[Paper Review] Resonant X-Ray Diffraction Study of Strongly Spin-Orbit-Coupled Mott Insulator CaIrO3

Kenya Ohgushi, Jun‐ichi Yamaura+7 more|arXiv (Cornell University)|Aug 23, 2011

Advanced Condensed Matter Physics4 citations

TL;DR

This study uses resonant x-ray diffraction at the Ir L-edges to demonstrate that CaIrO3 exhibits a spin-orbit-coupled J_eff = 1/2 state and a striped antiferromagnetic order with moments along the c-axis. The results confirm the theoretical prediction of a unique superexchange interaction in the J_eff = 1/2 state, highlighting the universal role of spin-orbit coupling in Ir-based Mott insulators regardless of lattice topology or coordination geometry.

ABSTRACT

We performed resonant x-ray diffraction experiments at the $L$ absorption edges for the post-perovskite-type compound CaIrO$_{3}$ with $(t_{2g})^5$ electronic configuration. By observing the magnetic signals, we could clearly see that the magnetic structure was a striped order with an antiferromagnetic moment along the c-axis and that the wavefunction of a $t_{2g}$ hole is strongly spin-orbit entangled, the $J_{ m eff} =1/2$ state. The observed spin arrangement is consistent with theoretical work predicting a unique superexchange interaction in the $J_{ m eff} =1/2$ state and points to the universal importance of the spin-orbit coupling in Ir oxides, irrespective of the local coordination and lattice topology. We also propose that the non-magnetic resonant scattering is a powerful tool for unraveling an orbital state even in a metallic iridate.

Motivation & Objective

• To investigate the magnetic and orbital structure of the post-perovskite CaIrO3, a strongly spin-orbit-coupled Mott insulator.
• To determine whether the J_eff = 1/2 state, previously observed in layered perovskites like Sr2IrO4, is realized in a different structural environment with edge-shared IrO6 octahedra.
• To test the theoretical prediction of a unique superexchange interaction in the J_eff = 1/2 state, particularly the suppression of antiferromagnetic interaction in edge-shared bonds and its replacement by a highly anisotropic ferromagnetic interaction.
• To establish resonant x-ray scattering as a probe for orbital and spin-orbit entangled states in non-magnetic or metallic iridates.

Proposed method

• Performed resonant x-ray diffraction (RXD) at the Ir L2 and L3 edges using linearly polarized x-rays at beamline BL19LXU, SPring-8.
• Measured intensity dependence on incident x-ray polarization (σ and π) and energy near the L-edge absorption edges to access magnetic and orbital contributions.
• Used energy-resolved RXD to isolate scattering contributions from intermediate states involving t2g and eg orbitals, enabling identification of the J_eff = 1/2 state via intensity ratios.
• Analyzed the anisotropic tensor susceptibility (ATS) scattering intensity to probe the wavefunction character of the t2g hole, particularly the deviation from pure J_eff = 1/2.
• Applied theoretical modeling based on Jackeli and Khaliullin's quantum compass model to interpret the observed magnetic structure and spin canting.
• Used mean-field theory including J1 (corner-shared), J2 (edge-shared), and Dzyaloshinskii-Moriya (D-M) interactions to explain the striped-type magnetic order and weak ferromagnetism.

Experimental results

Research questions

• RQ1 Does the J_eff = 1/2 state exist in CaIrO3, a post-perovskite with edge-shared IrO6 octahedra, despite its different coordination geometry compared to Sr2IrO4?
• RQ2 What is the magnetic structure of CaIrO3, and how does it relate to the predicted superexchange interactions in the J_eff = 1/2 state?
• RQ3 Can resonant x-ray diffraction detect orbital and spin-orbit entangled states in a non-magnetic or metallic iridate system?
• RQ4 Why is the observed uniform magnetic moment (0.07 μB/Ir) significantly smaller than the theoretical expectation (1 μB/Ir) for a fully localized J_eff = 1/2 state?
• RQ5 How do spin-orbit coupling and lattice topology jointly influence the magnetic ground state in Ir-based Mott insulators?

Key findings

• The magnetic structure of CaIrO3 is a striped antiferromagnetic order with moments aligned along the c-axis, confirmed by resonant x-ray diffraction.
• The t2g hole in CaIrO3 is in a J_eff = 1/2 state, with the A coefficient in the wavefunction constrained to 0.87 < A < 1.0, indicating a slightly modified J_eff = 1/2 state.
• The observed spin arrangement is consistent with the theoretical prediction of a unique superexchange interaction: antiferromagnetic via corner-shared bonds and ferromagnetic, anisotropic via edge-shared bonds.
• The weak ferromagnetic moment along the b-axis (0.07 μB/Ir) is explained by spin canting due to Dzyaloshinskii-Moriya interactions and anisotropic axes in the quantum compass model.
• The canted angle of ~4° derived from the observed moment is significantly smaller than the theoretical mean-field prediction of ~23°, suggesting effects from quantum fluctuations or deviations from pure J_eff = 1/2.
• Anisotropic tensor susceptibility (ATS) scattering is established as a powerful probe for orbital character in metallic iridates, even in the absence of long-range magnetic order.