AmericanChemicalSociety.com

Student recruitment and training. Two undergraduate students were involved in the project during Year 2 and have contributed important findings and methodological improvements. Specifically, an Environmental Studies-Chemistry major worked on this project as his senior thesis research, which he presented and defended May 2008. Because he received academic credit, he was not supported by a PRF stipend. His work included several visits to Argonne National Laboratory’s Advanced Photon Source (APS) to collect microtomography images, visits that exposed him to cutting-edge technology that few undergraduates have the opportunity to use. He also worked at our home institution to refine an aqueous-phase miscible displacement experimental system, complete with online/real-time UV-Vis Spectrometric detection and system mass-logging. His work was awarded high honors by the Environmental Studies Program and honors by the Department of Chemistry and Biochemistry. A second student, a Computer Science major, provided much-needed computer programming expertise to further automate several steps in processing and analysis of the Synchrotron X-ray Microtomography images that my chemistry student and I collected over the course of the year. This work was the first time he was asked to apply his computer science coursework and training to scientific problems. He gained valuable experience communicating across disciplines as he carefully explained his approaches, as well as the challenges he encountered to an audience that had little programming experience.

Experimental Activities and Findings

Automated REV Analyses. Work during Award Year 2 included ongoing development of procedures for assessing whether the small volume imaged within the larger porous media system is sufficiently large to capture a representative elementary volume (REV) with respect to the air-water interfacial areas. We had previously used an expanding-cube REV method to a limited set of natural and model porous media to measure interfacial area REV and determine whether/how the REV depends on system properties (e.g., grain size, grain size distribution, water saturation, etc.). Although this approach was promising, we developed a newer REV analysis method that better reflects the choices of scale presented to the researcher as one plans imaging experiments. Specifically, samples are not cubes (as reflected in our earlier REV method), but rather are cylindrical to accommodate the rotational imaging procedure. The imaged sample height is also constrained by the beam height. Accordingly, our new approach to determining image-based REV utilizes a cylinder of fixed height with an expanding radius. This new method essentially asks the practical question researchers need to answer: what diameter cylindrical sample tube is large enough to yield representative data? This new REV method will be applied to a suite of previously collected images in the coming year.

Synchrotron X-ray Microtomographic Imaging. During Award Year 2, three visits (two with students) to Argonne National Laboratory’s Advanced Photon Source (APS) to conduct X-ray microtomography imaging of unsaturated natural sandy porous media. APS X-ray beamtime is awarded via a competitive proposal process. Images were collected utilizing the newer methods developed in Year 1 and focused on expanding our knowledge of porous media interfacial area-REV, as well as examine the contributions of capillary (menisci of partially filled pores) and adsorbed film contributions to total interfacial area. Specifically, early images were obtained for systems wetted to the desired moisture saturation (Sw) by mechanical mixing of media and water at the desired ratio. Although total interfacial areas obtained at a given Sw have been found to be independent of wetting method, it was anticipated that the relative contributions of film and capillary interfaces. Thus, during Year 2 all images were obtained for systems wetted by first saturating the sample and then draining to the desired Sw. Image analysis will continue in the coming year.

Miscible displacement tracer experiments. Experimental activities at our home institution yielded the most important findings on the project to-date. Real-time system mass logging was added to the aqueous-phase interfacial tracer experimental system in order to confirm and quantify what I and others had superficially acknowledged, that surfactant-induced decreases in surface tension may alter fluid distributions in unsaturated systems, thereby, may alter AI measurements. In fact, our recent experiments revealed that soil water saturation (Sw) decreased 26%, 25%, and 6% for systems originally at 71, 55, and 34% Sw, respectively, upon introduction of a surfactant pulse. For example, a sand system originally at 71% Sw was drained to 45% Sw simply through the action of surfactant-induced drainage. This behavior was reproducible, as shown in the TOC graphic for a series of surfactant-water cycles. Each time the surfactant concentration increases (lower line; secondary y-axis), column Sw decreases (upper line; primary y-axis); whereas, when the surfactant is eluted from the column, the water saturation rebounds. Our experimental surfactant concentration, and indeed the specific surfactant (sodium dodecyl benzene sulfonate, SDBS) used, are consistent with those routinely used for AI-measurements, so it is likely that this effect occurred in other experiments and was either not detected or not reported.

Although the trends shown in the TOC graphic are consistent with general knowledge of capillary forces, the magnitude of these effects was larger than anticipated. Our work has confirmed this behavior, but the specific ramifications of this behavior on AI estimates are less clear. Measurement of AI via the dynamic surfactant method involves measurement of its retardation factor (R), the ratio of its travel time to the travel time of a non-retained tracer. Conventionally, the increased travel time of the surfactant relative to the non-retained tracer is attributed to surfactant adsorption to the solid phase and to the air-water interface. The contribution from solid-phase sorption is directly measured and readily accounted for, whereas AI is estimated simply by the additional travel time in excess of that caused by solid-phase adsorption. This renders the AI-measurement susceptible to error should factors other than solid and interfacial adsorption influence surfactant travel time. Our future work seeks to examine these two interconnected mechanisms using a combination of experimental work and through a newly developed collaboration, numerical modeling of unsaturated surfactant flow.