AN ATTEMPT TO RECONCILE INTERPRETATIONS OF ATOMIC VIBRATIONS AND SI-29 NMR DATA IN GLASSES

The infrared reflection spectra of lithium aluminosilicate glasses (join xLiAlO2-SiO2 (Li/Al congruent-to 1) with 0 < x < 1, and Li2Si2O5-xAl2O3 (Li/Si congruent-to 1) with 0 < x < 1), in the entire spectral range (20-4000 cm-1), are reported for the first time. Results are compared with spectra of sodium silicate and aluminosilicate, and calcium aluminosilicate glasses. Frequency shifts versus Al composition are interpreted in terms of mode coupling. Data obtained by three different spectroscopies, infrared reflectivity, Raman scattering and nuclear magnetic resonance spectroscopy (NMR), are compared and discussed. It is shown that these spectroscopies are complementary rather than redundant. Three examples are discussed in detail. The results are consistent with the concept of non-bridging oxygen, whereas infrared modes are explained in terms of number of oxygen shared by two network former cations or by one modifier and one former cation. The second example is an indication of increase of bond angle disorder with Al content deduced from infrared linewidth which may explain the Si-29 MAS-NMR linewidth variation. Also, the silicon effective charge deduced from infrared reflectivity spectroscopy and the Si-29 chemical shift are compared. Results show linear correlations and illustrate that chemical shift is sensitive to several contributions and not just the Coulombic field as the effective charge.