betaFeedback

WorksPhysics and Astronomy

Spin Torque Generated by Valley Hall Effect in WSe2

Abstract

Monolayer transition metal dichalcogenides are promising materials for spintronics due to their robust spin-valley locked valence states, enabling efficient charge-to-spin conversion via valley Hall effect with non-equilibrium spins possessing long spin diffusion lengths of hundreds of nanometers. In this work, we show that the injection of a pure valley current, induced by valley Hall effect in a WSe2 monolayer, imparts a spin torque on the magnetization of an overlaid Fe or CoFe in a tunneling structure. The torque efficiency is found to be comparable to that in conventional perpendicular magnetic tunnel junctions and can be further optimized with valley Hall angle in WSe2. The valley nature of the spin torque gives rise to out-of-plane damping-like torques in a current-in-plane configuration, vanishing charge transport perpendicular-to-the-plane as well as torque efficiency tunable through gating.

Introduction

The ability to electrically manipulate the magnetization of a ferromagnetic thin film with perpendicular magnetic anisotropy in an efficient manner is envisioned to enable unprecedented technological advances through the implementation of in-memory computing technologies [1, 2, 3, 4, 5, 6]. Deterministic field-free switching of a perpendicular magnetization requires incident spin currents with a non-vanishing spin component aligned with the perpendicular anisotropy axis. State-of-the-art approaches to the generation of such out-of-plane spins include the charge-to-spin conversion through unconventional spin Hall effects in low-symmetry non-magnets [7, 8, 9] and the injection of spins by additional ferromagnet layers in fully perpendicular magnetic tunnel junctions [1, 10, 11, 12, 13]. Recently, a new subfield of spintronics has emerged [14, 15, 16, 17] which exploits the valley degree of freedom of electrons in two-dimensional materials. Here, we examine the spin torque generated through a valley polarization on an overlaid ferromagnet, and its efficiency in switching a perpendicular magnetization. Addressing this question is key in evaluating the potential of valley physics for spintronics.

In monolayer transition-metal dichalcogenides (TMDs), the spin-orbit interaction induces a spin-valley locking, with K and K′ valleys supporting opposite spin states as required by time-reversal symmetry. Spin-resolved photoemission spectroscopy measurements have revealed valley-dependent out-of-plane spin polarized valence states in TMDs [18, 19, 20, 21, 22]. Subsequently, transport experiments unambiguously demonstrated that a flow of out-of-plane spins in WSe2_2 monolayers can be electrically generated through the valley Hall effect (VHE) [23], a topological Hall response driven by the finite and opposite Berry curvatures in the two valleys [23, 24, 25, 26, 27]. In these reports, an out-of-plane spin/valley polarization of 70% was observed at the edges of p-doped monolayer WSe2_2 due to VHE [27]. A similar electrically generated interfacial spin polarization of 38% was also observed in WSe2_2/graphene heterostructure [23]. Spin-valley locked states in TMDs are also long-lived [28, 29] with large out-of-plane spin diffusion lengths of hundreds of nanometers [25]. These features indicate the potential of spin-valley locking physics to enable unprecedented applications in spintronics.

In this letter, we address the feasibility of utilizing the spin-valley locking of TMDs to induce reversal of an adjacent perpendicular magnetization. By performing transport calculations on a TMD/insulator/ferromagnetic tunnel junction, we show the existence of a sizable spin torque (ST) acting on the magnetization of a Fe or CoFe slab originating solely from a non-equilibrium valley polarization in a WSe2_2 monolayer. We found that the torque efficiency due to spin-valley locking depends sensitively on the WSe2_2 doping levels and on the lattice misalignment with the ferromagnet and is comparable to that in perpendicular magnetic tunnel junction. The fact that the ST arises from the flow of out-of-plane spins constitutes a novel mechanism suitable for switching a perpendicular magnetization. Our findings point toward the utilization of spin-valley locking physics of TMDs in novel efficient quantum spintronic devices.

