Comment on `The electronic structure of CaCuO₂and SrCuO₂'

Abstract

Abstract. Recent electronic structure calculations for the title compounds performed by Wu et. al. [1] are critically reconsidered, applying high precision fullpotential bandstructure methods. It is shown that the bandstructure calculations presented by the authors contain several important inconsistencies, which make their main conclusions highly questionable.In a recent paper Wu et. al.[1] presented bandstructure calculations for the quasi onedimensional CuO-chain compound SrCuO 2 and the quasi two-dimensional material CaCuO 2 , both being of prototypical character and therefore of general interest. Wu et. al. used a full-potential linear combination of atomic orbitals method [2] in the framework of the local spin density approximation (LSDA) and included on-site Coulomb interaction corrections (LSDA+U ). The authors of Ref. [1] claim that on the basis of their full-potential band structure experimental findings can be well fit with an U of 5 eV, significantly smaller than U values reported in previous calculations [3].However, there are obvious inconsistencies and important differences between the calculations of Ref.[1] and previous studies [3,4,5,6], concerning (i) the proper symmetry in k-space, (ii) the widths and the orbital character of the shown bands, (iii) the total (DOS) as well as the partial densities of states (PDOS). Therefore, we reinvestigated the electronic structures of CaCuO 2 and SrCuO 2 using two independent, well basis converged full-potential bandstructure methods to find out whether or not the differences mentioned above could be understood as a consequence of the differences between a full-potential [1] and the earlier non-fullpotential calculations [3,5,6]. We carried out LSDA bandstructure calculations for CaCuO 2 within a full-potential minimum-basis local-orbital scheme (FPLO) [7] and within a full-potential linearized augmented plane wave (FLAPW) scheme [8], both in scalar relativistic versions. (We note that relativistic effects are in the order of 0.1 eV only.) In the FPLO-scheme, modified Ca 3d, 4s, 4p, (Sr 5s, 5p ,4d), Cu 3d, 4s, 4p, and O 2s, 2p, 3d states were used as valence states for CaCuO 2 (SrCuO 2 ), the lower lying states were treated as core states. The WIEN97-code [8] employs local orbitals (LO) to relax linearisation errors and to treat the O-2s and semicore Cu-3p and Ca-3s, 3p states. Well converged basis sets of over 500 APW functions plus LOs were used. The radii of the atomic spheres in the latter case were 1.8 a.u. for all atoms. The basic calculations were performed with 125 and 90 k-points in the FPLO-scheme and in the WIEN97-code, respectively, for the irreducible part of the Brillouin zone using the tetrahedron method. We itasize that the numerical convergence (with respect to the number of k-points N k , the valence basis set, the potential and the density

Magnetoelectronic properties of a ferrimagnetic semiconductor: The hybrid cupromanganiteCaCu3Mn4OWeht1, Pickett2 2001Phys. Rev. B844560View full textAdd to dashboardCite

“…We will show here that the semiconducting behavior is well reproduceed just considering GGA within density functional theory. Discrepancies between our results and those of Wu et al are due to approximations made in their methods, and possibly to their choice of basis set (see [9]).…”

Optical studies of gap, hopping energies, and the Anderson-Hubbard parameter in the zigzag-chain compoundSrCuO2

“…3 in the energy range between -7 eV and 10 eV. The electron density of states is calculated with 180 k points in the irreducible part of the BZ and it is in an agreement with previous publications [10,16]. Because the main properties depend on the electrons at the Fermi level, we paid a special attention to the energy dispersions near E F .…”

“…Quite recently, Rosner et al [16] calculated the electronic structure of SrCuO 2 using the LSDA method, also. They found that the gap in the antiferromagnetic phase of SrCuO 2 , reported in previous publications [11,12], is probably artificial and appears due to too small number of the k-points in the Brillouine zone taken in the computation procedure.…”

“…III.A. The reason why we use this method is to obtain the data necessary for the determination of the correlated electronic structure and to compare our results with the previously published ones [10,11,12,16]. In Sec.…”

Fermi surfaces of diborides: MgB2andZrB2

“…[28,29] One method is the full potential linearized augmented plane wave (FLAPW) method [30] as implemented in the WIEN97 code. [31] The other method is full potential local orbital code (FLPO) [32] based on local orbitals optimized to minimize the electronic energy.…”