Nature of oxygen dopant-induced states in high-temperature ${\mathrm{Bi}}_2{\mathrm{Sr}}_2\mathrm{Ca}{\mathrm{Cu}}_2{\mathrm{O}}_{8+x}$ superconductors: A photoemission investigation

We investigate the nature of oxygen dopant-induced electronic states in the high-temperature ${\mathrm{Bi}}_2{\mathrm{Sr}}_2\mathrm{Ca}{\mathrm{Cu}}_2{\mathrm{O}}_{8+x}$ superconductor by measuring photoemission spectra of underdoped, optimally doped, and overdoped samples using a wide photon energy range $(15–100\mathrm{eV})$. We find a small and broad nondispersive peak at $-0.8\mathrm{eV}$ associated with oxygen dopants. Surprisingly, the detailed analysis of the resonance profile suggests a mixing with Cu states. Particularly, the ${\mathrm{A}}_{1g}$ symmetry revealed by polarization-dependent experiments and the comparison with the single-layered ${\mathrm{Bi}}_2{\mathrm{Sr}}_2\mathrm{Cu}{\mathrm{O}}_{6+x}$ indicates that the oxygen dopant-induced states are mixed to the superconducting $\mathrm{Cu}{\mathrm{O}}_2$ planes through the oxygen dopant—apical oxygen $\mathrm{O}2p_z—\mathrm{Cu}3d_{3z^2-r^2}$ channel.

Figure 1

(Color online) (a) A ${\mathrm{k}}_F$-centered EDC of an OD71 sample along M-Y. (b) A zoom of the M-Y (thin plain line) and $\Gamma \text{−}\mathrm{Y}$ (dotted line) EDCs at ${\mathrm{k}}_F$ compared to the weighted average of the STS typical (60%) and ${\mathrm{O}}_{\delta }$ site (40%) spectra (thick line) from Ref. 2. (c) Weighted averages (20% steps) of the typical and ${\mathrm{O}}_{\delta }$ site STS spectra from Ref. 2. The left vertical dashed line and right vertical dotted line correspond to the $-0.96\mathrm{eV}$ STS peak and to the slope change from the weighted curve (third curve from the bottom), respectively. (d) The second derivative intensity plot of an OP90 sample along $\Gamma \text{−}\mathrm{M}$.

Figure 2

(Color online) (a) The ${\mathrm{k}}_F$-centered EDCs of the UD70 (bottom), OP90 (middle) and OD71 (top) samples EDCs obtained along $\Gamma \text{−}\mathrm{Y}$. Inset: background (dotted curve) associated with the OD71 EDC and the subtraction of the two curves (thick line). (b) Doping dependence of the $-0.8\mathrm{eV}$ peak spectral weight (circles), empirical relation (plain) between ${\mathrm{T}}_c$ and the doping x given in Ref. 6 $({\mathrm{T}}_c^{\max }=95\mathrm{K})$, and proportional relation between the spectral weight and x (dashed line).

Figure 3

(Color online) The nodal ${\mathrm{k}}_F$-centered EDCs are given in (a) for an OD71 sample using the high photon energy range $(28–100\mathrm{eV})$ and in (b) for an OD66 sample using the low photon energy range $(15–23\mathrm{eV})$. The corresponding weight of the $-0.8\mathrm{eV}$ peak is given in panels (c) and (d) for the high and low photon energy ranges, respectively.

Figure 4

(Color online) (a) Photon energy dependence of the nodal quasi-particle peak weight. Inset: Comparison between the nodal ${\mathrm{k}}_F$ EDCs obtained at $16.7\mathrm{eV}$ (dotted line) and $22.7\mathrm{eV}$ (plain line). (b) Polarization dependence of the $-0.8\mathrm{eV}$ spectral weight. (c) Comparison between the Bi2201 (OD7) and Bi2212 (OP91) $M$-centered EDCs. (d) Calculated total and ${\mathrm{O}}_{a}2p_z$ density of states for Bi2201 and Bi2212.