An outline is given of the model for some high-temperature superconductors which assumes that the carriers are holes in the (hybridized) oxygen 2p band and form 'spin polarons' with the moments on the copper atoms. A comparison is made with observations of spin polarons in Gda-xvxS4 and with the properties of Lal_xSr~VOa in relation to those of La2_xSrxCuO4. It is assumed, following several authors, that in the superconductors the polarons form bipolarons, which are bosons, and a comparison is made with some other treatments of this hypothesis. It is proposed that in many such superconductors the boson, essentially a pair of these holes, moves in an impurity band, and that normally all the polarons (fermions) form bipolarons; the fermions repel each other on the same site (positive Hubbard U) but attract when on adjacent sites; the critical temperature T~ is then that at which the Bose gas becomes non-degenerate. In such materials a nondegenerate gas of bosons would carry the current above T~ as first suggested by Alexandrov et al. (1986). The linear increase in the resistivity above T~ is explained on this hypothesis. The effective mass of the bipolaron is, we believe, large (~20-30me). The copper 3d 9 moments in the superconducting range resonate between their two orientations as a consequence of the motion of the carriers, as they do in the description by Brinkman and Rice (1970) of highly correlated metals. Spin polarons, we believe, form only when this is so, but not in the antiferromagnetic range of x. A discussion is given of the resistivity above T~, thermopower above T~, and of the nature of the superconducting gap as shown by tunnelling. We confine our discussion to the materials containing copper, excluding for instance cubic Bal_xKxBiO3, and possibly any superconductor containing bismuth, where the bosons may be Bi 3+.
Contents PAGE