A New Interpretation of the Mullite Problem

Abstract

THE published phase diagrams for the Al₂O₃–SiO₂ system have been many and varied. Bowen and Grieg¹ maintained that mullite of invariable composition 3Al₂O₃.2SiO₂ (71.8 weight per cent Al₂O₃) is the only stable alumino-silicate found at high temperatures in this system. Several workers have since found homogeneous mullites containing from 75 to as much as 82.6 per cent Al₂O₃, and it is at present generally accepted that continuous mullite solid solutions exist at least over the range 3Al₂O₃.2SiO₂–2Al₂O₃.SiO₂ (71.8–77.3 per cent Al₂O₃). Some investigators also claim the existence of mullites intermediate in composition between silli-manite (Al₂O₃.SiO₂) and 3Al₂O₃.2SiO₂, although continuous solid solution between these compositions has never been authenticated. More recently, the melting behaviour of mullite has been disputed, some workers adhering to Bowen and Grieg's belief that mullite melts incongruently, whereas others maintain that no decomposition into corundum and liquid occurs during melting, and that therefore mullite melts congruently.