The dissertation is concerned with the effects of oil spills on the environment, using the 2010 Gulf of Mexico event, which motivated the selection of this research topic, as a case study. It examines the spill from both a social and biological assessment in which policy, technology, and economics direct oil spill response and remediation.

The dissertation is partly based on material collected during two years of fieldwork in Southern Louisiana. The Deepwater Horizon case study includes qualitative research grounded in a participatory action research (PAR) approach. The PAR strategy includes collective inquiry and experimentation through direct experience.

Although historically the effort to mitigate the effects of the Deepwater Horizon spill was the greatest cleanup response to an oil spill, the effort only affected 24% of the oil released in the Gulf. The fate of the remaining oil is unknown. Natural gas was not included in the spill discharge metric, nor will recovered oil (skimming and siphoning) be deducted from the fine that will be assessed on the responsible party, British Petroleum. Response strategies, such as the use of chemical dispersants and in-situ burning, did not remove oil, but instead contributed to the cumulative pollution in the environment. This case study revealed an opportunity to create legislation that motivates increased investment in technologies and response strategies that support the removal of the oil from the environment.

Trough the Deepwater Horizon case study, I also explored alternative spill response technologies and approaches to remediation and restoration. More than a dozen alternative technologies were evaluated and adopted during the 87-day oil spill event. The technologies evaluated included advancements for oil removal -- skimming and shoreline cleanup. Furthermore, for the first time, an oiled marsh was set aside for the purpose of conducting applied oil remediation and restoration research. Through a multi-institutional collaboration, we designed and implemented a restoration project on set-aside marsh in Louisiana. This project abandoned the use of cultivars and instead embraced genetically diverse, locally adapted plants for shoreline restoration. Included in the marsh project was a plant propagation innovation which utilized composted bagasse, a waste product of the Louisiana sugar cane industry, as a growth medium. The bagasse adds valuable organic material to the oil-impacted marsh and proved to be a viable propagation medium for smooth cordgrass (Spartina alterniflora) plants.

Additional soil remediation research, funded by the Chevron Corporation, investigated the use of vermiremediation for crude oil-impacted soils. Analysis of vermitea, the liquid extract from vermicompost, indicated the presence of biosurfactant producing hydrocarbonoclastic bacteria, allowing for the increased solubility of hydrophobic compounds adsorbed to soil. Additional research and field-scale experiments are required to optimize vermiremediation and demonstrate the potential for scaling and adoption.

My research supports the use of natural attenuation of oil-contaminated soil though the adoption of strategies which help to maintain the existing ecosystem. My research findings elucidate the critical limitations of current conventional oil spill response technologies and reveal the environmental tradeoffs that occur during response decision-making. The dissertation demonstrates the need for additional investment in technology innovation and for broader response strategies and preparation for future oil spills.

The abundance and reactivity of 2:1 phyllosilicate minerals (e.g. mica, vermiculite, smectite, illite, and pyrophyllite) have made them a subject of much research. The permanent structural and variable edge charge that these minerals possess provides a diverse set of modes through which surface reactions occur. Surface reactions due to permanent charge occur at the basal surface or in the interlayer of the phyllosilicate sheet. These reactions are well characterized and understood at multiple length scales in geochemistry (i.e. from the angstrom to the micron scale). This understanding arises from the detailed atomic-scale description of the bulk mineral structure and the origin of the permanent charge in isomorphic substitutions within the mineral structure. Variable charge arises from acid-base reactions at the edges of the 2:1 phyllosilicates sheets. These edge surface reactions are important in the retention of ions, stabilization of soil organic matter, rheological properties and colloidal behavior of clay minerals, and the dissolution kinetics of the 2:1 phyllosilicates. Despite the importance of the edge to the reactivity and other surface properties of these minerals, an atomistic description of the edge structure is lacking.

Periodic bond chain (PBC) theory identifies the dominant edges and nanoparticulate morphologies of 2:1 phyllosilicate minerals. Initial atomic structures of the 2:1 phyllosilicate edges and pyrophyllite nanoparticle defined by PBC theory were simulated using molecular dynamics (MD). The equilibrium edge structures reveal that the chemistry of the aluminum (Al) at the edge-water interface plays an important role in the inherent disorder of the 2:1 phyllosilicate edges. The edge Al atoms respond to increases in local surface charge by decreasing the number of coordinating oxygen atoms. The edge Al atoms assume four different polyhedral morphologies (i.e. octahedral, square pyramidal, trigonal bipyramidal, and tetrahedral Al polyhedra). The development of a trigonal bipyramidal Al structure indirectly creates an inverted Si tetrahedron that is consistent with a near-forgotten model of smectite structure. This non-planar and irregular structure of the mineral edge creates an interfacial water structure that is distinct from that of the 2:1 phyllosilicate basal surface. The water self-diffusion coefficient at the surface is two orders of magnitude less at the interface than in the bulk water. Monovalent cation complexes at the edge interface almost exclusively form inner-sphere surface complexes. These cationic edge surface complexes can form at any one of three non-equivalent vacancies at the 2:1 phyllosilicate edge. Dimeric surface complexes begin to form with increasing cation surface excess as the pH increases.

