betaFeedback

WorksPhysics and Astronomy

Tuning spin currents in collinear antiferromagnets and altermagnets

Abstract

Spin current generation through non-relativistic spin splittings, found in uncompensated magnets and d-wave altermagnets, is desirable for low-power spintronics. Such spin currents, however, are symmetry forbidden in conventional collinear antiferromagnets and higher-order altermagnets. Using spin point group analysis, we demonstrate that finite spin currents can be induced in these materials via magnetoelectric, piezomagnetic, and piezomagnetoelectric-like couplings. We utilize electric fields, strain, and their combinations to drive symmetry-lowering phase transitions into uncompensated magnetic or d-wave altermagnetic states, thereby enabling finite spin conductivity in a broader class of magnetic materials. We further substantiate this framework using density functional theory and Boltzmann transport calculations on representative magnetic materials - KV2Se2O, RuF4 , Cr2O3 , FeS2 , and MnPSe3 - spanning these different cases. The charge-to-spin conversion ratio reaches up to almost 100% via uncompensated magnetism and about 40% via d-wave altermagnetism under realistic conditions, highlighting the effectiveness of this approach for efficient spin current generation.

Introduction

The generation of pure spin currents is of fundamental interest in spintronics. In most materials, spin currents are typically achieved via the spin-Hall effect driven by spin-orbit interaction (SOI) [1, 2]. However, SOI, which gives rise to spin currents, also results in spin dephasing mechanisms, such as D’yakonov-Perel [3] and Elliott-Yafet processes [4, 5]. Such dephasing mechanisms limit the coherence length of the spin currents, thereby constraining their practical applications. Recently, collinear compensated magnets [6, 7, 8, 9] and noncollinear antiferromagnets [10, 11, 12] have been shown to generate highly spin-polarized currents through exchange fields. The charge-to-spin conversion efficiency and coherence length of such nonrelativistic spin currents can exceed those of SOI-driven mechanisms by orders of magnitude [13, 14, 15].

In the context of symmetry, collinear compensated magnets are classified as antiferromagnets (AFMs) or altermagnets (AMs), depending on the symmetries that connect opposite spin sublattices [8]. In AFMs, the opposite spin sublattices are related by spatial inversion (PP) and translation (τ\tau), leading to spin-degeneracy and zero spin conductivity. In contrast, AMs have mirror-rotation symmetries connecting opposite spin sublattices, and are further categorized by their planar or bulk dd-, gg-, and ii-wave orders. Spin current generation has been theoretically predicted in dd-wave AMs, such as RuO2_2 [13]. However, the altermagnetic ground state of RuO2_2 is still under debate [15]. Unfortunately, spin conductivity remains forbidden in higher-order AMs (gg- and ii-wave) due to additional spin-degenerate nodal surfaces. Nevertheless, one can use strain and electric fields to manipulate the altermagnetic order to allow for spin currents [16, 17, 18].

For instance, recent works used uniaxial strain to induce gg- to dd-wave altermagnetic phase transitions in MnTe [16] and CrSb [17]. Other researchers showed that an electric field applied to the antiferromagnetic MnPSe3_3 can induce a transition to either a dd-wave [19] or an ii-wave AM phase [20]. While most such studies have focused on altermagnetic spin splittings, a general framework for generating nonrelativistic spin currents via magnetic phase transitions has remained unexplored.

In this article, we provide a general approach based on spin group symmetry formalism to induce spin currents through magnetoelectric, piezomagnetic, and piezomagnetoelectric-like coupling. We show that electric (Ei\mathcal{E}_i) and strain (ηjk\eta_{jk}) fields and their combination (Eiηjk\mathcal{E}_i\eta_{jk}) can be utilized to induce spin-currents in AFMs and higher-order AMs via reduction of their respective symmetries. By considering the lower effective spin point groups (SPGs) under external stimuli for all possible AFMs and higher-order AMs, we determine the correlation between spin current and Ei\mathcal{E}_i, ηjk\eta_{jk}, and Eiηjk\mathcal{E}_i\eta_{jk}. We compute the spin-polarized conductivity for several representative examples of collinear compensated magnets using DFT and Boltzmann transport simulations. We also determine the spin conductivity for the reduced SPGs when the materials are subjected to an external electric field along an arbitrary direction.

Computational details

In the nonrelativistic case, conductivity can be described through the spin-up (sks_k) and spin-down (−sk-s_k) channels with spin polarization along the kk direction. Spin-polarized conductivity, σijsk\sigma_{ij}^{s_k}, is defined as Jisk=σijskEjJ_i^{s_k} = \sigma_{ij}^{s_k}\mathcal{E}_j, where Ej\mathcal{E}_j is the electric field along the jj-direction and JiskJ_i^{s_k} is the current along the ii-direction. Within the Boltzmann transport theory, the conductivity σijsk\sigma_{ij}^{s_k} (contributed by EnkE_{nk}) is expressed as [21]

σijsk(EF)=−e2τ8π2ℏ3∑n∫∂Enksk∂ki∂Enksk∂kj∂f0∂Enksk d3k(1)\sigma_{ij}^{s_k}(E_F)=-\frac{e^2\tau}{8\pi^2\hbar^3}\sum_n\int\frac{\partial E_{nk}^{s_k}}{\partial k_i}\frac{\partial E_{nk}^{s_k}}{\partial k_j}\frac{\partial f^0}{\partial E_{nk}^{s_k}}\,d^3k \tag*{(1)}

Here, τ\tau is the relaxation time, and f0f^0 is the Fermi-Dirac distribution function. Within Eq. (1), the net charge and spin conductivity, respectively, are defined as

σij=∑k(σijsk+σij−sk)\sigma_{ij}=\sum_k\left(\sigma_{ij}^{s_k}+\sigma_{ij}^{-s_k}\right)
σijk=ℏ2e(σijsk−σij−sk).(2)\sigma_{ij}^{k}=\frac{\hbar}{2e}\left(\sigma_{ij}^{s_k}-\sigma_{ij}^{-s_k}\right). \tag*{(2)}

The spin-polarized conductivity σijsk\sigma_{ij}^{s_k} transforms under time-reversal (TT) as Tσijsk→σij−skT\sigma_{ij}^{s_k}\to\sigma_{ij}^{-s_k}. Accordingly, the charge and spin conductivity tensors are even and odd under time-reversal symmetry (TT), respectively. We initialize the spin-quantization axis along the zz-direction. Since the effect of SOI is expected to be negligible, it is excluded from the simulation [13]. The absence of SOI means there is no spin canting, and consequently, σijx/y=0\sigma_{ij}^{x/y}=0. This can also be seen from Equation (1): σijsk\sigma_{ij}^{s_k} transforms as sksikjs_ks_ik_j. The spin-only symmetry [Cz(π)∣∣E][C_z(\pi)||E] transforms σijsx/sy\sigma_{ij}^{s_x/s_y} to σij−sx/−sy\sigma_{ij}^{-s_x/-s_y}, which implies that the net spin current, σijx/y\sigma_{ij}^{x/y}, must vanish while the charge current can be nonzero. Therefore, under nonrelativistic collinear magnetism, the only allowed nonzero spin conductivity components are σijz\sigma_{ij}^{z}. The shape of σijz\sigma_{ij}^{z} is further restricted by the nontrivial spin group symmetry.

