betaFeedback

WorksPhysics and Astronomy

Magnetic precession induced spin accumulation in collinear antiferromagnets

Abstract

Generating and characterizing uniform and staggered spin polarization in antiferromagnets is one of the key challenges for antiferromagnetic spintronic technology. Here, we perform perturbative theory, group-theoretical symmetry analysis, low energy and ab initio simulations to propose that the magnetic precession near the equilibrium magnetic axis could generate finite uniform and staggered spin polarization at the opposite magnetic sublattices (referring to total magnetic and Néel vector generation) in a single AFM semiconductors. This response does not require the heterojunction setup and could eliminate the lattice mismatch issues at the junction. Through scrutinizing all symmetrically-protected vanishing magnetic moment groups and especially focusing on parity-time (PT ) invariant groups, we identify the symmetry constraints that describe the staggered spin accumulation responses, and disclose their fieldlike and dampinglike characters. This unravels a hidden spin accumulation mode in AFM semiconductors. Furthermore, we simulate such an effect using a perturbative approach and suggest that electric gate field and Floquet light-dressing can effectively manipulate these responses.

Introduction

The modern information storage and memory largely rely on spintronics science and technology, in which the engineering and manipulation of spin polarization is of great importance. Spin pumping has emerged as one of the central mechanisms, serving as a versatile platform to explore nonequilibrium spin generation and providing a dynamic pathway to generate spin angular momentum from magnetization dynamics. Conventional spin pumping usually occurs at the heterojunction between heavy metals and ferromagnets, as the magnetic moments in ferromagnets are easier to manipulate and detect, while antiferromagnetic (AFM) systems were largely overlooked as they exhibit zero net magnetization. Recent years have witnessed a great interest on AFM spintronics owing to their immune to stray field and their ultrafast kinetics [1, 2, 3, 4, 5, 6, 7, 8], yet an effective Néel vector polarization generation (scattered spin generation at the opposite magnetic sublattices) is one of its main obstacles for efficient information read and write technology.

In this work, we present a theoretical analysis and perform ab initio calculations to show that magnetic precession could dynamically pump staggered spin polarization (Néel vector pumping) in a single semiconducting AFM system, without the need of heterojunction setup. The conventional magnetic precession dynamics generating spin polarization has been widely studied in ferromagnetic systems which are usually metallic in their conductivity. Recent efforts have been devoted into electrical toggling (staggered) spin polarizations in AFMs [9, 10, 11, 12, 13]. The contributions usually require a sizable Fermi surface with strong magnetoelectric coupling, while the Fermi sea contributions are, even though symmetrically allowed, much weaker than the Fermi surface contributions [10]. Here, we propose another route to staggered spin polarization accumulation that could arise in intrinsic AFMs from the Fermi sea contributions. By performing perturbative response approach and magnetic group-theoretical investigations, we show that they could host a hidden and staggered spin accumulation under magnetic precession. Our low-energy model simulation results predict that the sublattice-opposite moments in AFMs can produce a staggered spin response even when their net magnetization vanishes. Based upon magnetic group theoretical analysis, we scrutinize all PTPT invariant magnetic point groups (MPGs) and unravel the symmetric operations for the Néel vector. Then we elucidate the symmetry-constrained uniform and staggered spin accumulation under magnetic precession. Note that this response is different from the prior antiferromagnetic spin pumping that typically involve interfacial effects and generate spin currents [14, 15]. Furthermore, this is also different from the magnetoelectric coupling [16] or spin-orbit torque [17, 18] that describe the electric field or current induced effective magnetic field. Here, we focus on the adiabatic torque (T⃗=m^×∂tm^\vec{\mathcal{T}}=\hat{\boldsymbol{m}}\times\partial_t\hat{\boldsymbol{m}}) induced (staggered) spin accumulation in a single phase material through perturbative responses. In order to illustrate the process, we perform first-principles calculations in a bilayer MnBi2_2Te4_4 and show that the sublattice-dependent spin accumulations can also be manipulated under electric and optical fields.

