leaching Procedure) based leachability by 26%-96%. Column breakthrough tests indicated that the nanoparticles were deliverable in the sediment/soil columns under moderate injection pressure. However, once the external pressure was removed, the delivered nanoparticles remained virtually immobile under typical groundwater flow conditions. When the Hg contaminated soil and sediment were treated with 52 to 95 pore volumes of a 500 mg/L FeS nanoparticle suspension, water-leachable Hg was reduced by 90%-93% and TCLP-leachable Hg was reduced by 65%-91%. The results warrant further field demonstration of this promising in situ remediation technology. During the 2010 Deepwater Horizon oil spill, ~2.1 million gallons of dispersants were applied to the surface and well head to break up oil slicks. However, it remains unknown how the dispersants affect the environmental fate and transport of persistent oil components such as polycyclic aromatic hydrocarbons (PAHs) in the Gulf Coast ecosystems. Effects of a model oil dispersant (Corexit EC9500A) on sorption/desorption of phenanthrene were investigated with two marine sediments. Kinetic data revealed that the presence of the dispersant at 18 mg/L enhanced phenanthrene uptake by up to 7%, whereas the same dispersant during desorption reduced phenanthrene desorption by up to 5%. Sorption isotherms confirmed that at dispersant concentrations of 18 and 180 mg/L phenanthrene uptake progressively increased for both sediments. Furthermore, the presence of the dispersant during desorption induced remarkable sorption hysteresis. The effects were attributed to added phenanthrene affinity and capacity due to sorption of the dispersant on the sediments. Dual-mode models adequately simulated sorption isotherms and kinetic data in the presence of the dispersant. Water accommodated oil (WAO) and dispersant-enhanced WAO increased phenanthrene sorption by up to 22%. Effects of the oil dispersant Corexit EC9500A on UV-mediated photodegradation of pyrene in the Gulf Coast seawater were investigated. The presence of 18 and 180 mg/L of the dispersant increased the first-order photodegradation rate by 5.5% and 17%, respectively, within 360 min, and the dispersant also reduced or eliminated pyrene volatilization. By combining the individual first-order rate laws for volatilization and photodegradation, we proposed an integrated kinetic model that was able to adequately predict the overall degradation data. Mechanistic studies indicated that superoxide radicals play a predominant role, and 1-hydroxypyrene was the main intermediate in the degradation process with and without the dispersant, suggesting that electrons are transferred from excited pyrene to oxygen. The dispersant enhanced the formation of the superoxide radicals. In the presence of 18 mg/L of the dispersant, the first-order photodegradation rate increased with increasing ionic strength and temperature, decreased with increasing HA concentration, but remained independent of solution pH. The results are important for understanding the roles of oil dispersants on environmental fate of spilled oil and persistent oil components.