Prof Radu Coldea

Abstract:

Na$_2$IrO$_3$, a honeycomb 5$d^5$ oxide, has been recently identified as a potential realization of the Kitaev spin lattice. The basic feature of this spin model is that for each of the three metal-metal links emerging out of a metal site, the Kitaev interaction connects only spin components perpendicular to the plaquette defined by the magnetic ions and two bridging ligands. The fact that reciprocally orthogonal spin components are coupled along the three different links leads to strong frustration effects and nontrivial physics. While the experiments indicate zigzag antiferromagnetic order in Na$_2$IrO$_3$, the signs and relative strengths of the Kitaev and Heisenberg interactions are still under debate. Herein we report results of ab initio many-body electronic structure calculations and establish that the nearest-neighbor exchange is strongly anisotropic with a dominant ferromagnetic Kitaev part, whereas the Heisenberg contribution is significantly weaker and antiferromagnetic. The calculations further reveal a strong sensitivity to tiny structural details such as the bond angles. In addition to the large spin-orbit interactions, this strong dependence on distortions of the Ir$_2$O$_2$ plaquettes singles out the honeycomb 5$d^5$ oxides as a new playground for the realization of unconventional magnetic ground states and excitations in extended systems.

Abstract:

We have investigated experimentally and theoretically the series (Na$_{1-x}$Li$_{x}$)$_{2}$IrO$_{3}$. Contrary to what has been believed so far, only for $x\leq0.25$ the system forms uniform solid solutions. For larger Li content, as evidenced by powder X-ray diffraction, scanning electron microscopy and density functional theory calculations, the system shows a miscibility gap and a phase separation into an ordered Na$_{3}$LiIr$_2$O$_{6}$ phase with alternating Na$_3$ and LiIr$_2$O$_6$ planes, and a Li-rich phase close to pure Li$_{2}$IrO$_{3}$. For $x\leq 0.25$ we observe (1) an increase of $c/a$ with Li doping up to $x=0.25$, despite the fact that $c/a$ in pure Li$_{2}$IrO$_{3}$ is smaller than in Na$_{2}$IrO$_{3}$, and (2) a gradual reduction of the antiferromagnetic ordering temperature $T_{N}$ and ordered moment. The previously proposed magnetic quantum phase transition at $x\approx 0.7$ may occur in a multiphase region and its nature needs to be re-evaluated.

Abstract:

We report inelastic neutron scattering measurements on Na2IrO3, a candidate for the Kitaev spin model on the honeycomb lattice. We observe spin-wave excitations below 5 meV with a dispersion that can be accounted for by including substantial further-neighbor exchanges that stabilize zig-zag magnetic order. The onset of long-range magnetic order below 15.3 K is confirmed via the observation of oscillations in zero-field muon-spin rotation experiments. Combining single-crystal diffraction and density functional calculations we propose a revised crystal structure model with significant departures from the ideal 90 deg Ir-O-Ir bonds required for dominant Kitaev exchange.