CuMnAs: Antiferromagnetic Metal for Spintronics

• Antiferromagnetic metal CuMnAs is a Mn-based compound with high Néel temperature and a tunable semimetallic-to-semiconducting electronic structure, making it central to spintronics research.
• Its multiple crystal polymorphs—orthorhombic, tetragonal, and hexagonal—exhibit distinct magnetic orders and electronic responses, aiding tailored device performance.
• State-of-the-art techniques enable electrical, optical, and thermal control of the Néel vector, paving the way for switching mechanisms and novel topological functionalities in memory and logic devices.

Antiferromagnetic metal CuMnAs is a class of Mn-based compounds that display robust antiferromagnetic order, high Néel temperature, and a tunable semimetallic to semiconducting electronic structure. Both bulk and epitaxial thin film forms of CuMnAs—including orthorhombic, tetragonal, and hexagonal polymorphs—have become reference materials in the development of antiferromagnetic spintronics. These compounds exhibit electrically and optically switchable order parameters via spin–orbit torques and magnetoelastic coupling, and possess a range of topological and nonlinear electronic responses.

1. Crystal Structure, Phase Stability, and Sample Preparation

CuMnAs exists in several crystallographically distinct structures. The stable orthorhombic phase (Pnma) is realized in bulk CuMnAs and CuMnP, while a metastable tetragonal phase (P4/nmm) can be stabilized via epitaxial growth on lattice-matched III–V substrates (e.g., GaAs, GaP) or on Si (Wadley et al., 2014). A recently identified hexagonal phase, Cu$_{0.82}$Mn$_{1.18}$As, broadens the structural landscape (Karigerasi et al., 2019).

Polymorph	Space Group	Magnetic Structure	Key Features
Orthorhombic	Pnma	Collinear AFM (bc-plane)	Semimetal–semi.
Tetragonal	P4/nmm	In-plane/L$_{Néel}$ AFM	Spintronic active
Hexagonal	P6$_3$/mmc	120° Triangular AFM	In-plane isotropy

Phase-pure tetragonal CuMnAs is challenging to synthesize in bulk due to a strong tendency for phase separation and competition from the orthorhombic structure; an excess of Cu is often required to avoid orthorhombic contamination (Karigerasi et al., 2022). Thin films are conventionally grown using molecular beam epitaxy (MBE), but industry-compatible routes such as magnetron sputtering have also enabled high-quality, electrically switchable films (Matalla-Wagner et al., 2019).

2. Antiferromagnetic Order and Magnetic Transitions

Both orthorhombic and tetragonal CuMnAs are robust collinear antiferromagnets with high Néel temperatures, in contrast to the weak (T$_N$ ≈ 50 K) antiferromagnetic order of cubic semi-Heusler CuMnSb (Maca et al., 2011). In the tetragonal phase, neutron diffraction and XMLD confirm in-plane staggered ordering of Mn moments transforming within the $\Gamma_{5}^{-}$ representation (Wadley et al., 2014).

Energy differences between FM and AFM configurations—calculated via DFT as E$_{\mathrm{FM}}$ – E$_{\mathrm{AFM}}$ ≈ 109 meV/f.u. (tetragonal) and ≈ 241 meV/Mn-atom (orthorhombic)—predict $T_N$ well above room temperature (Maca et al., 2011, Wadley et al., 2014). Monte Carlo simulations based on DLM-extracted exchange integrals yield T$_N$ ≈ 480–495 K, strongly corroborated by experiment (Slutsker et al., 2017, Maca et al., 2018).

In both tetragonal and orthorhombic phases, the AFM transition is accompanied by magnetoelastic coupling: lattice parameters exhibit discontinuities at T$_N$ and at lower-temperature transitions to weak ferromagnetic or canted states, evidenced by both XRD and calorimetry (Karigerasi et al., 2022, Zhang et al., 2017, Emmanouilidou et al., 2018). Some compositions (Cu$_{1.18}$Mn$_{0.82}$As, CuMn$_{0.964}$As$_{1.036}$) show further transitions to weak ferromagnetism near room temperature.

3. Electronic Structure and Topological Properties

DFT calculations for orthorhombic CuMnAs indicate that it lies at the brink of a semimetal–semiconductor transition, exhibiting a small but finite density of states (DOS) at E$_F$ (Maca et al., 2011). The carrier concentration is high and predominantly p-type, but mobility is anomalously low and effective masses are large (m* ≈ 1.5 of m$_e$), signifying “massive fermion” behavior rather than Dirac/Weyl-like dispersion (Zhang et al., 2017).