Density functional theory calculations were performed using the plane-wave ultrasoft pseudopotential [22] approach implemented in Quantum ESPRESSO [23]. The generalized gradient approximation with Hubbard correction (PBE+U) was used to treat exchange correlations self consistently [24]. The Kubo formalism, combined with DFT, was used to determine the charge and spin conductivities. To do so, we constructed a tight-binding Hamiltonian in the atom-centered Wannier basis [25] using the Wannier90 code [26]. In the single-particle energy eigenstate representation, the spin conductivity within the Kubo formalism is given by:

σijk=−eℏ2ΓV∫d3k(2π)3∑n⟨nk∣J^ik∣nk⟩⟨nk∣v^j∣nk⟩×δ(Enk−EF)(3)\sigma_{ij}^{k}=-\frac{e\hbar}{2\Gamma}V\int\frac{d^3k}{(2\pi)^3}\sum_n\langle nk|\hat{J}_i^k|nk\rangle\langle nk|\hat{v}_j|nk\rangle\times\delta(E_{nk}-E_F) \tag*{(3)}

where J^ik=12{v^i,s^k}\hat{J}_i^k=\frac{1}{2}\{\hat{v}_i,\hat{s}_k\} is the spin current operator. Γ=ℏ2τ\Gamma=\frac{\hbar}{2\tau} is a constant scattering rate, and is inversely proportional to the relaxation time, τ\tau. The replacement of J^ik\hat{J}_i^k with v^i\hat{v}_i in Eq. (3) gives the charge conductivity. The conductivity tensors were obtained by solving Eq. (3) as a postprocessing step of DFT calculations on a dense 3503350^3 kk-mesh for 3D materials and 3502350^2 kk-mesh for 2D materials, using the Wannier-Linear-Response code [13]. We made use of the Findsym [27], Bilbao Crystallographic Server [28], Findspingroup [29], and Tensor Sym-

metry [30] packages to obtain the nontrivial SPGs and the corresponding shape of σijz\sigma_{ij}^{z}.

Symmetry Analysis

Collinear magnets are more accurately described using spin group symmetry operations of the form [R1∣∣R2][R_1||R_2], where R1R_1 and R2R_2 act on spin and real space, respectively [32]. These systems exhibit symmetries that fall into 32 SPGs associated with FM order, 22 SPGs corresponding to conventional AFM order, and 36 SPGs belong to AM order [29]. The spin group operations transform σsk\sigma^{s_k} as

D(R2)σD(R1)skD(R2)†=σsk(4)\mathbf{D}(R_2)\sigma^{\mathbf{D}(R_1)s_k}\mathbf{D}(R_2)^\dagger=\sigma^{s_k} \tag*{(4)}

The spin conductivity component, σijz\sigma_{ij}^{z}, will be allowed if the SPG does not contain any symmetry operation enforcing σijsz=σij−sz\sigma_{ij}^{s_z}=\sigma_{ij}^{-s_z}. Since the [C2∣∣P][C_2||P] symmetry in AFMs leads to σsk=σ−sk\sigma^{s_k}=\sigma^{-s_k}, the spin current σk\sigma^k is forbidden for all AFM SPGs, as schematically shown in Figure 1(a) for SPGs containing [C2∣∣P][C_2||P] and [C2∣∣My][C_2||M_y]. The [C2∣∣P][C_2||P] symmetry can be broken by applying an electric field

Schematic of electric field and strain-induced spin conductivity

Figure 1. Schematic of electric field and strain-induced spin conductivity. (a) Spin-polarized Fermi surface of an antiferromagnet with [C2∣∣P][C_2||P] symmetry. The red and blue isosurfaces correspond to the opposite spin directions. Fermi surfaces of an antiferromagnet under an applied electric field when: (b) the effective spin group of the material possesses a symmetry connecting opposite spin sublattices (e.g., dd-wave altermagnet with [C2∣∣My][C_2||M_y]), and (c) no such symmetry exists (ferromagnetic or compensated ferrimagnetic). (d) Fermi surface of a higher-order planar gg-wave state. Fermi surfaces of a higher-order gg-wave altermagnet (d) without strain and under a strain field ηij\eta_{ij} when: (e) the effective spin group retains a symmetry connecting opposite spin sublattices (e.g., dd-altermagnet with [C2∣∣Mxy][C_2||M_{xy}]), and (f) no such symmetry remains. In the plots, Fermi surfaces are obtained from the minimal two-band models derived in Ref. [31], with strain effects incorporated using ki→(δij+ηij)kjk_i\to(\delta_{ij}+\eta_{ij})k_j.

(Ei\mathcal{E}_i) since [C2∥P]Ei→−Ei[C_2\Vert P]\mathcal{E}_i \to-\mathcal{E}_i. The applied electric field (Ei\mathcal{E}_i) also breaks the [E/C2∥Mi][E/C_2\Vert M_i] and [E/C2∥Cnj (j≠i)][E/C_2\Vert C_{n j}\ (j \ne i)] symmetries. Overall, the spin conductivity component, σijz\sigma_{ij}^{z}, becomes allowed if the electric field breaks all spin group symmetries that enforce σijz=σij−z\sigma_{ij}^{z}=\sigma_{ij}^{-z}. Table I lists the electric field-induced transitions of antiferromagnetic SPGs to either FM or AM SPGs. If the reduced effective SPG is an FM or a dd-wave AM, a net spin current becomes allowed. In such a case, spin-polarized conductivity can be considered to be bilinear in the electric field. This coupling between spin current and electric field is reminiscent of magnetoelectric coupling. For instance, AFM SPG 12/2m1^{2}/{}_{2}m (two-fold axis along yy-direction) reduces to the effective AM SPG of 2m^{2}m under an electric field either applied along Ex\mathcal{E}_x or Ez\mathcal{E}_z. This case corresponds to the spin-polarized Fermi surface are shown schematically in Figure 1(b). Also, AFM SPG 12/2m1^{2}/{}_{2}m reduces to an effective FM SPG of 121^{2} under electric field Ey\mathcal{E}_y [spin-polarized Fermi surfaces in Figure 1(c)]. Table I also lists the cases where the applied electric field reduces the AFM SPG to that of either a gg- or an ii-wave AM. For example, Ez\mathcal{E}_z applied to a 14/2m2m2m1^{4}/{}_{2}m^{2}m^{2}m or 26/1m2^{6}/{}_{1}m AFM, and

SPGsEx\mathcal{E}_xEy\mathcal{E}_yEz\mathcal{E}_z
21ˉ^{2}\bar{1}11^{1}111^{1}111^{1}1
221/1m^{2}2^{1}/{}_{1}m1m^{1}m22^{2}2 (d)1m^{1}m
122/2m^{1}2^{2}/{}_{2}m2m^{2}m (d)12^{1}22m^{2}m (d)
1m2m1m^{1}m^{2}m^{1}m222m^{2}2^{2}m (d)1m212m^{1}m^{2}1^{2}m (d)1m2m2^{1}m^{2}m^{2} (d)
1m1m2m^{1}m^{1}m^{2}m221m2m^{2}2^{1}m^{2}m (d)1m222m^{1}m^{2}2^{2}m (d)1m1m12^{1}m^{1}m^{1}2 (d)
2m2m2m^{2}m^{2}m^{2}m122m2m^{1}2^{2}m^{2}m (d)2m122m^{2}m^{1}2^{2}m (d)2m2m12^{2}m^{2}m^{1}2 (d)
142/2m^{1}4^{2}/{}_{2}m2m^{2}m (d)2m^{2}m (d)14^{1}4
241/2m^{2}4^{1}/{}_{2}m2m^{2}m (d)2m^{2}m (d)24^{2}4 (d)
142/1m1m1m^{1}4^{2}/{}_{1}m^{1}m^{1}m221m2m^{2}2^{1}m^{2}m (d)1m222m^{1}m^{2}2^{2}m (d)141m1m^{1}4^{1}m^{1}m
241/1m2m2m^{2}4^{1}/{}_{1}m^{2}m^{2}m221m2m^{2}2^{1}m^{2}m (d)1m222m^{1}m^{2}2^{2}m (d)241m2m^{2}4^{1}m^{2}m (d)
142/2m2m2m^{1}4^{2}/{}_{2}m^{2}m^{2}m122m2m^{1}2^{2}m^{2}m (d)2m122m^{2}m^{1}2^{2}m (d)142m2m^{1}4^{2}m^{2}m (g)
23ˉ^{2}\bar{3}11^{1}111^{1}113^{1}3
23ˉ111m^{2}\bar{3}^{1}1^{1}m1m^{1}m22^{2}2 (d)13111m^{1}3^{1}1^{1}m
23ˉ112m^{2}\bar{3}^{1}1^{2}m2m^{2}m (d)12^{1}213112m^{1}3^{1}1^{2}m (g)
261/1m^{2}6^{1}/{}_{1}m1m^{1}m1m^{1}m26^{2}6 (g)
162/2m^{1}6^{2}/{}_{2}m2m^{2}m (d)2m^{2}m (d)16^{1}6
162/1m1m1m^{1}6^{2}/{}_{1}m^{1}m^{1}m221m2m^{2}2^{1}m^{2}m (d)1m222m^{1}m^{2}2^{2}m (d)161m1m^{1}6^{1}m^{1}m
261/1m2m2m^{2}6^{1}/{}_{1}m^{2}m^{2}m222m1m^{2}2^{2}m^{1}m (d)1m121m^{1}m^{1}2^{1}m (d)261m2m^{2}6^{1}m^{2}m (g)
162/2m2m2m^{1}6^{2}/{}_{2}m^{2}m^{2}m122m2m^{1}2^{2}m^{2}m (d)2m122m^{2}m^{1}2^{2}m (d)162m2m^{1}6^{2}m^{2}m (i)
2m23ˉ^{2}m^{2}\bar{3}222m2m^{2}2^{2}m^{2}m (d)3m122m^{3}m^{1}2^{2}m (d)2m2m12^{2}m^{2}m^{1}2 (d)
2m23ˉ1m^{2}m^{2}\bar{3}^{1}m242m1m^{2}4^{2}m^{1}m (d)242m1m^{2}4^{2}m^{1}m (d)242m1m^{2}4^{2}m^{1}m (d)
2m23ˉ2m^{2}m^{2}\bar{3}^{2}m142m2m^{1}4^{2}m^{2}m (g)142m2m^{1}4^{2}m^{2}m (g)142m2m^{1}4^{2}m^{2}m (g)

