The sorption of methyl and ethyl alcohol by silica gels

Abstract

The original object of the present research was to investigate the sorption of methyl alcohol by silica gel, in order to compare the isothermals with those of water and ethyl alcohol, which had been determined by Lambert and Foster. Owing to the fact that the supply of gel used for this earlier work had been exhausted, it as necessary to examine a number of other specimens. One of these, which like the original gel, was a commercial product, was found to take up a larger amount at saturation, and also to exhibit hysteresis phenomena which were not observed with the original gel. Unfortunately it was impossible to find a sample giving exactly the same shape of isothermal as the original gel, although some specimens known to belong to the same batch gave very similar curves. One of these was selected for the present work, and an examination has been made of the isothermals of methyl and ethyl alcohols on these two types of silica gel at 25° C. The fact that one specimen showed hysteresis provided a suitable occasion further to confirm the view, always maintained by Lambert and the author, that these phenomena are in no way due to incomplete removal of permanent gases or other impurities from the systems. In previous work with silica and ferric oxide gels, evacuation was carried out at 150° C, which is actually the optimum temperatures of activation for the latter gel, the adsorptive capacity of which decreases rapidly when higher temperatures are employed. Silica gels, however, will withstand a much higher temperature with comparatively little diminution in capacity, and in the present work it has been shown that the hysteresis phenomena are not affected by raising the temperature of the initial evacuation to 250° C or even to 350° C. It has also been shown that the phenomena persist after continued "flushing out" of the gel with methyl or ethyl alcohol vapour. (McBain suggested that the hysteresis observed by Lambert and Clark with benzene on ferric oxide gel would not have ocucrred if the gel had been flushed out with benzene before the isothermals were determined.)