betaFeedback

WorksPhysics and Astronomy

Topological Spin-Orbit Torque in Ferrimagnetic Weyl Semimetal

Abstract

The spin-orbit torque (SOT) in a compensated ferrimagnetic Weyl semimetal, Ti2MnAl{\rm Ti}_{2}{\rm MnAl}, is studied by the linear response theory. We elucidate that the SOT driven by all the occupied electronic states is present in magnetic Weyl semimetal, unlike in conventional metallic magnets. Around the energy of the Weyl points, we find that such an SOT is dominant and almost independent of the disorder. The emergence of the SOT in Ti2MnAl{\rm Ti}_{2}{\rm MnAl} can be understood from the structure of the mixed Berry curvature around the Weyl points, which is similar to that of the ordinary Berry curvature.

Introduction

**Introduction—**Efficient electrical control of electron spins [1, 2, 3, 4] is one of the fundamental aims of spintronics. Spin-orbit coupling (SOC) plays a significant role by correlating the electron spin and orbital motion. The SOC enables the electrical generation of spin polarization in heterostructures [3, 5, 6, 7, 8, 9, 10] or bulk noncentrosymmetric crystals [11, 12, 13]. The generated spin polarization acts as an effective magnetic field for magnetic moments, whose effect is called the spin-orbit torque (SOT). The SOT has been intensely studied as a promising method for functional control of magnetization [5, 14, 15, 16, 17]. However, to obtain a large SOT, one needs to inject a large longitudinal current, which suffers from the energy dissipation by the Joule heating. Such an SOT is termed the “dissipative SOT”. The problem of dissipation is crucial in metals with a large density of states (DOS) near the Fermi level EFE_\mathrm{F}, where longitudinal conductivity σxx\sigma_{xx} is proportional to the DOS. The dissipation comes from the scattering of a large number of electrons around EFE_\mathrm{F} by disorders and phonons. To avoid the dissipation, we need systems that can generate the SOT with a small longitudinal current.

To suppress the dissipation, we focus on the characteristic transport properties of magnetic Weyl semimetals (MWSMs). The MWSMs feature linear dispersion and point nodes called Weyl points (WPs) [18, 19, 20]. Since the DOS near the energy of the WPs is negligibly small, σxx\sigma_{xx} tends to be suppressed, leading to the small dissipation in the longitudinal transport. The SOT in bulk MWSMs has been studied based on the continuum model with isotropic spin-momentum locking and broken inversion symmetry [21, 22]. The SOT obtained in these studies needs a finite DOS by shifting EFE_\mathrm{F} away from the WPs, similar to σxx\sigma_{xx}, corresponding to the dissipative SOT. By contrast, non-dissipative transport phenomena, such as the anomalous Hall effect, arise from the nontrivial band topology in MWSMs. The band topology is characterized by the Berry curvature (BC) from the WPs in momentum space [23, 24]. The anomalous Hall conductivity σyx\sigma_{yx} corresponds to the integral of the BC over all the occupied electronic states below EFE_\mathrm{F}. Such a mechanism is termed the “topological effect”.

The existence of the SOT driven by all the occupied electronic states below EFE_\mathrm{F} has been pointed out in several metallic magnets [11, 12, 13, 25]. We here refer to such a non-dissipative SOT as the “topological SOT”. Theoretically, the topological SOT can be characterized by a mixed Berry curvature (MBC) [26]. The MBC is the extension of the BC in momentum space to the composite parameter space spanned by momentum and magnetization. In the systems analyzed so far, the magnitude of topological SOT is smaller than the dissipative SOT. To enhance the topological SOT with low Joule heating, materials showing a large MBC are desired.

In this letter, we find a large topological SOT in a model of MWSM Ti2_2MnAl. Ti2_2MnAl is one of the candidate systems of ideal MWSMs with compensated ferrimagnetic ordering [27]. In Ti2_2MnAl, both the time-reversal and inversion symmetries are broken. We show that, in Ti2_2MnAl, the topological SOT is significantly larger than dissipative SOT near the energy of WPs with a typical disorder strength. Compared with the SOTs studied in the bulk of noncentrosymmetric antiferromagnets [11, 12, 13], we find a giant effective magnetic field in Ti2_2MnAl. These properties enable us to electrically control the compensated ferrimagnetic ordering with low dissipation in the bulk system. By analyzing the distribution of the MBC, we find that the origin of the large topological SOT is the MBC enhanced around the WPs. With such a large topological SOT, we expect the potential utility of noncentrosymmetric MWSMs, including Ti2_2MnAl, for integrated spintronics devices.

Model

Before proceeding to our analysis, first we explain the electronic properties of the ferrimagnetic Weyl semimetal Ti2_2MnAl [27]. The crystal structure of Ti2_2MnAl is an inverse Heusler type with space group Fˉ43m\bar{F}43m [28, 29] as shown in Figure 1(a). Magnetic moments at two Ti sites anti-parallelly align to that at the Mn site, showing zero net magnetization. The first-principles calculations of the electronic band structure predict that the WPs are located near EFE_\mathrm{F}, where the DOS is negligibly small. These properties lead to the large anomalous Hall angle θAHE=σyx/σxx\theta_\mathrm{AHE}=\sigma_{yx}/\sigma_{xx} [27]. Since not only the ends_mid=1 time-reversal symmetry but also the inversion symmetry is broken due to its crystalline structure, the SOT is permitted in bulk.

Crystal structure of Ti₂MnAl and energy dispersion around the Weyl points

Figure 1. (a) Crystal structure of Ti2_2MnAl. Ti and Mn are responsible for the compensated ferrimagnetic ordering. (b) The energy dispersion around the WPs on ky=0k_y = 0 plane.

For our calculations, we introduce a tight-binding model of Ti2_2MnAl [30]. In reality, all dd orbitals mainly contribute to the electronic structure near EFE_{\mathrm{F}}. For our minimal model to describe the low-energy band structure around EFE_{\mathrm{F}}, we consider a single isotropic orbital for two Ti and Mn as the basis to satisfy the symmetry of the crystalline structure. Here, to distinguish two different Ti sites, we use the notations Ti1 and Ti2 as shown in Figure 1(a). We neglect the orbital at the Al site for simplicity, while the effect of Al is substantially incorporated via SOC as discussed later. Our Hamiltonian HH consists of a hopping term, on-site energy, SOC, and exchange coupling:

