Role of lone pair electrons in determining the optoelectronic properties of BiCuOSe

The electronic structure of the oxychalcogenides LaCuOSe and BiCuOSe has been studied using O $K$-edge x-ray emission spectroscopy, x-ray absorption spectroscopy, and density functional theory, in order to examine the effects of the M${}^{3+}$ ion configurations. The known distortion of the BiO layers in BiCuOSe compared to the LaO layers in LaCuOCh; the significantly smaller band gap of BiCuOSe (0.9 eV) compared to LaCuOSe (2.8 eV); and similar hole transport properties of the two compounds are explained in terms of the electron lone pairs associated with the Bi $d^{10}{s}^2{p}^0$ electronic configuration. The Bi $6s$ orbitals are chemically active and form bonding and antibonding states with the oxygen $2p$ orbital. The structural distortion facilitates the interaction between the $6p$ orbital with $6s$ via the antibonding state. For BiCuOSe, the majority of the Bi $6s$ orbital character (i.e., the bonding state) lies below the valence band, with the antibonding state lying below the valence band maximum (VBM). The similar hole transport properties between the two compounds is a consequence of the Bi $6s$ contributing little to the Cu $3d$–Se $4p$ derived VBM. Finally, the band gap narrowing of BiCuOSe compared to LaCuOSe is mostly due to the low energy of the unoccupied Bi $6p$ orbitals along with the upshift of the VBM due to the presence of the O $2p$–Bi $6s$ antibonding states.

Figure 1

(Top) The DFT calculated PDOS of the valence and conduction band regions of BiCuOSe. The energy axis is referenced with respect to the VBM, such that zero eV is the VBM. The major orbital contribution is highlighted; the * denotes unoccupied orbital state. (Bottom) The corresponding calculated PDOS for LaCuOSe.

Figure 2

(Top) BiCuOSe O $K$-edge XES (solid circles) and core-hole corrected XAS (open circles) data plotted against the second derivatives (underneath) of the corresponding least-squares fits (lines). The separation between the band edges is shown (vertical dashed lines), as determined from the separation of the maxima of the second derivatives. The inset displays a magnified region of the XES and corresponding peak fit associated with O 2$p$–Bi 6$s$ bonding state. (Bottom) The corresponding plot for LaCuOSe; note the lack of spectral weight in the inset.

Figure 3

(Top) BiCuOSe XES data plotted against Gaussian broadened DFT calculated O $2p$ and Bi $6s$ PDOS. The DFT has been rigidly shifted for the calculated VBM (zero binding energy) to match the experimentally determined VBM at 529.38 eV (photon energy). (Bottom) LaCuOSe XAS data plotted against the Gaussian broadened O $2p$ and La $5p$ PDOS, rigidly shifted by 528.54 eV (i.e., to match the experimentally determined VBM of LaCuOSe). The locations of the bonding (B) and antibonding states (AB) are identified in each case.

Figure 4

(Top) BiCuOSe XAS data plotted against Gaussian broadened DFT calculations of Bi $6p$ and O $2p$ contributions. (Bottom) LaCuOSe XAS data plotted against Gaussian broadened DFT calculations of La $5d$, Cu $4s$, and O $2p$ contributions.