## Abstract

A quantitative investigation has been made of ion-exchange diffusion and equilibria in the felspathoids, basic cancrinite (M$_{2}$O.Al$_{2}$O$_{2}$.2$\cdot $4SiO$_{2}$.0$\cdot $6 M OH.xH$_{2}$O), basic sodalite (M$_{2}$O.Al$_{2}$O$_{3}$.2$\cdot $5SiO$_{2}$.0$\cdot $34 MOH.xH$_{2}$O) and K and Rb analcite (M$_{2}$O.Al$_{2}$O$_{3}$.4SiO$_{2}$). The isotherm contours were of three kinds: an 'ideal' form obeying the mass action law K = (B$_{c}$A$_{s}$/A$_{c}$B$_{s}$) (Na-Li, Na-K, Li-K and Na-Ag in basic sodalite); a sigmoid form obeying the equation log$_{10}$ K = log$_{10}$ (B$_{c}$A$_{s}$/A$_{c}$B$_{s}$) + C(1-2B$_{c}$) where the constant C takes a negative value (Na-Li in basic cancrinite); and a form exhibiting hysteresis (Na-Ag and Li-Ag in basic cancrinite and K-Rb in analcite). The hysteresis was shown to be due to limited mutual solid solubility of the end-members of the exchange and to an associated difficulty in nucleating crystallites of one growing phase on or in a matrix of the other. This effect is most strikingly found in the Rb-K exchange in analcite, for which various scanning loops were traversed. A quantitative approach to the theory of the above types of exchange isotherm has been given, and applied to the present and to earlier results obtained with crystalline exchangers. This gives a possible theoretical basis of the equation log$_{10}$ K = log$_{10}$ (B$_{c}$A$_{s}$/A$_{c}$B$_{s}$) + C(1-2B$_{c}$) and for Kielland's equation log$_{10}$ f$_{Ac}$ = CB$_{c}^{2}$. Examination of selectivity coefficients shows that the crystalline exchangers may possess very high selectivities towards one alkali metal ion as against another or for heavy metal ions such as Ag or T1. These selectivities may change radically as one crystal is substituted for another. Ion sieve effects, partial or complete, were observed and can bring about a Donnan membrane hydrolysis of salts even of strong acids or bases, such as CsCl. Comparison has been made of the exchange diffusion coefficients evaluated in these laboratories for several kinds of crystal. A feature of exchanges where the temperature variable has been studied is the normally small temperature coefficient of the exchange equilibria and so a small value of the heat of exchange. A model has been proposed which regards the reactions as an interchange of ions between two inert dielectrics. This interpretation provides a simple explanation of the observed behaviour.