Constitution of Sodium, Potassium, and Lithium Phosphate Glasses

The chain‐length distributions in solutions of sodium phosphate glasses were reinvestigated using improved paper chromatographic techniques. The range of compositions was extended to glasses of higher Na/P ratios by using a through‐flame technique for their preparation. The distributions obtained confirmed the earlier work with the exception that cyclic phosphates were found in solutions of glasses having or approaching the metaphosphate composition despite prolonged heating of the glasses in the molten condition. The earlier observation was that prolonged heating eliminated the cyclic phosphates. As before, no orthophosphate was found, as predicted by Van Wazer, and neither the Poisson nor the “random reorganization” distributions fitted the data for both high‐ and low‐average chain lengths. Although some sort of rearrangement is occurring in the melt, it may not be a completely random process; a rapid change of chemical properties with chain length of the short polymers may have to be taken into account. This was confirmed by similar studies of potassium and lithium phosphate glasses, whose distributions were found to be different enough to demonstrate a cation effect. Consequently, more than stoichiometry is involved.