[Paper Review] Multiple Energy Scales and Anisotropic Energy Gap in the Charge-Density-Wave Phase of Kagome Superconductor CsV3Sb5

K. Nakayama, Yongkai Li+6 more|arXiv (Cornell University)|Apr 16, 2021

Topological Materials and Phenomena29 citations

TL;DR

The study uses high-resolution ARPES to reveal SOC-induced Dirac gaps and a strongly anisotropic, multi-scale CDW gap in CsV3Sb5, with distinct gaps on saddle-point, Dirac, and Γ/AP electron-pocket bands.

ABSTRACT

Kagome metals AV3Sb5 (A = K, Rb, and Cs) exhibit superconductivity at 0.9-2.5 K and charge-density wave (CDW) at 78-103 K. Key electronic states associated with the CDW and superconductivity remain elusive. Here, we investigate low-energy excitations of CsV3Sb5 by angle-resolved photoemission spectroscopy. We found an energy gap of 70-100 meV at the Dirac-crossing points of linearly dispersive bands, pointing to an importance of spin-orbit coupling. We also found a signature of strongly Fermi-surface and momentum-dependent CDW gap characterized by the larger energy gap of maximally 70 meV for a band forming a saddle point around the M point, the smaller (0-18 meV) gap for a band forming massive Dirac cones, and a zero gap at the Gamma-centered electron pocket. The observed highly anisotropic CDW gap which is enhanced around the M point signifies an importance of scattering channel connecting the saddle points, laying foundation for understanding the nature of CDW and superconductivity in AV3Sb5.

Motivation & Objective

• Identify low-energy electronic states associated with CDW and superconductivity in CsV3Sb5.
• Characterize SOC-induced gaps at Dirac points in the band structure.
• Map the momentum and band dependence of the CDW gap across the 2D Brillouin zone.
• Relate observed CDW gaps to Fermi surface features and potential scattering channels.

Proposed method

• Perform high-resolution ARPES on CsV3Sb5 single crystals across temperatures above and below TCDW.
• Vary photon energy to probe different kz slices and distinguish bulk from surface features.
• Compare experimental band dispersions with first-principles calculations including SOC.
• Extract EDCs and symmetrized EDCs to identify energy gaps at key kF points.
• Analyze CDW gap as a function of momentum angle θ and assign gaps to saddle-point, Dirac, and Γ/AP pockets.

Experimental results

Research questions

• RQ1 What are the energy gaps opening at Dirac points due to SOC in CsV3Sb5?
• RQ2 How does the CDW gap vary across the Fermi surface and with momentum (k) in CsV3Sb5?
• RQ3 Which bands (saddle-point, Dirac, Γ/AP electron pocket) participate in the CDW, and with what magnitudes?
• RQ4 Is there evidence for a 3D component of CDW and how does it relate to kz dependence?

Key findings

• Dirac-crossing points exhibit an energy gap of 50–70 meV attributed to SOC.
• CDW gaps are strongly momentum-dependent with multiple energy scales: saddle-point band up to ~70 meV, Dirac band 0–18 meV, and Γ/AP electron pocket showing no gap.
• A hump around 70 meV in EDCs at the saddle-point kF point indicates a CDW gap opening on the saddle-point band.
• The Dirac-band gap (~16 meV at some points) is anisotropic and can vanish along certain directions (ΓK/AL line).
• CDW gap is absent on the Γ/AP electron pocket across measured cuts.
• Anisotropy places maximum gaps near θ ≈ 0° (ΓM/AL) and minimum near θ ≈ ±30° (ΓK/AH) for both saddle-point and Dirac bands.
• STM/STS and optical spectroscopy signatures (broad hump ~70 meV, V-shaped DOS, Drude weight suppression) align with ARPES-observed CDW scales.
• The data support inter-band scattering between saddle points via Q=(π,0) as a major energy-gain mechanism for CDW.
• A 3D CDW component (2×2×2) remains puzzling and may require further kz-selective studies.
• The results imply that superconductivity may be influenced by the largely gapped saddle-point band and CDW gap nodes on Dirac bands.