|
Abstract
Structural data for glasses
along the join 0.1[xNa2O-(1-x)Al2O3]-0.9SiO2+2 mol % P2O5, obtained
from multinuclear NMR and ab initio shielding calculations, have
been combined with Raman spectroscopic data on the glasses at
25°C and melts to temperatures above 1200°C. The structural
interpretation of those data is consistent with isolated PO4
and P2O7 complexes together with QnP (n=1-4) species to be present
in both glasses and melts. The QnP species consist of one PO4
group linked to the silicate network by 1 to 4 oxygen bridges
(corresponding to n=1-4 in the Q1P species).
The principal solution mechanism
of phosphorus in these glasses and melts can be described with
a total of 6 different schematic expressions. These expressions
are consistent with the observation from 29Si and 31P MAS NMR
that, for constant P2O5 content, increasing alumina/alkali ratio
leads to a decreased effect of phosphorus on the silicate polymerization.
Temperature-dependent interaction
between phosphate solute and silicate solvent above the glass
transition temperature has been identified. The types of interactions
were separated into three composition ranges. For peralkaline
melts, the phosphate species become more polymerized and the
silicate species less polymerized, with increasing temperature:
2PO4 + Q4Si ¤ P2O7 + Q3Si, and P2O7 + 5Q4Si ¤ 2Q1P
+ 3Q3Si, where QnSi denotes speciation of silicate and QnP speciation
of phosphate. The DH for these two reactions are 140-190 kJ/mol
and about 65 kJ/mol, respectively. For compositions near meta-aluminate,
the abundance of individual QnP species is temperature-dependent
at temperatures above the glass transition,
2Q3P ¤ Q2P + Q4P,
with DH=13-19 kJ/mol. The
abundance of silicate species is not affected by temperature
for those compositions. For peraluminous melts, the phosphate
species become less polymerized and the silicate species more
polymerized with increasing temperature above the glass transition,
Q4P + Q3Si ¤ Q3P +
2Q4Si, with DH=13-23 kJ/mol.
The structural data obtained
in the high-temperature regime indicate that the melt structures
resemble those of the glasses although the abundance of individual
melt species is temperature-dependent. Therefore, for peralkaline
and peraluminous melts, properties that depend on melt polymerization
become more pronounced if melt data rather than structural data
of glasses were employed. For compositions near meta-aluminate,
where neither phosphate nor silicate polymerization vary with
temperature, properties that depend on melt polymerization are
likely less sensitive to temperature.
|