H=−∑⟨ij⟩αβstαβCiαs†Cjβs+∑iαsϵαCiαs†Ciαs+i2λSOCa2∑⟨ij⟩αss′Ciαs†[(d1αij×d2αij)⋅σ]ss′Cjαs′−∑iαss′JαCiαs†(σ⋅n^)ss′Ciαs′.(1)\begin{aligned} H={}&-\sum_{\langle ij\rangle\alpha\beta s}t_{\alpha\beta}C_{i\alpha s}^{\dagger}C_{j\beta s}+\sum_{i\alpha s}\epsilon_{\alpha}C_{i\alpha s}^{\dagger}C_{i\alpha s}\\ &+i\frac{2\lambda_{\mathrm{SOC}}}{a^{2}}\sum_{\langle ij\rangle\alpha ss'}C_{i\alpha s}^{\dagger}\left[\left(\mathbf{d}_{1}^{\alpha ij}\times\mathbf{d}_{2}^{\alpha ij}\right)\cdot\boldsymbol{\sigma}\right]_{ss'}C_{j\alpha s'}\\ &-\sum_{i\alpha ss'}J_{\alpha}C_{i\alpha s}^{\dagger}\left(\boldsymbol{\sigma}\cdot\hat{\mathbf{n}}\right)_{ss'}C_{i\alpha s'}. \tag*{(1)} \end{aligned}

Here, Ciαs†C_{i\alpha s}^{\dagger} denotes an electron creation operator of sublattice α=Ti1,Ti2\alpha= \mathrm{Ti1}, \mathrm{Ti2}, and Mn with the spin s=↑,↓s = \uparrow,\downarrow at ii site. We briefly explain each term in the following (see Appendix A for details of our tight-binding model).

The first term includes the inter- and intra-sublattice hoppings. The second term denotes the on-site energy ϵα\epsilon_{\alpha} for each sublattice. The third term describes the effect of SOC due to the local breaking of inversion symmetry for the intra-sublattice hoppings. d1,2αij\mathbf{d}_{1,2}^{\alpha ij} are the two nearest-neighbor hopping vectors from the site ii to jj of the sublattice α\alpha, which are introduced according to the Fu-Kane-Mele model [31]. We consider SOC for the Ti2-Ti2 and Mn-Mn hoppings with the same strength λSOC\lambda_{\mathrm{SOC}}. We neglect the SOC for the Ti1-Ti1 for simplicity. The fourth term represents the exchange coupling between conduction electron spin σ\boldsymbol{\sigma} and the ferrimagnetic ordering. The direction of the ferrimagnetic ordering is represented by the unit vector n^\hat{\mathbf{n}}, and is fixed to the zz-direction in the following calculations. The ferrimagnetism is encoded in the mean-field coupling strength JαJ_{\alpha} for each sublattice α\alpha, where the signs of JTi1J_{\mathrm{Ti1}} and JTi2J_{\mathrm{Ti2}} are taken opposite to JMnJ_{\mathrm{Mn}}.

Once the model is given, we recall the electronic structure around EFE_{\mathrm{F}}. The energy spectrum obtained by the Bloch Hamiltonian of HH, denoted as H(k)\mathcal{H}(\mathbf{k}) exhibits the WPs as shown in Figure 1(b). Red and blue points correspond to the WPs with positive (+) and negative (−) chiralities, respectively. We obtain eight WPs on the kx=0k_x = 0 plane, as well as on the ky=0k_y = 0 and kz=0k_z = 0 planes, and thus 24 WPs in total. We determine EFE_{\mathrm{F}} from the 4/6 filling condition. In this study, we focus on the behavior of SOT around the energy of the WPs, −E0≤EF≤E0-E_0 \leq E_{\mathrm{F}} \leq E_0, where E0=50E_0 = 50 meV characteries the energy scale exhibiting the Weyl dispersion around the WPs.

Electrically induced spin density—We now proceed to the analysis of the SOT in this model. The conduction electron spin driven by the electric field acts as an effective magnetic field and exerts the SOT on the magnetic moments at Ti1, Ti2, and Mn. In the linear response regime, the spin density at sublattice α=(Ti1,Ti2,Mn)\alpha= (\mathrm{Ti1}, \mathrm{Ti2}, \mathrm{Mn}) is calculated by using the Kubo formula [4]:

⟨σμα⟩Ex=iℏlim⁡q→0∑n≠m∫kfnk−fmk+qEnk−Emk+q+iηSnm,k+qα,μJmn,k+qx.(2)\frac{\langle\sigma_{\mu}^{\alpha}\rangle}{E_x}=i\hbar\lim_{q\to0}\sum_{n\ne m}\int_{\mathbf{k}}\frac{f_{nk}-f_{mk+q}}{E_{nk}-E_{mk+q}+i\eta}S_{nm,k+q}^{\alpha,\mu}J_{mn,k+q}^{x}. \tag*{(2)}

Here, the index μ=(x,y,z)\mu=(x,y,z) denotes the direction of the induced spin density, whereas the direction of the external electric field EE is fixed to the xx-direction. We define the spin matrix element Snm,k+qα,μ=⟨nk∣σ^μα∣mk+q⟩S_{nm,\mathbf{k}+\mathbf{q}}^{\alpha,\mu}=\langle n\mathbf{k}|\hat{\sigma}_{\mu}^{\alpha}|m\mathbf{k}+\mathbf{q}\rangle, where σ^μα\hat{\sigma}_{\mu}^{\alpha} is the electron spin operator at the sublattice α\alpha, and ∣nk⟩|n\mathbf{k}\rangle is the eigenstate with eigenenergy EnkE_{nk} of H(k)\mathcal{H}(\mathbf{k}). Jmn,k+qx=⟨mk+q∣(−ev^x)∣nk⟩J_{mn,\mathbf{k}+\mathbf{q}}^{x}=\langle m\mathbf{k}+\mathbf{q}|(-e\hat{v}_{x})|n\mathbf{k}\rangle is the current matrix element where −ev^=−eℏ∂H∂k-e\hat{v}=-\frac{e}{\hbar}\frac{\partial H}{\partial\mathbf{k}} is the current operator. fnk=1/(e(Enk−EF)/kBT+1)f_{nk}=1/(e^{(E_{nk}-E_{\mathrm{F}})/k_{\mathrm{B}}T}+1) is the Fermi-Dirac distribution function, with T=0T=0 here. The integral is taken over the Brillouin zone, with ∫k=(2π)−3∫d3k\int_{\mathbf{k}}=(2\pi)^{-3}\int d^{3}k. For the momentum integration, the momentum space is split into the mesh of 250 × 250 × 250 for all calculations in this paper. We introduce the impurity effect by the first Born approximation, so that the self energy part η\eta in Eq. (2) is replaced by ℏ/2τ\hbar/2\tau with the relaxation time τ\tau.