Table I. Electric field-induced transition of antiferromagnetic SPG to ferromagnetic (or uncompensated magnetic) and altermagnetic SPG. The first column denotes the parent SPG, and the second, third, and fourth columns denote the effective SPG under the application of electric field Ex\mathcal{E}_x, Ey\mathcal{E}_y, and Ez\mathcal{E}_z, respectively. The superscripts 1 and 2 preceding each symmetry operation indicate whether the spin sublattices are connected to the same or opposite spin sublattices. The SPG without any symmetry connecting opposite spin sublattices (SPGs with only superscript 1) are ferromagnetic. For the altermagnetic SPGs, the spin-orders (dd, gg, and ii) are also indicated in parentheses.

Ei (i=x,y,z)\mathcal{E}_i\ (i=x,y,z) applied to a 2m3ˉ2m^{2}m\bar{3}^{2}m AFM results in gg-wave AMs. Also, the 16/2m2m2m^{1}6/{}_{2}m^{2}m^{2}m AFM reduces to an ii-wave AM under the application of Ez\mathcal{E}_z. In such cases, a higher order (i.e., trilinear) electric field may be required to induce a finite spin conductivity [33].

Unlike the electric field that breaks the [C2∥P][C_2\Vert P] symmetry of AFMs, the strain field (ηij\eta_{ij}, with i,j=x,y,zi,j=x,y,z), which transforms like kikjk_i k_j, remains invariant under [C2∥P]ηij→ηij[C_2\Vert P]\eta_{ij}\to\eta_{ij}. Therefore, ηij\eta_{ij} is not sufficient to induce spin currents in AFMs. However, strain can be used to induce phase transitions in gg- and ii-wave AMs to obtain either an FM or a dd-wave AM. For example, AM SPG 142/m2m1^{4}2/m^{2}m is a gg-wave altermagnet (see Figure 1(d)). The [E∥C4z][E\Vert C_{4z}], [E∥Mx][E\Vert M_x], and [E∥Mxy][E\Vert M_{xy}] symmetries would forbid a spin current in such a material. However, the application of the shear strain ηxy\eta_{xy} breaks [C2∥Mx][C_2\Vert M_x], leading to effective SPG 2m122m^{1}2 with spin-polarized Fermi surface similar to a dd-wave AM as shown in Figure 1(e). In addition, a simultaneous application of ηxz\eta_{xz} or ηyy\eta_{yy} along with ηxy\eta_{xy}, may break [C2∥Mxy][C_2\Vert M_{xy}] as well, leading to an effective ferromagnetic SPG that has no symmetry connecting opposite spin sublattices. Therefore, in this case, the spin polarized Fermi surface is similar to that for an FM, as shown in Figure 1(f). Table 2 lists the effective SPGs of gg- and ii-wave AMs under linear strain (ηij\eta_{ij}) by collecting spin group symmetry operations [R1∥R2][R_1\Vert R_2] such that R2ηij→ηijR_2\eta_{ij}\to\eta_{ij}.

