返回文章

Figure 1.  (a) Electronic band structure and projected density of states of CsCr₃Sb₅ without SOC at 5 GPa. The Fermi level is set to zero. The dashed black lines denote the band structure derived from the effective tight-binding model. (b) Crystal structure of CsCr₃Sb₅. Orbital-resolved density of states projected onto the FS of CsCr₃Sb₅ (c) without SOC and (d) with SOC (»_SOC = 0.025 eV), obtained from the effective tight-binding model. The color-coded circles in (a), (c), and (d) represent orbital contributions. The symbols α, β, and γ denote three distinct bands.

Figure 2.  Phase diagram depicting the leading instability channels within the RPA framework: (a) without SOC and (b) with SOC (»_SOC = 0.025 eV). The solid lines represent the phase boundaries, while the dashed lines indicate the crossover between the A_2g and E_2g pairing symmetry channels. SC and DW denote the superconducting and density-wave instabilities, respectively. (c) The leading eigenvalues λ of the linearized gap equations as a function of Hubbard interaction U.

Figure 3.  (a) Superconducting gap function in the A_2g channel for U = 1.0853 eV, J_H/U = 0.02, and λ_SOC = 0 eV, corresponding to the largest eigenvalue λ = 1.0 of the linearized gap equations. (b), (c) Two degenerate gap functions in the E_2g channel for U = 0.9964 eV, J_H/U = 0.2, and λ_SOC = 0 eV, also with λ = 1.0. (d), (g) Maximum eigenvalues λχmax of the particle–hole susceptibility χ^(q,0). The black line denotes the boundary of the first Brillouin zone, and the white arrows mark the wave vectors corresponding to susceptibility maxima. (e), (h) Static spin susceptibilities and eigenvalues λχmax, λχ0max. Both bare and dressed spin susceptibilities are scaled by 1/18, with the dressed susceptibilities further shown on a logarithmic scale. Arrows highlight the positions of the prominent peaks. The inset in (e) shows a zoomed-in view of the bare susceptibilities. (f), (i) Singlet component of the pairing kernel for the A_2g and E_2g channels, respectively. The parameters sets (J_H/U, U, λ_SOC) used in (d), (e), (f) and (g), (h), (i) are the same as those in (a) and (b), (c), respectively.

Figure 4.  (a), (b) Two degenerate gap functions in the E_2g channel. (c) Maximum eigenvalues of the particle–hole susceptibility χ^(q,0). The first Brillouin-zone boundary is outlined in black, and the dominant scattering vectors are indicated by white arrows. (d) Singlet component of the pairing kernel projected onto the leading E_2g channel. (e) Static spin susceptibilities plotted together with the leading eigenvalues λχmax (λχ0max). For this combined visualization, the bare spin susceptibilities are divided by 18, while the dressed spin susceptibilities are plotted as the logarithm of their values following the same scaling. Arrows mark the positions of the maxima. The inset shows a zoomed-in view of the bare susceptibilities. All results are for U = 0.98675 eV, J_H/U = 0.2, and »_SOC = 0.025 eV, with largest eigenvalue λ = 1.0.

Figure 5.  Dominant components |v_lν(q)|² of the eigenvector corresponding to the largest eigenvalue λχmax of the particle–hole susceptibility χ^(q,0) at Q₁ and Q₂ as a function of J_H/U for U = 0.9964 eV in the absence of SOC. The blue solid (dashed) lines denotes −D(q) (see main text) at Q₁ (Q₂), plotted against the right-hand y-axis as indicated by the blue arrow. (b) Dressed orbital (χ^O) and spin (χ^S) susceptibility for U = 0.98675 eV, J_H/U = 0.2, and »_SOC = 0.025 eV. The dressed orbital susceptibility χ^O is scaled by a factor of 20 for visual comparison.