Article pubs.acs.org/est
Size-Dependent Reactivity of Magnetite Nanoparticles: A FieldLaboratory Comparison Andrew L. Swindle,*,⊥ Andrew S. Elwood Madden,† Isabelle M. Cozzarelli,‡ and Mourad Benamara§ ⊥
Department of Geology, Wichita State University, Wichita, Kansas 67260, United States School of Geology and Geophysics, University of Oklahoma, Norman Oklahoma 73019, United States ‡ U.S. Geological Survey, Reston, Virginia 20192, United States § Nano-Bio Materials Characterization Facility, University of Arkansas, Fayetteville, Arkansas 72701, United States †
S Supporting Information *
ABSTRACT: Logistic challenges make direct comparisons between laboratory- and field-based investigations into the size-dependent reactivity of nanomaterials difficult. This investigation sought to compare the sizedependent reactivity of nanoparticles in a field setting to a laboratory analog using the specific example of magnetite dissolution. Synthetic magnetite nanoparticles of three size intervals, ∼6 nm, ∼44 nm, and ∼90 nm were emplaced in the subsurface of the USGS research site at the Norman Landfill for up to 30 days using custom-made subsurface nanoparticle holders. Laboratory analog dissolution experiments were conducted using synthetic groundwater. Reaction products were analyzed via TEM and SEM and compared to initial particle characterizations. Field results indicated that an organic coating developed on the particle surfaces largely inhibiting reactivity. Limited dissolution occurred, with the amount of dissolution decreasing as particle size decreased. Conversely, the laboratory analogs without organics revealed greater dissolution of the smaller particles. These results showed that the presence of dissolved organics led to a nearly complete reversal in the size-dependent reactivity trends displayed between the field and laboratory experiments indicating that size-dependent trends observed in laboratory investigations may not be relevant in organic-rich natural systems.
■
INTRODUCTION Research investigating nanoparticles, solid materials having physical dimensions ranging from a few nm to about 100 nm, has increased over the past several years and is projected to continue to grow in the future.1,2 Much of the interest in nanoparticles is driven by the fact that physical and chemical properties of these materials often vary as a function of particle size.3 Properties that have been shown to vary based on particle size include solubility, conductivity, magnetic properties, and aggregation.4,5 The cause of these variations is at least partially due to the fact that as the dimensions of a mineral particle are reduced, a greater percentage of the atoms exist on the surface of the particle, giving nanoparticles a very high surface area to volume ratio.6 Adding to the intrigue of nanoparticles is that among some materials, properties not only vary from the nanoto the bulk scale, but also vary continuously as a function of particle size, and lead to changes in reactivity that cannot be explained through surface area alone. Size-dependent mineral reactivity has been reported for a number of processes such as dissolution,7−9 redox reactions,10−14 sorption,15−18 crystal growth and phase transformation,19 colloidal transport,20,21 or even the effect of nanoparticle solubility on the saturation state of other system phases.22 These studies illustrate not only that properties of a given mineral can change as the mineral changes sizes in the nanoscale, but also that all other things being equal, © XXXX American Chemical Society
the smaller particle is expected to be the most reactive. This relationship between particle size and reactivity suggests that the chemical behavior of a given nanoparticle in the environment can depend significantly on its size. Additionally, the potential to maximize the desired property of a given particle by altering its size has led to the study of nanoparticles for use in a number of industrial applications, but raises concerns about the bioavailability and bioaccumulation of these materials. Mixed-valence nanoscale iron oxides, which tend to have high surface areas and sorption capacities, have been suggested as potential reductants for a number of environmental contaminants such as heavy metals, radionuclides, and chlorinated aliphatics.23−27 Magnetite is of particular interest in this regard as it possesses an intrinsic mechanism by which electrons can be shuttled from structural ferrous iron atoms to the particle surface, potentially allowing for greater contaminant remediation than ferric oxides such as goethite, which have no equivalent mechanism to reduce contaminants.28 Indeed, laboratory investigations have shown nanoscale magnetite to Received: January 14, 2014 Revised: July 25, 2014 Accepted: September 9, 2014
A
dx.doi.org/10.1021/es500172p | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
be a viable reductant of a number of contaminants,24,28−30 and that the capacity of magnetite to sorb and/or reduce various contaminants generally increases as particle size decreases.23,31 However, in many remediation settings, the contaminant of concern may be present in concentrations significantly below those of naturally occurring inorganic species such as sodium, magnesium, calcium, chloride, bicarbonate, and sulfate. The potential presence of dissolved and/or particulate organic compounds along with the activity of a microbiological community can further complicate matters in a field setting. Depending on the geochemistry of the system, organic compounds of varying chemical characteristics can occupy reactive surface sites on mineral particles, act as a catalyst for microbial reduction of ferric iron, or even directly reduce surficial ferric iron leading to particle destabilization due to an excess of ferrous iron.32−34 While synthetic magnetite nanoparticles have been extensively studied in the laboratory, how the size-dependent reactivity of the particles will manifest under field conditions is not well understood. The objective of this study was to investigate whether or not size-dependent reactivity of mineral nanoparticles in a field experiment was consistent with laboratory analog results using the specific example of magnetite dissolution. For the field experiment, magnetite nanoparticles were emplaced in groundwater monitoring wells at the U.S. Geological Survey (USGS) research site at the Norman Landfill, while groundwater data collected during the field investigation was used to prepare synthetic groundwater for the laboratory analogs. These results will help elucidate how the size-dependent reactivity of nanoparticles manifests in a field setting, which in turn will increase our understanding of how these materials will behave in the subsurface.