Computational Methods

We perform density functional theory (DFT) calculations in the Vienna ab initio simulation package [19, 20] within the projector augmented-wave method [21]. Generalized gradient approximation (GGA) method within the solid-state Perdew-Burke-Ernzerhof (PBE) form [22, 23] is adopted, and the Brillouin zone integration is performed using the Monkhorst-Pack special kk mesh scheme [24]. The strong correlation in the Mn-dd is treated using the Hubbard UU method with U=5.34 eVU=5.34\,\mathrm{eV}. Convergence criteria of total energy and force component are set to be 1×10−7 eV1\times10^{-7}\,\mathrm{eV} and 1×10−3 eV/A˚1\times10^{-3}\,\mathrm{eV}/\mathring{\mathrm{A}}, respectively. Self-consistent spin-orbit coupling (SOC) is included within all calculations. We fit the Hamiltonian using the maximally localized Wannier functions within the Wannier90 package [25], and then the susceptibility functions are evaluated by the home-built codes. Convergences of the computational approaches have been well-tested.

Results

Uniform and staggered spin accumulation under magnetic precession

To describe the dynamical uniform total and staggered spin accumulation induced by collective magnetic precession, we start from the general total electronic Hamiltonian of an AFM system [26]

H=H0+J∑i[m^i(t)⋅σ](1)H=H_0+J\sum_i\left[\hat{m}^{i}(t)\cdot\boldsymbol{\sigma}\right] \tag*{(1)}

where H0=−ℏ22m∇2+V(r)H_0=-\frac{\hbar^2}{2m}\nabla^2+V(\mathbf{r}) is spin-independent part of Hamiltonian that includes kinetic and Coulomb interactions. Here, m^i(t)\hat{m}^{i}(t) denoting the time and site (i)(i)-dependent magnetization unit vector. One could breakdown its effect into exchange and spin-orbit coupling (SOC) effect, such as Jm^i(t)=(−1)iJexS^i(t)/ℏ+λsoL^i/ℏJ\hat{m}^{i}(t)=(-1)^iJ_{\mathrm{ex}}\hat{S}^{i}(t)/\hbar+\lambda_{\mathrm{so}}\hat{L}^{i}/\hbar, where S^i(t)\hat{S}^{i}(t) and L^i\hat{L}^{i} being the site-dependent spin and orbital operators, respectively. σ\boldsymbol{\sigma} denotes the spin-1/2 Pauli matrix vector. Here, we treat the exchange and SOC effects as a whole since SOC is always present in realistic materials. Then the site-dependent spin torque is [27, 28, 29, 30, 31]

TiS=1iℏ[σ,Hi]=Jm^i×σ.(2)T_i^{S}=\frac{1}{i\hbar}[\sigma,H_i]=J\hat{m}^{i}\times\sigma. \tag*{(2)}

When the magnetic moment is localized near the magnetic sites, such as in 3d transition metal-based magnetic materials, one could show that the time-derivative of Hamiltonian is [32, 33, 34]

∂tH=J∑i[m^i(t)×σ]⋅[m^i(t)×∂tm^i(t)]=∑iTiS⋅Di.(3)\partial_t H=J\sum_i[\hat{m}^{i}(t)\times\sigma]\cdot[\hat{m}^{i}(t)\times\partial_t\hat{m}^{i}(t)]=\sum_i T_i^{S}\cdot\mathcal{D}_i. \tag*{(3)}

This will be used to evaluate the spin accumulation over time. Note that here we assume a slow time variation of the magnetization precession, so that the adiabatic condition is well satisfied. The change of spin is

∂t⟨Sα⟩=∑n,mfn[⟨∂tn∣m⟩⟨m∣Sα∣n⟩+⟨n∣Sα∣m⟩⟨m∣∂tn⟩]=∑n,m[fn⟨n∣Sα∣m⟩⟨m∣∂tn⟩+fm⟨n∣Sα∣m⟩⟨∂tm∣n⟩]=∑n,m(fn−fm)⟨n∣Sα∣m⟩⟨m∣∂tn⟩.(4)\begin{aligned} \partial_t\langle S_\alpha\rangle &=\sum_{n,m}f_n[\langle\partial_t n|m\rangle\langle m|S_\alpha|n\rangle+\langle n|S_\alpha|m\rangle\langle m|\partial_t n\rangle]\\ &=\sum_{n,m}[f_n\langle n|S_\alpha|m\rangle\langle m|\partial_t n\rangle+f_m\langle n|S_\alpha|m\rangle\langle\partial_t m|n\rangle]\\ &=\sum_{n,m}(f_n-f_m)\langle n|S_\alpha|m\rangle\langle m|\partial_t n\rangle. \tag*{(4)} \end{aligned}

