Two Dimensional Transport Characteristics of Surface Stabilized Zero-valent Iron Nanoparticles in Porous Media S. R. KANEL, R. R. GOSWAMI, T. P. CLEMENT,* M. O. BARNETT, AND D. ZHAO Department of Civil Engineering, Auburn University, Auburn, Alabama 36849 Received July 18, 2007. Revised manuscript received November 13, 2007. Accepted November 15, 2007. Zero-valent iron nanoparticles (INP) were synthesized and stabilized using poly acrylic acid (PAA) to yield stabilized INP (S- INP). A two-dimensional physical model was used to study the fate and transport of the INP and S-INP in porous media under saturated, steady-state flow conditions. Transport data for a nonreactive tracer, INP, and S-INP were collected under similar flow conditions. The results show that unstabilized INP cannot be transported into groundwater systems. On the other hand, the S-INP can be transported like a tracer without significant retardation. However, the S-INP plume migrated downward as it moved horizontally in the physical model, indicating that small density gradients have significant influence on two- dimensional transport. The variable-density groundwater flow model SEAWAT was used to model the observed density- driven transport patterns. This is the first time a two- dimensional transport data set is reported for demonstrating the multidimensional transport characteristics of nanoparticles. The data shows the importance of density effects, which cannot be fully discerned using one-dimensional, column experiments. Finally, we also demonstrate that the numerical model SEAWAT can be used to predict the density-driven transport characteristics of S-INP in groundwater aquifers. Introduction Iron nanoparticles are used in a variety of areas for magnetic/ electronic, catalytic, and biomedical applications (1). In the environmental area, nanoscale iron materials have been widely researched to explore their potential for treating contaminated soil and groundwater (2). Among available iron nanoparticles, zero-valent iron nanoparticles (INP) have attracted significant interest due to their ability to reduce a variety of environmental contaminants. For example, INP have been found to degrade chlorinated hydrocarbons such as trichloroethene (TCE), tetrachloroethene (PCE), and carbon tetrachloride (3, 4). In addition, environmental contaminants such as perchlorate (5), nitrate (6), and metals such as Cr(VI) (7, 8), lead, nickel, mercury (2), arsenic (9, 10), and U(VI) (11) can be transformed using INP. The INP can also produce hydroxyl radicals in the presence of oxygen to oxidize a variety of organic contaminants such as carbothioate herbicide/molinate (12) and benzoic acid (13). Despite its high reactivity, the natural tendency of INP to aggregate, due to its magnetic properties (14), may severely limit our ability to be deliver INP into deep porous media formations (3). To overcome this limitation, various surface modification and particle stabilization strategies have been developed by using different types of additives such as surfactant (Tween-20) (15), poly acrylic acid (PAA) (16, 17), carboxymethyl cellulose (CMC) (18), cellulose acetate (19) starch (20), noble metals (21), and oil emulsions (22). A majority of these studies used batch experiments to dem- onstrate the additive’s potential to stabilize the INP. However, the transport dynamics of stabilized INP can only be tested under dynamic flow conditions. Only a few studies have explored the transport behavior of S-INP in soil columns. Schrick et al. (2004) studied PAA-stabilized INP and its reactivity in a glass burette (17); Kanel et. al (2007) studied surfactant (Tween 20) stabilized INP (15) and PAA-stabilized INP (16) in a sand column and in a glass-bead packed column, respectively. They also studied the reactivity of various forms of stabilized INP for removing arsenic species. All of the above INP transport studies were limited to one–dimensional analysis. To the best of our knowledge, transport of S-INP under two-dimensional flow conditions has not been re- ported in the literature. Furthermore, there have been no studies on numerical modeling of the observed transport characteristics of INP in groundwater systems. In this study, we hypothesize that two-dimensional physical models can be used to unravel the multidimensional transport dynamics of S-INP, which may be influenced by small density gradients. We use a novel experimental setup to demonstrate the importance of density effects while injecting nanoparticles into saturated aquifer formations. We compare the two-dimensional transport data of S-INP and INP plumes against a tracer plume to demonstrate the efficiency of the stabilization process. Finally, we use the numerical model SEAWAT to test whether the observed S-INP plume can be conceptually modeled as a density-driven conservative plume. Materials and Methods All the chemicals used in the experiments were reagent- grade. Chemicals such as NaBH 4 and PAA were obtained from Sigma-Aldrich Chemical Co. (Sigma-Aldrich, St. Louis, MO). Ferrous iron (FeSO 4 .7H 2 O), was obtained from Fisher Chemical Company (Fisher Scientific, Fairlawn, NJ). The porous media selected for this study was A-110 silica beads obtained from Potters Industries (Malverne, PA). The mean bead diameter was 1.1 mm with a variation of ( 0.1 mm. The porous medium properties were estimated using methods reported in our previous work (23). The average porosity of the packed system was estimated to be 0.385. The average hydraulic conductivity was estimated to be 1050 m/day from in situ flow and head measurements. The value of longitudinal dispersivity was estimated to be 1 mm from tracer experi- ments. Non reactive dye (FD&C Red 40) was used as an optical tracer in all the experiments. Transport characteristics of this dye have been verified in previous experiments where it has been used to visualize the movements of nonreactive solutes (23). Ultrapure (18 Ωcm) deionized water purified by a Barnstead purification system was used to prepare all nanoparticles suspensions. A two-dimensional flow container, shown in Figure 1, was used as the physical model to conduct experiments. The dimensions of the flow container are: 50 cm (length) × 2 cm (width) × 28.5 cm (height). Two chambers (5 cm wide) were * Corresponding Author e-mail: clement@auburn.edu. Environ. Sci. Technol. 2008, 42, 896–900 896 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 3, 2008 10.1021/es071774j CCC: $40.75  2008 American Chemical Society Published on Web 12/21/2007