Theoretical method. Figure 1(a) displays a WSe2_2/insulator/CoFe stack as a prototypical example of the tunneling junction studied in this work. In this system, a non-equilibrium valley population induced in the WSe2_2 portion underneath the CoFe slab gives rise to vertical tunneling valley currents. This situation can be realized by positioning the CoFe slab on one of the side arms of the WSe2_2 crossbar and taking advantage of the VHE to produce a local non-equilibrium valley density [See Fig. 1(b)]. The associated non-equilibrium valley chemical potential difference, δμ=μK−μK′\delta\mu= \mu_{K} - \mu_{K^{\prime}}, is well described by the drift-diffusion approach [30], which renders

δμ=eIσxxθH1+θH2,(1)\delta\mu= \frac{eI}{\sigma_{xx}}\frac{\theta_H}{1+\theta_H^2}, \tag*{(1)}

where II is the electric field-accompanying in-plane charge current and θH\theta_H the valley Hall angle of WSe₂, defined as θH=σH/σxx\theta_H = \sigma_H/\sigma_{xx} where σH\sigma_H and σxx\sigma_{xx} are the valley Hall and longitudinal charge conductivities, respectively. Our drift-diffusion analysis indicates that Eq. (1) is the optimized valley potential induced by VHE in a crossbar geometry. The optimization takes place in the limit where the crossbar arm dimensions, W1W_1, W2W_2 and LL, become smaller than the spin diffusion length in WSe₂, where δμ\delta\mu becomes independent of the crossbar dimensions. In the supplementary material, we present numerical results showing the validity of Eq. (1) and how the optimization comes about in the large spin/valley diffusion length limit [30]. The electric field is assumed to be applied along the xx direction and the longitudinal charge conductivity is σxx=(4e2/h)(ηβ)−1ln⁡[1+exp⁡(βμ2D)]\sigma_{xx} = (4e^2/h)(\eta\beta)^{-1}\ln[1 + \exp(\beta\mu_{2D})], where μ2D\mu_{2D} characterizes the doping levels of WSe₂ and β=1/kBT\beta= 1/k_BT, with kBk_B being the Boltzmann constant and TT the temperature [31]. The valley polarization is defined as (nK−nK′)/(nK+nK′)×100%(n_K - n_{K'})/(n_K + n_{K'}) \times100\% with valley-dependent carrier density nK(K′)=(m∗/βπℏ2)ln⁡[1+exp⁡(β[μ2D±δμ/2])]n_{K(K')} = (m^*/\beta\pi\hbar^2)\ln[1 + \exp(\beta[\mu_{2D} \pm\delta\mu/2])] where m∗m^* is the effective mass of the WSe₂ valence states.

Figure 1(c) reveals that the total electrically generated valley polarization induced at the side arms of the crossbar is a non-monotonic function of θH\theta_H with a maximum at θH≈1\theta_H \approx1 for all doping levels, μ2D\mu_{2D}. Similar non-monotonic behavior was reported in Ref. [32]. Substantial valley polarizations are easily achieved at smaller μ2D\mu_{2D} due to the fact that valley depopulation takes place with greater ease at lower μ2D\mu_{2D} owing to the larger resistivity. The vertical flow of such non-equilibrium valley-polarized electrons can be addressed through a quantum mechanical tunneling approach, as described in the following.

We employ the Bardeen transfer Hamiltonian formalism [33, 34] to describe the tunneling process between the WSe₂ monolayer and ferromagnetic thin film. In Bardeen’s approach, quantum tunneling is treated perturbatively with transition rates being fully described by the electronic ground state of isolated contacts [35, 36]. The electronic states were obtained with the pseudopotential/plane-wave method employed in QUANTUM ESPRESSO [37] and subsequently converted to maximally-localized Wannier function basis with the WANNIER90 package [38]. Figure 2(a) and (Figure 2b) show the spin-resolved band structure and spin density-of-states of a WSe₂ monolayer and a 0.7 nm thick CoFe slab, respectively. The valence states of the WSe₂ monolayer display giant spin splittings of ≈468\approx468 meV, in agreement with previous results [18, 21], with nonequivalent valleys K and K′ hosting opposite out-of-plane spin polarizations. This strong spin-valley locking is better appreciated in the bottom panel of Figure 2(a), where we show the momentum-resolved spin density-of-states for the energy indicated by the horizontal dashed line in the band structure plot. The Fermi level momentum-resolved spin density-of-states of the CoFe slab, shown at bottom panel of Figure 2(b), indicates the presence of electron states throughout the whole Brillouin zone including momenta coinciding the K and K′' valleys. These finite vertical tunneling currents require momentum matched states in the two layers.

