Key points are not available for this paper at this time.
A discussion is given of metal-insulator transitions in VO2, Ti2O3 and also in Ti2-x V x O3, which at low temperatures becomes metallic with increasing x. Although these are band-crossing transitions, there being no magnetic moments on the metallic ions, a large discontinuity is expected in the number of current-carriers in VO2 as the degree of pairing of the V atoms is varied. We relate this to the first-order transition observed as the temperature is raised. In Ti2O3 electrons in the conduction band appear to be much heavier than holes in the valence band and are probably small polarons. The addition of ∼2% of V2O3 appears to push Ti2O3 across the metal-insulator transition, again with a discontinuous change in the number of carriers, the alloy being a semimetal with heavy electrons and light holes. The experiments of Sjöstrand and Keesom (1973) are discussed; the reason why the heavy electron polarons form a degenerate gas at helium temperatures, rather than a Wigner crystallization, is that the Coulomb interaction between them is strongly screened by the light holes. The situation is compared with that in Fe3O4 and Ti4O7.
Mott et al. (1974) studied this question.