Perspectives on the Theory of the New High T c Superconducting Oxides
- PDF / 614,597 Bytes
- 5 Pages / 604.8 x 806.4 pts Page_size
- 116 Downloads / 203 Views
MRS BULLETIN/JANUARY 1989
superconductivity has escaped after so many years. Despite the existing understanding of the properties of oxides, much work remains to be done before we have a good grip on the many-body theory of these strongly correlated syst e m s . In particular, the quasi t w o dimensional structure of the copper oxide planes has a profound influence on electronic and magnetic properties. It is natural to seek an explanation of superconductivity in the cuprates as yet another manifestation of the special nature of physics in two dimensions. Ironically, this feature of the problem also bedevils applications by limiting the critical current density attainable in ceramic samples, which do not have aligned C u 0 2 planes. Electronic Structure and Magnetism To get a feeling for what is going on, it is useful to start from simple valence counting arguments, which work quite well for the relatively ionic environment of the C u 0 2 planes in L a ^ S r , C u 0 4 and YBa 2 Cu 3 0 6 + x . Consider first La 2 Cu0 4 , which would have the ionic configuration La^ + Cu + 04 _ if there were only closed-shell configurations. Charge balance requires one extra hole per formula unit and, in a simple localized picture, it would reside on the low-energy Cu site to produce Cu 2+ . If there is a sufficiently small overlap of orbitals, the situation does not change too much. The relatively strong Coulomb interaction still enforces localization, but the occupancy of the Cu site is slightly reduced. Such a
structure has an energy gap for charge excitations, mainly governed by the energy to transfer a hole onto an oxygen site. The remaining degrees of freedom are local m o m e n t s — t h e spins of the odd holes in Cu 2 + . They are coupled by superexchange, an effect of the kinetic energy due to overlap. The holes have a relatively large zero point energy because they are confined to a small region. This region is larger, a n d the zero point energy is lower, if a hole is surrounded by opposite spins, for then it may make excursions onto the neighboring Cu sites without violating the exclusion principle. Thus it is to be expected that La 2 Cu0 4 should be an antiferromagnet, as indeed it is. 4 When a concentration x of Sr2+ is substituted for La 3+ , there is an additional density x of holes in the C u 0 2 planes. In view of the strong Coulomb repulsion associated with double occupancy of the relatively small Cu(3d) states, Cu 3+ is an i m p r o b a b l e c o n f i g u r a t i o n , a n d the a d d e d holes will go onto the 0 ( 2 p ) states. 5 Since the holes are few in number, there is plenty of room for them to move and form a band of states lying within the gap for Cu charge excitations. Thus doping with Sr produces a very low density of holes moving on oxygen sites. They are the bearers of the supercurrent. This picture is remarkably well supported by high-energy spectroscopy 6 and by neutron scattering studies of antiferromagnetism. In particular, x-ray absorption n e a r - e d g e spectroscopy 7 shows that the copper r
Data Loading...