Altermagnetic PatternSPGsηxx\eta_{xx}ηyy\eta_{yy}ηzz\eta_{zz}ηxy\eta_{xy}ηxz\eta_{xz}ηyz\eta_{yz}
Planar gg-wave14/1m2m2m{}^{1}4/{}^{1}m{}^{2}m{}^{2}m2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)2m2m1m{}^{2}m{}^{2}m{}^{1}m (d)14/1m2m2m{}^{1}4/{}^{1}m{}^{2}m{}^{2}m (g)2m2m1m{}^{2}m{}^{2}m{}^{1}m (d)22/2m{}^{2}{}^{2}/{}^{2}m (d)22/2m{}^{2}{}^{2}/{}^{2}m (d)
Planar gg-wave142m2m{}^{1}4{}^{2}m{}^{2}m2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)142m2m{}^{1}4{}^{2}m{}^{2}m (g)2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)2m{}^{2}m (d)2m{}^{2}m (d)
Planar gg-wave142222{}^{1}4{}^{2}2{}^{2}2222212{}^{2}2{}^{2}2{}^{1}2 (d)222212{}^{2}2{}^{2}2{}^{1}2 (d)142222{}^{1}4{}^{2}2{}^{2}2 (g)222212{}^{2}2{}^{2}2{}^{1}2 (d)22{}^{2}2 (d)22{}^{2}2 (d)
Planar gg-wave14222m{}^{1}4{}^{2}2{}^{2}m222212{}^{2}2{}^{2}2{}^{1}2 (d)222212{}^{2}2{}^{2}2{}^{1}2 (d)14222m{}^{1}4{}^{2}2{}^{2}m (g)2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)22{}^{2}2 (d)22{}^{2}2 (d)
Bulk gg-wave132m11{}^{1}3{}^{2}m{}^{1}122/2m{}^{2}2/{}^{2}m (d)22/2m{}^{2}2/{}^{2}m (d)132m11{}^{1}3{}^{2}m{}^{1}1 (g)11ˉ{}^{1}\bar{1}11ˉ{}^{1}\bar{1}22/2m{}^{2}2/{}^{2}m (d)
Bulk gg-wave132m11{}^{1}3{}^{2}m{}^{1}12m{}^{2}m (d)2m{}^{2}m (d)132m11{}^{1}3{}^{2}m{}^{1}1 (g)11{}^{1}111{}^{1}12m{}^{2}m (d)
Bulk gg-wave132211{}^{1}3{}^{2}2{}^{1}122{}^{2}2 (d)22{}^{2}2 (d)132211{}^{1}3{}^{2}2{}^{1}1 (g)11{}^{1}111{}^{1}122{}^{2}2 (d)
Bulk gg-wave26/2m{}^{2}6/{}^{2}m22/2m{}^{2}2/{}^{2}m (d)22/2m{}^{2}2/{}^{2}m (d)26/2m{}^{2}6/{}^{2}m (g)22/2m{}^{2}2/{}^{2}m (d)11ˉ{}^{1}\bar{1}11ˉ{}^{1}\bar{1}
Bulk gg-wave26{}^{2}62m{}^{2}m (d)2m{}^{2}m (d)26ˉ{}^{2}\bar{6} (g)2m{}^{2}m (d)11{}^{1}111{}^{1}1
Bulk gg-wave26{}^{2}622{}^{2}2 (d)22{}^{2}2 (d)26{}^{2}6 (g)22{}^{2}2 (d)11{}^{1}111{}^{1}1
Bulk gg-wave26/2m1m{}^{2}6/{}^{2}m{}^{1}m2m1m2m{}^{2}m{}^{1}m{}^{2}m (d)2m1m2m{}^{2}m{}^{1}m{}^{2}m (d)26/2m1m{}^{2}6/{}^{2}m{}^{1}m (g)22/2m{}^{2}2/{}^{2}m (d)12/1m{}^{1}2/{}^{1}m22/2m{}^{2}2/{}^{2}m (d)
Bulk gg-wave262m1m{}^{2}6{}^{2}m{}^{1}m2m1m22{}^{2}m{}^{1}m{}^{2}2 (d)2m1m22{}^{2}m{}^{1}m{}^{2}2 (d)262m1m{}^{2}6{}^{2}m{}^{1}m (g)22{}^{2}2 (d)1m{}^{1}m2m{}^{2}m (d)
Bulk gg-wave262212{}^{2}6{}^{2}2{}^{1}2221222{}^{2}2{}^{1}2{}^{2}2 (d)221222{}^{2}2{}^{1}2{}^{2}2 (d)262212{}^{2}6{}^{2}2{}^{1}2 (g)22{}^{2}2 (d)12{}^{1}222{}^{2}2 (d)
Bulk gg-wave26ˉ2m12{}^{2}\bar{6}{}^{2}m{}^{1}22m122m{}^{2}m{}^{1}2{}^{2}m (d)2m122m{}^{2}m{}^{1}2{}^{2}m (d)26ˉ2m12{}^{2}\bar{6}{}^{2}m{}^{1}2 (g)2m{}^{2}m (d)12{}^{1}22m{}^{2}m (d)
Bulk gg-wave26ˉ211m{}^{2}\bar{6}{}^{2}1{}^{1}m221m2m{}^{2}2{}^{1}m{}^{2}m (d)221m2m{}^{2}2{}^{1}m{}^{2}m (d)26ˉ211m{}^{2}\bar{6}{}^{2}1{}^{1}m (g)2m{}^{2}m (d)1m{}^{1}m22{}^{2}2 (d)
Planar ii-wave16/1m2m2m{}^{1}6/{}^{1}m{}^{2}m{}^{2}m2m2m1m{}^{2}m{}^{2}m{}^{1}m (d)2m2m1m{}^{2}m{}^{2}m{}^{1}m (d)16/1m2m2m{}^{1}6/{}^{1}m{}^{2}m{}^{2}m (d)21/1m{}^{2}1/{}^{1}m22/2m{}^{2}2/{}^{2}m (d)22/2m{}^{2}2/{}^{2}m (d)
Planar ii-wave162m2m{}^{1}6{}^{2}m{}^{2}m2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)162m2m{}^{1}6{}^{2}m{}^{2}m (i)12{}^{1}22m{}^{2}m (d)2m{}^{2}m (d)
Planar ii-wave162222{}^{1}6{}^{2}2{}^{2}2222212{}^{2}2{}^{2}2{}^{1}2 (d)222212{}^{2}2{}^{2}2{}^{1}2 (d)162222{}^{1}6{}^{2}2{}^{2}2 (i)12{}^{1}222{}^{2}2 (d)22{}^{2}2 (d)
Planar ii-wave16ˉ2m22{}^{1}\bar{6}{}^{2}m{}^{2}22m221m{}^{2}m{}^{2}2{}^{1}m (d)2m221m{}^{2}m{}^{2}2{}^{1}m (d)16ˉ2m22{}^{1}\bar{6}{}^{2}m{}^{2}2 (i)1m{}^{1}m22{}^{2}2 (d)2m{}^{2}m (d)
Bulk ii-wave1m13ˉ2m{}^{1}m{}^{1}\bar{3}{}^{2}m24/1m1m2m{}^{2}4/{}^{1}m{}^{1}m{}^{2}m (d)24/1m1m2m{}^{2}4/{}^{1}m{}^{1}m{}^{2}m (d)24/1m1m2m{}^{2}4/{}^{1}m{}^{1}m{}^{2}m (d)2m2m1m{}^{2}m{}^{2}m{}^{1}m (d)2m2m1m{}^{2}m{}^{2}m{}^{1}m (d)2m2m1m{}^{2}m{}^{2}m{}^{1}m (d)
Bulk ii-wave243m{}^{2}4{}^{3}m24/2m{}^{2}4/{}^{2}m (d)24/2m{}^{2}4/{}^{2}m (d)24/2m{}^{2}4/{}^{2}m (d)2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)2m2m12{}^{2}m{}^{2}m{}^{1}2 (d)
Bulk ii-wave2432{}^{2}4{}^{3}2242222{}^{2}4{}^{2}2{}^{2}2 (d)242222{}^{2}4{}^{2}2{}^{2}2 (d)242222{}^{2}4{}^{2}2{}^{2}2 (d)222212{}^{2}2{}^{2}2{}^{1}2 (d)222212{}^{2}2{}^{2}2{}^{1}2 (d)222212{}^{2}2{}^{2}2{}^{1}2 (d)

Table 2. Tuning altermagnetism by the application of strain. The first and second columns denote the altermagnetic pattern and the parent SPG, respectively. The latter six columns denote the effective SPG of the parent space group after the application of the strain field ηij\eta_{ij}.

Our symmetry analysis shows how external stimuli, which break spin group symmetries that enforce zero spin conductivity, can be used to allow spin conductivity in a larger group of materials. It should be further noted that the symmetry-lowering stimuli in Table I and Table 2 can be applied simultaneously to obtain spin conductivities in conventional magnets through piezomagnetoelectric-like (Eiηjk\mathcal{E}_i\eta_{jk}) couplings, where electric and strain fields act in concert. For example, AFM SPG 14/2m2m2m1^{4}/{}_{2}m^{2}m^{2}m allows spin conductivity under the application of Ezηij\mathcal{E}_z\eta_{ij}, but not when Ez\mathcal{E}_z or ηij\eta_{ij} are used separately.

DFT Analysis

In this section, we use DFT-based calculations to investigate spin current generation in selected materials that are representative of the distinct symmetry cases summarized in Tables I and II. We begin by analyzing the spin conductivity in both bulk and two-dimensional dd-wave AMs, which intrinsically support finite spin currents. Subsequently, we consider prototypical materials to explore three key mechanisms for spin current generation: (i) an electric field–driven transition from antiferromagnetic (AFM) to uncompensated magnetic order, (ii) a strain-induced symmetry transition from gg-wave AM to dd-wave AM, and (iii) a combined electric field and strain-mediated transition from ii-wave to dd-wave altermagnetism. Note that in accordance to Neumann’s principle, we only need to consider point group operations for the response tensors that represent macroscopic/bulk properties [34]. Hence, we have omitted the translation part from the symmetry operations. Also, we have omitted the spin-only symmetry group “∞m1\infty m1” from our discussion of SPGs as this continuous symmetry only enforces conditions on the transverse spin polarization response (σijx/y=0\sigma_{ij}^{x/y}=0), but imposes no conditions on σijz\sigma_{ij}^{z}. Additionally, we define the charge-to-spin conversion ratio as 2eσijz/ℏσij2e\sigma_{ij}^{z}/\hbar\sigma_{ij} [16]. In the limiting case, when energy falls into a region without bands, the charge conductivity σij\sigma_{ij} approaches zero, and this ratio becomes mathematically indeterminate. To address this, we set the conversion ratio to zero whenever σij\sigma_{ij} falls below a defined threshold in the charge-to-spin conversion versus energy plots. Note that in semiconductors, hole or electron doping is required to observe spin transport. A low-level doping concentration (≈2%\approx2\%) is sufficient to provide a stable carrier density while maintaining the compensated magnetic order [35].