To understand the origin of the electrically induced spin density, we decompose Eq. (2) into two parts as

⟨σμα⟩disEx=2τ∑n∫k∂fnk∂E Snn,kα,μJnn,kx.(3)\frac{\langle\sigma_{\mu}^{\alpha}\rangle^{\mathrm{dis}}}{E_x}=2\tau\sum_n\int_{\mathbf{k}}\frac{\partial f_{nk}}{\partial E}\,S_{nn,\mathbf{k}}^{\alpha,\mu}J_{nn,\mathbf{k}}^{x}. \tag*{(3)}
⟨σμα⟩topEx=−2ℏ∑n≠m∫kfnkIm⁡[Snm,kα,μJmn,kx](Enk−Emk)2+(ℏ2τ)2.(4)\frac{\langle\sigma_{\mu}^{\alpha}\rangle^{\mathrm{top}}}{E_x}=-2\hbar\sum_{n\ne m}\int_{\mathbf{k}}f_{nk}\frac{\operatorname{Im}\left[S_{nm,\mathbf{k}}^{\alpha,\mu}J_{mn,\mathbf{k}}^{x}\right]}{(E_{nk}-E_{mk})^{2}+\left(\frac{\hbar}{2\tau}\right)^{2}}. \tag*{(4)}

(3) includes only the electronic states near EFE_{\mathrm{F}} from ∂fnk/∂E\partial f_{n\mathbf{k}}/\partial E, whereas (4) describes the contribution from all the electronic states below EFE_{\mathrm{F}} from fnkf_{n\mathbf{k}}. Thus, we classify ⟨σμ⟩dis\langle\sigma_\mu\rangle^{\mathrm{dis}} as the contribution to the “dissipative SOT”, and ⟨σμ⟩top\langle\sigma_\mu\rangle^{\mathrm{top}} as the “topological SOT”. We show later that ⟨σμ⟩top\langle\sigma_\mu\rangle^{\mathrm{top}} in our model is indeed described by the topological feature called the MBC.

Based on the above formulae, we study the EFE_{\mathrm{F}}-dependence of the induced spin densities. Figure 2(a) and Figure 2(b) respectively show the energy dependences of ⟨σμ⟩dis\langle\sigma_\mu\rangle^{\mathrm{dis}} and ⟨σμ⟩top\langle\sigma_\mu\rangle^{\mathrm{top}} with a fixed relaxation time τ=3ℏ/E0\tau= 3\hbar/E_0. In Figure 2(a), the magnitudes of ⟨σμ⟩dis\langle\sigma_\mu\rangle^{\mathrm{dis}} on all the three sites are suppressed around EFE_{\mathrm{F}}, namely, around the energy of the WPs. Such an energy dependence is similar to the results obtained with the minimal Weyl Hamiltonian with the isotropic spin-momentum locking in continuum [21, 22]. Contrary, as shown in Figure 2(b), ⟨σμ⟩top\langle\sigma_\mu\rangle^{\mathrm{top}} for all sublattices are always nonzero around EFE_{\mathrm{F}} and nearly independent of EE. Such a topological response was not seen in the minimal Weyl Hamiltonian [21, 22], which is the artifact of the isotropic spin-momentum locking structure in the Fermi sea around each WP. Since the SOC structure and the WP distribution of Ti2_2MnAl considered here are anisotropic in momentum space, it is capable of showing the finite topological response ⟨σμ⟩top\langle\sigma_\mu\rangle^{\mathrm{top}}. We find that, in response to EE applied to the xx-direction, ⟨σα⟩dis\langle\sigma^\alpha\rangle^{\mathrm{dis}} points to the xx-direction while ⟨σα⟩dis\langle\sigma^\alpha\rangle^{\mathrm{dis}} to the yy-direction.

Figure 2. The energy dependences of (a) the dissipative and (b) topological spin densities on each sublattice when the electric field is applied to the +x+x direction. (c) ⟨σα⟩top\langle\sigma_\alpha\rangle^{\mathrm{top}} are schematically shown by the arrows on each sublattice α=Ti1\alpha= \mathrm{Ti1} (yellow), Ti2\mathrm{Ti2} (red), and Mn\mathrm{Mn} (blue). These spin densities are additive on the compensated ferrimagnetic ordering (green) and act as an effective magnetic field hSOT\mathbf{h}_{\mathrm{SOT}}.

The topological response leads to the highly efficient conversion from current to spin density, because of the low current density. To compare the spin density generated per unit current on each sublattice with the conventional Rashba system, we evaluate a dimensionless index ζα=evF⟨σyα⟩top⟨jx⟩dis.\zeta^\alpha= e v_{\mathrm{F}} \frac{\langle\sigma_y^\alpha\rangle^{\mathrm{top}}}{\langle j_x \rangle^{\mathrm{dis}}}. Here, ⟨jx⟩dis\langle j_x \rangle^{\mathrm{dis}} is the longitudinal current flowing in the xx-direction, which is calculated by the formula similar to (3), by replacing σ^μ\hat{\sigma}_\mu with −ev^x-e\hat{v}_x. vF=6.3×104v_{\mathrm{F}} = 6.3 \times10^4 m/s is the Fermi velocity of the Weyl cone. With E=EFE = E_{\mathrm{F}}, we find these dimensionless indices ζTi1=5.6×103\zeta^{\mathrm{Ti1}} = 5.6 \times10^3, ζTi2=5.0×103\zeta^{\mathrm{Ti2}} = 5.0 \times10^3, and ζMn=4.2×103\zeta^{\mathrm{Mn}} = 4.2 \times10^3. These values are about 10410^4 times larger than the typical value 0.120.12 of the two-dimensional Rashba system (see Appendix B). In other words, in Ti2_2MnAl, we can reduce the current by the factor of ∼10−4\sim10^{-4} from that in the Rashba systems to induce the same magnitude of spin density.

Effective magnetic field and spin-orbit torque

Above results show that, near EFE_{\mathrm{F}}, the signs of spin densities at Ti1 and Ti2 and that at Mn are opposite, as schematically shown in Figure 2(c). These spin densities exert torques on the magnetic moments at the Ti sites and the Mn sites oppositely to each other. During the dynamics of these magnetic moments, violation of their compensation is predicted to be energetically prohibited [30]. Thus, we treat the compensated ferrimagnetic ordering n^\hat{\mathbf{n}} as the only dynamical variable in this system. We consider the effective magnetic field acting on n^\hat{\mathbf{n}} in the linear response regime [32]:

hSOT(r)=⟨∂H∂n^⟩=∑αJα⟨σα(r)⟩.(5)\mathbf{h}_{\mathrm{SOT}}(\mathbf{r}) = \left\langle\frac{\partial H}{\partial\hat{\mathbf{n}}} \right\rangle= \sum_\alpha J_\alpha\langle\boldsymbol{\sigma}^\alpha(\mathbf{r}) \rangle. \tag*{(5)}

Here, ⟨σα⟩\langle\boldsymbol{\sigma}^\alpha\rangle is given by the formula in Eq. (2). Note that hSOT\mathbf{h}_{\mathrm{SOT}} has a unit of energy density. In later discussions, we will convert it to the unit of tesla and compare it with other systems. This hSOT\mathbf{h}_{\mathrm{SOT}} exerts the SOT on n^\hat{\mathbf{n}}, expressed as TSOT=n^×hSOT\mathbf{T}_{\mathrm{SOT}} = \hat{\mathbf{n}} \times\mathbf{h}_{\mathrm{SOT}}. Figure 3(a) depicts the effective magnetic field for dissipative contribution hSOTdis\mathbf{h}_{\mathrm{SOT}}^{\mathrm{dis}} (blue arrow) from ⟨σα⟩dis\langle\boldsymbol{\sigma}^\alpha\rangle^{\mathrm{dis}} and topological one hSOTtop\mathbf{h}_{\mathrm{SOT}}^{\mathrm{top}} (red arrow) from ⟨σα⟩top\langle\boldsymbol{\sigma}^\alpha\rangle^{\mathrm{top}}.

