Intrinsic suppression of topological thermal Hall effect in an exactly solvable quantum magnet

TL;DR

This study investigates the topological thermal Hall effect in a quantum magnet, revealing that inter-bosonic interactions suppress the effect, challenging the applicability of non-interacting topological models to bosonic systems.

Contribution

The paper provides the first experimental and theoretical analysis showing that interactions in bosonic topological systems significantly suppress the thermal Hall effect, contrasting with non-interacting predictions.

Findings

Measured no discernible thermal Hall conductivity in SrCu₂(BO₃)₂

Calculated sign of conductivity is negative and much smaller than previous models

Inter-bosonic interactions strongly influence topological transport properties

Abstract

In contrast to electron (fermion) systems, topological phases of charge neutral bosons have been poorly understood despite recent extensive research on insulating magnets. The most important unresolved issue is how the inevitable inter-bosonic interactions influence the topological properties. It has been proposed that the quantum magnet SrCu$_2$(BO$_3$)$_2$ with an exact ground state serves as an ideal platform for this investigation, as the system is expected to be a magnetic analogue of a Chern insulator with electrons replaced by bosonic magnetic excitations (triplons). Here, in order to examine topologically protected triplon chiral edge modes in SrCu$_2$(BO$_3$)$_2$, we measured and calculated the thermal Hall conductivity $\kappa_{xy}$. Our calculations show that the sign of $\kappa_{xy}$ is negative, which is opposite to the previous calculations, and its magnitude is 2$\pi$ times smaller. No discernible $\kappa_{xy}$ was observed, which is at most 20-30% of our calculations if present. This implies that even relatively weak inter-particle interactions seriously influence the topological transport properties at finite temperatures. These demonstrate that, in contrast to fermionic cases, the picture of non-interacting topological quasi-particles cannot be naively applied to bosonic systems, calling special attention to the interpretation of the topological bosonic excitations reported for various insulating magnets.