Figure 2. The first principles spin-resolved band structure (top panel) and spin density-of-states (bottom panel) of a WSe₂ monolayer and a 0.7 nm thick CoFe slab are shown in panels (a) and (b), respectively. The horizontal dashed line in the band structure plots indicate the energy used in the corresponding spin density-of-states calculations. The valence states of WSe₂ display a valley-dependent out-of-plane spin polarization (SzS_z) that can be extracted through valley Hall effect [23].

The spin-space components of the tunneling charge-current density reads

Jss′=eℏ∫dε∫dkT(2π)2∫dkB(2π)2[f(kT)−f(kB)]Tss′(ε,kT,kB)(2)J^{ss'} = \frac{e}{\hbar} \int d\varepsilon\int\frac{d\mathbf{k}_{T}}{(2\pi)^2} \int\frac{d\mathbf{k}_{B}}{(2\pi)^2} [f(\mathbf{k}_{T}) - f(\mathbf{k}_{B})]T^{ss'}(\varepsilon,\mathbf{k}_{T},\mathbf{k}_{B}) \tag*{(2)}

where f(kT(B))=[1+exp⁡(β[ε−μ(kT(B))])]−1f(\mathbf{k}_{T(B)}) = [1+\exp(\beta[\varepsilon-\mu(\mathbf{k}_{T(B)})])]^{-1} is the Fermi-Dirac distribution of the top (bottom) electrode and Tss′(ε,kT,kB)T^{ss'}(\varepsilon,\mathbf{k}_{T},\mathbf{k}_{B}) are the spin space components of the momentum-dependent tunneling rates. The momentum-dependent chemical potentials are μ(kB)=μ0+δμ(kB)\mu(\mathbf{k}_{B})=\mu_0+\delta\mu(\mathbf{k}_{B}) and μ(kT)=μ0\mu(\mathbf{k}_{T})=\mu_0, where μ0\mu_0 is the equilibrium chemical potential of the whole structure and δμ(kB)\delta\mu(\mathbf{k}_{B}) equals +δμ/2+\delta\mu/2 (−δμ/2-\delta\mu/2) at the K (K′') valley due to the VHE. The function δμ(kB)\delta\mu(\mathbf{k}_{B}) is related to the electronic properties of WSe2_2 and applied in-plane current through Eq. (1).

The tunneling rate components appearing in Eq. (2) are

Tss′(ε,kT,kB)=2π∑n,m∣Mnmss′(kT,kB)∣2δ(ε−εnskT)δ(ε−εms′kB)(3)T^{ss'}(\varepsilon,\mathbf{k}_{T},\mathbf{k}_{B}) = 2\pi\sum_{n,m} \left|M_{nm}^{ss'}(\mathbf{k}_{T},\mathbf{k}_{B})\right|^2 \delta(\varepsilon-\varepsilon_{n s \mathbf{k}_{T}})\delta(\varepsilon-\varepsilon_{m s' \mathbf{k}_{B}}) \tag*{(3)}

with matrix elements expressed in terms of the electronic eigenstates of isolated top and bottom electrodes through

Mnmss′(kT,kB)=ℏ22m∫z0dS(∂ψnskT∂zψms′kB†−ψnskT∂ψms′kB†∂z)(4)M_{nm}^{ss'}(\mathbf{k}_{T},\mathbf{k}_{B}) = \frac{\hbar^2}{2m}\int_{z_0} dS\left(\frac{\partial\psi_{n s \mathbf{k}_{T}}}{\partial z}\psi_{m s' \mathbf{k}_{B}}^\dagger-\psi_{n s \mathbf{k}_{T}}\frac{\partial\psi_{m s' \mathbf{k}_{B}}^\dagger}{\partial z}\right) \tag*{(4)}

where the surface integral is performed halfway inside the barrier at z0z_0 and mm is the effective mass of electrons in the insulating region [30]. We use the embedded Green’s function formalism for computing the matrix element [34]. In Ref. [30] we show how this approach can be used to model tunneling through an amorphous tunnel barrier with incommensurate contacts.

