Solute Rejection by Porous Glass Membranes

摘要

It was found that inclusion of 0.3 g/l AlCl3, 6H20 as a feed additive in hyper-filtration tests was sufficient to stabilise a porous glass membrane (batch A-1) with relatively high volume flux and modest solute rejection characteristics. One-tenth of this concentration was not sufficient. Further testing was directed toward treatment of porous glass membranes with concentrated aluminium chloride solution for a limited time, followed by hyperfiltration with a dilute feed additive. As little as 0.003 g/l AlCl3,6H20 stabilised batch A-1 for 80 h; while batch A-5, with lower volume flux and higher solute rejection characteristics, could be stabilised with no feed additive but only treatment with a concentrated aluminium solution every 100 h. Other experiments showed a similar equilibrium silica solubility, about 110 parts/million, in all solutions regardless of AlCl3, HCl, NaCl, or urea addition. The test which showed rates of silica dissolution revealed, however, that both low pH and aluminium lowered the rate of solution. The effectiveness of porous glass membrane stabilisation by aluminium treatment or reduced pH was also reflected in greatly reduced silica concentration of the effluent from hyper- filtration tests. Treatment of porous glass membranes with ferric and zirconyl chlorides was also tested, but aluminium chloride was the most effective in reducing silica loss. Chemical analysis of treated and tested membranes showed that aluminium remained in the pore structure in concentrations comparable to that of reactive silica hydroxyls. This is greater than the apparent concentration of ion exchange sites, and considerably less than total free hydroxyl concentration. Since suitable aluminium treatment or feed additive maintains membrane flux, rejection, and pore size characteristics, despite loss of soluble silica in the effluent, it appeared that added aluminium was retained in the membrane structure in amounts equivalent to the dissolved silica.