POLYMERIZATION AND DISPROPORTIONATION REACTIONS IN SILICATE GLASSES: NEW EVIDENCE FOR "FREE" OXIDE
2Q3 → Q2 + Q4 (1); 2Q2 → Q1 + Q3 (2); 2Q1 → Q0 + Q2 (3)
Comparison of the best resolved 29Si NMR2 and Raman3 spectra for ~40 mol% Na2O glass show that both spectra are dominated by two resolved bands: one band from Q2 and one larger, broader band from Q3. Estimation of the small amount of Q4 (~2%) from static 29Si NMR enabled Stebbins2 to calculate a disproportionation constant of 0.011 for Eq. (1). The recently published 40% BaO and K2O Raman spectra4,5 are remarkably similar to the 40% Na2O spectrum, implying the same assignments for all three glasses- one band from each of Q2 and Q3. The Raman spectra from 50 mol% Li2O, Na2O and K2O glasses and melts4,5 are again all qualitatively similar, displaying one band for each of Q1 (weak), Q2 (strong) and Q3 (medium) species. A requirement of Eq.2 is that at 50 mole% alkali oxide, Q1 = Q3 from mass balance considerations, NBO/Si = 2, and “free“oxide (FO) = 0. These spectra all demonstrate that Q3>Q1, so Eq.2 does not hold and it cannot be the only reaction relating the Q species. Polymerization reactions such as:
2Q2 = 2Q3 + FO (4)
must be considered. Indeed, for these 50 mol% glasses and melts, NBO/Si = ≤1.8, and mol% FO is present. Rxn. 4 dominates for the K metasilicate glass, yielding NBO/Si of 1.5 and a large FO of 7-8%, in agreement with O 1s XPS and the 29Si NMR spectra.
1) Virgo et al. Science 208, 1371 (1980). 2) Stebbins, Nature 330, 465 (1987). 3) Nesbitt et al. Chem. Geol. 562 , 129940 (2022). 4) Moulton et al. Chem. Geol.586, 120611 (2021). 5) Koroleva et al. Minerals 13, 94 (2023).