Experimental Evidence of Néel-order-driven Magneto-optical Kerr Effect in an Altermagnetic Insulator

TL;DR

Problem: the origin of the large MOKE in hematite is unclear due to mixed contributions from Néel order, net magnetization, and external fields. Approach: integrate broadband MOKE measurements with magnetometry, symmetry analysis, and first-principles calculations, using photon-energy dependence and single-domain behavior to disentangle contributions; express the transverse optical conductivity as $\vec{\sigma}^A = \boldsymbol{\alpha}\cdot\vec{N} + \boldsymbol{\beta}\cdot\vec{M} + \boldsymbol{\gamma}\cdot\vec{H}$ and compare first- vs higher-order SOC effects. Findings: the Néel-order term $\boldsymbol{\alpha}\cdot\vec{N}$ dominates the MOKE signal, while $\boldsymbol{\beta}\cdot\vec{M}$ and $\boldsymbol{\gamma}\cdot\vec{H}$ are negligible in the visible range; first-principles calculations show $\alpha N$ larger than $\beta M_S$ and the field-related contribution is small, with the Néel effect arising at first order in SOC and magnetization at higher order. Significance: establishes the altermagnetic origin of the large MO response in hematite and motivates exploration of altermagnetic insulators for zero-field optical devices and integrated photonic applications.

Abstract

The magneto-optical Kerr effect (MOKE) is investigated in hematite, a collinear antiferromagnetic insulator, across a broad wavelength spectrum. By combining the optical measurements with magnetometry results, we unambiguously demonstrate that the Néel-order contribution dominates the MOKE signal, while contributions from net magnetization and external magnetic fields are negligible. This conclusion is quantitatively supported by first-principles calculations, and qualitatively by a symmetry analysis that the Néel contribution appears at the first order in spin-orbit coupling while the magnetization contribution starts only at the third order. This study clarifies the altermagnetic origin of the pronounced MOKE in hematite, underscoring the potential of altermagnets as a promising new class of magneto-optical materials.

Figures (4)

• Figure 1: (a) Illustration of field-sweeping transverse optical conductivity $\sigma^A_{i}$ results. (b) Crystal structure of hematite in hexagonal lattice and definition of Cartesian coordinate system $xyz$, where $x\parallel [11\bar{2}0]$, $y\parallel [\bar{1}100]$ and $z\parallel [0001]$. (c), (d), (e) Magnetic symmetries and corresponding $\vec{\sigma}^A$ of hematite induced by Néel order in rhombohedral lattice, when the $\vec{N}$ is oriented along $x$, $y$ and $z$ directions. The relative positions of Fe atoms are marked in (b). The glide plane $g$ consists of mirror reflection against the gray plane and translation along the $c$ axis, denoted by black arrows. $\hat{T}$, $C_3$ and $C_2$ represent the time reversal, three-fold rotation and two-fold rotation, respectively.
• Figure 2: (a), (b) Magnetic hysteresis of hematite along the $x$ and $z$ directions at room temperature. (c), (d) MOKE configuration of hematite $(11\bar{2}0)$ and $(0001)$ samples. The incident angle is 0.2 deg. (e), (f) $\sigma^A_{x}$ and $\sigma^A_{z}$ of hematite at the wavelength of 550 nm, when $H$ is applied along $x$ and $z$ directions, respectively. The length of error bar is 0.72 S/cm.
• Figure 3: (a) $H_x$ dependency of $\sigma^A_{x}$ at different wavelengths, measured in the hematite $(11\bar{2}0)$ sample. Light lines correspond to the real part, while dark lines correspond to the imaginary part. (b),(c) Intercept and slope of $\sigma^A_{x}$ spectrum. (d),(e),(f) Calculated Néel-order-induced conductivity $\alpha N$, magnetization-induced conductivity $\beta_{xx}M_S$ and field-induced conductivity $\gamma_{xx}H_0$, where $M_S=2.13\times10^3\ \rm A/m$, $H_0=1\ \rm T$.
• Figure 4: (a) $H_z$ dependency of $\sigma^A_{z}$ at different wavelengths, measured in the hematite $(0001)$ sample. Light lines correspond to the real part, while dark lines correspond to the imaginary part. (b),(c) Intercept and slope of $\sigma^A_{z}$ spectrum.