The total tunneling charge-current and out-of-plane polarized spin currents are Jc=J‾+ΔJcos⁡(θ)J_c=\overline{J}+\Delta J\cos(\theta) and Qz=Q‾z+ΔQzcos⁡(θ)Q^z=\overline{Q}^z+\Delta Q^z\cos(\theta), respectively, where θ\theta is the angle between the magnetization of the ferromagnetic slab and the zz axis. We have defined J‾=(J↑↑+J↑↓+J↓↑+J↓↓)/2\overline{J}=(J^{\uparrow\uparrow}+J^{\uparrow\downarrow}+J^{\downarrow\uparrow}+J^{\downarrow\downarrow})/2, ΔJ=(J↑↑−J↑↓+J↓↑−J↓↓)/2\Delta J=(J^{\uparrow\uparrow}-J^{\uparrow\downarrow}+J^{\downarrow\uparrow}-J^{\downarrow\downarrow})/2, Q‾z=(ℏ/4e)[J↑↑+J↑↓−J↓↑−J↓↓]\overline{Q}^z=(\hbar/4e)[J^{\uparrow\uparrow}+J^{\uparrow\downarrow}-J^{\downarrow\uparrow}-J^{\downarrow\downarrow}] and ΔQz=(ℏ/4e)[J↑↑−J↑↓−J↓↑+J↓↓]\Delta Q^z=(\hbar/4e)[J^{\uparrow\uparrow}-J^{\uparrow\downarrow}-J^{\downarrow\uparrow}+J^{\downarrow\downarrow}]. The damping-like torque acting on the magnetization of the ferromagnetic slab is τ=τm^×(m^×z^)\boldsymbol{\tau}=\tau\hat{\mathbf{m}}\times(\hat{\mathbf{m}}\times\hat{\mathbf{z}}) where m^\hat{\mathbf{m}} is the unit vector along the magnetization direction and τ=(Qz(0)−Qz(π))/2=ΔQz\tau=(Q^z(0)-Q^z(\pi))/2=\Delta Q^z [39]. In the following, we discuss the main features of the ST as well as the tunneling charge and spin current originating from a valley polarization in the WSe2_2.

Results and discussions. We initially take the zigzag direction of the WSe2_2 monolayer to be aligned with the [100] direction of the Fe and CoFe slabs, an assumption that will be relaxed later. Figure 3(a) displays the energy dependence of the damping-like torque acting on the magnetization of a Fe (red curve) or CoFe (blue curve) slab. We have taken the in-plane charge current to be I=1μAI=1\mu\mathrm{A} and the ST was evaluated at θ=π/2\theta=\pi/2, since its angular dependence is fully predetermined. As shown, all contributions to the ST take place within a small energy window around the equilibrium chemical potential, where states with energy above and below μ0\mu_0 contribute positively. Hence, a finite integrated ST acts on the magnetization of the ferromagnetic slab originating solely from the non-equilibrium valley-polarized electrons in the WSe2_2 monolayer. Figure 3(a) also indicates a relatively stronger (weaker) ST in the WSe2_2/insulator/Fe(CoFe) tunnel junction, which we will revisit later.

Figure 3. (a) Energy-resolved spin torque acting on the magnetization of an Fe (red) and CoFe (blue) slab. The energies are measured in relation to the equilibrium chemical potential of the system μ0\mu_0. (b) WSe2_2 chemical potential dependence of the spin torque. Here, μ2D\mu_{2D} characterizes the doping levels of WSe2_2 only, i.e., the chemical potential of the ferromagnet is kept constant at μ0\mu_0 in these calculations. The energy-resolved tunneling charge-current passing through the insulating barrier is shown in panel (c) for both systems. (d) Momentum-resolved non-equilibrium chemical potential of the WSe2_2 monolayer.

