Improved Prediction of Young's Modulus of Fluorine-Containing Glasses Using MAS-NMR Structural Data

The model developed by Makishima and Mackenzie (MM) may yield reasonable estimates for the E-modulus of a range of glasses. In the MM model the bonding enthalpy and packing densities present in the compounds that form the glass are taken as input for the calculation. This study shows that a more accurate estimate can be obtained by incorporating in the model structural information from MAS-NMR data. Specifically, we have determined by means of the impulse excitation technique (IET) the E-modulus for ionomer glasses with composition 4.5SiO23Al2O31.5P2O53MO2MF2, where M denotes the alkaline earth metal (M = Mg, Ca, Sr, or Ba). The MAS-NMR structural analysis shows that substitution of calcium by barium or strontium results in a disrupted network, whereas magnesium leads to a more packed network. In this study we will show how a higher coordination state of the aluminum as determined by 27Al MAS-NMR can be taken into account in the model. This leads to rather small corrections of the estimates for these particular glasses. In contrast, the 19F MAS-NMR study shows the presence of AlFM(n) or AlF and SiFM(n) types of environment in the glass network. AlF and SiF bonds are not accounted for in the E-modulus estimate by the MM model. We will show how by incorporating the new bonding of F with Al and Si a significantly improved estimate of the E-modulus is obtained compared with the original model.