Previous work in the field of salt effect in vapor‐liquid equilibrium and the theory involved are reviewed briefly. Vapor‐liquid equilibrium data at atmospheric pressure are reported for the systems methanol‐water, ethanol‐water, and n‐propanol‐water, each saturated with a variety of common inorganic salts. A simple equation is proposed for the representation of salt effect. An attempt is made to ascertain the theoretical significance of this equation, and in so doing, to relate salt effect in vapor‐liquid equilibrium to the properties of the system components.

No abstract

A semi-empirical test for the thermodynamic consistency of isobaric liquid-vapor equilibrium data for ethanol-water saturated with a salt is proposed. The test is based upon an adaptation of the Herington method ( I ) in which the ternary system is treated as a special binary. The test is applied both to full concentration range data and to data which are incomplete a t the high alcohol end. For data which are incomplete a t the high water end, a less severe test is employed.Literature data for 23 systems composed of ethanol, water, and an inorganic salt at saturation were tested. By the criterion of the proposed test, fourteen were deemed consistent, six were borderline, and three were pronounced inconsistent.All consistent isobaric liquid-vapor equilibrium data must satisfy the equation Generally heat of mixing data are unavailable, and so for binary isobaric systems Herington (1 ) suggested a semiempirical procedure in which the area above and the area below the abscissa are measured, the areas ( 2 ) are summed and are subtracted one from the other ( I ) . The percentage deviation is defined asHerington showed that if the data are consistent.* J is defined asThe only previous attempt to test systems containing two liquids and a salt was by the "slope" method. Kogan The calculated and experimental values should lie on two parallel curves. The method was tested by Kogan on isobaric data for three systems containing ethanol, water, and a salt at saturation. The biggest drawback of the "slope" method is that it is insensitive. An alternative method is presented for testing the consistency of salt data which is based upon Herington's test. METHOD OF TESTING ISOBARIC SALT DATAThe method makes the basic assumption that the ternary system can be treated as a binary composed of ethanol saturated with salt as one component and water saturated with salt as the other component. When these t w o solutions are mixed it is found in every case that, where the solubility of the salt in the boiling alcohol-water mixtures has been measured, precipitation occurs and hence the final solution is always saturated with salt. The activity coefficients of ethanol and water are calculated from Equation (6) : with one modification. The saturation vapor pressure pi0 is not applicable when the solvent is saturated with salt. Instead, for water the vapor pressure of the saturated salt solution pz' is substituted for pZo. For ethanol these vapor pressures are not available and so a correction to the saturation vapor pressure is applied by multiplying by the ratio of the vapor pressure of ethanol saturated with salt IIT to the vapor pressure of pure ethanol p l~O at the salt solution boiling point. This ratio c is assumed independent of temperature. Incomplete data are treated somewhat differently than complete data and so each will be considered separately. March, 1972

A salt dissolved in a mixed solvent is capable, through preferential association or other structure‐related effects in the liquid phase, of altering the composition of the equilibrium vapor phase. Hence salt effect on vapor‐liquid equilibrium relationships provides a potential technique of extractive distillation for difficult separations. The literature pertaining to salt effect in vapor‐liquid equilibrium and in extractive distillation over the period 1966‐present is reviewed, and also the remainder of work up to 1966 not covered in the Part I review.