Experimental and theoretical study of the electronic structure of single-crystal ${\mathrm{BaBiO}}_3$

High quality single crystals of ${\mathrm{BaBiO}}_3$ were grown by congruent melting technique and characterized with x-ray diffraction, x-ray photoemission, and transport property studies. The perovskite oxide ${\mathrm{BaBiO}}_3$ is a negative charge transfer gap high $T_c$ oxide parent superconducting compound exhibiting self-doping of holes into the oxygen $2p$ band. We study the low energy scale valence and conduction bands in detail from both a theoretical perspective as well as through x ray, absorption/emission, and photoelectron spectroscopies. X-ray spectroscopy verifies the results of density functional theory (DFT) regarding the overall band structure featuring strong O $2p$ character of the empty antibonding combination of the hybridized Bi $6s$ and O $2p$ states. From the analysis of the core level line shapes we conclude that the dominant O $2p$-Bi $6s$ hybridization energy scale determines the low energy scale electronic structure. This analysis provides further insight into the importance of self-doped oxygen $2p$ states in this high $T_c$ family of oxides.

Figure 1

(a) Photograph of the ${\mathrm{BaBiO}}_3$ single crystal. (b) X-ray back-reflection Lauegram of the grown crystal with the x rays oriented along the [100] crystallographic direction in a cubic unit cell.

Figure 2

Main: Oxygen $K$-edge XAS and RXES spectra measured on a ${\mathrm{BaBiO}}_3$ single crystal and the DFT(LDA) oxygen $p$ projected density of states of ${\mathrm{BaBiO}}_3$. The energy has been shifted by $-528.5$ eV for the experimental axis to facilitate comparison with the calculation. Inset: The oxygen $K$-edge XAS in a wider energy range.

Figure 3

Single-crystal ${\mathrm{BaBiO}}_3$ oxygen $1s$ XPS results on the left and the corresponding carbon $1s$ peaks on the right. The spectra from top to bottom show the progression as the cleaning procedure is optimized. The top is on the as grown single crystal and the bottom is after the optimized cleaning procedure described in the text. The removal of the carbon peak confirms the effectiveness of the procedure. The oxygen spectrum loses most of its structure with the progression of the cleaning but remains slightly asymmetric. The curves are shifted along the “Intensity” axis.

Figure 4

A schematic concept of the involved hybridization terms. There are oxygen $p$ orbitals at the corners and bismuth $s$ at the center of the octahedron. (a) $t_{{\text{O}}_p{\text{-Bi}}_{6s}}=\sqrt{6}{t}_{sp\sigma }$. (b) $\sigma$ and $\pi$ oxygen $p$ hopping terms. (c) [and (a)] $t_{pp}=1/2(t_{pp\sigma }-t_{pp\pi })$. (d) $t_{pp}^{\prime }=1/2(t_{pp\sigma }+t_{pp\pi })$.

Figure 5

The average number of holes as a function of (a) Bi $6s$/O $p$ hybridization energy ($t_{sp\sigma }$) for $\mathrm{\Delta }=-4$ eV and (b) the charge transfer energy $\mathrm{\Delta }$ for $t_{sp\sigma }=2.1$ eV.

Figure 6

Calculated oxygen $1s$ XPS spectra with all the hybridization terms as zero for different core-hole potential ($Q$) values. The curves are shifted along the “Intensity” axis.

Figure 7

Calculated oxygen $1s$ XPS spectra for different Bi $6s$/O $p$ hybridization energies ($t_{sp\sigma }$), for $t_{pp\pi }=-0.04$ eV, $t_{pp\sigma }=0.63$ eV, $U=3$ eV, $Q=3.6$ eV, and $E_{\text{CF}}=1$ eV. Multiplets lose their intensities significantly with turning on the bismuth-oxygen hybridization. The curves are shifted along the “Intensity” axis.