(619c) High Performance Surface Nano-Structured Reverse Osmosis Membranes for Seawater and Brackish Water Desalination | AIChE

(619c) High Performance Surface Nano-Structured Reverse Osmosis Membranes for Seawater and Brackish Water Desalination

Authors 

Chen, Y. - Presenter, University of California, Los Angeles
Cohen, Y., UCLA
Reverse osmosis (RO) based desalination is the dominant technology for water desalination of seawater, brackish water and in water reuse applications. Although RO desalination is a mature technology, there are various challenges including perm-selectivity tradeoff of RO membranes, membrane fouling and mineral scaling. Accordingly, the present study presents an approach to address the above challenges via scalable membrane surface modification technology. The approach is based on atmospheric pressure plasma-induced graft polymerization (APPIGP) for surface nano-structuring (SNS) of polyamide (PA) RO thin film composite (TFC) membranes with a layer of tethered poly(acrylic acid) (PAA). The APPIGP approach comprises of surface activation with atmospheric pressure plasma (based on He) using an impinging plasma jet from a plasma unit on a robotic arm. The plasma head was set to scan the flat sheet base PA coupon membrane surface at a prescribed rate for a treatment exposure period of 1-2 seconds via single or multiple scans. Graft polymerization of the acrylic acid (AA) onto the surface activated PA membrane was then conducted in a reactor at 70oC over a period of about 1 hr (initial monomer concentration ([M]o) of 21 vol% and pH of 6). The resulting SNS-PAA-PA membranes were then tested with respect to permeability and salt rejection performance. The SNS-PAA-PA membrane exhibited both greater salt rejection (37.5-40% lower salt transport coefficient (B)), and higher water permeability (by 5-21%) relative to commercial RO membranes of the same range of salt rejection. The SNS-PAA-PA membrane also demonstrated excellent removal of nitrate, boron, arsenite, and arsenate with corresponding rejections of 98.0, 86.8, 94.5 and 99.4% which were higher (by 1-10%) relative to the base PA membrane.

Fouling stress tests were conducted with bovine serum albumin (BSA) and sodium alginate as model foulants, whereby flux decline was monitored for operation at a constant transmembrane pressure and initial flux that was identical for all fouling tests. The SNS-PAA-PA membranes demonstrated lower fouling propensity relative to commercial membranes and with flux recovery of 96.3-100% upon simple freshwater flush. Membrane mineral scaling tests were also conducted with calcium carbonate and calcium sulfate as model mineral scalants. These scaling tests were performed with model feed solution and under operating conditions such that the saturation indices for gypsum and calcite at the membrane surface were about 2 and 14, respectively. Reduced scaling propensity, quantified via flux decline (for operation at a prescribed transmembrane pressure), was demonstrated for the SNS-PAA-PA membrane. These tests revealed about 11% and 15% flux decline for the SNS-PAA-PA membrane with model foulants of gypsum and calcium carbonate, respectively, relative to ~15-19% and 17-21% flux decline for the commercial membrane having similar range of permeability and salt rejection. The SNS-PAA-PA membrane exhibited ~94-100% permeability recovery upon a simple freshwater flush, up to 8.8% improved relative to the Base-PA membrane. It is also noted that while salt rejection for the commercial PA membranes decreased by ~1.5-5.7% and 4.4-13.5% after the Gypsum and Calcium Carbonate scaling tests, respectively, while the SNS-PAA-PA membrane retained its salt rejection performance. The results of the present study and potential scalability of APPIGP suggest that there is merit in further optimization of the SNS approach for the development of high-performance RO elements for seawater and brackish water desalination.