The proposed research will both extend the PI's work on novel, adsorption-based methods to reduce fouling of microfiltration (MF) and ultrafiltration (UF) membranes, and explore a more general water treatment approach - dubbed microgranular adsorptive filtration (ìGAF) - that has emerged from the prior research. Microgranular adsorptive filtration is, essentially, the miniaturization of packed bed processes such as granular media filtration or GAC adsorption to a scale where the whole treatment process takes place in a thin (<500 µm) layer of powder-sized particles that have been pre-deposited on a membrane surface. In such processes, the membrane serves primarily as the support for the powder, rather than in its conventional role as the primary agent for pollutant removal.
The research will investigate three implementations of ìGAF: filtration to remove colloidal and particulate matter (i.e., miniaturization of a rapid sand filter); adsorption to remove trace inorganic and organic contaminants (miniaturization of an ion exchange bed or GAC contactor); and adsorption to remove NOM (replacement of an enhanced coagulation process with a miniaturized packed bed process).
The experimental procedures will consist of depositing a layer of the microgranular media on a membrane, passing water through the layer and membrane at a constant flow rate, and determining both the contaminant removal efficiency and the buildup of hydraulic resistance. At the end of a treatment step, the membrane will be backwashed to flush the media out of the system, and a new cycle will be initiated. Parameters to be varied include the thickness of the deposited layer, the hydraulic application rate, the length of the treatment step, and the backwashing time and intensity. Process performance will be assessed based on the contaminant removal efficiency, the pressure buildup during individual treatment steps, and the long-term stability of that pressure profile (i.e., the reversibility of any fouling that occurs). The potential applications of ìGAF are enormous, but have not been recognized in the past for two reasons. First, the tendency to focus on using a single technology to address a single water treatment goal has led to the use of membranes only as tools for contaminant removal. And second, until recently, the cost of membranes has made their use simply as supports for other media economically prohibitive. In prior research on the use of adsorbents to control membrane fouling, the PI found that the adsorbents were efficiently removing the contaminants that the membrane was intended to treat. That led to the realizations that (1) the most efficient treatment approach was to pack as much adsorbent on the membrane as possible, and to use the membrane strictly as a support for the adsorbent, and (2) ìGAF has many potential applications beyond fouling control. These applications, which include particle removal, trace contaminant removal, and pre-treatment of water prior to desalination, can potentially reduce the size, energy demands, and cost of numerous water treatment processes, while simultaneously making them more efficient. Preliminary tests of a few applications have been encouraging.
The intellectual merit of the proposed work derives from the understanding that it will promote of the behavior of packed layers on membranes in general, and adsorbent layers in particular. In the past, cake layers have been investigated almost exclusively in the context of how severely they foul membranes. This project will introduce the idea that cake layers can sometimes be enormously beneficial to membrane processes, and in the process will demonstrate how they can solve specific current problems associated with membrane fouling, desalination, and control of trace contaminants.
The most significant broader impact of the work will be the dissemination of the whole idea of microgranular adsorptive filtration. Several applications of ìGAF will be investigated in the research, but many others will undoubtedly be recognized and pursued by other researchers who have never previously thought about ìGAF as a realistic process. Beyond this, the research will advance the arsenal of possible treatment technologies for improving the quality of impaired water, will train new professionals in the field, and engage K-12 students in thinking about water supply and water quality.
In this research project, we investigated the performance of a new material (Heated Aluminum Oxide Particles, HAOPs) that has potential applications for removing contaminants from drinking water, especially when applied in conjuction with existing water treatment membranes. HAOPs were shown to remove much of the Natural Organic Matter (NOM) that is present in all drinking water sources and that is problematic because, if not removed, it can react with chlorine to form carcinogenic compounds. In addition, NOM is a major cause of membrane plugging (‘fouling’) when membranes are used in water treatment. We found that, if a thin layer of HAOPs is deposited on membranes before they are used in the filtration step, the NOM that causes fouling is very efficiently removed by the HAOPs before the water reaches the membrane, thereby preventing fouling. The ability of the HAOPs layer to collect NOM is apparent in the attached image, which shows the surface of the HAOPs layer at various times during a test. The HAOPs themselves are pure white, but as the test proceeds, they turn yellow, then golden, and ultimately deep brown at they collect more and more NOM. The benefits of NOM removal by HAOPs for membrane treatment processes are potentially very dramatic. At present, most ‘low-pressure’ membranes used for drinking water treatment (those used to remove microorganisms and other small particles, but not dissolved materials like salts) have to be cleaned at a frequency of about once per hour. The cleaning step removes the particles and NOM that have accumulated on the membrane surface. Without this step, the holes in the membrane become plugged, and much more energy is required to force water through the membrane than when it is clean. However, the cleaning step takes the membrane out of service temporarily, requires a substantial energy input itself, and increases wear and tear on the membrane as well as the entire mechanical system that runs the process. Our tests suggest that, in many cases, the use of HAOPs can reduce the required cleaning frequency to once per day or even once per several days. The research conducted in this NSF project demonstrated the technical features of HAOPs and the feasibility of using them to remove NOM from water and to dramatically reduce membrane fouling. The significance of the effort was recognized by the largest, drinking-water-focused professional organization in the US, the American Water Works Association, who selected the PhD dissertation describing the work as the top dissertation of 2012 in the water treatment field. In addition, the work has attracted interest from the Office of Naval Research, which is supporting follow-up efforts to demonstrate the process at a larger scale, and from private investors interested in commercializing it.
