Water composite membranes

Interfacial polymerization membranes are less appHcable to gas separation because of the water swollen hydrogel that fills the pores of the support membrane. In reverse osmosis, this layer is highly water swollen and offers Httle resistance to water flow, but when the membrane is dried and used in gas separations the gel becomes a rigid glass with very low gas permeabiUty. This glassy polymer fills the membrane pores and, as a result, defect-free interfacial composite membranes usually have low gas fluxes, although their selectivities can be good. 

The first reverse osmosis modules made from cellulose diacetate had a salt rejection of approximately 97—98%. This was enough to produce potable water (ie, water containing less than 500 ppm salt) from brackish water sources, but was not enough to desalinate seawater efficiently. In the 1970s, interfacial composite membranes with salt rejections greater than 99.5% were developed, making seawater desalination possible (29,30) a number of large plants are in operation worldwide. 

An excellent review of composite RO and nanofiltration (NE) membranes is available (8). These thin-fHm, composite membranes consist of a thin polymer barrier layer formed on one or more porous support layers, which is almost always a different polymer from the surface layer. The surface layer determines the flux and separation characteristics of the membrane. The porous backing serves only as a support for the barrier layer and so has almost no effect on membrane transport properties. The barrier layer is extremely thin, thus allowing high water fluxes. The most important thin-fHm composite membranes are made by interfacial polymerization, a process in which a highly porous membrane, usually polysulfone, is coated with an aqueous solution of a polymer or monomer and then reacts with a cross-linking agent in a water-kniniscible solvent. 

Early ultrafiltration membranes had thin surface retentive layers with an open structure underneath, as shown in Fig. 20-62. These membranes were prone to defects and showed poor retention and consistency. In part, retention by these membranes would rely on large retained components in the feed that polarize or form a cake layer that plugs defects. Composite membranes have a thin retentive layer cast on top of a microfiltration membrane in one piece. These composites demonstrate consistently high retention and can be integrity-tested by using air diffusion in water. 

For this reason much work has been done at the ALZA Corporation and elsewhere to increase the water permeation rates by various technologies. For example, ALZA scientists utilized a composite membrane in the development of their first commercial product with this technology [21,22], In this system they first applied a CA membrane containing a high concentration of porosigens. A second dense membrane containing only CA was added. In this way the overall fluid permeability was increased, since the thickness of the dense portion of the film could be proportionately reduced. 

B.-B. Li et al. [64] have studied the separation of EtOH-H20 solutions by pervaporation (PV) using chitosan (CS), poly (vinyl alcohol)-poly(acrylonitrile) (PVA-PAN) and chitosan-poly(vinyl alcohol)/poly(acrylonitrile) (CS-PVA/PAN) composite membranes. It was found that the separation factor of the CS-PVA/PAN composite membrane increased with an increase of PVA concentration in the CS-PVA polymer from 0 to 40 wt%. With an increase in the membrane thickness from 12 to 18 pm, the separation factor of the CS-PVA/PAN composite membrane increased and the permeation flux decreased. With an increase of ethanol-water solution temperature, the separation factor of the CS membrane decreased and the permeation flux of the CS membrane increased while the separation factor and the permeation flux of PVA/PAN and CS-PVA/PAN composite membranes increased. 

Another possible solution to the problem of analyzing multiple-layered membrane composites is a newly developed method using NMR spin-lattice relaxation measurements (Glaves 1989). In this method, which allows a wide range of pore sizes to be studied (from less than 1 nm to greater than 10 microns), the moisture content of the composite membrane is controlled so that the fine pores in the membrane film of a two-layered composite are saturated with water, but only a small quantity of adsorbed water is present in the large pores of the support. It has been found that the spin-lattice relaxation decay time of a fluid (such as water) in a pore is shorter than that for the same fluid in the bulk. From the relaxation data the pore volume distribution can be calculated. Thus, the NMR spin-lattice relaxation data of a properly prepared membrane composite sample can be used to derive the pore size distribution that conventional pore structure analysis techniques... 

Of the existing flat-sheet RO membranes, cellulose acetate membranes of the Loeb-Sourirajan type give the best results because their open microporous substrate minimizes internal concentration polarization. Conventional interfacial composite membranes, despite their high water permeabilities and good salt rejections, are not suitable for PRO because of severe internal concentration polarization. 