The ST efficiency in a ferromagnetic element of area AFM≈196 nm×196 nmA_{\mathrm{FM}}\approx196\,\mathrm{nm}\times196\,\mathrm{nm}, defined as ξ=AFM(e/ℏ)τ/I\xi=A_{\mathrm{FM}}(e/\hbar)\tau/I, due to a valley polarization of 70% is found to be ξCoFe≈0.07\xi_{\mathrm{CoFe}}\approx0.07 and ξFe≈0.2\xi_{\mathrm{Fe}}\approx0.2 for CoFe and Fe, respectively, for typical WSe2_2 hole concentrations of p≈2.75×1011 cm−2p\approx2.75\times10^{11}\,\mathrm{cm}^{-2}. For the sake of comparison, the spin transfer torque efficiency in perpendicular magnetic tunnel junctions range in the interval 0.01–0.1 with a slight voltage dependence, as estimated from the typical critical switching current density 10610^6 A/cm2^2 and thermal stability factor of ≈60\approx60 [13, 40, 41]. Therefore, such valley-induced torque efficiency is large enough to excite magnetization dynamics or reversal of a perpendicular ferromagnetic thin film. We also emphasize that the valley-induced torque efficiency can be further optimized with the valley Hall angle and doping concentrations, as discussed next.

Our results also indicate that the ST efficiency depends sensitively on the doping levels of WSe2_2, μ2D\mu_{2D}, as measured from the valence band top. This is shown in Figure 3(b) for the Fe (red symbols) and CoFe (blue symbols) cases. Here, the ST is a monotonically decreasing function of μ2D\mu_{2D} with a stronger dependence at lower doping levels for both ferromagnetic slabs. This behavior is a signature of the valley polarization dependence of the ST and can be understood in the low temperature limit as follows: the Fermi-Dirac distribution can approximated as f(ϵ−μT(B))≈Θ(μT(B)−ϵ)f(\epsilon-\mu_{T(B)}) \approx\Theta(\mu_{T(B)}-\epsilon), where Θ\Theta is the Heaviside function, leading to τ∝δμ\tau\propto\delta\mu. By explicitly writing δμ\delta\mu in terms of μ2D\mu_{2D} through Eq. (1) we find τ/I∝1/μ2D\tau/I \propto1/\mu_{2D}, in agreement with the behavior of Figure 3(b). This result derives exclusively from the valley physics of WSe2_2 monolayer and suggests that the ST efficiency can be modulated through gating, a feature that has no analogue in conventional spintronics.

Figure 3(c) shows that for both ferromagnetic slabs, states with energies above and below μ0\mu_0 contribute oppositely to the vertical total charge-current density JcJ_c. This implies in the scenario of Figure 1(a): The ST acting on the ferromagnet originates from an incoming flow of spin up electrons and an outflow of spin down electrons such that the net tunneling charge current vanishes while a net out-of-plane polarized spin current penetrates the ferromagnet. This is possible by the non-equilibrium valley-dependent chemical potential established in the WSe2_2 monolayer as shown in Figure 3(d), where the chemical potential of electrons with momenta at the vicinity of the K (K′') is slightly lower (higher) than that of the ferromagnet, leading to a vertical outflow (inflow) of K (K′') valley-polarized electrons.

Figure 1. (a) Representative WSe₂/insulator/CoFe tunnel junction with vertical tunneling valley/spin currents induced by an in-plane electric field in the WSe₂ monolayer. (b) The valley Hall effect causes a non-equilibrium valley imbalance in the transverse side arms of a WSe₂ crossbar in response to a longitudinal in-plane electric field. The induced valley voltage is optimized when the dimensions W1W_1, W2W_2 and LL are smaller than the spin/valley diffusion length in WSe₂. (c) Field-induced non-equilibrium valley polarization in WSe₂ as a function of the valley Hall angle at several equilibrium WSe₂ doping levels μ2D\mu_{2D}.

