Deep-lying semi-Dirac fermions in hexagonal close-packed cadmium
Alaska Subedi, Kamran Behnia
TL;DR
Problem: identify deep-lying semi-Dirac fermions in a real metal and explain their origin. Approach: perform first-principles LDA calculations on hexagonal Cd to reveal linearly dispersing out-of-plane bands near $-3$ eV, with hybridization between $s$ and $p_z$ orbitals; orbital analysis shows direction-dependent mixing. Key findings: a lens-shaped electron pocket with cross-section $A(k_z)$ linearly dependent on $k_z$ and slope $dA/dk_z oughly 8.73$ Å$^{-1}$, matching Sondheimer oscillation data; comparison with hypothetical hcp Ag shows chemistry controls the semi-Dirac dispersion and tensile strain can tune the inversion. Significance: demonstrates a concrete solid-state realization of semi-Dirac fermions governed by orientation-dependent hybridization and strain engineering, with observable signatures in quantum oscillations including the Sondheimer effect.
Abstract
Semi-Dirac fermions are massless in one direction and massive in the perpendicular directions. Such quasiparticles have been proposed in various contexts in condensed matter. Using first principles calculations, we identify a pair of semi-Dirac bands anti-crossing at $-3$ eV below the Fermi level in the electronic structure of hexagonal close-packed cadmium. The linear out-of-plane dispersion is kept up to the Fermi level. We demonstrate that the dichotomy between the linear and quadratic dispersions is driven by an orientation-sensitive hybridization between the $s$ and $p_z$ orbitals. The upper semi-Dirac band produces a lens-shaped nonellipsoidal Fermi sheet whose cross-section area has a $k$-dependence that is in excellent agreement with the experimentally measured period of Sondheimer oscillations.
