Circular Dichroism in the Angle-Resolved Photoemission Spectrum of the High-Temperature ${\mathrm{Bi}}_2{\mathrm{Sr}}_2{\mathrm{CaCu}}_2{\mathbf{O}}_{8+\delta }$ Superconductor: Can These Measurements Be Interpreted as Evidence for Time-Reversal Symmetry Breaking?

We report first-principles computations of the angle-resolved photoemission response with circularly polarized light in ${\mathrm{Bi}}_2{\mathrm{Sr}}_2{\mathrm{CaCu}}_2{\mathrm{O}}_{8+\delta }$ for the purpose of delineating contributions to the circular dichroism resulting from distortions and modulations of the crystal lattice. Comparison with available experimental results shows that the measured circular dichroism from antinodal mirror planes is reproduced in quantitative detail in calculations employing the average orthorhombic crystal structure. We thus conclude that the existing angle-resolved photoemission measurements can be understood essentially within the framework of the conventional picture, without the need to invoke unconventional mechanisms.

Figure 1

Theoretically obtained photointensity in orthorhombic Bi2212 for emission from the Fermi energy using 21 eV light incident normally to the surface. (a) LH polarized light; (b) RH polarized light; and (c) dichroic intensity obtained by taking the difference (LH-RH) of the maps in (a) and (b). Black denotes low and white denotes high intensity.

Figure 2

Computed dichroic intensity at and near the $M(\pi ,0)$ symmetry point. (a) Dichroic intensity as a function of binding energy for momenta along the black line in Fig. 1c. The horizontal axis denotes $k_{\parallel }$ distance from the $M$ point, where ${\mathbf{k}}_{\parallel }=0$ is the $M$ point. (b) The same as (a), except this panel corresponds to a very small initial state width to highlight various spectral features (AB, BB, SB) discussed in the text. (c) Spectral weight integrated over the binding energy range of 0–0.2 eV [pair of black horizontal lines in (a)] for RH and LH polarized light. (d) Normalized dichroic signal $D_n$, energy-integrated over the 0–0.2 eV binding energy range, for the orthorhombic and tetragonal structures.

Figure 3

Energy-integrated dichroic weight $D_n$ at the $M$ point as a function of geometric distortion. The horizontal axis gives the fractional (in percent) movement of the atoms in going from the tetragonal to the experimental orthorhombic structure at 12 K. Smooth lines have been drawn through the computed points.

Figure 4

(a) Computed energy-integrated intensity in Bi2212 near the $M$ point for the experimentally refined orthorhombic structures at 12 (solid blue lines) and 298 K (dashed red lines) of Ref. 20. The horizontal axis is the same as in Fig. 2 except for a narrower momentum range with $k_{\parallel }=0$ corresponding to the $M$ point. Results are for LH and RH polarized light at 21 eV. (b) Dichroic signal obtained by taking the difference LH-RH from the results in (a).