dd-wave AMs

KVSe2_2O belongs to a broader class of materials crystallizing in the orthorhombic crystal structure with a space group of P4/mmmP4/mmm [see Figure 2(a)]. KVSe2_2O is a room temperature metallic altermagnet with quasi-one-dimensional Fermi sheets extending along the kxk_x and kyk_y direction, and quasi-2D Fermi pockets centered around the X and Y points as seen in Figure 2(b). The SPG of KVSe2_2O is 24/1m2m^{2}4/{}^{1}m{}^{2}m, containing [C2∣∣C4z][C_2||C_{4z}], [E∣∣P][E||P], [E∣∣Mz][E||M_z], [E∣∣Mx][E||M_x], and [C2∣∣Mxy][C_2||M_{xy}]. Following Neumann’s principle, the symmetry operations of this material’s SPG are adopted by its spin anisotropy. Using the transformation rules of Eq. (1) and Eq. (2), [C2∣∣C4z][C_2||C_{4z}] transforms σyysx=σxxsx\sigma_{yy}^{sx}=\sigma_{xx}^{sx}, σxysx=−σyxsx\sigma_{xy}^{sx}=-\sigma_{yx}^{sx}, σxxsz=σyysz\sigma_{xx}^{sz}=\sigma_{yy}^{sz}, and σzysz=−σzxsz\sigma_{zy}^{sz}=-\sigma_{zx}^{sz}. Therefore, [C2∣∣C4z][C_2||C_{4z}] enforces σxxsz=σyysz=σxxsz=σzzsz=0\sigma_{xx}^{sz}=\sigma_{yy}^{sz}=\sigma_{xx}^{sz}=\sigma_{zz}^{sz}=0. Furthermore, [C2∣∣Mz][C_2||M_z] transforms σxysx→−σyxsx\sigma_{xy}^{sx}\to-\sigma_{yx}^{sx}, and hence, σxysx=σyxsx=0\sigma_{xy}^{sx}=\sigma_{yx}^{sx}=0 (Eq. (2)). Following a similar procedure for all symmetry operations, the symmetry allowed components of the spin conductivity tensor are σyysz\sigma_{yy}^{sz} and σzxsz\sigma_{zx}^{sz}, with σzysz=−σzxsz\sigma_{zy}^{sz}=-\sigma_{zx}^{sz}. Figure 2(c) shows the spin conductivity, σzxsz\sigma_{zx}^{sz}, for the different scattering rates. The spin conductivity is inversely proportional to Γ\Gamma within the considered range of 1 meV to 100 meV. The charge-to-spin conversion efficiency (2eℏσijkσij\frac{2e}{\hbar}\frac{\sigma_{ij}^{k}}{\sigma_{ij}}) is 76%, which can reach almost 100% with electron doping of 0.5 eV [see Figure 2(d)], where the Fermi pockets around the X and Y points vanish and spin currents are entirely governed by the Fermi sheets.

Crystal and magnetic structures, Fermi surface, and conductivity of KVSe$_2$O

Figure 2. (a) The crystal and magnetic structures of KVSe2_2O. (b) The Fermi surface of KVSe2_2O projected on the kz=0k_z = 0 plane, calculated without spin-orbit coupling. The red and blue curves denote spin-up and spin-down Fermi surfaces, respectively. (c) The charge and spin conductivity as a function of the scattering rate (Γ\Gamma). (d) The spin conductivity and corresponding charge-to-spin conversion ratio as a function of the Fermi energy.

RuF4_4 is a well known 2D dd-wave AM crystallizing in space group P21/cP2_1/c [see Figure 3 (a)]. The SPG of RuF4_4 is 22/2m{}^{2}2/{}^{2}m with spin symmetry operations [C2∣∣My][C_2||M_y] and The longitudinal and transverse spin conductivity for an arbitrary direction of the electric field for KVSe2_2O and RuF4_4 is plotted in § 1 of the SM.

Crystal and magnetic structures, Fermi surface, and conductivity of RuF$_4$

Figure 3. (a) The crystal and magnetic structures of RuF4_4. (b) The Fermi surface of RuF4_4 projected on the kz=0k_z = 0 plane, calculated without spin-orbit coupling. (c) The spin conductivity and (d) corresponding charge-to-spin conversion ratio as a function of the Fermi energy.

[E∥\parallelP] [see Figure 3 (b)]. [C2∥_2\parallelMy_y] transforms σiisz=−σii−sz\sigma_{ii}^{s_z} = -\sigma_{ii}^{-s_z}, therefore σiisz=0\sigma_{ii}^{s_z} = 0. Also, [C2∥_2\parallelMy_y] leads to σxysz=−σxy−sz\sigma_{xy}^{s_z} = -\sigma_{xy}^{-s_z}, σyxsz=−σyx−sz\sigma_{yx}^{s_z} = -\sigma_{yx}^{-s_z}, σxzsz=σxz−sz\sigma_{xz}^{s_z} = \sigma_{xz}^{-s_z}, σzxsz=σzx−sz\sigma_{zx}^{s_z} = \sigma_{zx}^{-s_z}, σyzsz=−σyz−sz\sigma_{yz}^{s_z} = -\sigma_{yz}^{-s_z}, and σzysz=−σzy−sz\sigma_{zy}^{s_z} = -\sigma_{zy}^{-s_z}. Therefore, the nonzero components of the spin conductivity tensor are σxysz\sigma_{xy}^{s_z}, σyxsz\sigma_{yx}^{s_z}, σyzsz\sigma_{yz}^{s_z}, and σzysz\sigma_{zy}^{s_z}. Since RuF4_4 is a semiconductor, the charge conductivity is zero at the Fermi energy. Figure 3(c) shows the spin conductivity, σxysz\sigma_{xy}^{s_z}, of the hole-doped RuF4_4 as a function of the Fermi energy. The maximum charge-to-spin conversion efficiency is 80% when the hole doping moves the Fermi level below the intrinsic Fermi level by ∼0.9\sim0.9 eV [Figure 3(c)].

Note that although both KVSe2_2O and RuF4_4 are dd-wave AMs, their respective responses to the applied electric field differ from each other. This is owing to the different spin anisotropies of their Fermi surfaces. In the case of RuF4_4, the bands are spin-degenerate along the xx and yy directions, while for KVSe2_2O, the bands are spin-degenerate along the x+yx+y and x−yx-y directions. Hence, if the applied electric field is in the xx direction, KVSe2_2O will show longitudinal spin conductivity while RuF4_4 will exhibit transverse spin conductivity. On the other hand, KVSe2_2O will exhibit transverse spin conductivity, and RuF4_4 will show longitudinal spin conductivity when an electric field is applied along the x+yx+y or x−yx-y direction.

Ez\mathcal{E}_z induced AFM to uncompensated magnetism

Bulk Cr2_2O3_3 is a well-known magnetoelectric antiferromagnet that crystallizes in the Rˉ3c\bar{R}3c space group symmetry. The SPG of Cr2_2O3_3 is $2\bar{3}1^{\prime}1m$ with the following spin symmetry operations: [C2∥_2\parallelP], [E∥\parallelMy_y], [E∥\parallelC_{3z}$], and [C2∥_2\parallelS3z_{3z}] [see Figure 4(a)]. Since the [C2∥_2\parallelP] symmetry operation transform σijsz\sigma_{ij}^{s_z} to σij−sz\sigma_{ij}^{-s_z}, spin conductivity is forbidden in bulk Cr2_2O3_3. The application of Ez\mathcal{E}_z breaks [C2∥_2\parallelS3z_{3z}] and [C2∥_2\parallelP], while preserving [E∥\parallelMy_y] and [E∥\parallelC3z_{3z}] symmetries. This reduces the original SPG to a ferromagnetic 13ˉ1′1m1\bar{3}1^{\prime}1m SPG (see Table I). To apply the electric field, we used a [1] film of Cr2_2O3_3 with a 20 Å vacuum separating the film from its image in the zz-direction. This structure was previously found to result in the most stable surface [36]. Since the electric field, Ez\mathcal{E}_z, breaks the [C2∥_2\parallelP] symmetry, which is the only symmetry that connects the opposite spin sublattices, it leads to a Zeeman-like spin splitting of the conduction and valence bands, as seen in Figure 4(b). Despite the presence of spin splitting, which is similar to that of a ferromagnet, our DFT calculations reveal a net zero magnetic moment. This apparent contradiction can be attributed to the semiconducting nature of Cr2_2O3_3, which gives rise to zero net magnetic moment according to the Luttinger’s criterion. In a stoichiometric compound, Luttinger’s criterion requires that the net magnetization be equal to NμBN \mu_B (NN=integer). When at least one spin channel is gapped (i.e. the material is either a semiconductor or a half metal), and the orbital moment is quenched, this results in zero net magnetic moment for small perturbations. The [E∥My][E\parallel M_y] and [E∥C3z][E\parallel C_{3z}] symmetries allow spin conductivity components – σxxz\sigma_{xx}^{z}, σyyz\sigma_{yy}^{z}, and σzzz\sigma_{zz}^{z} – to be non-zero, with σxxz=σyyz\sigma_{xx}^{z}=\sigma_{yy}^{z}. It should be pointed out that since Jxz=σxxzExJ_x^z=\sigma_{xx}^{z}E_x and σxxz\sigma_{xx}^{z} itself depends on EzE_z, we get that Jx∝ExEzJ_x\propto E_xE_z. In other words, the spin conductivity is at least second order in the electric field. Figure 4(c) plots the spin conductivity, σxxz\sigma_{xx}^{z}, for the electron- and hole-doped Cr2_2O3_3 as a function of Fermi energy with and without an external electric field (Ez=0.1E_z=0.1 V/Å). The spin conductivity σxxz\sigma_{xx}^{z} switches signs on going from negative values of Fermi energy (corresponding to hole-doping) to the positive values (electron doping). This is owing to the opposite spin magnetization of the topmost valence and lowest conduction bands. As seen in Figure 4(c), the maximum value of σxxz\sigma_{xx}^{z} is ∼750(ℏ/e)\sim750(\hbar/e)S/cm for the electron-doped material. Figure 4(d) plots the charge-to-spin conversion ratio as a function of Fermi energy. It is clearly seen that the charge current is fully spin polarized around the CBM and VBM, as only one type of spin-conduction channel contributes to the conductivity. On the other hand, away from the CBM and VBM, the charge-to-spin conversion ratio decreases as both spin conduction channels start contributing.

