Comment on “Spectroscopic studies of oxygen speciation in potassium silicate glasses and melts”

Abstract

Recent O1s XPS studies suggest that significant free oxide (O2– or K–O–K) is present in potassium silicate glasses, in contrast with what was concluded from 29Si and 17O solid-state NMR data. An alternative peak assignment of the Qn peaks in the 29Si NMR spectra was proposed to bring the 29Si NMR data in line with the O1s data, but this reassignment (i) is not supported by any evidence other than the resulting agreement between the NMR and XPS data, (ii) is at odds with the spectral properties of the bands (peak position and width), (iii) ignores the strong evidence for the original peak assignment, and (iv) results in highly implausible equilibrium constants for the Qn speciation reactions. More likely causes for the apparent discrepancy between the XPS and NMR data are the incorrect estimation of the precision of the bridging oxygen content from the O1s XPS data, the systematic overestimation of the bridging oxygen content from the O1s XPS data, and (or) the incorrect estimation of the accuracy of the compositional analysis, which was based on the precision rather than the accuracy of the electron microprobe analysis. Thus, the available evidence strongly suggests that the original assignment of the 29Si NMR spectra is correct. Neither the 29Si NMR nor the O1s XPS data support the presence of significant amounts of free oxide in potassium silicate glasses with K2O/SiO2 ≪ 2.