The last equation holds since ∂t⟨m∣n⟩=⟨∂tm∣n⟩+⟨m∣∂tn⟩=0\partial_t\langle m|n\rangle=\langle\partial_t m|n\rangle+\langle m|\partial_t n\rangle=0. fnf_n refers to the Fermi-Dirac distribution of electronic structure at band nn. We omit the kk dependence and site index (i)(i) for clarity reason. The time-derivation ∣∂tn⟩|\partial_t n\rangle can be expressed using ∂tH\partial_t H

∣∂tn⟩=∣n⟩⟨n∣∂tn⟩+∑m≠n∣m⟩⟨m∣∂tH∣n⟩En−Em+iη.(5)|\partial_t n\rangle=|n\rangle\langle n|\partial_t n\rangle+\sum_{m\ne n}\frac{|m\rangle\langle m|\partial_t H|n\rangle}{E_n-E_m+i\eta}. \tag*{(5)}

Here, EnE_n refers to the eigenenergy of band nn and η\eta is a universal phenomenological broadening factor to incorporate disorder, impurities, and environmental effects. Therefore, under magnetic precession within the adiabatic strong exchange limit, the spin accumulation is proportional to the Gilbert-damping D⃗\vec{\mathcal{D}} that serves as a generalized force for spin accumulation Figure 1. The linear response perturbation formula is phenomenologically written as

Schematic plot of antiferromagnetic bipartite lattice with collinear equilibrium magnetic configuration

Figure 1. Schematic plot of antiferromagnetic bipartite lattice with collinear equilibrium magnetic configuration along zz, where the magnetic precession of the two sublattices AA and BB lead to the dynamics of both total spin (∂t⟨sm⟩\partial_t\langle s_m\rangle) and staggered spin (∂t⟨sn⟩\partial_t\langle s_n\rangle) accumulation. The Gilbert damping D⃗A\vec{\mathcal{D}}^{A} and D⃗B\vec{\mathcal{D}}^{B} are plotted. Here, sm=(sA+sB)/2s_m=(s^{A}+s^{B})/2 and sn=(sA−sB)/2s_n=(s^{A}-s^{B})/2 are the uniform and staggered spin polarization, respectively. Here, the magnetic precession angles that deviate from the equilibrium magnetic easy axis are exaggerated for clarity reason.

∂t⟨Sα⟩=∑α,βχαβSDβ−⟨Sα⟩τ.(6)\partial_t\langle S_\alpha\rangle=\sum_{\alpha,\beta}\chi_{\alpha\beta}^{S}\mathcal{D}_\beta-\frac{\langle S_\alpha\rangle}{\tau}. \tag*{(6)}

Here, we include the phenomenological dissipation effect using an effective magnetic precession lifetime τ\tau. This indicates that the spin accumulation does not infinitely increase as time and would saturate at δ⟨Sα⟩=τχαβSDβ\delta\langle S_\alpha\rangle=\tau\chi_{\alpha\beta}^{S}\mathcal{D}_\beta. This is similar as the injection current generation (intrinsic ballistic current) under light irradiation [35, 36, 37, 38], but physically distinct from the standard SOT scenario [39] where the torque is driven by an external electric field rather than the magnetic dynamics itself.

In this work, we mainly focus on the susceptibility function tensor χ↔\overleftrightarrow{\chi} perturbatively, and leave the whole magnetic dynamical effect in future works. Therefore, we explicitly emphasize that the theoretical framework in this work is established only in the adiabatic (low-frequency) regime, where the precession frequency ω\omega is assumed to be much smaller than the characteristic electronic energy scales. More strictly, the condition ℏω≪Eg\hbar\omega\ll E_g (ω\omega the magnetic precession frequency) is required to suppress potential interband excitations.

