No abstract

procedure proposed for the comparison of potentials in different solvents. However, if a given procedure is to have any possible general validity, it must generate self-consistent and reasonable predictions. This is a necessary, but not sufficient, condition for its validity. Probable uncertainties in different procedures can be identified, as was done in this communication and elsewhere (6, 18), although it is difficult to gauge the magnitude of the uncertaintie's. In certain cases the applicability of a given procedure to a given pair of solvents can be questioned because it generates predictions that are qualitatively in conflict with strong experimental evidence.Two examples of such evidence will be given. First, as discussed above, there are strong polarographic indications that the alkali metal ions are solvated more strongly by water than by either sulfolane or acetonitrile. The same conclusion applies to the remaining solvents listed in Table IV. Likewise, nuclear magnetic resonance and infrared studies indicate that in acetonitrile-water mixtures lithium ion is preferentially solvated by water over virtually the whole range of solvent composition (27). Any procedure which predicts the opposite, is open to criticism for comparisons of potentials in these solvents with those in water. However, it (27) Unpublished results, this laboratory, 1968. may be more appropriate for other pairs of solvents; we believe this to be the case for the procedure based on the ferrocene potential (6). Second, Hittorf transference measurements for silver nitrate in acetonitrile-water mixtures show directly that silver ion is preferentially solvated by acetonitrile, and nitrate ion by water (28). Again, any procedure which fails to predict both of these experimental facts is not appropriate for comparisons between acetonitrile and water. It is clear that at present the reliance which can be placed on any given approach available for the comparison of potentials in different solvents depends on the extent to which its predictions can be tested by independent experimental evidence. ACKNOWLEDGMENTWe thank C. E. Wilson of this department for his valuable advice and experimental help during the polarographic measurements.

No abstract

No abstract

No abstract