Quasiparticle interference and spectral function of the UTe$_2$ superconductive surface band

We compute the (0-11) surface spectral function, the surface density of states (DOS), and the quasiparticle interference (QPI) patterns, both in the normal state and superconducting (SC) state of UTe$2$. We consider all possible non-chiral and chiral order parameters (OPs) that could in principle describe the superconductivity in this compound. We describe the formation of surface states whose maximum intensity energy depends on the nature of the pairing. We study also the QPI patterns resulting from the scattering of these surface states. We show that the main feature distinguishing between various OPs is a QPI peak that is only observed experimentally in the superconducting state. The energy dispersion and the stability of this peak is consistent among the non-chiral OPs only with a $B{3u}$ pairing. Moreover, $B_{3u}$ is the only non-chiral pairing that shows a peak at zero energy in the DOS, consistent with the experimental observations.

💡 Research Summary

This paper presents a comprehensive theoretical study aimed at resolving the long-standing debate over the superconducting order parameter symmetry in the heavy-fermion superconductor UTe2. The research employs calculations of surface spectral functions and quasiparticle interference (QPI) patterns on the experimentally relevant (0-11) surface to distinguish between all possible triplet pairing symmetries.

The authors construct a four-orbital tight-binding model for UTe2, incorporating two uranium and two tellurium orbitals per unit cell. The model parameters are tuned to reproduce two cylindrical Fermi surface sheets aligned along the c-axis, consistent with quantum oscillation experiments. Within this model, they consider the four possible odd-parity (triplet) irreducible representations of the D2h point group (Au, B1u, B2u, B3u) for a single-component order parameter, as well as six possible chiral combinations formed by pairing two non-chiral representations with a π/2 phase difference.

The core methodology involves computing the surface Green’s function for the (0-11) surface by treating it as the limit of a plane-like impurity with infinite potential. This surface Green’s function is then used within the T-matrix formalism to calculate the QPI patterns generated by a point-like impurity located on that surface. This approach allows for a direct simulation of scanning tunneling microscopy (STM) experiments.

The key findings are twofold. First, the analysis of the surface density of states (DOS) reveals a crucial distinguishing feature: among the non-chiral pairings, only the B3u symmetry produces a pronounced peak at zero energy (E=0). All other non-chiral pairings (Au, B1u, B2u) result in surface spectral weight maxima at finite energies, leading to a double-peak structure in the DOS with a dip at zero energy. The presence of a zero-bias peak in experimental STM spectra is a strong point in favor of the B3u scenario.

Second, and more decisively, the calculated QPI patterns provide a fingerprint for the pairing symmetry. While many QPI peaks are common to different pairings or even the normal state, the study identifies a specific QPI peak whose appearance and stability are unique to the superconducting state. The energy dispersion and robustness of this particular peak are consistent with experimental observations only for the B3u order parameter. The chiral pairings considered yield QPI patterns incompatible with the experimental data.

The paper concludes that the combination of a zero-bias peak in the surface DOS and the characteristic, stable QPI peak observed only in the superconducting state provides compelling theoretical evidence that UTe2 is a single-component, non-chiral triplet superconductor with B3u pairing symmetry. This result rules out earlier proposals of chiral two-component order parameters and distinguishes B3u from the other possible non-chiral symmetries, offering a resolution to a central question in the understanding of this intriguing material.