A study of ion exchange equilibrium (1990)
Type of ContentTheses / Dissertations
Thesis DisciplineChemical Engineering
Degree NameDoctor of Philosophy
PublisherUniversity of Canterbury. Chemical and Process Engineering
AuthorsAddison, Paul A.show all
A study has been made of methods for the characterization and prediction of ion exchange equilibria. Methods which took account of the thermodynamic non-idealities present in ion exchange systems were found to provide the best results. A characterization method which used the extended Debye-Hückel relation and the Wilson equation to estimate solution and resin phase activity coefficients, respectively, was developed. A simplifying reciprocal relationship between a pair of Wilson interaction parameters was hypothesized and was found to give results which were not significantly different from those yielded by the more complicated expression, even though the number of parameters had been reduced from three to two. The reciprocity hypothesis was able to be justified in terms of commonly made approximations. By its application to the characterization of a number of ternary equilibria, the Hála constraint relating three pairs of Wilson interaction parameters, previously only applied to vapour-liquid equilibrium, was found to be equally useful for ion exchange purposes. Predictions of ternary ion exchange equilibria were successfully made using the Wilson parameters and thermodynamic equilibrium constants obtained from binary characterizations. Only slight reductions in the quality of prediction were evident when the reciprocal relation and the Hála constraint were included in the predictive algorithm, even though the number of parameters used was reduced from eight to four. When data from only two of the three constituent binary systems were used to predict a ternary system, best results were obtained when the two pairs used both contained the most favourably selected ion in the ternary system. A CSTR technique for the measurement of ion exchange equilibrium isotherms was designed, built and tested. The method had the particular advantages that it was continuous and enabled on-line acquisition and analysis of results by computer. Only one experimental run was necessary to obtain a complete binary ion exchange isotherm. The uncertainty in computed resin phase compositions was found to compare very favourably with those arising from other methods. Measurements were made of the kinetics of the CSTR process and of the transient behaviour of the ion selective electrode used to ascertain solution phase compositions. Both of these were found to have a significant effect at low concentrations, and appropriate adjustments to measured data were made. It was recommended that these corrections may be avoided by use of a lower feed flowrate at the expense of a longer run time, necessitating the installation of a pump capable of delivering flowrates below those used in this work. Reproducible results were obtained for the Na+/H+ exchange on Dowex 50W-X8 at 0.1 N total solution concentration and a temperature of 20°C. The results did not, however, agree with some previously published data for the same system. This discrepancy was thought to be due to a difference in the mean resin particle size between the two sets of experiments. Batch measurements were made, and confirmed the binary isotherm measured by the CSTR.