To further explore the nature and behavior of the ST, we study how its magnitude is affected by a lattice misorientation. This is done by rotating the [100] orientation of the ferromagnetic layer by an angle ϕ\phi with respect to the zigzag direction of the WSe2_2 monolayer, which is maintained fixed along the xx direction. The angular dependence of the torque efficiencies acting on the CoFe and Fe slabs are shown in Figure 4(a) and (d), respectively. The results reveal an oscillatory behavior with a 30∘30^\circ period for both cases, where the maximum-to-minimum torque ratio, τmax⁡/τmin⁡\tau_{\max}/\tau_{\min}, is approximately 1.5 and 3 for the CoFe and Fe cases, respectively. The stronger angular dependence in the Fe case is due to more Fermi surface states at momenta coinciding with the K and K′' valleys as compared to CoFe system. This is shown in panels (b) and (e) of Figure 4, where we show the Fermi level momentum-resolved density-of-states of CoFe and Fe, respectively, at ϕ=0∘\phi= 0^\circ. The superposed hexagonal Brillouin zone of the WSe2_2 monolayer highlights the location of K and K′' valleys. As seen in Figure 4(e), the two valleys located at ky=0k_y = 0 Å−1^{-1} coincide with momentum space density-of-states hot spots of Fe, giving rise to a larger tunneling spin current. Such Fermi surface matching is less ideal for the CoFe system. Figure 4(b) and (Figure 4e) show the hotspot misalignments giving rise to the maximum ST efficiency in CoFe and Fe systems, respectively, while Figure 4 (c) and (Figure 4f) display the configurations giving rise to the minimum ST efficiency [30] for more discussion.

Figure 4. Dependence of the spin torque efficiency due to a lattice misalignment. The lattice misalignment is quantified by the angle ϕ\phi between the [100] direction of the ferromagnetic layer relative to the zigzag direction of the WSe2_2 monolayer, which is held fixed along the xx direction. Panel (a) shows the angular dependence of the torque efficiency in the WSe2_2/insulator/CoFe system while panels (b) and (c) show the momentum-resolved density-of-states map of the CoFe at ϕ=0∘\phi= 0^\circ and ϕ=7.5∘\phi= 7.5^\circ, respectively. Panels (d), (e) and (f) display similar results but for the WSe2_2/insulator/Fe system, where the misalignment angles in the two maps are taken to be ϕ=0∘\phi= 0^\circ and ϕ=15∘\phi= 15^\circ. For visualization purposes, we superposed the hexagonal Brillouin zone of WSe2_2, highlighting the valleys.

Conclusions. We have shown that a valley Hall effect-induced non-equilibrium valley polarization in a WSe2\mathrm{WSe}_{2} monolayer results in a spin torque (ST) acting on the magnetization of a ferromagnetic thin film in WSe2\mathrm{WSe}_{2}/insulator/ferromagnetic tunnel junctions. The valley-induced damping-like torque arises from the flow of out-of-plane spin polarized electrons and, therefore, is suitable for exciting magnetization dynamics of thin films with perpendicular magnetic anisotropy, a novel approach that does not require the presence of additional ferromagnetic elements or mirror-broken spin Hall materials. The valley-induced ST was shown to display an efficiency comparable to that in perpendicular magnetic tunnel junctions for typical valley polarizations reported from experiments, with further room for improvement through the valley Hall angle and doping concentrations in the WSe2\mathrm{WSe}_{2} monolayer, a feature that enables one to modulate the ST efficiency through gating. Finally, we demonstrate how the lattice misalignment and different magnetic layers affect the spin torque efficiency.

Acknowledgments. This material is based upon work supported by Intel Corporation from the University Center (Valleytronics) program. We acknowledge the Minnesota supercomputing institute (MSI) for providing the computational resources. We thank Punyashloka Debashis at Intel Corporation for helpful discussions.

