Are Spin Density Waves Involved in High-Temperature Superconductivity?

Abstract

In the La2-xBaxCu04, La2-xSrxCu04, and M1Ba2Cu30y compounds, where M is a transition metal such as yttrium and y is apparently between 7 and 6, the quasi-two-dimensional metallic planes of repeated Cu02 have a perfectly-nested square Fermi surface when there are exactly two extra electrons per Cu02, in a tight-binding band structure. In this case one expects formation of an electron charge density wave (CDW) or spin density wave (SDW). The Fermi surface is significantly different from a simple square in more sophisticated band-structure theory,1,2 however in this case saddle-points (Van Hove density of states singularities) in E versus k may be responsible for a density wave, as suggested by Rice and Scott3 for the 2H crystal form of several transition metal di-chalcogenides. In this case, when the number of extra electrons per Cu02 unit is somewhat less than two, e.g. 1.85 for x = 0.15 in the lanthanum compounds, or somewhat more than two for M1Ba2Cu30y, density wave formation may be controlled by the Van Hove singularities, with no increase or even a decrease in the electrical resistivity of the CDW or SDW state from the normal-metal state3: the high-velocity parts of the Fermi surface are not gapped by the density wave.