The implications of the dense random packing (DRP) model are discussed, with an expectation that it may represent a possible structure model for perfect amorphous solids. The topological definition of the DRP is argued on the basis of Sadoc's idea that the DRP is an assembly of spheres, close-packed in a curved non-Euclidean space and mapped onto the linear Euclidean space. Thermal properties of the DRP, especially excess entropy, are also discussed in this context. Discussions are evidenced by the observed changes in structures and thermal properties of metallic glasses due to particle bombardments and cold works. Similar arguments prove to apply for systems such as amorphous silicon and amorphous silica.
The recent applications of a conventional transmission electron microscope (TEM) and an analytical electron microscope (AEM) to studying ceramic grain boundaries are reviewed. Dislocation structures at small angle grain boundaries and the difficulty in observing general high angle grain boundaries with TEM are presented by using examples of micrographs observed. Elemental analysis data of Sialon and phase separated glass with AEM are also described.
To clarify the formation mechanism of snow polycrystals, which have the misorientation of c-axis at 70.5°, the possibility of formation of a cubic ice nucleus (diamond structure) is discussed on the basis of the homogeneous nucleation theory for supercooled water formed in the phase of supersaturated vapour. From the difference of specific interfacial energy between a hexagonal ice and a cubic ice, it follows that a critical cubic nucleus has a smaller activation energy than a critical hexagonal nucleus below a critical temperature ; namely, Ostwald's step rule (Stufenregel) holds for a change from cubic ice to hexagonal ice below a critical temperature.