Previous page     Contents     Next page

Based on limited spectroscopic investigations (NMR, Raman, IR and XANES) it is clear that the solution behaviour of phosphorus in aluminosilicate melts is strongly dependent on composition, in particular, the alkali/aluminium ratio. A tentative model of the structural role of phosphorus has been suggested which describes the formation of alkali phosphate complexes in peralkaline melts (x = Na/Na + Al > 0.5), the polymerisation of these complexes to more extended chains in metaluminous (x = 0.5) and peraluminous (x < 0.5) melts and the formation of AlPO₄ complexes in the latter systems. In consequence, the addition of P₂O₅ should lead first to a higher polymerisation and at a crucial ratio x to a depolymerisation of the aluminosilicate network.

Previously published spectroscopic investigations do not provide unambiguous results, and a comprehensive picture of the solution mechanism of phosphorus in aluminosilicate melts. Therefore, in conjunction with physical property measurements (see 3.8d), ³¹P MAS NMR investigations on the system xNa₂O:(1-x)Al₂O₃:2Si₂O containing up to 7 mol% P₂O₅ were performed to monitor the formation of various P-bearing species. The compositions studied ranged from x=1 to x=0.44; the respective ³¹P MAS NMR spectra are shown in Fig. 3.8-5. In the spectra of the peralkaline samples with x=1....0.7 the resonances of Na₃PO₄ monomers (15 ppm) and Na₄P₂O₇ dimers (2 ppm) can be identified. The relative intensities of these peaks is a function of the P₂O₅ content as proposed by the above mentioned model. Furthermore, the relative intensity of two signals (shoulders at 7 ppm and -7 ppm) increases as x decreases; these two resonances are tentatively assigned to end- and middle groups of more extensive chains (NaPO₃)_n. The spectra of the peralkaline samples show clearly that the Na-P complexes become more polymerised with increasing P₂O₅ content and decreasing peralkalinity. The spectra of met- and peraluminous samples contain only one broad, almost featureless peak which does not vary significantly as a function of phosphorus content. The chemical shifts increase as a function of the aluminium content and approach the range typical of AlPO₄ species for the samples with a peraluminous composition. Conclusions drawn from the ³¹P MAS NMR investigations agree well with those inferred from macroscopic property measurements, but a number of questions still need to be investigated in detail. Further work will concentrate on an unambiguous peak assignment including two-dimensional NMR experiments, quantitative determinations (e.g. calculation of average (NaPO₃)_n chain lengths), the investigation of the influence of experimental conditions such as quench rate and traces of water within the melts, and the observation of the network-formers silicon and aluminium.
 

Fig. 3.8-5: 31P MAS NMR spectra of xNa2O:(1-x)Al2O3:2Si2O glasses containing 2.5 mol% P2O5.