Environ. Sci. Technol. 2009, 43, 5769–5775
Measurement of Colloid Mobilization and Redeposition during Drainage in Quartz Sand J O N A T H A N W . B R I D G E , * ,† A. LOUISE HEATHWAITE,‡ AND STEVEN A. BANWART† Cell-Mineral Research Centre, Kroto Research Institute, North Campus, University of Sheffield, Broad Lane, Sheffield, S3 7HQ, U.K., and Centre for Sustainable Water Management, Lancaster Environment Centre, Lancaster University, LA1 4YQ, U.K.
Received February 27, 2009. Revised manuscript received June 18, 2009. Accepted June 19, 2009.
Movement of wetting and drying fronts through the vadose zone can mobilize and transport colloid particles but the mechanisms are not fully understood. We used mesoscale (mm-dm) fluorescence imaging to measure mobilization of 1.9 µm diameter carboxylate-latex microspheres during drainage in quartz sand. Experiments were performed at ionic strengths of 2-50 mM and drainage rates of 1.0-0.2 mL min-1. Colloids were mobilized and transported steadily at a sharp decrease in pore saturation marking the drying front. The mobilization rate varied directly with the initial immobile particle concentration. The mobilization rate constant varied inversely with ionic strength and directly with drainage rate. Peak mobile particle concentration at the drying front varied nonmonotonically, and the mobilization efficiency decreased with distance traveled by the drying front, at high ionic strengths. These findings constitute evidence for particle redeposition from the drying front as drainage progresses, which we propose is a key factor in the observed variations with ionic strength and drainage rate in the total number of particles removed during drainage. The measured outcomes of particle mobilization during a drainage event are sensitive to the distributions of immobile particles prior to drainage and dependent on the length scales over which the drainage event is observed.
Introduction Pore water flow in the vadose zone is characterized by cycles of wetting and drying. These transient conditions create intermittent variations in capillary pressure, pore flow velocity, pore water content (saturation), and the surface area of air-water interfaces (AWIs) within the soil. Colloid particles and microbes can be mobilized and transported by transient flows (1-8). Colloidal materials are environmentally important because they can act as vectors for sorbing contaminants (9) or are contaminants themselves (e.g., microbial pathogens). An understanding of the processes controlling their mobilization and transport under transient flows is critical to improving transport models and risk assessment for the vadose zone (10) and * Corresponding author phone: +44 (0)114 222 5798; fax: +44 (0)114 222 5701; e-mail:
[email protected]. † University of Sheffield. ‡ Lancaster University. 10.1021/es900616j CCC: $40.75
Published on Web 07/06/2009
2009 American Chemical Society
connected groundwater resources as well as in bioremediation engineering (11). The presence of the air phase plays a fundamental role in particle transport in partially saturated pores (12). Some studies (5, 13) have suggested that moving AWIs are a primary mechanism for particle mobilization and transport during transient flow. Gomez-Suarez et al. (14) and others (e.g., ref 2) have shown that movement of air bubbles across a surface can “scour” particles if the attractive capillary forces associated with the air interface exceed the attractive electrostatic and van der Waals interactions (DLVO forces, ref 3) associated with the solid surface. The dominance of capillary over DLVO forces leads to significant mobilization of particles trapped in thin water films (15) as these films thicken and disappear during wetting (3, 7). However during drying this process is unlikely to occur, as thin films and associated AWIs increase in extent during pore emptying. Crist et al. (16, 17) and others (18-20) have demonstrated that static AWIs act as sinks for particle deposition in unsaturated pores. Nevertheless particles are mobilized during drainage events (e.g., refs 4, 5) and the mechanisms responsible are not well understood. In this work, we used mesoscale time lapse fluorescence imaging (TLFI) (21, 22) to quantify the mobilization of colloid particles within quartz sand during drainage at several ionic strengths and drainage rates. TLFI enables simultaneous, independent time-series measurements of pore saturation and particle concentrations in millimeter-scale unit volumes throughout a decimeter-scale porous media domain. The objectives were (i) to make direct measurements of colloid mobilization and transport relative to drying front movement; (ii) to describe how particle mobilization contributes to cumulative particle transport at the centimeter-scale as drainage progresses; (iii) to provide observations to support a conceptual model of particle mobilization during drainage; and (iv) to identify the critical factors determining transport outcomes at macro-scales.
Experimental Procedure Ultraviolet-Fluorescence Imaging Apparatus. Fluorescence imaging is one of several experimental methods for mesoscale (millimeter-decimeter length scale) visualization of solute or colloid transport in translucent porous media (reviewed by ref 23). The ultraviolet (UV)-fluorescence imaging apparatus used here has been applied to investigate both saturated and unsaturated porous media, and was described in detail previously (21, 22). Briefly, a thin-bed flow chamber (10 cm width × 20 cm height × 0.67 cm depth) is filled with quartz sand and illuminated with UV light (400 nm) with an intensity proportional to their mass per unit volume (21, 22). This fluorescence is captured using a CCD camera (8-bit color, Hitachi-Denshi UK) at regular time intervals during flow through the sand pack. Materials. Thin-bed flow chambers were custom-built using Perspex and quartz glass as described by ref 21. The porous medium was 20/30 mesh (713 ( 23 µm diameter) Ottawa quartz sand (>99.8% SiO2) used as supplied with a porosity n of 0.35 ( 0.03 (22). The fluorescent solute (10-5 M disodium fluorescein aqueous solution adjusted to a range of ionic strengths with NaCl) and colloids (red-fluorescent 1.9 µm diameter latex microspheres with carboxylate surface groups) used were described in detail by ref 21. Both colloids and quartz sand were negatively charged at the ambient pH of the fluorescein solutions (5.9 ( 0.5, ref 21). Calculations VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5769
of interaction energies using extended DLVO theory (12) indicated significant repulsive energy barriers, i.e., unfavorable (24) conditions for both particle-particle aggregation and particle-sand attachment at ionic strengths
