Investigating Iron Mineral Transformation in Hydrocarbon Contaminated Sediments Using Detailed Mineral Magnetism

Abstract:

Continuous monitoring of the natural attenuation of hydrocarbons at contaminated sites using magnetic susceptibility (k) has been proposed as a means for observing iron mineral transformations. At the National Crude Oil Spill Fate and Natural Attenuation Research Site in Bemidji, MN, elevated k signals were found to be transient, thereby making long-term magnetic susceptibility monitoring uncertain. Furthermore, the iron mineral phases acting as transformation reactants and products associated with this long-term decrease in magnetic susceptibility remain largely unknown. To address these ambiguities, we collected detailed mineral magnetism measurements, such as hysteresis loops, backfield curves, isothermal remanent magnetizations, and first-order reversal curves (FORC) on core samples retrieved from the site and fresh magnetite packs installed at different depths within the contaminated and uncontaminated aquifer. Our results show that the magnetite packs display decreases in saturation magnetization (Ms) with time, where samples within the plume showed a 90% decrease compared to 18% for samples outside the plume over a twelve-month period. This loss in magnetization was accompanied by increases in bulk coercivity. In addition, low temperature magnetometry on the magnetite pack samples showed no evidence for goethite, hematite, iron sulfide, or siderite. The only expression during low temperature magnetometry was that of magnetite, as indicated by a prominent Verwey transition. However, samples within the plume showed gradual smearing of the Verwey transition and higher recovered remanence during thermal cycling which is indicative of maghemitization. Thus, we hypothesize that within and outside the plume, maghemitization is one of the processes occurring at this site. The precipitous decrease in saturation magnetization within the plume, as well as the absence of any other iron minerals, suggests that dissolution of magnetite may also be taking place particularly at the center of the plume. However, more laboratory investigations are needed to confirm these findings and reveal the complete nature of the hydrological processes involved and the paired oxidative and reductive biogeochemical reactions within a naturally attenuating hydrocarbon plume.