Discovering topological phases in gray-Tin

Non-trivial topological phases often emerge in narrow-gap semiconductors with a delicate blend of spin-orbit coupling and electron correlation. The diamond-lattice allotrope of Sn ($α$-Sn) exemplifies this behavior, hosting multiple topological phases that can be tuned by small distortions in the lattice. Despite rapid experimental progress, theoretical descriptions of $α$-Sn lack predictive power and rely mainly on tight-binding models and density functional theory with uncontrolled approximations. We employ first-principles fully self-consistent, relativistic GW (scGW) to overcome these limitations. The scGW recovers the experimentally observed zero-gap semiconductor and the strain-induced topological insulator and Dirac semimetal phases, while also predicting new trivial and topological insulators and a Dirac semimetal phase, further demonstrating the versatility of $α$-Sn for band engineering. Additionally, we propose a robust diagnostic of topological behavior based on a combined analysis of band and orbital-occupation dispersions, tailored for correlated methods where standard mean-field-based topological invariants fall short. Our findings pave the way for studying a broad class of topological materials using accurate first-principles methods beyond density functional theory.

💡 Research Summary

This paper presents a comprehensive first‑principles study of gray‑tin (α‑Sn), a diamond‑lattice narrow‑gap semiconductor whose electronic structure is highly sensitive to spin‑orbit coupling (SOC) and lattice distortions. The authors employ a fully relativistic, fully self‑consistent GW (scGW) approach, implemented with the exact two‑component (X2C) formalism, to overcome the well‑known deficiencies of density‑functional theory (DFT) in describing α‑Sn. Unlike DFT, which often predicts an unphysical negative Γ–L band gap and shows strong dependence on the chosen exchange‑correlation functional, scGW yields a zero‑gap semimetal at the experimental lattice constant (≈6.49 Å) and correctly reproduces the subtle evolution of the band structure under both isotropic and anisotropic strain.

Key technical achievements include: (i) a systematic treatment of electron‑electron correlation via orbital‑dependent screened Coulomb interactions, (ii) inclusion of scalar relativistic effects and SOC within the X2C framework, and (iii) a robust post‑processing pipeline that transforms the Green’s function to a symmetrized atomic‑orbital (SAO) basis, allowing the extraction of both spectral functions and orbital occupation numbers. The authors demonstrate that scGW converges independently of the DFT starting point, eliminating the “one‑shot” dependence that plagues G₀W₀ calculations.

The isotropic‑strain analysis reveals a rich sequence of phases as the lattice constant is reduced from 6.7 Å to 6.4 Å. At the experimental lattice constant the 5s and 5p orbitals undergo a double inversion driven by SOC, producing the well‑known topological‑insulator (TI) surface states. When the lattice is compressed below ≈6.55 Å, the 5s band moves above the Fermi level, opening a trivial semiconductor gap. In the narrow window 6.55–6.60 Å, the renormalized 5s and 5p bands cross linearly at the Fermi level, forming a three‑dimensional Dirac semimetal (3D‑TDS). Further compression (≈6.4 Å) restores a trivial insulating state with an X‑shaped valence band.

Uniaxial strain along the