Theoretical proposals for Dirac semimetallicity require preservation of specific symmetries (e.g., screw axis S$_{2z}$); in real crystals, symmetry breaking due to magnetic easy axis selection or canting gaps the Dirac points and generates a more conventional semimetallic state (Zhang et al., 2017, Emmanouilidou et al., 2018). Step-edge As deficiency in thin films leads to zigzag reconstructions, enhancing spin polarization and creating localized edge states with elevated magnetic moments—a critical factor for topological responses and surface magnetism (Nguyen et al., 2019).

In the orthorhombic phase, nodal lines and gapped Dirac points close to E$_F$ contribute strong (spin) Berry curvature, giving rise to a prominent, anisotropic spin Hall effect ($\sigma^{S,γ}_{\alpha\beta}$), while anomalous Hall conductivity is symmetry-forbidden in the ground state, but can be large under external magnetic fields (Huyen et al., 2021).

4. Defects, Phase Separation, and Magnetotransport

Systematic ab initio calculations and XRD analysis identify low-energy Mn and Cu vacancies and antisite defects as the dominant defect types in tetragonal CuMnAs (Máca et al., 2017, Maca et al., 2018). Their presence substantially affects resistivity, with Mn$_{\text{Cu}}$ antisites producing particularly pronounced virtual bound states at E$_F$. Experimentally observed room-temperature and low-temperature resistivities (≈160 µΩ·cm and 90 µΩ·cm, respectively) are in good agreement with defect-informed modeling (Maca et al., 2018).

High-resolution diffraction reveals nanoscale phase separation—coherent stripe-like domains with compositionally- and magnetically-distinct regions—arising via (pseudo-)spinodal decomposition, especially in non-stoichiometric and Cu-rich samples (Karigerasi et al., 2022). This phase separation modifies both exchange interactions and local anisotropy, resulting in inhomogeneous transport and magnetic behavior.

5. Electrical, Optical, and Thermal Control of the Néel Vector

CuMnAs thin films enable current-induced reorientation of the Néel vector via the intrinsic Néel-order spin–orbit torque (NSOT), arising due to the inversion-partner symmetry of the Mn sublattices (Grzybowski et al., 2016, Matalla-Wagner et al., 2019). The action of NSOT is to generate local effective fields on the sublattices ($\mathbf{H}^{(A)}_{\text{eff}} = +\chi J$, $\mathbf{H}^{(B)}_{\text{eff}} = -\chi J$), which rotate the Néel vector ($\mathbf{L}$) perpendicular to the applied current.

Switching is clearly detected through anisotropic magnetoresistance (AMR) and planar Hall effects. Thermal activation governs both the domain reorientation and relaxation: the energy barrier for switching (E$_B = K_{4\parallel} V_g$) is surmounted via a combination of NSOT and local Joule heating (Matalla-Wagner et al., 2019, Omari et al., 2021). Both orthogonal and polarity-driven current pulse schemes effect switching with distinct retention, reproducibility, and signal-to-noise properties (Omari et al., 2021).

Terahertz-pulse-induced ultrafast switching can be optically gated via transient photoconductivity in the substrate, allowing for spatially-localized suppression of magnetic reorientation with femtosecond precision and ~100 nm resolution (Heitz et al., 2021). Magneto-Seebeck microscopy (MSE) further extends the read-out toolkit, providing table-top imaging of 90°/180° domain wall motion (Janda et al., 2020).

6. Magnetic Domain Engineering, Anisotropy, and Device Integration

Patterning and strain engineering are effective levers for domain engineering in CuMnAs thin films (Reimers et al., 2023). Lithographically-defined edges impose surface anisotropy, while magnetoelastic coupling (destressing energy due to strain incompatibility) propagates the edge effect microns into the film. The effective domain wall width varies exponentially with distance from the patterned edge:

$$\frac{1}{d_\mathrm{DW}^2} = \frac{1}{D^2} (1 - a\, e^{-x/el} )$$

This interplay is captured by a combined model for exchange, magnetocrystalline, edge, and destressing energies. Compared to Mn$_2$Au, which has a much higher bulk anisotropy, CuMnAs shows pronounced susceptibility to shape and strain effects.

