Magnetization dynamics fingerprints of an excitonic condensate t2g4 magnet

Nitin Kaushal, Jacek Herbrych, Gonzalo Alvarez, Elbio Dagotto

Research output: Contribution to journalArticlepeer-review

6 Scopus citations

Abstract

The competition between spin-orbit coupling (SOC) λ and electron-electron interaction U leads to a plethora of novel states of matter, extensively studied in the context of t2g4 and t2g5 materials, such as ruthenates and iridates. Excitonic magnets—the antiferromagnetic state of bounded electron-hole pairs-are prominent examples of phenomena driven by those competing energy scales. Interestingly, recent theoretical studies predicted that excitonic magnets can be found in the ground state of SOC t2g4 Hubbard models. Here we present a detailed computational study of the magnetic excitations in that excitonic magnet, employing one-dimensional chains (via density matrix renormalization group) and small two-dimensional clusters (via Lanczos). Specifically, first we show that the low-energy spectrum is dominated by a dispersive (acoustic) magnonic mode, with extra features arising from the λ=0 state in the phase diagram. Second, and more importantly, we found a novel magnetic excitation forming a high-energy optical mode with the highest intensity at wave-vector q→0. In the excitonic condensation regime at large U, we also have found a novel high-energy π mode composed solely of orbital excitations. These features do not appear all together in any of the neighboring states in the phase diagram and thus constitute unique fingerprints of the t2g4 excitonic magnet, of importance in the analysis of neutron and resonant inelastic x-ray scattering experiments.

Original languageEnglish
Article number235135
JournalPhysical Review B
Volume104
Issue number23
DOIs
StatePublished - Dec 15 2021

Bibliographical note

Publisher Copyright:
©2021 American Physical Society

Fingerprint

Dive into the research topics of 'Magnetization dynamics fingerprints of an excitonic condensate t2g4 magnet'. Together they form a unique fingerprint.

Cite this