Crystal and magnetic structures of Cr₂O₃, spin-polarized band structure, spin conductivity, and charge-to-spin conversion ratio

Figure 4. (a) The crystal and magnetic structures of Cr2_2O3_3. The blue and red spheres are used for Cr and O atoms, respectively. (b) Spin-polarized band structure of [1] film of Cr2_2O3_3 with and without the electric field, EzE_z. (c) The spin conductivity and (d) corresponding charge-to-spin conversion ratio as a function of the Fermi energy with and without EzE_z.

ηxy\eta_{xy} induced gg-wave to dd-wave transition

FeS2_2 belongs to a large class of planar 2D materials that crystallize in the P4/mbmP4/mbm space group [Fig. 5(a)]. FeS2_2 is a planar gg-wave AMAM with nontrivial SPG 1/4∣m2m2m1/4|m^2m^2m containing [C2∣Mx][C_2|M_x], [C2∣Mxy][C_2|M_{xy}], [C2∣Mxy][C_2|M_{xy}], [E∣∣C4z][E||C_{4z}], and [E∣∣Mz][E||M_z]. As discussed previously, the symmetry operation [C2∣∣Mx][C_2||M_x] suppresses the following spin conductivity components: σxxz\sigma_{xx}^{z}, σyyz\sigma_{yy}^{z}, σzzz\sigma_{zz}^{z}, σyzz\sigma_{yz}^{z}, and σzyz\sigma_{zy}^{z}. Additionally, the mirror symmetry, [E∣∣Mz][E||M_z], eliminates σxzz\sigma_{xz}^{z} and σzxz\sigma_{zx}^{z}, while [C2∣∣Mxy][C_2||M_{xy}] forbids σxyz\sigma_{xy}^{z} and σyxz\sigma_{yx}^{z}. Therefore, spin current is completely forbidden in a planar gg-wave AM. Consistent with the symmetry analysis, DFT calculations confirm a vanishing spin conductivity [see the solid black line in Figure 5(e)]. The gg-wave SPG 1/4∣m2m2m1/4|m^2m^2m of FeS2_2 reduces to a dd-wave SPG (2m2m1m2m^2m^1m) through application of ηxx\eta_{xx}, ηyy\eta_{yy}, or ηxy\eta_{xy} strains, whereas it can be reduced to a dd-wave SPG (22/2m2^2/2m) through an application of ηxz\eta_{xz} or ηyz\eta_{yz} strains [see Table II]. As a representative case, we use the shear strain, ηxy\eta_{xy}, of 2%. The structural changes produced by this strain can be seen in Figure 5 (b), where we have exaggerated the strain percentage to 20% for the sake of representation only. The SPG of FeS2_2 under ηxy\eta_{xy} strain is 2m2m1m2m2m1m, with spin symmetry operations [C2∥Mxy][C_2\Vert M_{xy}], [C2∥Mxy′][C_2\Vert M_{xy}^{\prime}], and [E∥Mz][E\Vert M_z]. Therefore, 2m2m1m2m2m1m allows for non-zero σxxz\sigma_{xx}^z and σyyz\sigma_{yy}^z components, with σyyz=−σxxz\sigma_{yy}^z=-\sigma_{xx}^z. Figure 5(c) and Figure 5(d) show the altermagnetic phase transition from planar gg-wave to planar dd-wave upon the application of the strain. The maximum spin conductivity value is calculated to be upto 2800 ℏ\hbar S/cm as seen in 5(e). This occurs at an energy of -1.45 eV, where the Fermi surface is mostly composed of spin-polarized sheets along the xx and yy directions [Figure 5(d) and Figure 5(e)]. These Fermi sheets are similar to those for the case of KVSe2_2O [Figure 2(b)].

Crystal and magnetic structures, constant energy contours, spin conductivity, and charge-to-spin conversion ratio of FeS₂

Figure 5. (a) The crystal and magnetic structures of an FeS2_2 monolayer (a) without and (b) with strain field, ηxy\eta_{xy}. The magnetic structures in (a) and (b) are shown by superimposing the ⊙\odot and ⊗\otimes symbols on the Fe atoms, representing spins that point in +z+z (i.e out of the plane of paper) and −z-z (into the plane of paper) directions, respectively. Also, in order to make the strain-induced distortions visible to the naked eye, we have exaggerated the strain to 20% in (b). The constant energy contours for (c) unstrained and (d) strained FeS2_2 for EF−1.45E_F - 1.45 eV with maximum charge-to-spin conversion. The (e) spin conductivity and (f) charge-to-spin conversion ratio as a function of energy.

5(d) and 5(e)]. These Fermi sheets are similar to those for the case of KVSe2_2O [Figure 2(b)]. As seen in Figure 5(f), the maximum spin-to-charge conversion ratio is 50%, which is smaller than that obtained for KVSe2_2O due to the Fermi pockets around X and Y, which are slighter bigger than those of KVSe2_2O [compare Figure 2(b) and Figure 5(d)]. Recall from subsection A that transverse spin conductivity will be observed when the electric field is applied along the x±yx\pm y directions [see section I of SM].

