Portal de Periódicos da CAPES

Hyperfiltration studies XIV Porous tubes precoated with filteraids as supports for dynamically formed membranes

1968; Elsevier BV; Volume: 5; Issue: 3 Linguagem: Inglês

10.1016/s0011-9164(00)80108-3

1873-4464

James S. Johnson, Kurt A. Kraus, Samuel M. Fleming, H. D. Cochran, J.J. Perona,

Membrane Separation Technologies

A surface favorable for dynamic formation of hyperfiltration membranes can be attained by thinly coating a coarsely porous material with particulate or fibrous substances. The coating is deposited by circulating a suspension of filteraids under pressure. The technique is illustrated with salt-filtering membranes formed dynamically of hydrous oxide and polymeric vinyl pyridine additives and with membranes formed from feed constituents of waste liquors generated in the sulfite pulping process. With this configuration, permeability may be restored by backwashing, after which the coating and membrane may be formed. We have earlier reported that salt-filtering films can be formed dynamically on porous bodies by circulating past them feeds containing certain additives(2/2-4). Membranes formed of polyelectrolytes, which might be expected to give ion-exchange layers, frequently have permeabilities which are extremely high in comparison with conventional detachable membranes — for example, cellulose acetate. Their salt rejection is lower than that of cellulose acetate, particularly at high feed concentrations. However, with brackish waters for which rejection requirements may not be high, they might have practical usefulness in the hyperfiltration, or reverse osmosis, desalination process. Membranes formed of non-electrolyte additives may have higher rejection, but usually seem to require supports of smaller and more homogeneous pores than those of polyelectrolytes; the layers probably need to be thinner for acceptable permeability (5). We have found that many solutions requiring treatment, for example, pulp-mill wastes (6) and sewage effluents (7), already contain membrane-forming materials which make it possible to eliminate much objectionable matter by hyperfiltration without the necessity of putting in additives. Realization of the practical possibilities of this class of membranes obviously requires the availability of suitable porous supports. The supports must be strong enough to contain the pressures necessary, must have pore sizes suitable for the membrane-forming material in question, and should be cheap, i.e., of low cost per unit production rate and of long lifetime. Pore sizes necessary for different additives or feeds vary, but usually fall in the tenth to a few microns range, and the pore sizes should be reasonably homogeneous. Even a small fraction of pores much larger than the median may be deleterious, if they present a gap too wide to be bridged by the additive, since a disproportionate fraction of the product stream will be feed which passed through the large pores unaffected. A favorable configuration would have a thin layer of the preferred pore size on which to lay the membrane, on top of a support having much larger pores to pass the product with minimum pressure drop. A procedure which one might predict would attain this configuration would be to circulate a slurry of fine particles under pressure past a body containing large pores. We have found that membranes can be formed on a surface layer deposited by this procedure. Combinations of many different porous materials with many different suspensions (commercially available as filteraids) have been effective. This procedure extends the usable pore sizes into those of commercial filters, of which there appear to be many more available from five micron pores up than in the range of sizes particularly suitable for membrane formation. Perhaps most important, membranes, along with the filteraid precoat, should be removable by backwashing, if permeability becomes undesirably low because of fouling or other reasons. The results obtained to date represent only a survey of the many possible porous bodies, filteraids, additives, and experimental conditions. It is improbable that the combinations so far tried are near the optimum for the membranes in question. We are encouraged enough by our results, however, to feel a reliminary report is warranted.