One of the most remarkable peculiarities of refractory carbides and nitrides is their ability to have a considerable and varying concentration of vacancies in metal and nonmetal sublattices: see the monographs by Toth (1971), Samsonov and Vinitsky (1976) and Kosolapova (1986). Many properties of these phases are very sensitive to vacancy concentration. For example, vacancy ordered M5C5 and M8C7 phases can easily be formed, vacancy ‘clusterisation’ is rather typical in carbides and nitrides. The formation of defects is often accompanied by local crystal lattice distortions, see Venables, Kahn and Lye (1968), de Novion, Lorenzelli and Costa (1966), Onozuka (1978, 1982), Landesman, Christensen and de Novion (1985), Gusev and Rempel (1984, 1989) and Metzger, Peisl and Kaufmann (1983). Apart from structural parameters, many other properties of refractory phases – mechanical, magnetic, superconducting, catalytic, etc. (see Toth (1971), Samsonov and Vinitsky (1976), Metzger et al (1983)) – change considerably in the homogeneity region.

Such unusual behaviour gives an incentive to active study of the influence of vacancies on the properties of carbides and nitrides using both the experimental and theoretical methods of solid-state physics and chemistry. So far, a large number of studies on the electronic states of nonstoichiometric compounds have been made using X-ray emission, electron, optical spectroscopy and other methods. The results of these have been discussed in several review papers: Zhurakovsky (1976), Meizel, Leonhard and Sargan (1981), Ivanovsky, Gubanov, Kurmaev and Shveikin (1983). These experimental results provide a reliable basis for testing the theoretical models proposed for vacancy-containing phases.