Non-Stoichiometric Compounds Derived From Extended Defects

Abstract

The non-stoichiometric compounds that we describe in this chapter are closely correlated with the classical non-stoichiometric compounds derived from point defects discussed in Chapter 1. For the past twenty years precise structural analyses on complex binary and ternary compounds have been carried out using X-ray and neutron diffraction techniques. Moreover, owing to the striking development of the resolving power of the electron microscope crystal structures can be seen directly as structure images. As a result, it has been shown that most complex structures can be derived by introducing extended defects regularly into a mother structure. A typical example is a ‘shear structure’, which is derived by introducing planar defects of anion rows into the mother lattice. A ‘block structure’ is derived by introducing two groups of planar defects. ‘Vernier structures’, ‘micro-twin structures’, ‘intergrowth structures’, and ‘adaptive structures’ are also described in detail in this chapter. At the beginning of 1950, Professor A. Magnéli’s group in Sweden started a systematic study of the crystal structures of the oxides of transition metal elements such as Ti, V, Mo, and W, mainly by X-ray diffraction techniques. As a result, they confirmed the existence of the homologous compounds expressed by VnO2n–1; TinO2n–1 etc. (n = 2, 3, 4, . . .) and also predicted that the crystal structure of these compounds could be derived from a mother structure, ‘rutile’. Figure 2.1 shows the X-ray powder diffraction patterns (CuKα) of compounds TiOx between Ti2O3 (x = 1.5) and TiO2 (x = 2.0).3 This clearly indicates the convergence of the diffraction patterns to that of TiO2 (rutile) with increasing x, which is why the Magnéli school predicted the mother structure to be rutile. This prediction was verified by the structure determinations of Ti5O95 and VnO2n–1.6 These compounds are called Magnéli phases after the main investigator, and similar compounds have been discovered.