These MD simulations of the 2:1 phyllosilicate edges, nanoparticles, and cationic edge surface complexes demonstrate the complexity of the edge surface structure and reactivity. These insights from MD simulations provide support for interpretations of experimental results that heretofore could not be tested. In addition, these results demonstrate that a high-quality set of atomic interaction parameters derived from ab initio simulations of atomic clusters can be used to predict complex mineral structures.

Mercury (Hg) is an environmental contaminant that is distributed globally. In anoxic sediments, such as those found in estuaries and wetlands, divalent Hg [Hg(II)] is transformed by sulfate reducing bacteria into the biomagnifying toxicant monomethyl mercury. Incorporation of Hg(II) into the solid phase or reduction to zerovalent Hg [Hg(0)] render mercury unavailable to methylating bacteria. The goal of this dissertation is to determine whether the nanocrystalline iron sulfide mineral mackinawite (FeS), which is found in anoxic sediments, will limit Hg(II) bioavailability through adsorption or reduction. In this dissertation the interactions between Hg(II) and FeS are investigated in batch reactors as a function of pH (6 - 8) and reaction time (1- 24 hours) at environmentally relevant Hg:FeS ratios (2 × 10^-3 - 4 × 10^-4) using cold vapor atomic fluorescence spectrometry (CVAFS) and Hg L_III-edge extended X-ray absorption fine structure (EXAFS) spectroscopy. To gain further insight into the mechanism of interaction between Hg(II) and FeS, the FeS substrate is characterized using transmission electron microscopy (TEM), X-ray diffraction (XRD), and Fe L_{2,3-edge and K-edge and S K-edge X-ray absorption spectroscopy (XAS), providing the number and type of surface sites on FeS available to react with Hg.}

The composition of the FeS substrate used throughout this dissertation is characterized through measurement of the aqueous Fe(II) and S(-II) concentrations in suspensions of FeS as a function of time and these concentrations are compared to those predicted by an equilibrium model of FeS solubility. However, the solubility product constant of FeS reported in the literature varies widely, as do the type of aqueous Fe(II)-S(-II) complexes and their associated equilibrium constants. Thus, in order to construct an equilibrium model for FeS solubility, a critical review of the FeS solubility product constant and of the Fe(II)-S(-II) complexes is performed that informs a sensitivity analysis of the equilibrium model to the input equilibrium expressions. It is found that only one type of aqueous Fe(II)-S( II) complex reproduces the solubility of FeS as a function of pH as has been described by previous researchers, despite the numerous complexes that have been proposed. Furthermore, the equilibrium model is most sensitive to the choice of solubility product constant. The measured concentrations of aqueous Fe(II) are often found to be more than two standard deviations greater than the average of the predicted aqueous Fe(II) concentration, based on all solubility product constants, especially at pH 7 and pH 8. Additionally, the aqueous Fe(II) concentrations decrease more slowly than would be predicted by the kinetics of FeS precipitation. Lastly, the concentration of aqueous S(-II) is lower than predicted based on any of the solubility product constants for FeS. Thus, it is concluded that, despite the appearance of reflections in the X-ray diffraction patterns that correspond only to FeS, additional solid-phase Fe and S species are present in the suspensions.

