Evolution of the electronic structure in Ta$_2$NiSe$_5$ across the structural transition revealed by resonant inelastic x-ray scattering

TL;DR

This study uses resonant inelastic X-ray scattering to investigate the electronic structure and phase transition in Ta$_2$NiSe$_5$, revealing a direct band gap at low temperature and structural influences on the transition.

Contribution

It provides element-specific insights into the electronic structure across the phase transition, highlighting the largely structural nature of the transition with minimal excitonic contribution.

Findings

Confirmation of a direct band gap at low temperature

Persistence of conduction and valence band features across the transition

Evidence supporting a primarily structural phase transition

Abstract

We utilized high-energy-resolution resonant inelastic X-ray scattering (RIXS) at both the Ta and Ni $L_3$-edges to map out element-specific particle-hole excitations in Ta$_2$NiSe$_5$ across the phase transition. Our results reveal a momentum dependent gap-like feature in the low energy spectrum, which agrees well with the band gap in element-specific joint density of states calculations based on ab initio estimates of the electronic structure in both the low temperature monoclinic and the high temperature orthorhombic structure. Below $T_c$, the RIXS energy-momentum map shows a minimal gap at the Brillouin zone center ($\sim$ 0.16 eV), confirming that Ta$_2$NiSe$_5$ possesses a direct band gap in its low temperature ground state. However, inside the gap, no signature of anticipated collective modes with an energy scale comparable to the gap size can be identified. Upon increasing the temperature to above $T_c$, whereas the gap at the zone center closes, the RIXS map at finite momenta still possesses the gross features of the low temperature map, suggesting a substantial mixing between the Ta and Ni orbits in the conduction and valence bands, which does not change substantially across the phase transition. Our experimental observations and comparison to the theoretical calculations lend further support that the phase transition and the corresponding gap opening in Ta$_2$NiSe$_5$ is largely structural by nature with possible minor contribution from the putative exciton condensate.