References

rent.

References

  1. [1]H. Meng and J.-P. Wang, Appl. Phys. Lett. 88, 172506 (2006).
  2. [2]S. Mangin, D. Ravelosona, J. A. Katine, M. J. Carey, B. D. Terris, and E. E. Fullerton, Nat. Mater. 5, 210 (2006).
  3. [3]M. Gajek, J. J. Nowak, J. Z. Sun, P. L. Trouilloud, E. J. O’Sullivan, D. W. Abraham, M. C. Gaidis, G. Hu, S. Brown, Y. Zhu, et al., Appl. Phys. Lett 100, 132408 (2012).
  4. [4]D. Zhang, M. Bapna, W. Jiang, D. Sousa, Y.-C. Liao, Z. Zhao, Y. Lv, P. Sahu, D. Lyu, A. Naeemi, Tony Low, S. A. Majetich, and J.-P. Wang, Nano Lett. 22, 2, 622 (2022).
  5. [5]A. Sebastian, M. L. Gallo, R. K.-Aljameh and E. Eleftheriou, Nature Nanotechnology 15, 529 (2020).
  6. [6]W. Zhang, B. Gao, J. Tang, P. Yao, S. Yu, M.-F. Chang, H.-J. Yoo, H. Qian and H. Wu, Nature Electronics 3, 371 (2020).
  7. [7]D. MacNeill, G. M. Stiehl, M. H. D. Guimaraes, R. A. Buhrman, J. Park and D. C. Ralph, Nat. Phys. 13, 300 (2017).
  8. [8]Felix Xue, C. Rohmann, J. Li, V. Amin, and P. M. Haney, Phys. Rev. B 102, 014401 (2020).
  9. [9]S. Shi, J. Li, C.-H. Hsu, K. Lee, Y. Wang, L. Yang, J. Wang, Q. Wang, H. Wu, W. Zhang, G. Eda, G. Liang, H. Chang, H. Yang, Advanced Quantum Technologies 4, 2100038 (2021).
  10. [10]H. Suto, T. Nagasawa, T. Kanao, K. Yamada and K. Mizushima, Sci. Rep. 9, 19543 (2019).
  11. [11]D. J. P. de Sousa, P. M. Haney, D. L. Zhang, J. P. Wang, and Tony Low, Phys. Rev. B 101, 081404(R) (2020).
  12. [12]H. Liu, D. Bedau, J. Z. Sun, S. Mangin, E. E. Fullerton, J. A. Katine, A. D. Kent, Journal of Magnetism and Magnetic Materials 358, 233 (2014).
  13. [13]J. Z. Sun, IBM Journal of Research and Development, 50, 81 (2006).
  14. [14]H. Yuan, M. S. Bahramy, K. Morimoto, S. Wu, K. Nomura, B.-J. Yang, H. Shimotani, R. Suzuki, M. Toh, C. Kloo, X. Xu, R. Arita, N. Nagaosa and Y. Iwasa, Nature Physics 9, 563 (2013).
  15. [15]V. T. Renard, B. A. Piot, X. Waintal, G. Fleury, D. Cooper, Y. Niida, D. Terguthra, A. Fujiwara, Y. Hirayama and K. Takashina, Nature Communications 6, 7230 (2015).
  16. [16]L. A. Benítez, J. F. Sierra, W. S.Torres, A. Arrighi, F. Bonell, M. V. Costache and S. O. Valenzuela, Nature Physics 14, 303 (2018).
  17. [17]R. Ahammed and A. De Sarkar, Phys. Rev. B 105, 045426 (2022).
  18. [18]Z. Y. Zhu, Y. C. Cheng, and U. Schwingenschlögl Phys. Rev. B 84, 153402 (2011).
  19. [19]R. Suzuki, M. Sakano, Y. J. Zhang, R. Akashi, D. Morikawa, A. Harasawa, K. Yaji, K. Kuroda, K. Miyamoto, T. Okuda, et al., Nat. Nanotech. 9, 611 (2014).
  20. [20]J. M. Riley, F. Mazzola, M. Dendzik, M. Michiardi, T. Takayama, L. Bawden, C. Granerød, M. Leandersson, T. Balasubramanian, M. Hoesch, et al., Nat. Phys., 10, 835 (2014).