Mackinawite (space group P4/nmm; a = b = 3.674, c = 5.033; α = β = γ = 90°) consists of edge-sharing FeS tetrahedra that form sheets that are stacked along the c-axis and bonded through van der Waals interactions between the S atoms on adjacent sheets. The particle size of FeS is fundamental to its reactivity, however it is poorly described. The average size of FeS particles is estimated as a function of pH (6 - 8), time (1 - 24 hours), and ionic strength (0.1 - 0.4 M) using a combination of Fe K-edge EXAFS spectroscopy, TEM and XRD. Transmission electron micrographs reveal FeS to have a sheet-like morphology and to be longer in the basal plane than in the direction of stacking. The particle sizes are found to be broadly distributed. X-ray diffraction yields the largest estimate of the average particle size (20 - 26 nm in the direction of the basal plane, 7.6 - 9.8 nm in the direction of stacking) and EXAFS spectroscopy is sensitive to the smallest particles (0.95 - 1.78 nm in the direction of the basal plane). Transmission electron micrographs exhibit particles that range in size from 9 - 39.1 nm (in the direction of the basal plane) and 2 - 5.6 nm (in the direction of stacking), with the average dimensions being 22.1 nm × 3.3 nm (basal plane × direction of stacking). Because each of the techniques employed is sensitive to different sizes within the distribution, in combination these techniques can be used to derive a particle size distribution. It was found that the distribution of particles was dominated by particles less than 2 nm (in the direction of the basal plane), which outnumbered larger particles by a factor of 1.5 - 13. The surface area of FeS particles is estimated to be 200 m²/g for the larger particles and 1000 m²/g for the smaller particles. The surface S(-II) site density is estimated to be 12 sites/nm², which, if the smaller surface area is applied, yields a surface S(-II) concentration of 0.47 mM.

The composition of the FeS substrate is examined using X-ray absorption spectroscopy to yield information on the oxidation state and surface speciation of the mineral as a function of pH (6 - 8), time (1 - 24 hours) and Hg(II) concentration (4 - 20 μM). The concentration of Fe(III), as determined by linear combination fitting of the X-ray absorption near edge structure (XANES) spectra, increases with increasing equilibration time between five minutes and 24 hours; as the FeS suspensions achieve steady state (between 5 and 24 hours), the concentration of Fe(III) reaches 10 - 30 %. Fitting of the EXAFS spectra reveal that the Fe(III) species is octahedrally coordinated to six O atoms and that a few (less than six) of these octahedra are linked together, sharing edges. No iron(III) oxides, such as ferrihydrite [Fe(OH)₃], goethite (α-FeOOH) or lepidocrocite (γ-FeOOH), are able to reproduce the sample EXAFS spectra. Furthermore, the measured aqueous Fe(II) and S( II) concentrations are inconsistent with the formation of crystalline iron(III) oxides. Lastly, the presence of Fe(III) corresponds to decreased ordering of FeS, which cannot be explained by the cancellation between Fe(II)- and Fe(III)-associated sine waves that comprise the EXAFS spectra. Thus, it is concluded that the Fe(III)-O species is associated with the FeS surface. The aqueous S(-II) concentrations in the FeS suspensions, which are diminished relative to those predicted to be in equilibrium with FeS, but elevated relative to those predicted to be in equilibrium with iron(III) oxides, indicate that S(-II) is oxidized in the FeS suspensions. The location of the S K-edge is shifted relative to the edges of other iron sulfide minerals, suggesting that oxidized S species, such as S(0) and polysulfides (S_n^2-), may exist in association with the FeS surface. Lastly, the S K-edge spectra exhibit a pre-edge peak that is not exhibited by the S K-edge spectra of macro-crystalline FeS. Because the presence of this feature is not related to a shift in the energy of the edge (none is observed), it is concluded that this pre-edge feature originates from a covalent, surface-bonding environment between S(-II) and Fe. Thus, although the surface of FeS can be approximated as consisting of unsaturated S atoms, which are available to adsorb metal ions, the surface also comprises oxidized species, including S(0), S_n^2-, and Fe(III)-O.

A fraction of the Hg(II) added to FeS suspensions is adsorbed; β-HgS does not precipitate. This conclusion is supported by the presence of covalently-bound surface sulfide groups and the high concentration of surface sulfide ligands (despite the presence of surface Fe(III)-O, S(0) and S_n^2-). Mercury(II) forms linear, trigonal and tetrahedral complexes with S(-II). Even though Hg(II)-S(-II) species, which are the dominant aqueous Hg(II) species in FeS suspensions and are present as surface complexes, are reduced only at high electron activities, the majority of the Hg(II) added to FeS suspensions is reduced to Hg(0), which forms a discrete phase. Although it is not possible to determine which species in FeS - Fe(II) or S(-II) - transfer electrons to Hg(II) because both Fe(III) and oxidized S species form in FeS suspensions, it can be concluded that neither aqueous Fe(II) or S(-II) on equilibrium with FeS are capable of reducing Hg(II)-S(-II) species. It is concluded that FeS will provide a mechanism for the generation of Hg(0) under the sulfidic conditions where other reduction mechanisms (microbial and photocatalytic) will not occur. Thus, FeS will remove Hg(II) from the aqueous phase, where it is available to methylating bacteria, through production of Hg(II)-S(-II) surface complexes and Hg(0).