Environ. Sci. Technol. 2010, 44, 7443–7449
Colloid-Facilitated Transport of Cesium in Vadose-Zone Sediments: The Importance of Flow Transients TAO CHENG AND JAMES E. SAIERS* School of Forestry and Environmental Studies, Yale University, 195 Prospect Street, New Haven, CT 06511
Received February 4, 2010. Revised manuscript received August 18, 2010. Accepted August 20, 2010.
Colloid-sized particles are commonly detected in vadosezone pore waters and are capable of binding chemicals with sorptive affinities for geologic materials. Published research demonstrates that colloids are capable of facilitating the transport of sorptive contaminants under conditions of steady pore water flow, when volumetric moisture content and pore water velocity are constant. Less is known about the role of colloids in governing contaminant mobility under transientflow conditions, which are characteristic of natural vadosezone environments. The objective of this study is to elucidate the influences of flow transients on the mobilization and transport of in situ colloids and colloid-associated contaminants. We conducted column experiments in which the mobilization of in situ colloids and 137Cs was induced by transients associated with the drainage and imbibition of 137Cs contaminated-sediments. Our results demonstrate that substantial quantities of in situ colloids and colloid-associated 137Cs are mobilized as volumetric moisture content declines during porous-medium drainage and as volumetric moisture content increases during porousmedium imbibition. We also find that the colloid-effect on 137Cs transport is sensitive to changes in pore water ionic strength. That is, the quantities of colloids mobilized and the capacity of the these colloids to bind 137Cs decrease with increasing ionic strength, leading to a decrease of the mass of 137Cs eluted from the columns during porous-medium drainage and imbibition.
Introduction Mineral colloids are ubiquitous in soil waters and groundwaters, and they adsorb dissolved contaminants. The colloids may carry the bound contaminants as they migrate through the subsurface and thus serve as a vector for contaminant transport (e.g., refs 1–4). This process is often referred to as “colloid-facilitated contaminant transport” and is especially important for the movement of contaminants that have low solubility or strong affinity for solid surfaces. Such contaminants include radionuclides, transition metal ions, and nonpolar organics (e.g., refs 3–8). The majority of studies on colloid-facilitated transport have focused on water-saturated systems with the goal of advancing understanding of contaminant migration through geologic materials that lie beneath the water table (e.g., refs 1, 3, and 6–8). Laboratory based observations demonstrate that the strength of the colloid effect on contaminant mobility is determined by the pore water concentrations of colloids, * Corresponding author phone: (203) 432-5121; fax: (203) 4325023; e-mail:
[email protected]. 10.1021/es100391j
2010 American Chemical Society
Published on Web 09/02/2010
colloid transport characteristics, the capacity of the colloids to adsorb the contaminant, and the rate of contaminant desorption from the colloids (1, 3, 5–8). Fewer studies on colloid-facilitated transport have centered on water-unsaturated conditions (e.g., refs 4 and 5), despite the recognition that substantial quantities of colloids are generated in the vadose zone (9, 10) and that many contaminants enter drinking-water aquifers following delivery through the unsaturated soils of the vadose zone (11, 12). Experiments conducted under steady, unsaturated flow reveal that contaminant mobility declines with volumetric moisture content owing to a decrease in colloid mobility and an increase in contaminant desorption from mobile colloids (5, 13). Flow in shallow vadose-zone environments occurs in response to rainfall and snowmelt and is often transient or, in other words, characterized by temporal variations in volumetric moisture content and pore water velocity. These flow transients mobilize colloids in high concentrations and hence provide optimal conditions for colloid-facilitated transport of vadose-zone contaminants (14–22). Nevertheless, the manner in which flow transients and interactions between flow transients and physicochemical properties of the soil-water system affect colloid-facilitated contaminant transport has received little attention and hence is poorly understood. The objective of this study is to investigate the effects of the mobilization of naturally occurring mineral colloids on the transport of the radionuclide 137Cs through unsaturated geologic materials. This work extends our previous research on the mobilization and transport of mineral colloids (14) by elucidating the role of these colloids in facilitating the transport of a sorbing contaminant. Our study is distinguished from most published studies on colloidfacilitated transport through its focus on colloid and contaminant mobilization and transport under transient-flow conditions, rather than steady-flow conditions. Transient pore water flow occurs during soil imbibition, such as at the onset of rainfall, and soil drainage, such as upon the cessation of rainfall. Furthermore, we explore the movement of colloids and a contaminant that are in situ, rather than colloids that are pre-equilibrated with a contaminant and applied to the tops of soil columns. Our experiments, then, are suitable for making inferences on the mobilization, in addition to the transport and retention, of colloids and 137Cs and thus more closely mimic the phenomenon of colloid-facilitated transport as it occurs in real vadose-zone environments.