Effective magnetic fields, response coefficients, density of states, and SOT efficiencies

Figure 3. (a) The effective magnetic fields form dissipative and topological response, acting on the compensated ferrimagnetic ordering n^\hat{n}. (b) The energy dependences of the response coefficients χxxdis\chi_{xx}^{\mathrm{dis}} and χyxtop\chi_{yx}^{\mathrm{top}}. (c) The density of states and one of the energy dispersion around WPs. The inversion symmetry breaking in Ti2MnAl\mathrm{Ti}_2\mathrm{MnAl} causes energy differences between WPs. (d) The energy dependences of efficiencies of SOT θxxdis\theta_{xx}^{\mathrm{dis}} and θyxtop\theta_{yx}^{\mathrm{top}}.

First, we investigate the EE dependences of the effective magnetic field with a fixed relaxation time τ=3ℏ/E0\tau= 3\hbar/E_0. We characterize the effective magnetic field induced by the electric field hSOTdis/top=χ^dis/topE\mathbf{h}_{\mathrm{SOT}}^{\mathrm{dis/top}} = \hat{\chi}^{\mathrm{dis/top}}\mathbf{E} by the response coefficient tensor χ^dis/top\hat{\chi}^{\mathrm{dis/top}}, which is given by

χμxdis/top=∑αJα⟨σμα⟩dis/topEx.(6)\chi_{\mu x}^{\mathrm{dis/top}} = \sum_\alpha J_\alpha\frac{\langle\sigma_\mu^\alpha\rangle^{\mathrm{dis/top}}}{E_x}. \tag*{(6)}

In Figure 3(b), the components of χ^dis\hat{\chi}^{\mathrm{dis}} and χ^top\hat{\chi}^{\mathrm{top}} are shown as functions of EE. Near EFE_{\mathrm{F}}, χ^top\hat{\chi}^{\mathrm{top}} is larger than χ^dis\hat{\chi}^{\mathrm{dis}} and shows a relatively moderate energy dependence. To characterize the efficiency of the SOT per current, we introduce the following quantity,

θμxdis/top=hSOT,μdis/top⟨jx⟩dis=χμxdis/topσxx.(7)\theta_{\mu x}^{\mathrm{dis/top}} = \frac{h_{\mathrm{SOT},\mu}^{\mathrm{dis/top}}}{\langle j_x \rangle^{\mathrm{dis}}} = \frac{\chi_{\mu x}^{\mathrm{dis/top}}}{\sigma_{xx}} . \tag*{(7)}

Here, σxx\sigma_{xx} is a longitudical conductivity and has an minimum around the energy of the WPs, reflecting the structure of DOS shown in Figure 3(c). Figure 3(d) shows the energy dependence of θxxdis\theta_{xx}^{\mathrm{dis}} and θyxtop\theta_{yx}^{\mathrm{top}}. In Figure 3(d), one can find that θyxtop\theta_{yx}^{\mathrm{top}} has a peak near EFE_{\mathrm{F}} and is dominant over θxxdis\theta_{xx}^{\mathrm{dis}}. The peak structure of θyxtop\theta_{yx}^{\mathrm{top}} arises from the combination of the large topological SOT and the small longitudinal conductivity around the energy of the WPs. By measuring hSOTh_{\mathrm{SOT}} in the unit of Tesla (see Appendix C), we find that θyxtop\theta_{yx}^{\mathrm{top}} is of the order of 10−13 Tm2/A10^{-13}\ \mathrm{Tm}^2/\mathrm{A} within the calculated energy range, reaching the value θyxtop=5.09×10−13 Tm2/A\theta_{yx}^{\mathrm{top}}=5.09\times10^{-13}\ \mathrm{Tm}^2/\mathrm{A} at EFE_{\mathrm{F}}. This value is 10-1000 times larger than the SOT efficiency measured in the heterostructures of ferromagnets and heavy metals [7, 8], and also those calculated in the bulk antiferromagnetic Mn2Au\mathrm{Mn}_2\mathrm{Au} [11, 12] and CuMnAs\mathrm{CuMnAs} [13]. Owing to the topological contribution, Ti2MnAl\mathrm{Ti}_2\mathrm{MnAl} can generate an SOT with the efficiency much higher than those previous reports.

Next, we consider the relaxation-time dependences of χxxdis\chi_{xx}^{\mathrm{dis}} and χyxtop\chi_{yx}^{\mathrm{top}} for non-doped and doped cases, which are shown in Figure 4(a) and Figure 4(b). In the non-doped case with the 4/6-filling, EFE_{\mathrm{F}} is located near the energy of the WPs, as shown in the inset of Figure 4(a). From the calculation results in Figure 4(a), we find that χyxtop\chi_{yx}^{\mathrm{top}} is almost independent of τ\tau, which implies the insensitivity to the disorder. Moreover, throughout the entire range of τ\tau we have calculated, χyxtop\chi_{yx}^{\mathrm{top}} remains larger than χxxdis\chi_{xx}^{\mathrm{dis}}. In contrast, χxxdis\chi_{xx}^{\mathrm{dis}} is proportional to τ\tau as seen from Eq. (3). With regard to magnitude of χxxdis\chi_{xx}^{\mathrm{dis}}, the DOS is small enough, and hence χxxdis\chi_{xx}^{\mathrm{dis}} becomes also small in comparison with χyxtop\chi_{yx}^{\mathrm{top}}.

Figure 4. Relaxation time dependences for response coefficients χxxdis\chi_{xx}^{\mathrm{dis}} and χyxtop\chi_{yx}^{\mathrm{top}} when (a) EFE_{\mathrm{F}} is located at the energy of the WPs and (b) EFE_{\mathrm{F}} is away from the energy of the WPs.

Then, we investigate the dependence of χxxdis\chi_{xx}^{\mathrm{dis}} and χyxtop\chi_{yx}^{\mathrm{top}} on τ\tau in doped case where EFE_{\mathrm{F}} is 50 meV above the energy of the WPs, which are shown in Figure 4(b). We find that both the τ\tau-dependence and magnitude of χyxtop\chi_{yx}^{\mathrm{top}} are almost unchanged from those in Figure 4(a). In contrast, χxxdis\chi_{xx}^{\mathrm{dis}} is still proportional to τ\tau, while its magnitude becomes larger. The magnitude of χxxdis\chi_{xx}^{\mathrm{dis}} becomes comparable to that of χyxtop\chi_{yx}^{\mathrm{top}} and exceeds it around τ=3.5×ℏ/E0\tau=3.5\times\hbar/E_0. This is because the DOS becomes larger when EFE_{\mathrm{F}} is shifted away from the energy of the WPs.

From the discussions above, we find that χxxdis\chi_{xx}^{\mathrm{dis}} is always proportional to τ\tau, with its magnitude dependent on EFE_{\mathrm{F}}, which is consistent with the previous study with the minimal Weyl Hamiltonian in continuum [21]. Our main finding in this study is that the topological response χyxtop\chi_{yx}^{\mathrm{top}} is present irrespective of τ\tau, and is almost independent of the choice of EFE_{\mathrm{F}}. Such a behavior of χyxtop\chi_{yx}^{\mathrm{top}} is related to the band geometry characterized by the MBC from the WPs, as we shall discuss in the following.

**Mixed Berry curvature and discussion—**So far we have found the topological SOT χyxtop\chi_{yx}^{\mathrm{top}} is not comparable to the dissipative SOT χyxdis\chi_{yx}^{\mathrm{dis}} in our model of Ti2MnAl\mathrm{Ti}_2\mathrm{MnAl}. To understand the origin of this χyxtop\chi_{yx}^{\mathrm{top}}, we study the MBC [26]. It is a BC extended in the composite parameter space (k,n^)(\boldsymbol{k},\hat{n}) spanned by the momentum k\boldsymbol{k} and the ferrimagnetic ordering n^\hat{n}. The yxyx components of the BC Ωyxkk\Omega_{yx}^{kk} and the MBC Ωyxnk\Omega_{yx}^{nk} for the occupied bands are expressed as

Ωyxkk(k)=2Im⁡∑n≠mfnk⟨nk∣∂kyH∣mk⟩⟨mk∣∂kxH∣nk⟩(Enk−Emk)2,\Omega_{yx}^{kk}(\boldsymbol{k})=2\operatorname{Im}\sum_{n\ne m}\frac{f_{nk}\langle n\boldsymbol{k}|\partial_{k_y}H|m\boldsymbol{k}\rangle\langle m\boldsymbol{k}|\partial_{k_x}H|n\boldsymbol{k}\rangle}{(E_{nk}-E_{mk})^2},
Ωyxnk(k)=2Im⁡∑n≠mfnk⟨nk∣∂n^yH∣mk⟩⟨mk∣∂kxH∣nk⟩(Enk−Emk)2.(8)\Omega_{yx}^{nk}(\boldsymbol{k})=2\operatorname{Im}\sum_{n\ne m}\frac{f_{nk}\langle n\boldsymbol{k}|\partial_{\hat{n}_y}H|m\boldsymbol{k}\rangle\langle m\boldsymbol{k}|\partial_{k_x}H|n\boldsymbol{k}\rangle}{(E_{nk}-E_{mk})^2}. \tag*{(8)}

The topological SOT [Eq. (4)] can be rewritten with the MBC [4, 26, 33],

hSOTtop=eℏ∫kΩ^nk(k) E,(9)h_{\mathrm{SOT}}^{\mathrm{top}} = e\hbar\int_k \widehat{\Omega}^{nk}(\mathbf{k})\,\mathbf{E}, \tag*{(9)}

in the clean limit τ→∞\tau\to\infty.

To understand the source of the MBC contributing to the large SOT found in Ti2_2MnAl, we compare the distributions of the BC −Ωyxkk-\Omega^{kk}_{yx} and the MBC −Ωyxnk-\Omega^{nk}_{yx} on the kxk_x-kyk_y plane in Figure 5. We find that the MBC shows peak structures around the WPs, which are similar to those of the ordinary BC contributing to the intrinsic anomalous Hall effect. We have checked that the distributions of the MBC on the other planes containing the WPs show similar peak structures. These numerical results suggest that, even in the absence of spin-momentum locking, the velocity operator ∂H∂k\frac{\partial H}{\partial k} and the effective magnetic field operator ∂H∂n^\frac{\partial H}{\partial\hat{n}} have similar structures near the WPs, yielding the peaks of both the ordinary and MBCs. As a result, the WPs serve as the origin not only of the intrinsic anomalous Hall effect but also of the topological SOT.

Figure 5. The mapping of the yxy x component of (a) the BC −Ωyxkk-\Omega^{kk}_{yx} and (b) the MBC −Ωyxnk-\Omega^{nk}_{yx} on kz=0.1[1/a]k_z = 0.1[1/a] at EFE_{\mathrm{F}}. In momentum space, the positions of the peaks/valleys of −Ωyxkk/nk-\Omega^{kk/nk}_{yx} correspond to the positions of the WPs.

Conclusion— Our study clarifies that the ferrimagnetic Weyl semimetal Ti2_2MnAl shows distinct behavior in the SOT. The previous studies on the SOT in MWSMs discussed only the dissipative SOT, but not the topological SOT [21, 22]. In contrast, the present study finds that ferrimagnetic Weyl semimetal Ti2_2MnAl shows the giant topological SOT, which is useful for dissipation-free switching of magnetization. The key element for this topological SOT is the MBC from all the occupied states. In Ti2_2MnAl, this topological quantity is notably enhanced around the WPs in momentum space. Moreover, due to the tiny carrier density around the energy of the WPs, the longitudinal current gets suppressed. As a result of these characteristics in Ti2_2MnAl, the efficiency of the effective magnetic field per the injected current reaches around 10–1000 times those estimated previously in noncentrosymmetric antiferromagnets. These results suggest that the design of the MBC in magnetic topological systems, such as MWSMs, is important for the further development of SOT-based spintronics devices.

Acknowledgments— This work was supported by JST CREST, Grant Nos. JPMJCR18T2 and by JSPS KAKENHI, Grant Nos. JP20H01830, JP22K03446, JP22K03538, and JP23K19194.

References

End matter

Appendix A · Details of the tight-binding model — We here explain the details of the hopping term and SOC term in our tight-binding model based on our previous study [30]. The primitive unit cell is the FCC type and the unit vectors are a1=a2(1,1,0)\mathbf{a}_1 = \frac{a}{2}(1,1,0), a2=a2(0,1,1)\mathbf{a}_2 = \frac{a}{2}(0,1,1) and a3=a2(1,0,1)\mathbf{a}_3 = \frac{a}{2}(1,0,1), where aa is a lattice constant and set a=1a = 1 for simplicity. In the model Hamiltonian [see Eq. (1)], the first term includes inter-sublattice hopping with strengths tTi1−Ti2t_{\mathrm{Ti1-Ti2}}, tTi2−Mnt_{\mathrm{Ti2-Mn}}, and tMn−Ti1t_{\mathrm{Mn-Ti1}}, and the intra-sublattice hopping with strength tTi1−Ti1t_{\mathrm{Ti1-Ti1}}, tTi2−Ti2t_{\mathrm{Ti2-Ti2}}, and tMn−Mnt_{\mathrm{Mn-Mn}}. The second term is the on-site energy and ϵα\epsilon_{\alpha} is its strength. The third term is SOC and originates from the intrinsic electric field, which is generated due to the local breaking of inversion symmetry. The dominant asymmetry comes from the imbalance between Ti1 and Al sites. We consider SOC for the Ti2-Ti2 and Mn-Mn hoppings with the same strength λSOC\lambda_{\mathrm{SOC}}, and neglect the SOC for the Ti1-Ti1 for simplicity. d1,2αij\mathbf{d}_{1,2}^{\alpha ij} are the two nearest-neighbor hopping vectors from the site ii to jj of the sublattice α\alpha, which was originally introduced in a Fu-Kane-Mele model [31]. The fourth term represents the exchange coupling between conduction electron spin and magnetic moment. σ\boldsymbol{\sigma} is the Pauli matrix vector representing electron spin. n^\hat{\mathbf{n}} is the unit vector representing the compensated ferrimagnetic ordering. JαJ_{\alpha} is the coupling strength on sublattice α\alpha.

The hopping parameters are set to tTi1−Ti2=420t_{\mathrm{Ti1-Ti2}} = 420 meV, tTi2−Mn=176.4t_{\mathrm{Ti2-Mn}} = 176.4 meV, $t_{\mathrm{Mn-Ti1}} = 504 meV, tTi1−Ti1=21t_{\mathrm{Ti1-Ti1}} = 21 meV, tTi2−Ti2=357t_{\mathrm{Ti2-Ti2}} = 357 meV, and tMn−Mn=−21t_{\mathrm{Mn-Mn}} = -21 meV. On-site energies ϵTi1=ϵTi2=ϵMn=−903\epsilon_{\mathrm{Ti1}} = \epsilon_{\mathrm{Ti2}} = \epsilon_{\mathrm{Mn}} = -903 meV. The strengths of the exchange coupling are JTi1=JTi2=294J_{\mathrm{Ti1}} = J_{\mathrm{Ti2}} = 294 meV, JMn=−714J_{\mathrm{Mn}} = -714 meV. The sign difference corresponds to the ferrimagnetic ordering. The strength of SOC is λSOC=−84\lambda_{\mathrm{SOC}} = -84 meV. The parameters are set so that the energy spectrum including the WPs, the DOS, and the strength of the anomalous Hall conductivity become consistent with first-principles calculations[27].

Spin density and its efficiency of 2D Rashba Model

2D Rashba Model explains the two dimensional electron gas (2DEG) with the Rashba type SOC. Calculating the SOT by this model, only the yx(xy)yx(xy) component of the dissipative contribution is finite. It is given by the following [34, 35],

⟨σy⟩disEx=2m∗eℏ2λτ(10)\frac{\langle\sigma_y \rangle^{\mathrm{dis}}}{E_x} = \frac{2m^*e}{\hbar^2}\lambda\tau \tag*{(10)}

Here, m∗m^* is the effective mass of 2DEG, and λ\lambda is the strength of Rashba type SOC. With the electron density nn, the DC conductivity is also given by ⟨jx⟩disEx=ne2m∗τ\frac{\langle j_x\rangle^{\mathrm{dis}}}{E_x} = \frac{ne^2}{m^*}\tau. The dimensionless index ζ\zeta which explains spin density generated per unit current is given by,

ζ=evF⟨σy⟩dis⟨jx⟩dis=2m∗2λvFℏ2n(11)\zeta= e v_{\mathrm{F}}\frac{\langle\sigma_y \rangle^{\mathrm{dis}}}{\langle j_x\rangle^{\mathrm{dis}}} = \frac{2m^{*2}\lambda v_{\mathrm{F}}}{\hbar^2 n} \tag*{(11)}

Here, as the effective mass m∗m^* and electron density nn of the 2DEG, the typical values of 0.068me0.068m_e and 1015 m−210^{15}\ \mathrm{m}^{-2} are chosen respectively. The strength of SOC is 1.51×1041.51 \times10^4 [36]. vFv_{\mathrm{F}} is the Fermi velocity, given by vF=ℏkF/m∗v_{\mathrm{F}} = \hbar k_{\mathrm{F}}/m^*, where Fermi wavevector kFk_{\mathrm{F}} is assumed 0.1 nm−10.1\ \mathrm{nm}^{-1}. Substituting these typical values into Eq. (11), one can obtain a value ζ=0.12\zeta= 0.12.

Estimation of effective magnetic field of Ti₂MnAl

Here, we explain details of the estimation of the SOT efficiency per current as defined by Eq. (7). First, we estimate the magnitude of the electric field in Ti2_2MnAl when a current is applied. Given a typical experimental current density of j=1010 A/m2j = 10^{10}\ \mathrm{A/m^2} and the longitudinal conductivity of Ti2_2MnAl at EFE_{\mathrm{F}}, σxx=0.24e2aℏ\sigma_{xx} = 0.24\frac{e^2}{a\hbar}, the electric field in Ti2_2MnAl is estimated to be 105 V/m10^5\ \mathrm{V/m}. This value is used in the following estimation of the magnitude of the effective magnetic field.

