Inelastic neutron scattering measurements of La1.85Sr0.15CuO4 show that a discontinuity in dispersion develops in the middle of the highest energy LO phonon branch at low temperatures. The result suggests dynamic short range unit cell doubling in the CuO2 plane along the direction of the Cu-O bond, such as charge ordering on every other row of oxygen. Such charge ordering is related to, but is different from, the stripe charge ordering observed for nonsuperconducting cuprates. The question of the influence of the electron-lattice interaction in the normal and superconducting electronic properties of high temperature superconductors remains open. Perhaps the most intriguing experimental evidence of a large and anomalous electron-lattice interaction comes from inelastic neutron scattering experiments on various high temperature superconducting materials [1‐ 6]. A signature of these materials is the strong renormalization of longitudinal high frequency (,80 meV) oxygen bond-stretching phonons in the CuO2 plane when holes are doped into the insulating compounds. The doping causes a large decrease in the frequency of the zone boundary s0.5, 0, 0d half-breathing phonon mode (15%‐ 20% softening compared to the insulating compound), indicating that there is a strong coupling of these phonons to the doped holes. This effect seems to be ubiquitous, and has been observed in single-crystal inelastic neutron scattering measurements of La22xSrxCuO4 [1], La2CuO41y [3], YBa2Cu3O61y [4,5], and even Ba12xKxBiO3 [6]. There are also indications of this effect from the observed softening of oxygen phonon bands in inelastic neutron scattering measurements of the phonon density of states in Pb2Sr2sCa,YdCu3O8 [7] and Li11xTi22xO4 [8]. It has proven quite difficult to interpret these results in terms of normal electron-phonon coupling in a metal, especially since the other measured phonon modes are, for the most part, unaffected (especially the oxygen breathing and quadrupole modes). Typical indications of electronphonon coupling, such as Kohn anomalies, seem to be inconsistent with the wave vector of the affected phonon. The Fermi surface should have strong nesting at 2kf which is close to q › s0.5, 0.5d, but not at s0.5, 0d, as demonstrated by the local-density approximation calculation [9] and ARPES measurements [10]. We have chosen to investigate these phonon phenomena more closely and have discovered that the anomalous nature of the bond-stretching branch is even more mysterious at low temperatures. The sample studied consisted of two single crystals of La1.85Sr0.15CuO4 which were comounted in an aluminum can filled with He exchange gas. Both samples were grown by the floating zone method and were obtained from the same batch. The total size of the sample is approximately 3.8 3 1.5 3 0.8 cm 3 . Previ
Read full abstract