An experimental study of some silicate based glasses
TOF neutron diffraction and multinuclear MAS NMR data are combined to investigate the local structure of three silicate based glass forming systems. The effects of experimental resolution on the structural parameters obtained from neutron diffraction data are considered using simulated correlation functions. The Gaussian fit parameters to a peak in T(r) report a smaller peak width and area than that of the Gaussian broadening. This is due to the contribution of the realspace resolution function. This effect is most dramatic for small values of (u 2) 1/2 and Qmax but is negligible for typical values. The experimental uncertainty in measuring Q is considered for TOF neutron diffraction data. ΔQ is considered constant for data measured at several scattering angles. Some of the scattering intensity is transferred to a tail on the low r side of each peak, the magnitude of which increases with ΔQ. The peak fit parameters (position, width and area) change with increasing ΔQ, both before and after removing the gradient at low r. The Q-space resolution at a fixed scattering angle is considered by broadening i(Q) with a Gaussian of width ΔQ/Q=constant. The peak fit parameters change with increasing ΔQ/Q but not significantly at the resolution quoted for the high angle detectors on LAD. The interpretation of experimental data in terms of glass structure is greatly dependant upon an accurate knowledge of the glass composition. It is considered necessary to determine the concentration of all the cation species by a single or combination of techniques. TOF neutron diffraction and 170 MAS NMR data are reported for several (SnO)x(Si02)1-x glasses and a partially crystallised (SnO)(Si02) sample. 29Si, 119Sn MAS NMR and a powder X-ray diffraction data are also presented for the partially crystallised sample. The tin is present as Sn2+ and is three co-ordinated to oxygen at all compositions. The (SnO)x(Si02)1-x glasses are thought to consist of a network of Sn03/3 and Si04/2 polyhedra. The tin polyhedra may form pairs or chains. The local order in the crystalline phase is thought to be similar to that in the glasses. There are two tin and two silicon sites in the crystal phase of approximate composition (SnO)4(Si02). It has not been possible to refine the crystal structure. 29Si MAS NMR data are presented for (Li20)x(SnO)1-x(Si02), (Na20)x(SnO)1-x (Si02) and (K20)x(SnO)1-x(Si02) glasses and TOP neutron diffraction data for (K20)x(SnO)1-x(Si02). The tin is present as Sn2+ and is three co-ordinated at all compositions. Replacing the tin with modifier oxide reduces the 29Si NMR chemical shift. The modifier cation is thought to depolymerise the tin silicate network, associating with the tin to maintain charge neutrality. 29Si and 31p MAS NMR and TOP neutron diffraction data are presented for (K20- 4Si02)x(P205)1-x, (K20-2Si02)x(P205)1-x and (Si02)x(P205)1-x glasses. Each glass contains SiIV and SiVI species, the potassium tetrasilicate-Pjt), glasses also contain SiV species. The proportion of silicon species changed with thermal history. Each glass is thought to consist of a modified phospho silicate network. The potassium disilicate-Pjo, glasses are thought to consist of a continuous network of Si(OP)3(OSi), Si(OP)4(OSi)2, (p-Q2)- and (P-Q4)+ species. The alkali-free phosphosilicate and potassium tetrasilicate-P2O5, glasses are thought to have similar structures but it is not possible to define the phosphate species. It was not possible to distinguish the different species by neutron diffraction. The phosphosilicate network is thought to be more disordered than modified phosphate and modified silicate networks.