Session W29: Unique Physical Properties and Characterization of Kitaev Spin Liquid Candidate Materials

Recent results on Kitaev interactions in Co based magnets.

Kitaev quantum spin liquids (QSLs) are exotic states of matter that are predicted to host Majorana fermions and gauge flux excitations. However, so far all known Kitaev QSL candidates are known to have appreciable non-Kitaev interactions that pushes these systems far from the QSL regime.   Co based magnets have been proposed to be perhaps a more ideal platform for realizing Kitaev QSLs.   In this talk I will show evidence for a Kitaev interactions in both the quasi-one-dimensional ferromagnet CoNb₂O₆ as well as the hexagonal magnet BaCo₂(AsO₄).   Although it is usually believed to be the best material realization of a 1D Ising chain, by combining terahertz spectroscopy and calculations we have shown that CoNb₂O₆ is well described by a model with bond-dependent interactions.  We call this model the ‘twisted Kitaev chain’, as these interactions are similar to those of the honeycomb Kitaev spin liquid. The ferromagnetic ground state of CoNb₂O₆ arises from the compromise between two axes. Owing to this frustration, even at zero field domain walls have quantum motion, which is described by the celebrated Su–Schriefer–Heeger model of polyacetylene and shows rich behavior as a function of field.   Most recently, we have shown also that the honeycomb cobalt-based Kitaev QSL candidate, BaCo₂(AsO₄)₂, has dominant Kitaev interactions. Due to only small non-Kitaev terms a magnetic continuum consistent with Majorana fermions and the existence of a Kitaev QSL can be induced by a small out-of-plane-magnetic field.  Our results demonstrate BaCo₂(AsO₄)₂, as a far more ideal version of Kitaev QSL compared with other candidates.   

Characterizing Extended Kitaev Models Under a Magnetic Field

The Kitaev honeycomb model and its extensions across dimensions, spin length, and additional interactions has been the subject of intense theoretical research. Initially an academic problem, the extended Kitaev model has found relevance in a large number of material candidates including the promising honeycomb α-RuCl₃.

Recently there has been interest in low-dimensional analogues of the extended Kitaev model, as they exhibit rich phase diagrams with SPTs, chiral order, higher-rank magnetism, and Luttinger liquids. In addition to being more amenable to numerical study, their phase diagrams have some remarkable similarities to the 2D honeycomb model and may provide further insights into the nature of the 2D phases.

Aligned with this program, in this talk I summarize recent numerical and analytical work on these extended Kitaev models. After discussing the anisotropic spin-S Kitaev chain, I go on to talk about quasi-1D Kitaev systems under magnetic fields, as well as a difficult corner of the honeycomb phase diagram thought

to be relevant for α-RuCl₃. Finally, I discuss a symmetry-based thermodynamic signature of topological phase transitions, applicable to a number of candidate materials.   

Neutron Scattering Study of the Kitaev Spin-Liquid Candidate Material BaCo2(AsO4)2

Recently, a new generation of Kitaev materials has been proposed that is based on the low spin configuration 3d⁷ Co²⁺. We present an inelastic single-crystal neutron scattering study of one such spin-liquid candidate material BaCo₂(AsO₄)₂ (BCAO). This material has attracted interest due to the observation of the suppression of magnetic order into a paragagnet-like phase above an in-plane critical field of H_ab≈0.55 T. This melting of order with field is reminiscent of α-RuCl₃, where the high field state is proposed to be a quantum spin-liquid.  Structural considerations and the small critical field make BCAO an appealing Kitaev spin-liquid candidate. The observed magnetic excitations in a fully field-polarized phase are modeled using linear spin-wave theory, and accompanying bulk characterization studies are used to determine the exchange interactions and place the material within the generalized phase diagram of the K-Γ-Γ'-J model.