Measuring Colloidal Travel Through Sand Using Fluorescence

Hank Hogan

The poet William Blake evoked the image of seeing the whole world in a grain of sand, but some researchers instead use entire beds of sand to model the movement of contaminants and microbes in soil. Sand is less complex than soil; still, measuring the movement of colloids through the particles has been difficult.

UV light illuminates solute (green) and colloid (red) tracers passing through sand beds, which are used as a simple model for soil. Here, colloids from a point source form a narrow plume as they travel downward through saturated sand. Images courtesy of Jonathan W. Bridge, University of Sheffield.

Now researchers from the University of Sheffield and from Lancaster University, both in the UK, have used a novel fluorescence imaging method to answer some questions about the process of colloid deposition in sand beds. Their results support a previously unconfirmed hypothesis that, for dilute solutions in dry soils, the interfaces between air and water are relatively more important than others.

“Only at lower water content does the proportion of air-water interfaces become sufficiently large to contribute significantly to the total colloid removal,” explained Steven A. Banwart, a professor of environmental engineering science at Sheffield. Jonathan W. Bridge, also of Sheffield, and A. Louise Heathwaite of Lancaster also were on the team.

This schematic shows the fluorescence imaging system. Fluorescent tracers are injected into a sand bed, where UV light excites them. The resulting emissions are captured by a CCD camera, with filters that allow the emission from various fluorophores to be measured independently. The intensity of the red/green fluorescence is directly proportional to the mass of each fluorophore per unit volume of sand. By making calibrated measurements of the fluorescence intensity, researchers can measure the tracer concentration and the pore saturation under controlled conditions.

The scientists showed that the salinity of pore water affects the relative importance of air-water interfaces for colloid removal in partially saturated sandy soil. In high-saline conditions, which promote colloid deposition on the surfaces of soil grains, they found that the presence of air-filled pores had little effect on colloidal removal.

As described in the Nov. 15, 2007, ASAP Edition of Environmental Science and Technology, the group used Sylvania blacklight blue fluorescent tubes to provide excitation at 350 nm, with an additional minor peak at 470 nm. This source excited fluorescent tracers that they injected into the sand bed — one with a green emission that they added to a solute and another with a red emission that they injected with the colloid. For detection, they used a Hitachi color CCD camera with filters to separate out the fluorescence signals.

Both the light source and the detector were placed in front of a 250 × 150-mm thin-bed flow cell that had a quartz window facing the lamp and camera. They placed the setup in a darkened room. With the lamp on, the UV light penetrated the quartz and caused the tracers to fluoresce. The resulting emission was visible. From a vantage point 900 mm in front of the cell, the camera captured time-lapse images of mass distribution as the colloid was pulsed into the sand, as well as of the fluorescent solute concentration.

The researchers showed that the solute fluorescence intensity was proportional to the pore saturation, allowing them to determine the latter with a simple intensity measurement. They also showed that the fluorescence images revealed colloid removal efficiency directly, something they had not been sure would be possible.

“It wasn’t clear if we would obtain sufficient temporal resolution in the time-lapse imaging, compared to the expected rates of colloid removal,” Banwart said.

The key was to make a series of fluorescence measurements at many points in the sand before, during and after a pulse of colloid. This allowed the quantification of both mobile and deposited colloid concentrations directly from the data. The group achieved high spatial resolution because each camera pixel covered 0.478 mm² of the flow cell.

The researchers plan to apply the imaging technique to answer other questions. Ultimately, they would like to quantitatively image biological function, gene expression and process rates for living microbes in model soils.

Contact: Jonathan W. Bridge, University of Sheffield; e-mail: j.bridge@shef.ac.uk.