EzηxxE_z\eta_{xx} induced AFM to dd-wave AM transition

MnPSe3_3 is a 2D antiferromagnet, which crystallizes in the P‾31m\overline{P}31m space group [Figure 6(a)]. The SPG of MnPSe3_3 is 232m112^3 2m^1 1, and spin current is forbidden due to spin-degenerate bands. MnPSe3_3 has been the subject of several reports that have studied its AFM to dd-wave AM transition under the application of a vertical electric field [[9, 37]. Under the application of an electric field (Ez\mathcal{E}_z), the SPG of MnPSe3_3 reduces from 232m112^3 2m^1 1 to 132m111^3 2m^1 1, with the latter containing [E∥C3z][E\Vert C_{3z}] and [C2∥Mx][C_2\Vert M_x] symmetries [see Table I and Figure 6(b)]. Although the 132m111^3 2m^1 1 SPG corresponds to the bulk gg-wave AM, the 2D nature of MnPSe3_3 (kz=0k_z=0) makes it a planar ii-wave AM [Figure 6(c)]. The ii-wave altermagnetic nature of monolayer MnPSe3_3 can be confirmed from the six-fold symmetric constant energy-contours in Figure 6(c). The spin currents are forbidden with and without the electric field due to underlying spin group symmetries [E∥C3z][E\Vert C_{3z}] and [C2∥Mx][C_2\Vert M_x]. This is due to the fact that [E∥C3z][E\Vert C_{3z}] enforces σxxz=σyyz\sigma_{xx}^z=\sigma_{yy}^z, σxyz=−σyxz\sigma_{xy}^z=-\sigma_{yx}^z, and σxzz=σyzz=σzxz=σzyz=0\sigma_{xz}^z=\sigma_{yz}^z=\sigma_{zx}^z=\sigma_{zy}^z=0, while the [C2∥Mx][C_2\Vert M_x] symmetry leads to σiiz=−σii−z\sigma_{ii}^z=-\sigma_{ii}^{-z}, σyzz=σzy−z\sigma_{yz}^z=\sigma_{zy}^{-z}, σzyz=σzy−z\sigma_{zy}^z=\sigma_{zy}^{-z}, σxyz=−σxy−z\sigma_{xy}^z=-\sigma_{xy}^{-z}, σyxz=−σyx−z\sigma_{yx}^z=-\sigma_{yx}^{-z}, σxzz=−σxz−z\sigma_{xz}^z=-\sigma_{xz}^{-z}, and σzxz=−σzx−z\sigma_{zx}^z=-\sigma_{zx}^{-z}. However, the 132m111^3 2m^1 1 SPG can be further reduced to the 2m2m SPG by using ηxy\eta_{xy}, ηyy\eta_{yy}, and ηyz\eta_{yz} [Table II]. On the other hand, an application of either ηxy\eta_{xy} or ηxz\eta_{xz} reduces the SPG to that of an uncompensated magnet, while the 132m111^3 2m^1 1 SPG remains unaltered by the uniaxial strain along the zz-direction (ηzz\eta_{zz}). Therefore, monolayer MnPSe3_3 AFM will undergo a magnetic transition to a dd-wave altermagnet under Ezηxx\mathcal{E}_z\eta_{xx}, Ezηyy\mathcal{E}_z\eta_{yy}, and Ezηyz\mathcal{E}_z\eta_{yz}. For illustrative purposes, we consider Ez(ηxx+ηyy)\mathcal{E}_z(\eta_{xx}+\eta_{yy}) with ηxx=−ηyy=0.02\eta_{xx}=-\eta_{yy}=0.02, resulting in the structure shown in Figure 6(b), and the corresponding constant energy contour in Figure 6(d). The new SPG of MnPSe3_3 under this strain is 2m^2m, containing the [C2∥Mx][C_2\Vert M_x] symmetry operation [see Tables I and II]. As mentioned earlier, the only nonzero components of the spin conductivity tensor are σxzy\sigma_{xz}^y, σyzx\sigma_{yz}^x, σzzy\sigma_{zz}^y, and σzyx\sigma_{zy}^x. The σxzy\sigma_{xz}^y and σzyx\sigma_{zy}^x components are forbidden due to the 2D nature of MnPSe3_3. Figure 6(e) and Figure 6(f) show the spin conductivity and charge-to-spin conversion ratio for MnPSe3_3 with electric field Ez\mathcal{E}_z and Ez(ηxx+ηyy)\mathcal{E}_z(\eta_{xx}+\eta_{yy}). While there is no spin current in the absence of strain, the strain-induced charge-to-spin conversion ratio reaches up to ∼\sim10%, which corresponds to a spin-splitting angle of approximately 11°. Here we have used the following expression for the spin-splitter

Crystal and magnetic structures, constant energy contours, spin conductivity, and charge-to-spin conversion ratio

Figure 6. (a) The crystal and magnetic structures of the monolayer MnPSe3_3 (a) without and (b) with applied strain. In (b), the strain, ηxx+ηyy\eta_{xx}+\eta_{yy} (with ηxx=−ηyy\eta_{xx}=-\eta_{yy}), is exaggerated to 10% to magnify the structural distortions, making them visible to the naked eye. Also, the magnetization in (a) and (b) are shown by superimposing the ⊙\odot and ⊗\otimes symbols on the Mn atoms, representing spins that point in +z+z (i.e out of the plane of paper) and −z-z (into the plane of paper) directions, respectively. The constant energy contours (at an energy of EF−1.02E_{\mathrm{F}}-1.02 eV) for (c) Ez=0.5E_z=0.5 V/Å and (d) Ez=0.5E_z=0.5 V/Å, and ηxx=−ηyy=0.02\eta_{xx}=-\eta_{yy}=0.02, showing the ii to dd wave transition. The (e) spin conductivity and (f) charge-to-spin conversion ratio as a function of energy.

@references 马会 angle: θij=arctan⁡(22eℏσijsσiic)\theta_{ij} = \arctan\left(2\frac{2e}{\hbar}\frac{\sigma^{s}_{ij}}{\sigma^{c}_{ii}}\right) [13, 17], with σijs\sigma^{s}_{ij} (σijc\sigma^{c}_{ij}) being the transverse spin conductivity (longitudinal charge conductivity). This spin-splitter angle is experimentally measurable [17, 38].

Conclusion

In this work, we have proposed a general framework to obtain nonrelativistic spin currents in collinear AFMs, and gg- and ii-wave AMs. Using symmetry analysis, we have demonstrated that spin conductivity can be induced through magnetoelectric, piezomagnetic, and piezomagnetoelectric-like couplings. The exact functional form of the spin-conductivity tensor is determined by the effective reduced SPG under the application of an external electric field and/or strain. By performing DFT calculations on realistic materials that were subjected to either electric fields, strains, or both, we have shown that the charge-to-spin conversion ratios in AFMs, as well as gg- and ii-wave AMs, can be as high as those seen for ferromagnets, compensated ferrimagnets, and dd-wave altermagnets. It should be pointed out that the external stimuli considered in our work are realistic and can be achieved experimentally. For example, the direction-dependent strain field can be obtained by transferring the exfoliating altermagnetic crystals onto a flexible substrate with pre-patterned metal pads [39, 40]. The feasibility of our proposed approaches is also confirmed by recent reports of in-plane strain and magnetic field-induced nematicity in AFM intercalated transition metal dichalcogenides [41]. In addition, recently, finite anomalous Hall conductivity was successfully measured in the semiconducting gg-wave AM, α\alpha-Fe2_2O3_3, with 1% Ti doping [35]. Hence, doping to induce charge conductivity, along with electric fields and/or strain to drive the symmetry-lowering phase transitions, can be used to observe spin conductivity in intrinsically-semiconducting AFMs and higher-order AMs. Our proposed design principles for inducing spin currents are general and will expand the range of materials available for ultrafast magnetic spintronics.

Acknowledgments

The research performed at Howard University was supported by Air Force Office of Scientific Research under award number FA9550-23-1-0679 and the National Science Foundation Grant No. DMR-1752840. This work used the Expanse cluster at SDSC through allocation PHY180014 from the Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support (ACCESS) program, which is supported by National Science Foundation Grants Nos. 2138259, 2138286, 2138307, 2137603, and 2138296, and Maryland Advanced Research Computing Center.

