(189e) The Impact Of Surface Topology On Scaling Propensity Of Polyamide Membranes

Membrane mineral salt scaling is one of the major factors that limit recovery in reverse osmosis desalination of brackish groundwater. Mineral salt scaling leads to flux decline and shortening of membrane useful life. The primary mineral salt scalants that are typically of concern in inland water desalination are calcium carbonate, calcium sulfate and barium sulfate, typically in the calcite, gypsum and barite forms, respectively. In order to develop predictive models of mineral scale development on RO membranes, there is a need to first assess the importance of membrane topology and surface chemistry on mineral salt scaling. Such information could also be beneficial to developing optimal criteria for designing RO/NF membranes of low scaling propensity. Accordingly, a systematic surface crystallization study was carried out, with gypsum as the model scalant, to evaluate the kinetics of surface crystallization on polyamide membranes of different surface topologies. The study was carried out using a quartz crystal microbalance (QCM) with polyamide membranes that were interfacially polymerized onto the surface of quartz crystals. With the developed methodology one can create differing surface topologies for the same polyamide membrane surface chemistry, thereby enabling isolation of the impact of membrane topology on surface crystallization. To examine the impact the surface topology, membrane surface roughness was varied through various pH adjustments of the underlying polyelectrolyte adhesion layer. The kinetics of surface crystallization was then studied in a QCM flow cell in which the accumulation of surface mineral mass was continuously monitored. The study was carried out with supersaturated feed solutions (with respect to gypsum), prepared by mixing the ionic ingredients just prior to delivery to the QCM flow cell in order to avoid bulk crystallization in a standing feed solution. The feed solution (in all experiments) was maintained at a supersautarion index of ~1.5 for which the bulk crystallization induction time was of the order of 9 hr. Therefore, bulk crystallization was avoided in this system given the much shorter convective residence time of the solution in the system (< 1 min), relative to the bulk crystallization induction time. As a consequence, crystallization was restricted to surface crystallization in the QCM flow cell. Imaging of the QCM surfaces by SEM and AFM provided detailed information on the surface crystal number density, crystal size and morphology. This information was used to examine the correlation between surface topology and gypsum surface crystallization kinetics. The kinetics of surface crystallization (i.e., time evolution of gypsum surface scale) was also analyzed using a single crystal growth model along with a population growth model. This information provided, for the first time, a direct measure of surface crystallization kinetics on RO membranes. The results of the study demonstrated that surface topology can have a significant impact on surface crystallization suggesting that there is merit in developing approaches of optimization membrane surface topology so as to reduce scaling propensity.