A preprint study reports that pressure is associated with a much wider spread of magnetic modes in the layered magnet Sr2IrO4. At ambient pressure, two modes differed by approximately 3 cm^-1, or about 0.37 meV. Under pressure, magnon energies spread from 20 to above 60 cm^-1, or 2.5 to 7.5 meV. Magnons are collective wave-like excitations of magnetism. The change was not uniform: top-plane and side-plane modes shifted in opposite directions with pressure and varied from spot to spot.
Researchers examined single crystals using angle-resolved Raman polarization measurements in ab-plane and out-of-plane geometries. The spectra covered 10 to 700 cm^-1, and the modes were followed to 12 GPa in a perforated-diamond anvil cell. The study also varied the laser-spot location, allowing spectra from different sample positions to be compared.
The effect depends on the geometry
In the ab-plane geometry, magnons were mainly found between 10 and 20 cm^-1, although some reached 40 cm^-1 above 8 GPa. Side-plane magnons occupied a higher range, from 20 to 80 cm^-1. Top-plane modes red-shifted with pressure, while side-plane modes blue-shifted, and the response varied significantly from spot to spot. The two scattering configurations were mutually exclusive.
Phonons, the crystal's lattice vibrations, behaved differently. Their energies did not vary significantly with laser-spot position and evolved approximately linearly with pressure in both geometries. Polarization analysis identified four A1g phonon modes at 188, 277, 334 and 561 cm^-1, while side-plane data verified Eg symmetry for the mode at 238 cm^-1.
Polarization favors a conventional reading
At ambient pressure, the 18.0 cm^-1 ab-plane magnon showed a four-fold polarization pattern. The pattern was consistent with B1g-type symmetry in D2h and had the same measured pattern as B2g in D4h. In xc side-plane geometry, the 18 cm^-1 mode retained a circular crossed-polarization response with no nodes, while its intensity varied by approximately plus or minus 20 percent. A global fit separated two xc-parallel modes at 18.4 and 21.3 cm^-1.
The authors say this low-energy Raman response can be accounted for by conventional magnons within the established magnetic symmetry and does not require additional quadrupolar order. That conclusion is an interpretation of the polarization data rather than a direct test of quadrupolar order.
A model ties the modes to tiny couplings
To connect the spectra to microscopic interactions, the researchers used linear spin-wave calculations. The model produced eight magnon modes. Four high-energy modes were nearly degenerate at approximately 40 meV, while the low-energy sector contained two even-parity Raman-active modes and two odd-parity Raman-silent modes.
The model was refined using Raman gaps of 18 cm^-1 for B1g and 21 cm^-1 for B3g. The reported values were Γ1 = 3.0 microelectronvolts, J1c = 12.0 microelectronvolts, J2c = -5.0 microelectronvolts and ∆c = 1.1 microelectronvolts. These are model parameters rather than direct measurements of local stacking configurations or interlayer couplings.
The calculations found that the B3g side-plane magnon was strongly sensitive to J1c and J2c, whereas the B1g in-plane magnon was nearly insensitive to those interlayer exchanges. The authors use this contrast to present Raman-active magnons as amplifiers of very small ground-state-selection interactions.
What the measurements leave open
The evidence has clear boundaries. The number of crystals, measured spots and total spectra is not reported, so sample coverage and reproducibility cannot be quantified. Chemical disorder, residual strain and pressure gradients are not independently separated from the pressure response, and local stacking configurations and interlayer couplings are inferred rather than directly measured. The authors also note that pressure modifies several interactions at once, so changing a single coupling in the model does not directly simulate the experimental pressure transition. A higher-pressure spin-flip transition makes magnons Raman-silent, limiting direct Raman tracking.
No inferential tests, confidence intervals or p-values are reported. The reported shifts and energy ranges are therefore descriptive measurements rather than formal statistical estimates of uncertainty.
The authors interpret the opposing geometry shifts, spot dependence and model sensitivity as evidence that local variations in weak interlayer interactions make competing magnetic stackings visible in the Raman spectrum. That interpretation remains specific to Sr2IrO4 and this experimental system, with the underlying local configurations still inferred rather than directly resolved.
The document is identified as arXiv:2608.25040v1 and dated 25 Aug 2026.
Paper data and sources
Original title: Raman magnon spectroscopy of local interactions and ground state selection in $\mathrm{Sr_2IrO_4}$
Authors: Xiang Li, Scott E. Cooper, Ahmed E. Fahmy et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-25
DOI: Not available
Original paper · Full text