Room temperature control of axial and basal antiferromagnetic anisotropies using strain

TL;DR

The paper addresses the challenge of deterministically controlling antiferromagnetic order in thin films by leveraging strain in free-standing $\alpha$-Fe$_2$O$_3$ membranes. The authors apply both isotropic and anisotropic in-plane strains and map the resulting nanoscale AFM order using XMLD-STXM, supported by a Landau model with magneto-elastic terms and MuMax3 micromagnetic simulations. They demonstrate that isotropic strain can reverse the axial Morin transition and favor the easy-plane state, while anisotropic strain induces a uniaxial basal anisotropy, reconfiguring trigonal domains yet preserving topological textures such as merons and antimerons.Overall, strain emerges as a versatile, reversible control knob for engineering multi-axial AFM anisotropies and textures at room temperature, with implications for reconfigurable AFM spintronic and magnonic devices.

Abstract

Antiferromagnetic materials are promising platforms for the development of ultra-fast spintronics and magnonics due to their robust magnetism, high-frequency relativistic dynamics, low-loss transport, and the ability to support topological textures. However, achieving deterministic control over antiferromagnetic order in thin films is a major challenge, due to the formation of multi-domain states stabilised by competing magnetic and destressing interactions. Thus, the successful implementation of antiferromagnetic materials necessitates careful engineering of their anisotropy. Here, we demonstrate strain-based robust control over multiple antiferromagnetic anisotropies and nanoscale domains in the promising spintronic candidate a-Fe2O3, at room temperature. By applying isotropic and anisotropic in-plane strains across a broad temperature-strain phase space, we systematically tune the interplay between magneto-crystalline and magneto-elastic interactions. We discover that strain-driven control steers the system towards an aligned antiferromagnetic state, whilst preserving topological spin textures, such as merons, antimerons and bimerons. We directly map the nanoscale antiferromagnetic order using linear dichroic scanning transmission X-ray microscopy integrated with in situ strain and temperature control. A Landau model and micromagnetic simulations reveal how strain reshapes the magnetic energy landscape. These findings suggest that strain could serve as a versatile control mechanism to reconfigure equilibrium or dynamic antiferromagnetic states on demand in a-Fe2O3, paving the way for next-generation spintronic and magnonic devices.

Figures (5)

• Figure 1: Strain distribution in a square membrane. a) Example of a simulated deflection profile for a square membrane (dimensions: 1$\times$1 mm$^2$). b-d) Corresponding maps of the strain component (b) the horizontal component $\varepsilon_{11}$, (c) the vertical component $\varepsilon_{22}$, and (d) the shear component $\varepsilon_{12}$. The red box marks the 50x50 $\mathrm{\mu}$m region about the centre within which imaging was performed. The strain is roughly uniform $\varepsilon_{11} \approx \varepsilon_{22}$ close to the centre of the membrane, and the shear component $\varepsilon_{12}$ is negligible. e) The symmetric membrane strain at the centre of a square membrane estimated from deflection using equation \ref{['eqn:StrainCali']}. The dashed line is the simulated symmetric strain as a function of the membrane deflection.
• Figure 2: Effect of isotropic strain. (a-d) 10x10 $\mu$m XMLD-STXM images obtained at 280 K on a membrane with an inherent (zero-strain) transition close to 309 K. Purple and orange colours correspond to OOP and IP Néel vector orientations respectively. (e) Phase diagram of the magnetic state as a function of temperature and strain. The solid black line is a linear fit of the transition temperature as a function of applied strain. The dashed blue curve is a higher-order (cubic) fit that better matches the data at higher strain. The black circles indicate the positions in the phase diagram corresponding to experimental data.
• Figure 3: Strain distribution in a rectangular membrane. a) Example of a simulated deflection profile for a rectangular membrane (dimensions: 1$\times$0.25 mm$^2$). b-d) Corresponding maps of the strain component across (b) the short, high-strain axis $\varepsilon_{11}$, (c) the long, low-strain axis $\varepsilon_{22}$, and (d) the shear component $\varepsilon_{12}$. The red box marks the 50x50 $\mathrm{\mu}$m region about the centre within which imaging was performed. The strain is roughly spatially uniform (but anisotropic) close to the centre of the membrane. e) Strain components at the membrane centre estimated from the measured deflection using equation \ref{['eqn:StrainCali']}. The dashed lines are the simulated strain components as a function of the membrane deflection.
• Figure 4: Effect of anisotropic strain. (a-c) Vector maps and (d-f) corresponding pixel-wise distribution of phase angle in the same region of an $\alpha$-Fe$_2$O$_3$ membrane near the centre of a rectangular Si$_3$N$_4$ holder. The images correspond to three different applied pressures: (a,d) 0 mbar unstrained state, (b,e) 400 mbar, $\varepsilon_{11} = 0.35\%$ and (c,f) 800 mbar, $\varepsilon_{11} = 1.23\%$. Red/green/blue colours in both the vector maps and pole plots correspond to the three trigonal domains and the local Néel vector orientation is indicated by white bars in the vector maps. Regions identified as topological merons and antimerons are shown by yellow circles and blue squares, respectively. The grey regions are defects on the sample surface. The white scale bar is 1 $\mathrm{\mu m}$ long.
• Figure 5: Micromagnetic simulations. Pole plots of the angular distribution for simulated merons with different values of strain-induced anisotropy $K_{\varepsilon}$ and basal plane anisotropy $K_\text{B}$.