Ion Exchange in Nonaqueous and Mixed Media

Abstract

V. SUMMARY AND CONCLUSION Enough experimental data on nonaqueous ion-exchange separations have been gathered to show that one is dealing with a field whose theoretical aspects are infinitely complicated. The separations can be based on several different types of chemical reactions. Apart from true ion exchange, sorption can take place by non-exchange reactions such as acid-base reactions or complex-formation, either by hydrogen bonding or by metal-ligand coordination. All these reactions can be influenced by a multitude of parameters such as solvent polarity, solvent-solute interaction by solvation or complex formation, solvent structure changes, solvent-resin matrix interaction, solvent-solvent interaction, acid-base properties of the solvent, non-exchange electrolyte invasion, and so on. Any one or several of these parameters may or may not be operative in a given case, and a given parameter can be of primary importance in one case and be only a minor contributing factor in another. In view of such complexity, one should not be surprised that the few attempts so far made to develop a theory of nonaqueous ion exchange were not too successful and could be applied only to a few rather simple cases. Even considering only one type of chemical reaction, the true ion exchange, the chances of ever arriving at a comprehensive theory appear to be pretty slim. An intriguing subject from a fundamental point of view is ion exchange in non-polar anhydrous solvents such as benzene, in which the degrees of ionic dissociation, even of strong electrolytes, are extremely low. Little fundamental work on this subject has been done so far, and still less has been done under well-defined experimental conditions. This is particularly the case with respect to the strict absence of water, which even in quite small amounts can have disproportionately great effects on ion-exchange systems in less polar solvents. A careful investigation of ion exchange in benzene or similar solvents under strictly anhydrous conditions would be an interesting subject, which could perhaps reveal the answer to the intriguing question whether ion exchange is possible only between free (dissociated) ions, or whether it can take place also between ion-pairs! This would be undoubtedly a valuable contribution to our limited knowledge on electrolytes in nonaqueous solutions. From the point of view of chemical analysis, nonaqueous ion exchange has already developed into a valuable enrichment of the analytical chemist's arsenal of separation methods. The best chances for further useful developments seem to involve separation methods based on complex-forming reactions. Many complexes form more easily in solvents less polar than water, and their ion-exchange behaviors can be influenced by modifying the solvent composition, which provides more versatility than can be expected in water. The separation of metals by anion exchange of their anionic complexes has already been well developed during the last 10 years, mostly by the use of water-containing mixed solvents. Non-exchange separations by complex formation, either by hydrogen bonding or by acid-base reaction or by ligand-exchange chromatography on resins having coordinating metal counterions, are methods particularly suitable for the use of non-polar solvents or solvents of low polarity. The macroporous resins absorb easily considerable amounts of these solvents and therefore should afford an impetus to the development of these unique possibilities of separation methods for various organic compounds. The most important and useful numerical factor for the separation of a given species is its distribution coefficient. There are many thousands of numerical values dispersed in the literature, some determined under well-defined conditions but others of only limited value because the experimental conditions were ill-defined. Quite often, indeed, one can hardly compare values from different sources because of differences in defining the coefficients. It would be most helpful for the practicing analytical chemist if reliable values were available in tables, as is the case for RF-values in chromatography or for absorption peak frequencies in spectroscopy. The following steps could help to attain this goal: 1. Establish by convention a definition of the distribution coefficient and perhaps determine the conditions for its determination. 2. Collect published data and catalogue those judged to be well-defined and of reasonable accuracy.