References

  1. [1]I. Žutić, J. Fabian, and S. D. Sarma, Spintronics: Fundamentals and applications, Reviews of modern physics 76, 323 (2004).DOI
  2. [2]J. Hirsch, Spin hall effect, Physical review letters 83, 1834 (1999).
  3. [3]M. Dyakonov and V. Perel, Spin relaxation of conduction electrons in noncentrosymmetric semiconductors, Soviet Physics Solid State, Ussr 13, 3023 (1972).
  4. [4]R. J. Elliott, Theory of the effect of spin-orbit coupling on magnetic resonance in some semiconductors, Physical Review 96, 266 (1954).
  5. [5]Y. Yafet, Solid state physics. vol. 14/edited by f. seitz and d. turnbull (1963).
  6. [6]S. Hayami, Y. Yanagi, and H. Kusunose, Bottom-up design of spin-split and reshaped electronic band structures in antiferromagnets without spin-orbit coupling: Procedure on the basis of augmented multipoles, Physical Review B 102, 144441 (2020).
  7. [7]L.-D. Yuan, Z. Wang, J.-W. Luo, and A. Zunger, Prediction of low-z collinear and noncollinear antiferromagnetic compounds having momentum-dependent spin splitting even without spin-orbit coupling, Physical Review Materials 5, 014409 (2021).
  8. [8]L. Šmejkal, J. Sinova, and T. Jungwirth, Beyond conventional ferromagnetism and antiferromagnetism: A phase with nonrelativistic spin and crystal rotation symmetry, Physical Review X 12, 031042 (2022).
  9. [9]S. Sheoran and P. Dev, Spontaneous anomalous hall effect in two-dimensional altermagnets, Physical Review B 111, 184407 (2025).
  10. [10]J. Železný, H. Gao, A. Manchon, F. Freimuth, Y. Mokrousov, J. Zemen, J. Mašek, J. Sinova, and T. Jungwirth, Spin-orbit torques in locally and globally noncentrosymmetric crystals: Antiferromagnets and ferromagnets, Physical Review B 95, 014403 (2017).
  11. [11]B. Flebus, Y. Tserkovnyak, and G. A. Fiete, Interfacial spin seebeck effect in noncollinear magnetic systems, Physical Review B 99, 224410 (2019).
  12. [12]S.-W. Cheong and F.-T. Huang, Altermagnetism with non-collinear spins, npj Quantum Materials 9, 13 (2024).
  13. [13]R. González-Hernández, L. Šmejkal, K. Výborný, Y. Yanagi, J. Sinova, T. Jungwirth, and J. Železný, Efficient electrical spin splitter based on nonrelativistic collinear antiferromagnetism, Physical Review Letters 126, 127701 (2021).
  14. [14]H. Vakili, E. Schwartz, and A. A. Kovalev, Spin-transfer torque in altermagnets with magnetic textures, Physical Review Letters 134, 176401 (2025).
  15. [15]A. Bose, N. J. Schreiber, R. Jain, D.-F. Shao, H. P. Nair, J. Sun, X. S. Zhang, D. A. Muller, E. Y. Tsymbal, D. G. Schlom, et al., Tilted spin current generated by the collinear antiferromagnet ruthenium dioxide, Nature Electronics 5, 267 (2022).
  16. [16]K. Belashchenko, Giant strain-induced spin splitting effect in mnte, ag-wave altermagnetic semiconductor, Physical Review Letters 134, 086701 (2025).DOI
  17. [17]B. Karetta, X. H. Verbeek, R. Jaeschke-Ubiergo, L. Šmejkal, and J. Sinova, Strain controlled g-to d-wave transition in altermagnetic crsb, arXiv preprint arXiv:2505.21293 (2025).DOI
  18. [18]L. Šmejkal, Altermagnetic multiferroics and altermagne-toelectric effect, arXiv preprint arXiv:2411.19928 (2024).
  19. [19]D. Wang, H. Wang, L. Liu, J. Zhang, and H. Zhang, Electric-field-induced switchable two-dimensional altermagnets, Nano Letters 25, 498 (2024).DOI
  20. [20]I. Mazin, R. González-Hernández, and L. Šmejkal, Induced monolayer altermagnetism in mp (s, se) _3 and fese, arXiv preprint arXiv:2309.02355 (2023).arxiv.org/abs/2309.02355
  21. [21]J. Callaway, Quantum theory of the solid state (Academic press, 2013).
  22. [22]D. Vanderbilt, Soft self-consistent pseudopotentials in a generalized eigenvalue formalism, Physical review B 41, 7892 (1990).pubmed.ncbi.nlm.nih.gov/9993096
  23. [23]P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni, I. Dabo, et al., Quantum espresso: a modular and open-source software project for quantum simulations of materials, Journal of physics: Condensed matter 21, 395502 (2009).DOI
  24. [24]J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Physical review letters 77, 3865 (1996).pubmed.ncbi.nlm.nih.gov/10062328
  25. [25]N. Marzari and D. Vanderbilt, Maximally localized generalized wannier functions for composite energy bands, Physical review B 56, 12847 (1997).DOI
  26. [26]A. A. Mostofi, J. R. Yates, Y.-S. Lee, I. Souza, D. Vanderbilt, and N. Marzari, wannier90: A tool for obtaining maximally-localised wannier functions, Computer physics communications 178, 685 (2008).DOI
  27. [27]H. T. Stokes and D. M. Hatch, Findsym: program for identifying the space-group symmetry of a crystal, Journal of Applied Crystallography 38, 237 (2005).DOI
  28. [28]M. I. Aroyo, J. M. Perez-Mato, C. Capillas, E. Kroumova, S. Ivantchev, G. Madariaga, A. Kirov, and H. Wondratschek, Bilbao crystallographic server: I. databases and crystallographic computing programs, Zeitschrift für Kristallographie-Crystalline Materials 221, 15 (2006).
  29. [29]X. Chen, J. Ren, Y. Zhu, Y. Yu, A. Zhang, P. Liu, J. Li, Y. Liu, C. L. Li, and Q. Liu, Enumeration and representation theory of spin space groups, Physical Review X 14, 031038 (2024).DOI
  30. [30]R.-C. Xiao, Y. Jin, Z.-F. Zhang, Z.-H. Feng, D.-F. Shao, and M. Tian, Tensorsymmetry: a package to get symmetry-adapted tensors disentangling spin-orbit coupling effect and establishing analytical relationship with magnetic order, arXiv preprint arXiv:2504.04424 (2025).DOI
  31. [31]M. Roig, A. Kreisel, Y. Yu, B. M. Andersen, and D. F. Agterberg, Minimal models for altermagnetism, Physical Review B 110, 144412 (2024).DOI
  32. [32]D. B. Litvin and W. Opechowski, Spin groups, Physica 76, 538 (1974).DOI
  33. [33]M. Ezawa, Third-order and fifth-order nonlinear spin-current generation in g-wave and i-wave altermagnets and perfectly nonreciprocal spin current in f-wave magnets, Physical Review B 111, 125420 (2025).DOI
  34. [34]A. Cracknell, Symmetry properties of the transport coefficients of magnetic crystals, Physical Review B 7, 2145 (1973).DOI
  35. [35]E. Galindez-Ruales, R. Gonzalez-Hernandez, C. Schmitt, S. Das, F. Fuhrmann, A. Ross, E. Golias, A. Akashdeep, L. Lünénbürger, E. Baek, et al., Revealing the altermagnetism in hematite via xmcd imaging of anomalous hall electrical transport, Advanced Materials , e05019 (2025).
  36. [36]A. L. Wysocki, S. Shi, and K. D. Belashchenko, Microscopic origin of the structural phase transitions at the cr 2 o 3 (0001) surface, Physical Review B—Condensed Matter and Materials Physics 86, 165443 (2012).DOI
  37. [37]S. Sheoran and S. Bhattacharya, Nonrelativistic spin splittings and altermagnetism in twisted bilayers of centrosymmetric antiferromagnets, Physical Review Materials 8, L051401 (2024).arxiv.org/abs/2310.19395
  38. [38]X. Tao, Q. Liu, B. Miao, R. Yu, Z. Feng, L. Sun, B. You, J. Du, K. Chen, S. Zhang, et al., Self-consistent determination of spin hall angle and spin diffusion length in pt and pd: The role of the interface spin loss, Science advances 4, eaat1670 (2018).DOI
  39. [39]J. Son, K.-H. Kim, Y. Ahn, H.-W. Lee, and J. Lee, Strain engineering of the berry curvature dipole and valley magnetization in monolayer mos 2, Physical review letters 123, 036806 (2019).DOI
  40. [40]S. Sheoran, M. Jain, R. Moulik, and S. Bhattacharya, Probing the uniaxial strain-dependent valley drift and berry curvature in monolayer mosi 2 n 4, Physical Review Materials 7, 114003 (2023).DOI
  41. [41]Z. Feng, W. Lu, T. Lu, F. Liu, J. R. Sheeran, M. Ye, J. Xia, T. Kurumaji, and L. Ye, Nonvolatile nematic order manipulated by strain and magnetic field in a layered antiferromagnet, arXiv preprint arXiv:2507.05486 (2025).arxiv.org/abs/2507.05486

Paper details

vlnc:W_p6t44vxj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 10

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents