The nucleation and crystallization behavior of a series of glasses based on 4.5SiO2-3Al2O3-YP2O5-3CaO-1.51CaF2 was studied. The parameter Y was varied to give calcium to phosphate ratios between one and two. All of the glasses studied crystallized firstly to fluorapatite (Ca5PO4)3F). The glass with a calcium to phosphate ratio of 1.67, corresponding to apatite, bulk nucleated to give fluorapatite (FAP). The glasses with calcium : phosphate ratios either less than that of apatite, or greater than that of apatite all exhibited surface nucleation of FAP. However, following a nucleation hold of one hour at approximately 50 K above the glass transition temperature these glasses exhibited bulk nucleation of FAP.

Fly ashes exist as a mixture of major amorphous phases and minor crystalline phases. For commercial applications, such as in concretes and for the production of zeolites, it would be desirable to be able to predict the reactivity of¯y ashes. The amorphous phase dominates degradation behaviour, because glasses have a higher potential energy than the equivalent crystal structure and the variation of bond angles and distances in a glass make the bond breakage easier. Despite the large quantities of¯y ash produced annually by coal-burning power plants, there have been very few studies investigating the microstructure and composition of the amorphous component. In particular, there has been little research undertaken in measuring the glass transition temperature (T g ), which can be directly correlated to the chemical reactivity of the glass phase. Thirteen European¯y ashes were used for the present study. Differential scanning calorimetry (DSC) was employed to determine the presence of transition temperatures and any other thermal events (exotherms or endotherms) in the glassy phase of the¯y ashes. Several different but distinct behaviours were evident in the DSC traces with T g values visible for six of the ashes. The results suggest that thermal analysis has potential as a technique for¯y ash characterisation.

The amorphous phase of¯y ash dominates degradation behaviour because glass has a higher potential energy than the equivalent crystal structure and the variation of bond angles and distances in a glass make the bond breakage easier. It would be advantageous to predict the presence and subsequent degradability of glass on the basis of the solid-state chemistry of the¯y ash. To this end, and inorganic polymer model was applied to a selection of European¯y ashes to determine the value known as cross-link density (CLD). A cross-link density value of less than two implies that the material is amorphous in nature and the lower the CLD below two, the greater the reactivity and solubility of the glass. Applying this model may facilitate the selection of the most suitable¯y ash for a particular recycling application where glass reactivity or dissolution rates are important. To check the applicability of the model to the glass phase of¯y ashes, CLD calculations have been performed by removing the contribution to the ash composition from the known crystal phases. The model would be then expected to give a maximum CLD value of two for all the materials. While this approach has been applied successfully to synthetic glasses and glass-ceramics in the past, only very limited applicability has been found with¯y ashes. This is believed to be due to the inherent heterogeneity of the glass phase in¯y ash.