Study of equilibrium in exchange between HC1 gas and a KBr surface

Yoshikata Koga
In the main part of this work, an equilibrium in anion exchange between HC1 gas and vacuum-sublimed KBr powder has been studied from 0° to 87°C. For total amounts of exchange corresponding to less than one surface layer, the equilibrium isotherms could be represented by a very simple empirical equation of the form, 1/R[subscript p]X = 1/R[subscript p]X[subscript m] + 1/K'Xm', --------------------------------(1) where X = amount exchanged normalized to BET surface, Xm and Xm' are paramaters of
more » ... change capacity with the same dimensions as X, K' = a dimensionless parameter related in some way to an equilibrium constant for exchange, and R[subscript p] = P[subscript HBr]/P[subscript HCI] in the gas phase at equilibrium. If Xm were independent of temperature and equal to unity or some whole number of layers, eq. (1) with Xm = Xm' would correspond to an ideal chemical equilibrium expression, K = R[subscript p] [x/(Xm - X)] --------------------------------(2). In practice Xm was found to be temperature-dependent (Xm[symbol omitted]exp(-4,600/RT)) and ranged 0.5 at 9.4°C to 3 at 87°C (rather scattered experimental points render the 0°C value uncertain). The quantity designated (K'Xm') in eq. (1) is obtained experimentally as a single constant, and its separation Into an equilibrium constant and an exchange capacity parameters depends on the interpretation of the results. (K' Xm') has approximately the same temperature dependence as Xm , but it is not clear in the first instance whether this is to be ascribed to K' or to Xm'. Various models are examined which may give rise to the above experimental results. Among them; (a) one in which co-operative interactions of first and second anion neighbours in a single layer are treated by a modified form of the quasi-chemical approximation seems to be able to explain isotherms for 0° and 9.4°C, (b) another in which three layers are taken into account in a simple manner is capable of explaining isotherms for 500 to 87°C. This latter is capable of explaining results at all [...]
doi:10.14288/1.0059928 fatcat:oc6ehz73knafzpddkrxboygloi