  21. [21]Di Xiao, G.-B. Liu, W. Feng, X. Xu, and W. Yao Phys. Rev. Lett. 108, 196802 (2012).
  22. [22]C. J. Ciccarino, T. Christensen, R. Sundararaman and P. Narang, Nano Lett. 18, 5709 (2018).
  23. [23]T. Y. T. Hung, A. Rustagi, S. Zhang, P. Upadhyaya, Z. Chen, InfoMat. 2, 968 (2020).
  24. [24]K. F. Mak, K. L. McGill, J. Park, P. L. McEuen, Science 344, 1489 (2014).
  25. [25]Z. Wu, B. T. Zhou, X. Cai, .k Cheung, G.-B. Liu, M. Huang, J. Lin, T. Han, L. An, Y. Wang, et al., Nat. Commun. 10, 611 (2019).
  26. [26]J. Lee, K. F. Mak and J. Shan, Nat. Nanotech. 11, 421 (2016).
  27. [27]E. Barré, J. A. C. Incorvia, S. H. Kim, C. J. McClellan, E. Pop, H.-S. P. Wong, and T. F. Heinz Nano Letters 19 (2), 770 (2019).
  28. [28]L. Yang, N. A. Sinitsyn, W. Chen, Y. Yuan, J. Zhang, J. Lou and S. A. Crooker Nature Physics 11, 830 (2015).
  29. [29]J. Kim, C. Jin, B. Chen, H. Cai, T. Zhao, P. Lee, S. Kahn, K. Watanabe, T. Taniguchi, S. Tongay, M. F. Crommie and F. Wang, Science Advances 3, e170051 (2017).
  30. [30]See Supplemental Material at http://xxxx for more details on the theoretical model.
  31. [31]In our calculations, we assume T = 1 K for the temperature, η = 0.1 eV for the spectral broadening and I = 1μA for the in-plane electric field-accompanying charge cur- rent.
  32. [32]H. Li, D. E. Nikonov, C.C. Lin, K. Camsari, Y.-C. Liao, C.-S. Hsu, A. Naeemi and I. A. Young, IEEE Journal on Exploratory Solid-State Computational Devices and Circuits,8, 10 (2022).
  33. [33]J. Bardeen, Phys. Rev. Lett. 6, 57 (1961).
  34. [34]D. Wortmann, H. Ishida, and S. Blügel, Phys. Rev. B 72, 235113 (2005).
  35. [35]M. T. Greenaway, E. E. Vdovin, A. Mishchenko, O. Makarovsky, A. Patané, J. R. Wallbank, Y. Cao, A. V. Kretinin, M. J. Zhu, S. V. Morozov, et al., Nat. Phys. 11, 1057 (2015).
  36. [36]R. M. Feenstra, D. Jena and G. Gu, Journ. Appl. Phys. 111, 043711 (2012).
  37. [37]P. Giannozzi, O. Andreussi, T. Brumme, O. Bunau, M. Buongiorno Nardelli, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, M. Cococcioni et al., J. Phys.: Condens. Matter 29, 465901 (2017).
  38. [38]A. M. Mostofi, J. R. Yates, G. Pizzi, Y.-S. Lee, I. Souza, D. Vanderbilt, and N. Marzari, Comput. Phys. Commun. 185, 2309 (2014).
  39. [39]I. Theodonis, N. Klio̱ssis, A. Kalitsov, M. Chshiev, and W. H. Butler, Phys. Rev. Lett. 97, 237205 (2006).
  40. [40]In these estimations, we assume α = 0.001 is the intrinsic damping of the ferromagnetic slab.
  41. [41]A. Kalitsov, W. Silvestre, M. Chshiev and J. P. Velev, Phys. rev. B 88, 104430 (2013).

Paper details

vlnc:W_csg829eq
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