Referring to Figure 2(b), we approximate that the spin densities from all electronic states below EFE_{\mathrm{F}} on each sublattice have almost the same magnitude: ⟨σTi1⟩top≈⟨σTi2⟩top≈−⟨σMn⟩top≡stop\langle\sigma^{\mathrm{Ti1}}\rangle^{\mathrm{top}} \approx\langle\sigma^{\mathrm{Ti2}}\rangle^{\mathrm{top}} \approx-\langle\sigma^{\mathrm{Mn}}\rangle^{\mathrm{top}} \equiv s^{\mathrm{top}}. With this assumption, the effective magnetic field hSOTh_{\mathrm{SOT}} defined by Eq. (5) is rewritten as,

hSOTtop=∑α=Ti1,Ti2,MnJα⟨σα⟩top=(JTi1⟨σTi1⟩top+JTi2⟨σTi2⟩top+JMn⟨σMn⟩top)≈(JTi1+JTi2−JMn)stop.(12)\begin{aligned} h_{\mathrm{SOT}}^{\mathrm{top}} = \sum_{\alpha=\mathrm{Ti1,Ti2,Mn}} J_\alpha\langle\sigma^\alpha\rangle^{\mathrm{top}} \\ &= \left(J_{\mathrm{Ti1}}\langle\sigma^{\mathrm{Ti1}}\rangle^{\mathrm{top}} + J_{\mathrm{Ti2}}\langle\sigma^{\mathrm{Ti2}}\rangle^{\mathrm{top}} + J_{\mathrm{Mn}}\langle\sigma^{\mathrm{Mn}}\rangle^{\mathrm{top}}\right) \\ &\approx(J_{\mathrm{Ti1}} + J_{\mathrm{Ti2}} - J_{\mathrm{Mn}})s^{\mathrm{top}}. \tag*{(12)} \end{aligned}

We note that hSOTh_{\mathrm{SOT}} defined here has the dimension of energy density, not magnetic field. To convert it to the dimension of magnetic field, tesla, we need to divide it by the magnetic moment for each site,

BSOTα=Jαa32mα⟨σα⟩.(13)B_{\mathrm{SOT}}^{\alpha} = \frac{J_\alpha\frac{a^3}{2}}{m_\alpha}\langle\sigma^\alpha\rangle. \tag*{(13)}

Here, a32\frac{a^3}{2} represents the volume of the conventional unit cell of Ti2_2MnAl, and mαm_\alpha denotes the magnitude of the magnetic moment residing on sublattice α\alpha. This BSOTαB_{\mathrm{SOT}}^\alpha can be expressed with hSOTh_{\mathrm{SOT}} as,

BSOTα,top=Jαa32mα⟨σα⟩top≈1mαJαa32stop=1mαJαJTi1+JTi2−JMna32hSOTtop.(14)\begin{aligned} B_{\mathrm{SOT}}^{\alpha,\mathrm{top}} = \frac{J_\alpha\frac{a^3}{2}}{m_\alpha}\langle\sigma^\alpha\rangle^{\mathrm{top}} \\ &\approx\frac{1}{m_\alpha}J_\alpha\frac{a^3}{2}s^{\mathrm{top}} \\ &= \frac{1}{m_\alpha}\frac{J_\alpha}{J_{\mathrm{Ti1}} + J_{\mathrm{Ti2}} - J_{\mathrm{Mn}}}\frac{a^3}{2}h_{\mathrm{SOT}}^{\mathrm{top}}. \tag*{(14)} \end{aligned}

For instance, the field strength at the Ti1 site reads ∣BSOTTi1,top∣≈1.15×10−3 T|B_{\mathrm{SOT}}^{\mathrm{Ti1,top}}| \approx1.15 \times10^{-3}\ \mathrm{T}, by substituting into Eq. (14) the quantities mTi1=1.44μBm_{\mathrm{Ti1}} = 1.44\mu_{\mathrm{B}} [28], JTi1=JTi2=294J_{\mathrm{Ti1}} = J_{\mathrm{Ti2}} = 294 meV, JMn=−714J_{\mathrm{Mn}} = -714 meV, lattice constant a=6.24×10−10 ma = 6.24 \times10^{-10}\ \mathrm{m} [28], the electric field in Ti2_2MnAl Ex=105 V/mE_x = 10^5\ \mathrm{V/m}, and the response coefficient at EFE_{\mathrm{F}}, χyxtop=1.36×10−2ea2\chi_{yx}^{\mathrm{top}} = 1.36 \times10^{-2}\frac{e}{a^2} obtained from Figure 3(b). As discussed in the main text, the spin densities of Ti1 and Ti2 have signs opposite to that of Mn. Therefore, the effective magnetic fields acting on the compensated ferrimagnetic order n\boldsymbol{n} are additive across the sublattices [see Figure 2(c)]. By calculating BSOTα,topB_{\mathrm{SOT}}^{\alpha,\mathrm{top}} for Ti2 and Mn and summing them, the magnetic field acting on the compensated ferrimagnetic ordering n\boldsymbol{n} is found to be BSOTn≈5.09×10−3 TB_{\mathrm{SOT}}^{n} \approx5.09 \times10^{-3}\ \mathrm{T}, or θyxtop=BSOT,yn/jx=5.09×10−13 T m2/A\theta_{yx}^{\mathrm{top}} = B_{\mathrm{SOT},y}^{n}/j_x = 5.09 \times10^{-13}\ \mathrm{T\,m^2/A} with respect to the longitudinal current jxj_x.

