Dimin Fan



Document Type


Degree Name



Oregon Health & Science University


Reducing iron minerals are of broad interest in biogeochemistry and environmental engineering, in part because of their roles in a variety of approaches to remediation of contaminated groundwater. Zerovalent iron (Fe(0)), which is predominantly of anthropogenic origin, and divalent iron minerals (Fe(II)), usually from naturally-occurring biogeochemical processes, together comprise a broad class of the reducing materials that are responsible for most of the contaminant removal by in-situ chemical reduction (ISCR). Regardless of how ISCR is implemented (e.g., injection of ZVI, dithionite, or polysulfide), the physical and chemical properties of the resulting reductants are strongly influenced by the biogeochemical transformations involving iron and sulfur that are at aqueous mineral interfaces. These processes operate from molecular scale to grain scale, and to pore scale in controlling the contaminant removal and the overall remediation performance. Therefore, from both fundamental and practical point of view, a more comprehensive understanding of the biogeochemical processes that control contaminant fate in ISCR and related scenarios is needed. The research presented in this dissertation is composed of two parts. Part I (Chapters 2 and 3) concerns the development of a combined chemical and biological remediation strategy involving the application of nano zerovalent iron (nZVI) to sequester pertechnetate (TcO4−), a common radionuclide oxyanion. Part II (Chapters 4 and 5) concerns the development of a general approach to characterize the thermodynamic, kinetic, and capacity aspects of a broad spectrum of reducing iron minerals in order to better understand their contributions to in-situ chemical reduction. Part I—Laboratory batch experiments were conducted to investigate the abiotic sulfidation of nZVI and sequestration of Tc by sulfidated nZVI to evaluate the prospect of remediating Tc-contaminated groundwater by adding nZVI to stimulate sulfidic conditions in situ. Complementary solid/surface characterization methods showed that up to 10% of added nZVI was converted to FeS after 24 h exposure to aqueous sulfide (at sulfide to iron ratios above 0.112). The sulfidated nZVI gave faster Tc removal rates than nZVI that was not exposed to sulfide. Evidence by transmission electron microscopy (TEM) showed that the majority of Tc was associated with FeS, and X-ray absorption spectroscopy (XAS) showed there was a complete shift in the Tc sequestration products from TcO2 to TcS2 as aqueous sulfide increased from 0 to 1 mM. The stability of the resulting reduced Tc sulfide phases to reoxidation was examined by exposing them to ambient air. Tc sequestered by sulfidated nZVI was reoxidized (to dissolved TcO4− in solution) significantly more slowly compared with the reduction products formed by non-sulfidated nZVI. The onset of Tc release occurred after complete consumption of FeS, indicating that oxidation was inhibited partially due to the presence of FeS serving as redox buffer. XAS characterization also revealed a solid state transformation process from TcS2 to TcO2 during oxidation, which also contributed to the inhibition of Tc reoxidation. Part II—A set of chemical reactive probes (CRPs) were used to characterize the thermodynamic and kinetic properties of reducing iron minerals, with the ultimate goal of using them to predict the abiotic natural attenuation of chlorinated ethenes (Chapter 4). After equilibration in aqueous media containing various combinations of iron minerals with adsorbed Fe(II), the redox speciation of the thermodynamic CRPs was measured spectrophotometrically, and the results were used to calculate mineral reduction potentials as a function of mineral type, mineral loading, and Fe(II) concentration. The potentials determined by this method (ECRP) were 150–200 mV more negative than the potentials measured by conventional oxidation-reduction potentials with a platinum electrode (EPt), consistent with the hypothesis that the latter method does not fully represent the mineral reduction potential. ECRP was shown to qualitatively correlate with the rate constants for reduction several model contaminants (4-chloronitrobenzene, carbon tetrachloride, and 2-chloroacetophenone). Further correlation analysis using reduction rate constants compiled from prior studies showed consistent reactivity patterns of three classes of contaminants—including chlorinated ethenes—across the whole range of ISCR relevant reducing mineral minerals, despite the experimental variability of different sources. These results suggest the possibility of using fast reacting CRPs to assess slow abiotic natural attenuation processes from both thermodynamic and kinetic perspectives. Indigo disulfonate (I2S), one of the thermodynamic CRPs used in Chapter 4, was applied as a kinetic and capacity probe to characterize nZVI modified by carboxymethyl cellulose (CMC-nZVI), one of most widely used nZVI formulations in ISCR (Chapter 5). The results suggested that I2S is specific to Fe(0) at neutral pHs, providing an operationally easy way to quantify the Fe(0) content in groundwater impacted by nZVI injection. The reaction stoichiometry of I2S to Fe(0) is close to 1.5, suggesting complete oxidation of Fe(0) to Fe(III) because reduction of I2S is a two-electron process. Under the experimental conditions, the kinetics of I2S reduction was described by the second-order kinetic model. I2S was further applied to quantify the corrosion rate of CMC-nZVI in anaerobic water as a function of initial nZVI concentration and pH to simulate likely geochemical scenarios upon nZVI emplacement. The results revealed orders of magnitude faster Fe(0) loss than conventional nZVI, suggesting chemical transformation of CMC-nZVI is sufficiently fast that its “reductant demand” should be a major consideration in design on any application of CMC-nZVI in remediation.




Institute of Environmental Health


School of Medicine

Available for download on Sunday, June 03, 2018