Limited testing on chlorine sensitivity of poly(ether/amidel and poly(ether/urea) thin film composite membranes have been reported by Fluid Systems Division of UOP [4]. Poly(ether/amide] membrane (PA-300] exposed to 1 ppm chlorine in feedwater for 24 hours showed a significant decline in salt rejection. Additional experiments at Fluid Systems were directed toward improvement of membrane resistance to chlorine. Different amide polymers and fabrication techniques were attempted but these variations had little effect on chlorine resistance [5]. Chlorine sensitivity of polyamide membranes was also demonstrated by Spatz and Fried-lander [3]. It is generally concluded that polyamide type membranes deteriorate rapidly when exposed to low chlorine concentrations in water solution. 

Table 6 lists several of the salient properties of this new composite membrane. When salt rejection was evaluated at different pressures in simulated seawater trials, potable water (containing less than 500 ppm dissolved salts) was generated at as low as 600 psi, with very good flux (12 gfd) at that pressure. In spiral-wound membrane element trials on actual 33,000 ppm seawater, potable water was obtained even at 500 psi, albeit at low flux. These results surpass by far the capabilities of any of the "NS" series of membranes. 

This thin-film-composite membrane has been found to have appreciable resistance to degradation by chlorine in the feed-water. Figure 2 illustrates the effect of chlorine in tap water at different pH values. Chlorine (100 ppm) was added to the tap water in the form of sodium hypochlorite (two equivalents of hypochlorite ion per stated equivalent of chlorine). Membrane exposure to chlorine was by the so-called "static" method, in which membrane specimens were immersed in the aqueous media inside closed, dark glass jars for known periods. Specimens were then removed and tested in a reverse osmosis loop under seawater test conditions. At alkaline pH values, the FT-30 membrane showed effects of chlorine attack within four to five days. In acidic solutions (pH 1 and 5), chlorine attack was far slower. Only a one to two percent decline in salt rejection was noted, for example, after 20 days exposure to 100 ppm chlorine in water at pH 5. The FT-30 tests at pH 1 were necessarily terminated after the fourth day of exposure because the microporous polysul-fone substrate had itself become totally embrittled by chlorine attack. 

In summary, the FT-30 membrane is a significant improvement in the art of thin-film-composite membranes, offering major improvements in flux, pH resistance, and chlorine resistance. Salt rejections consistent with single-pass production of potable water from seawater can be obtained and held under a wide variety of operating conditions (ph, temperature, pressure, and brine concentration). This membrane comes close to being the ideal membrane for seawater desalination in terms of productivity, chemical stability, and nonbiodegradability. 

Petersen, R.J. Larson R.E. Majerle, R.J. "Development of the FT-30 Thin-Film Composite Membrane for Desalting Applications," Technical Proceedings, 8th Ann. Conf. National Water Supply Improvement Assn., San Francisco, CA, July 6-10, 1980. 

Research effort at Albany International Research Co. has developed unit processes necessary for pilot scale production of several species of reverse osmosis hollow fiber composite membranes. These processes include spin-dope preparation, a proprietary apparatus for dry-jet wet-spinning of microporous polysul-fone hollow fibers, coating of these fibers with a variety of permselective materials, bundle winding using multifilament yarns and module assembly. Modules of the membrane identified as Quantro II are in field trial against brackish and seawater feeds. Brackish water rejections of 94+% at a flux of 5-7 gfd at 400 psi have been measured. Seawater rejections of 99+% at 1-2 gfd at 1000 psi have been measured. Membrane use requires sealing of some portion of the fiber bundle for installation in a pressure shell. Much effort has been devoted to identification of potting materials which exhibit satisfactory adhesion to the fiber while... 

Fig. 1. Water flux and NaCl rejection of several membrane types (10), where (D) represents seawater membranes, which operate at 5.5 MPa and 25°C ( ), brackish water membranes, which operate at 1500 mg/L NaCl feed, 1.5 MPa, and 25°C and (SSI) nanofiltration membranes, which operate at 500 mg/L NaCl feed, 0.74 MPa, and 25°C. A represents cellulose acetate—cellulose triacetate B, linear aromatic polyamide C, cross-linked polyether D, cross-linked fully aromatic polyamide E, other thin-film composite membranes F, asymmetric membranes G, BW-30 (FilmTec) H, SU-700 (Toray) I, A-15 (Du Pont) J, NTR-739HF (Nitto-Denko) K, NTR-729HF (Nitto-Denko) L, NTR-7250 (Nitto-Denko) M, NF40 (FilmTec) N, NF40HF (FilmTec) O, UTC-40HF (Toray) P, NF70 (FilmTec) Q, UTC-60 (Toray) R, UTC-20HF (Toray) and S, NF50 (FilmTec). To convert MPa to psi,...

---