Membranes ultrathin cellulose acetate

In 1966, Cadotte developed a method for casting mlcroporous support film from polysulfone, polycarbonate, and polyphenylene oxide plastics ( ). Of these, polysulfone (Union Carbide Corporation, Udel P-3500) proved to have the best combination of compaction resistance and surface microporosity. Use of the mlcroporous sheet as a support for ultrathin cellulose acetate membranes produced fluxes of 10 to 15 gfd, an increase of about five-fold over that of the original mlcroporous asymmetric cellulose acetate support. Since that time, mlcroporous polysulfone has been widely adopted as the material of choice for the support film in composite membranes, while finding use itself in many ultrafiltration processes. 

Figure 5.2 Schematic diagram of the float-casting of ultrathin cellulose acetate membranes.

The seminal discovery that transformed membrane separation fi-om a laboratory to an industrial process was the development in the 1960s of the Loeb-Sourirajan process to make defect-free ultrathin cellulose acetate membranes [1]. Loeb and Sourirajan were trying to use membranes to desalt water by reverse osmosis (RO). The concept of using a membrane permeable to water and impermeable to salt to remove salt from water had been known for a long time, but the fluxes of aU the membranes then available were far too low for a practical process. The Loeb-Sourirajan breakthrough was the development of an anisotropic membrane. The membrane consisted of a thin, dense polymer skin 0.2-0.5 pm thick sup-... 

Nodules are defined as spherical cells with a diameter of a few hundred angstroms that are compacted irregularly at the membrane surface. They can also be observed underneath the membrane surface when a cross-sectional picture is taken. Each nodule contains several tens of thousands of macromolecules. Schultz and Asunmaa were the first to report the observation of nodules on the surface of an ultrathin cellulose acetate membrane by electron microscope [1]. Figure 4.1 shows the picture taken by them. The nodular structure of the membrane surface is clearly seen with an average nodular diameter of 188 3 A. The same authors also took a picture of an asymmetric cellulose acetate membrane and found that it, too, had a nodular structure. Panar et al. [2] then observed the close monolayer packing of micelles with diameters from 400 to 800 A when a cross-sectional picture of an asymmetric aromatic polyamide-hydrazide membrane was taken (Fig. 4.2). The top monolayer covers a support layer where the spherical micelles are irregularly packed with void spaces of 75-100 A. They attributed the formation of the nodules to the micellar structure that was initially present at the surface of the polyamidehydrazide solution. 

The origin of thin-film-composite reverse osmosis membranes began with a newly formed research institute and one of its first employees, Peter S. Francis. North Star Research and Development Institute was formed in Minneapolis during 1963 to fill a need for a nonprofit contract research institute in the Upper Midwest. Francis was given the mission of developing the chemistry division through support, in part, by federal research contracts. At this time the Initial discoveries by Reid and Breton ( ) on the desalination capability of dense cellulose acetate membranes and by Loeb and Sourlrajan (,2) on asymmetric cellulose acetate membranes had recently been published. Francis speculated that improved membrane performance could be achieved, if the ultrathin, dense barrier layer and the porous substructure of the asymmetric... 

Ultrathin (about 400 A thick), polymer-blend membrane prepared from poly(styrenephosphonate diethyl ester) and cellulose acetate Size-quantized CdS particles generated in membranes... 

The first composite reverse osmosis membrane reported in the technical literature was developed by Peter Francis of North Star Research Institute in 1964 (4). This membrane was formed by float-casting an ultrathin film of cellulose acetate (CA) upon a water surface, removing the membrane from the water surface by lamination onto a pre-formed microporous support film and drying to bond the membrane to the support. This float-casting procedure has since been described in the technical literature for both flat sheet and tubular membranes ( 5, 6, T). 

Virtually the entire membrane manufacture today is based on laminate structures comprising a thin barrier layer deployed upon a much thicker, highly permeable support. Most are formed of compositionaUy homogeneous polysulfone, cellulose acetate, polyamides, and various fluoropolymers by phase inversion techniques in which ultrathin films of suitably permselective material are deposited on prefabricated porous support structures. Hydrophobic polymers as polyethylene, polypropylene, or polysulfone are often used as supports. A fairly comprehensive hst of microporous and ultrafiltration commercial membranes and produced companies are presented in Refs [107-109]. A review on inorganic membranes has been given in Ref. [110]. 

The first composite reverse osmosis membrane to be developed and described consisted of an ultrathin film of secondary cellulose acetate deposited onto a porous Loeb-Sourirajan membrane.3 The ultrathin film of cellulose acetate was fabricated by a water surface float-casting technique. This has been described to some extent in the published technical literature,4 5 and in considerable detail in several reports on government-funded research projects.3 6 Figure 5.2 illustrates this process schematically. 

This method was applied for in-situ formation of ultrathin (-3000 A) cellulose acetate (CA) phase inversion membranes on glassy carbon electrodes [26]. Aalami-Aleagha et al. [27] used the spraying method for the preparation of porous metallic membranes. This work focused on characterization of the metalhc membranes produced by this technique based on porosity, oxide content, and the pore size distribution. 

Cellulose nitrate and cellulose acetate (CA) were among the first asymmetric, reverse osmosis membranes to be produced [121]. Plummer et al. [122] described 13 specimen preparation methods for observation of CA membrane structures. They pointed out the lack of contrast in epoxy embedded sections and that one of the best stains, osmium tetroxide, reacts with the polymer. Freeze fractured membranes were found by these authors to be of questionable value. In our experience, if care is taken, SEM study of fractured membranes can provide an informative view of the structure even though some structures collapse, and their sizes cannot be accurately determined. A method found acceptable was ultrathin sectioning of gelatin embedded wet membranes (TEM). The structure of CA membranes was shown by replication [123] and SEM [124]. 

---