References

  1. [1]M. I. Dyakonov and V. Perel, Current-induced spin orientation of electrons in semiconductors, Phys. Lett. A 35 (1971).DOI
  2. [2]Y. K. Kato, R. C. Myers, A. C. Gossard, and D. D. Awschalom, Observation of the spin Hall effect in semiconductors, Science 306, 1910 (2004).pubmed.ncbi.nlm.nih.gov/15539563
  3. [3]J. Sinova, S. O. Valenzuela, J. Wunderlich, C. H. Back, and T. Jungwirth, Spin Hall effects, Rev. Mod. Phys. 87, 1213 (2015).DOI
  4. [4]A. Manchon, J. Železný, I. M. Miron, T. Jungwirth, J. Sinova, T. Thiaville, K. Garello, and P. Gambardella, Current-induced spin-orbit torques in ferromagnetic and antiferromagnetic systems, Rev. Mod. Phys. 91, 035004 (2019).DOI
  5. [5]I. Mihai Miron, G. Gaudin, S. Auffret, B. Rodmacq, A. Schuhl, S. Pizzini, J. Vogel, and P. Gambardella, Current-driven spin torque induced by the Rashba effect in a ferromagnetic metal layer, Nat. Mater. 9, 230 (2010).DOI
  6. [6]I. M. Miron, K. Garello, G. Gaudin, P.-J. Zermatten, M. V. Costache, S. Auffret, S. Bandiera, B. Rodmacq, A. Schuhl, and P. Gambardella, Perpendicular switching of a single ferromagnetic layer induced by in-plane current injection, Nature 476, 189 (2011).
  7. [7]T. Suzuki, S. Fukami, N. Ishiwata, M. Yamanouchi, S. Ikeda, N. Kasai, and H. Ohno, Current-induced effective field in perpendicularly magnetized ta/cofeb/mgo wire, Applied Physics Letters 98 (2011).DOI
  8. [8]S. Emori, U. Bauer, S.-M. Ahn, E. Martinez, and G. S. Beach, Current-driven dynamics of chiral ferromagnetic domain walls, Nature materials 12, 611 (2013).DOI
  9. [9]M. Jamali, K. Narayanapillai, X. Qiu, L. M. Loong, A. Manchon, and H. Yang, Spin-orbit torques in co/pd multilayer nanowires, Physical review letters 111, 246602 (2013).DOI
  10. [10]K. Garello, I. M. Miron, C. O. Avci, F. Freimuth, Y. Mokrousov, S. Blügel, S. Auffret, O. Boulle, G. Gaudin, and P. Gambardella, Symmetry and magnitude of spin-orbit torques in ferromagnetic heterostructures, Nature nanotechnology 8, 587 (2013).DOI
  11. [11]J. Železný, H. Gao, K. Vyb̌orný, J. Zemen, J. Mašek, A. Manchon, J. Wunderlich, J. Sinova, and T. Jungwirth, Relativistic néel-order fields induced by electrical current in antiferromagnets, Phys. Rev. Lett. 113, 157201 (2014).DOI
  12. [12]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 fer- romagnets, Phys. Rev. B 95, 014403 (2017).
  13. [13]P. Walch, B. Howells, J. Železný, C. Andrews, V. Hills, R. P. Campion, V. Novák, K. Olejník, F. Maccher mezzozi, S. Dhesi, et al., Electrical switching of an antiferromagnet, Science 351, 587 (2016).
  14. [14]A. Chernyshov, M. Overby, X. Liu, J. K. Furdyna, Y. Lyanda-Geller, and L. P. Rokhinson, Evidence for reversible control of magnetization in a ferromagnetic material by means of spin–orbit magnetic field, Nat. Phys. 5, 656 (2009).DOI
  15. [15]L. Liu, O. J. Lee, T. J. Gudmundsen, D. C. Ralph, and R. A. Buhrman, Current-induced switching of perpendicularly magnetized magnetic layers using spin torque from the spin Hall effect, Phys. Rev. Lett. 109, 096602 (2012).pubmed.ncbi.nlm.nih.gov/23002867
  16. [16]L. Liu, C.-F. Pai, Y. Li, H. Tseng, D. Ralph, and R. Buhrman, Spin-torque switching with the giant spin Hall effect of tantalum, Science 336, 555 (2012).DOI
  17. [17]S. Fukami, T. Anekawa, C. Zhang, and H. Ohno, A spin–orbit torque switching scheme with collinear magnetic easy axis and current configuration, Nat. Nanotechnol. 11, 621 (2016).DOI
  18. [18]S. Murakami, Phase transition between the quantum spin hall and insulator phases in 3d: emergence of a topological gapless phase, New Journal of Physics 9, 356 (2007).
  19. [19]X. Wan, A. M. Turner, A. Vishwanath, and S. Y. Savrasov, Topological semimetal and Fermi-arc surface states in the electronic structure of pyrochlore iridates, Phys. Rev. B 83, 205101 (2011).DOI
  20. [20]A. A. Burkov and L. Balents, Weyl semimetal in a topological insulator multilayer, Phys. Rev. Lett. 107, 127205 (2011).DOI
  21. [21]D. Kurebayashi, Y. Araki, and K. Nomura, Microscopic theory of electrically induced spin torques in magnetic weyl semimetals, journal of the physical society of japan 90, 084702 (2021).DOI
  22. [22]A. A. Burkov, M. Smith, A. Hickey, and I. Martin, Current-induced spin accumulation and magnetoresistance in chiral semimetals, Phys. Rev. B 108, 205115 (2023).DOI
  23. [23]N. P. Armitage, E. J. Mele, and A. Vishwanath, Weyl and dirac semimetals in three-dimensional solids, Rev. Mod. Phys. 90, 015001 (2018).DOI
  24. [24]D. Xiao, M.-C. Chang, and Q. Niu, Berry phase effect on electronic properties, Rev. Mod. Phys. 82, 1959 (2010).DOI
  25. [25]H. Kurebayashi, J. Sinova, D. Fang, A. Irvine, T. Skinner, J. Wunderlich, V. Novák, R. Campion, B. Gallagher, E. Vehstedt, et al., An antidamping spin–orbit torque originating from the berry curvature, Nature nanotechnology 9, 211 (2014).DOI
  26. [26]F. Freimuth, S. Blügel, and Y. Mokrousov, Spin-orbit torques in co/pt(111) and mn/w(001) magnetic bilayers from first principles, Phys. Rev. B 90, 174423 (2014).
  27. [27]W. Shi, L. Muechler, K. Manna, Y. Zhang, K. Koepernik, R. Car, J. van den Brink, C. Felser, and Y. Sun, Prediction of a magnetic Weyl semimetal without spin-orbit coupling and strong anomalous Hall effect in the Heusler compensated ferrimagnet Ti₂MnAl, Phys. Rev. B 97, 060406 (2018).
  28. [28]W. Skaftouros, K. Özdoğan, E. Şaşıoğlu, and I. Galanakis, Generalized slater-pauling rule for the inverse heusler compounds, Physical Review B 87, 024420 (2013).DOI
  29. [29]W. Feng, X. Fu, C. Wan, Z. Yuan, X. Han, N. Quang, and S. Cho, Spin gapless semiconductor like ti2mnal film as a new candidate for spintronics application, Phys. Status Solidi RRL 9, 641 (2015).DOI
  30. [30]T. Meguro, A. Ozawa, K. Kobayashi, and K. Nomura, Effective tight-binding model of compensated ferrimagnetic weyl semimetal with spontaneous orbital magnetization, Journal of the Physical Society of Japan 93, 034703 (2024).DOI
  31. [31]L. Fu, C. L. Kane, and E. J. Mele, Topological Insulators in Three Dimensions, Phys. Rev. Lett. 98, 106803 (2007).
  32. [32]Y. Tserkovnyak, A. Brataas, G. E. W. Bauer, and B. I. Halperin, Nonlocal magnetization dynamics in ferromagnetic heterostructures, Rev. Mod. Phys. 77, 1375 (2005).
  33. [33]C. Xiao, B. Xiong, and Q. Niu, Electric driving of magnetization dynamics in a hybrid insulator, Phys. Rev. B 104, 064433 (2021).DOI
  34. [34]V. M. Edelstein, Spin polarization of conduction electrons induced by electric current in two-dimensional asymmetric electron systems, Solid State Communications 73, 233 (1990).DOI
  35. [35]J.-i. Inoue, G. E. Bauer, and L. W. Molenkamp, Diffuse transport and spin accumulation in a rashba two-dimensional electron gas, Physical Review B 67, 033104 (2003).DOI
  36. [36]J. Nitta, T. Akazaki, H. Takayanagi, and T. Enoki, Gate control of spin-orbit interaction in an inverted in₀.₅₃ga₀.₄₇as/in₀.₅₂al₀.₄₈as heterostructure, Phys. Rev. Lett. 78, 1335 (1997).DOI

Paper details

vlnc:W_cwvbmqu8
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

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