landfill are sulfate-reduction and methanogenesis, which occur simultaneously in portions of the leachate plume.39,41 Iron in the landfill alluvium predominantly occurs as nanoscale hematite occluded from groundwater by clays,42,43 thus ironreduction is not favored over sulfate-reduction and the two TEAPs occur simultaneously in certain areas of the site.37,39,44 Historical groundwater data show that dissolved iron concentrations in the leachate plume exceed those measured in background wells, indicating that iron within the plume is being mobilized.45 Field Experiments. In order to install and then later retrieve nanoparticles from the subsurface, subsurface nanoparticle holders (SNHs) were designed and constructed (Figure 1). Each SNH is composed of chemically inert Techtron
Figure 1. Photograph of a subsurface nanoparticle holder (SNH) designed for use in this project. Arrows indicate the eight recessed compartments designed to hold TEM grids.
■
METHODS All water used in these experiments, both to prepare solutions and to rinse and store particles, was degassed ultra pure (UP, 18.2 MΩcm) water prepared by boiling UP water while sparging with ultrapure N2 gas for a minimum of 30 min. The water was then removed from the heat source, capped and transferred to an anaerobic chamber (95%/5% N2/H2). All chemicals used in these experiments were reagent grade. Particle Synthesis. Magnetite particles of three distinct size ranges (90 (±42) nm, 44 (±11) nm, and 6 (±2) nm, see Supporting Information (SI)) were synthesized using established methods35,36,23 as detailed in the SI. Following synthesis, magnetite particles were magnetically separated, washed a minimum of three times to remove excess salts, and then stored in an anaerobic chamber. Norman Landfill Research Site. The Norman Landfill is an environmental research site operated by the USGS under the Toxic Substances Hydrology Program since 1995.37−39 The site, located in Norman, Oklahoma, originally operated as a municipal waste landfill from approximately 1922 to 1985 when the site was closed. The Norman Landfill site has been studied for over 30 years and the hydrologic, geologic, and geochemical systems at the site have been well-characterized.40 The Norman Landfill is located on alluvial sediments along the Canadian River in central Oklahoma (SI). A leachate plume extends approximately 250 m downgradient from the landfill toward the river. Microbial degradation in the leachate plume has resulted in the formation of distinct biogeochemical zones as available terminal electron acceptors (TEAs) are depleted.37,39 The dominant TEA processes (TEAPS) at the
polyphenylene sulfide and allows a total of eight transmission electron microscopy (TEM) grids to be suspended in a groundwater monitoring well. Each TEM grid fits into a recessed compartment and is secured in place by a cover attached via several stainless steel machine screws. Conical openings in the holder and cover align with each of the grids so that both grid surfaces are exposed to the groundwater. A hole at the top of each SNH allows for a cable to be attached so that the devices can be lowered into, and retrieved from, groundwater monitoring wells. The SNHs have dimensions of 0.4″ × 0.4″ × 6.0″, as to be compatible with the 1″ OD monitoring wells at the Norman Landfill. A total of four SNHs were used for the field experiments. Each SNH was loaded with two TEM grids (copper grid with Formvar/carbon support film) of each particle size as well as two blank grids as control. Magnetite particles were attached to the TEM grids by mixing each of the washed particle slurries with isopropyl alcohol in a 1:1 ratio. A 7 μL aliquot of the slurry-alcohol mixture was then deposited on the TEM grid and allowed to dry for ∼30 s, after which a Kimwipe was used to remove the remaining solution. The grids were loaded into the SNHs in an anaerobic chamber and then transported to the site in a nitrogen-flushed plastic bag. A cluster installation of four temporary monitoring wells (SL35-A−D) was installed approximately 10 m southeast of existing landfill monitoring well MLS35−5 (SI). The SNHs were emplaced such that they spanned the 15.24 cm screened interval of the wells. The SNHs were then retrieved incrementally, at 5 days, 10 days, 21 days, B
dx.doi.org/10.1021/es500172p | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
width and area of individual particles in TEM images using ImageJ64 software (Figure 2). As will be discussed further in
and 30 days. After extraction, a Kimwipe was used to adsorb excess moisture from the TEM grid surfaces before the SNHs were placed in a nitrogen-flushed plastic bag. The SNHs were then transported to the anaerobic chamber where the TEM grids were removed, air-dried in the anaerobic chamber, and stored pending analysis. Groundwater Samples. Groundwater samples were collected from the cluster of temporary monitoring wells prior to, and at the conclusion of the field experiments. Groundwater samples were collected and analyzed as described in Cozzarelli et al. (2011) (SI). Groundwater parameters collected prior to the field experiments were used to confirm that the temporary wells were screened within the same biogeochemical zone (SI). Additionally, the combined data sets were used to prepare synthetic groundwater for the laboratory analog experiments. The groundwater is circumneutral (pH 6.8), and high in bicarbonate (28.9 mM), sodium (21.0 mM), calcium (4.4 mM), magnesium (4.0 mM), chloride (11.1 mM), and ammonium (4.8 mM), consistent with a leachate release. Full results and discussion of the groundwater sampling data are available in the SI. Synthetic Groundwater and Dissolution Experiments. Triplicate laboratory analog dissolution experiments were performed using synthetic groundwater based on the data collected during the field experiments. As the groundwater analysis revealed numerous cations and anions present in low concentrations (