Such elastic and anisotropic phenomena are vital considerations when designing antiferromagnetic memory and logic elements, especially given the critical role of equilibrium domain configuration in setting switching characteristics and device reproducibility.

7. Nonlinear Hall Effect, Topological Phenomena, and Quench Switching

The intrinsic nonlinear Hall effect (INHE) in tetragonal CuMnAs serves as a direct, relaxation time–independent probe of the Néel vector. The key second order conductivity tensor is (Wang et al., 2021):

$$\sigma^{(\alpha\beta\gamma)}_{\rm INH} = 2e^3 \sum_{n \neq m} \mathrm{Re} \int \frac{d^3k}{(2\pi)^3} \frac{v^\alpha_n {\cal A}^\beta_{nm}{\cal A}^\gamma_{mn}}{\epsilon_n - \epsilon_m} \frac{\partial f(\epsilon_n)}{\partial \epsilon_n} - (\alpha \leftrightarrow \beta)$$

This INHE can reach mA/V$^2$ and depends sensitively on both chemical potential and temperature, peaking near band edge crossings described by a tilted massive Dirac model. The non-Drude, T-odd INHE component allows detection of 180° Néel vector flips in PT-symmetric antiferromagnets, directly enabling electrical readout in spintronic devices. Symmetry analysis identifies 53 magnetic point groups supporting INHE.

Quench switching—the creation of metastable, resistive high-domain-wall-density states by rapid heating/cooling above/below T$_N$—has been demonstrated in both CuMnAs and Mn$_2$As (Olejník et al., 4 Nov 2024). The relaxation follows a superposition of stretched exponentials with time constants obeying an Arrhenius law, $\tau = \tau_0 \exp(E_b / k_BT)$, with $E_b$ proportional to $T_N$. The multilevel, long-lived (hours-scale at room temperature in Mn$_2$As) resistivity states open routes toward neuromorphic and analog memory functionalities.

• (Maca et al., 2011): "CuMn-V compounds: a transition from semimetal low-temperature to semiconductor high-temperature antiferromagnets"
• (Wadley et al., 2014): "Tetragonal phase of epitaxial room-temperature antiferromagnet CuMnAs"
• (Grzybowski et al., 2016): "Imaging current-induced switching of antiferromagnetic domains in CuMnAs"
• (Zhang et al., 2017): "Massive fermions with low mobility in antiferromagnet orthorhombic CuMnAs single crystals"
• (Emmanouilidou et al., 2018): "Spin-flop phase transition in the orthorhombic antiferromagnetic topological semimetal Cu0.95MnAs"
• (Maca et al., 2018): "Tetragonal CuMnAs alloy: role of defects"
• (Matalla-Wagner et al., 2019): "Electrical Néel-order switching in magnetron-sputtered CuMnAs thin films"
• (Nguyen et al., 2019): "Emerging edge states on the surface of the epitaxial semimetal CuMnAs thin film"
• (Karigerasi et al., 2019): "An in-plane hexagonal antiferromagnet in the Cu-Mn-As system, Cu$_{0.82}$Mn$_{1.18}$As"
• (Wang et al., 2019): "Spin flop and crystalline anisotropic magnetoresistance in CuMnAs"
• (Janda et al., 2020): "Magneto-Seebeck microscopy of domain switching in collinear antiferromagnet CuMnAs"
• (Huyen et al., 2021): "Spin and anomalous Hall effects emerging from topological degeneracy in Dirac fermion system CuMnAs"
• (Omari et al., 2021): "Low-Energy Switching of Antiferromagnetic CuMnAs/ GaP Using sub-10 Nanosecond Current Pulses"
• (Heitz et al., 2021): "Optically gated terahertz-field-driven switching of antiferromagnetic CuMnAs"
• (Wang et al., 2021): "Intrinsic nonlinear Hall effect in antiferromagnetic tetragonal CuMnAs"
• (Karigerasi et al., 2022): "High-resolution diffraction reveals magnetoelastic coupling and coherent phase separation in tetragonal CuMnAs"
• (Reimers et al., 2023): "Magnetic domain engineering in antiferromagnetic CuMnAs and Mn$_2$Au devices"
• (Olejník et al., 4 Nov 2024): "Quench switching of Mn2As"