Abstract
In solutions of sodium naphthalenide in 1,2-dimeth- oxyethane1,2 (DME) the ratio of the free-ion spectrum to that of the ion pair may vary in a manner not in harmony with the requirement of an ion-pairing equi- librium. For example, in two samples of naphthalenide concentration, 5X10-sM, the ratio R of unsplit to split lines was 8.9 and 1.4.
To test whether an uncontrolled source of sodium3 ion was repressing dissociation in some instances, an apparatus was constructed in which nonvolatile ma- terial could be trapped, and solvent and naphthalene redistilled to produce fresh solutions of radical anion. We expected that applying this trapping cycle would enhance the unsplit spectrum with respect to split spectrum. Just the opposite happened. For the first solution R=4.0. One trapping cycle reduced R to 0.05 and a second to zero. Analysis of the residues from all the solutions made in this apparatus yielded enough potassium to account for the unsplit spectrum observed. Similar results were obtained in other apparatuses.
These results are incomprehensible if the unsplit spectrum is that of the free ion. However, if the unsplit spectrum is that of potassium naphthalenide we may invoke the equilibrium