Experimental Section Overview. We measured the mobilization of in situ (not injected) colloids and 137Cs from 137Cs-contaminated sediments during porous-medium drainage and imbibition. Hanford Coarse Sand (HCS) was selected as the porous medium for these experiments because it has been previously characterized and used in published experiments on colloid transport (5, 8, 14–16, 21, 23). Although we used HCS, our experiments were not designed to mimic the hydrological conditions at Hanford, Washington, where infiltration rates and quantities are much lower than those tested in this study. Separate sets of duplicate experiments were conducted at pore water ionic strengths that spanned from 0.16 to 50 mM (Table 1) and that fall within the range of ionic strengths reported for natural soil waters. Although the experiments were characterized by periods of transient pore water flow, pore water ionic strength was held constant throughout the duration of each experiment. Experiments at an ionic strength of 2 mM were conducted in columns of two different lengths VOL. 44, NO. 19, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7443
TABLE 1. Experimental Conditions for the Column Tests: The Column Experiments Are Distinguished by (i) Pore Water Ionic Strength (I) and (ii) Column Length treatment ID I (mM) 1 2 3 4
0.16 2 50 2
column total pore Cs in the column length HCS dry volumea before drainage (cm) mass (g) (mL) (10-10 mol) 20.8 20.8 20.8 10.4
650 650 650 325
140 140 140 70
232 232 232 120
a The total pore volume includes the volume occupied by both pore water and pore air.
(Table 1) to evaluate the effects of colloid redeposition on the quantities of colloids and 137Cs delivered through the sand packs. Porous Medium. Uncontaminated (137Cs-free) Hanford Coarse Sand (HCS), collected from the 200E Submarine Pit of the DOE Hanford site, was used in this study as the vadosezone sediment. HCS is composed predominantly of quartz sand with lesser amounts of sodium/potassium feldspar, smectite, illite, chlorite, and kaolinite. The HCS has a median grain size of 1.7 mm, and mineral grains with dimensions between 1 mm and 5 mm compose 80% of its mass (see (23) for details). The HCS was air-dried, sieved through a 2 mm stainless-steel sieve, homogenized by manual mixing with a plastic spade, and stored in plastic buckets. The median grain size of the sieved sand was approximately 1.2 mm, based on the HCS grain size distribution information (23). Artificial Pore Waters. Three artificial pore waters of ionic strengths (I) equal to 0.16, 2, and 50 mM were made by adding the appropriate amount of NaCl to a 0.16 mM NaHCO3 solution, which was prepared using deionized water. 137Cscontaining pore waters were prepared by diluting a 137CsCl stock solution (Eckert & Ziegler Isotope Products, with a specific activity of 13.75 Ci/g Cs) to a total Cs concentration of 6.6 × 10-8 mol/L with the artificial pore waters. 137 Cs Contaminated HCS. Prior to initiation of the column experiments, the HCS was mixed with a 137Cs-containing artificial pore water (I ) 0.16, 2, or 50 mM) in an 8-L plastic container. The solid/solution ratio of the mixture equaled 1800 g HCS/L solution, and the Cs/HCS ratio was 3.66 × 10-11 mol Cs/g HCS. The sand-Cs mixture was stirred gently every a few hours and equilibrated overnight (18-20 h) before being emplaced in the laboratory columns. Following the overnight equilibration, supernatant samples were taken from the mixture and filtered through 0.2 µm pore-size nylon membrane filters (Acrodisc, Pall) to determine the dissolved 137Cs concentrations using a liquid scintillation counter. More than 99% of the Cs adsorbed to the HCS, as demonstrated by the low dissolved Cs concentration (