The susceptibility function is a second order tensor indexed by α\alpha and β\beta. Note that they are essentially denoted in the local coordinates, generated along ∂m^i/∂t\partial\hat{\mathbf{m}}^i/\partial t, (m^i×∂tm^i)(\hat{\mathbf{m}}^i\times\partial_t\hat{\mathbf{m}}^i), and m^i\hat{\mathbf{m}}^i. When the precession is not significantly away from the equilibrium axis (differ by a small angle θ∼0−5∘\theta\sim0-5^\circ), we approximately use the Cartesian coordinate of the equilibrium axis to conduct the analysis [34]. In AFMs with two opposite magnetic sublattices (AA and BB), the above expression in Eq. (6) can be projected onto them separately. Under a coherent magnetic precession with m^A≃−m^B\hat{\mathbf{m}}^A\simeq-\hat{\mathbf{m}}^B, one yields D⃗A≃D⃗B\vec{\mathcal{D}}^A\simeq\vec{\mathcal{D}}^B, giving rise to site-dependent (uniform and staggered) spin polarization pumping even in the absence of a net magnetization. This condition corresponds to a phase-locked precession where the sublattice spins rotate collectively and oppositely, producing identical damping fields. This refers to the regime where our perturbation theory is formulated under the strong inter-sublattice exchange limit and within the small precession angle. In this regime, the relevant AFM eigenmode enforces an approximately antiparallel alignment, so that the relative phase between the two sublattices is locked. Deviations from such an idealized situation, such as arising from elliptical precession trajectories, unequal and potentially anisotropic damping, or large cone angles, would break the linear response treatment and are not captured by this analysis. According to the above discussion and similar as in previous works [28, 34], the coefficient tensor χαβS\chi^S_{\alpha\beta} can be evaluated via

χαβS=ℏ∫[dk][∑n,m(fnk−fmk)⟨unk∣σα∣umk⟩⟨umk∣Tβ∣unk⟩Enk−Emk+iη](7)\chi^S_{\alpha\beta}=\hbar\int[dk]\left[\sum_{n,m}(f_{nk}-f_{mk})\frac{\langle u_{nk}|\sigma_\alpha|u_{mk}\rangle\langle u_{mk}|\mathcal{T}_\beta|u_{nk}\rangle}{E_{nk}-E_{mk}+i\eta}\right] \tag*{(7)}

If the torques on two magnetic sublattices are different, one can decompose them into uniform and staggered forms on the two magnetic sublattices, T=TA+TB\mathcal{T}=\mathcal{T}^A+\mathcal{T}^B and T~=TA−TB\widetilde{\mathcal{T}}=\mathcal{T}^A-\mathcal{T}^B, respectively. Here, we focus on the uniform torque (T\mathcal{T}) effect to generate non-staggered and staggered spin polarization in the main text. It dominates when m^A≃−m^B\hat{\mathbf{m}}^A\simeq-\hat{\mathbf{m}}^B (and D⃗A≃D⃗B\vec{\mathcal{D}}^A\simeq\vec{\mathcal{D}}^B, in-phase locked condition under strong exchange field limit). Deviations from this ideal condition, such as elliptical precession or unequal damping, can be approximately described by T~\widetilde{\mathcal{T}}. Its effect on spin accumulations are briefly reported in SM [40]. In principle, it allows us to separate the precession torque effects from the individual sublattices AA and BB. We emphasize that such a separation is valid only within the near-equilibrium perturbative framework, while highly nonlinear dynamical regimes lie beyond the scope of the present linear response analysis.

In Eq. (7), the integral is performed in the DD-dimensional first Brillouin zone, [dk]=dDk(2π)D[dk]=\frac{d^Dk}{(2\pi)^D}, and the value is per unit cell, volume of VucV_{\mathrm{uc}}. We further decompose such a response function into time-reversal T\mathcal{T} even and T\mathcal{T} odd contributions, respectively,

χαβS,even=ℏ∫[dk][∑n,m(fn−fm)ηIm⁡(σnm,αTmn,β)(En−Em)2+η2](8)\chi^{S,\mathrm{even}}_{\alpha\beta}=\hbar\int[dk]\left[\sum_{n,m}(f_n-f_m)\frac{\eta\operatorname{Im}(\sigma_{nm,\alpha}\mathcal{T}_{mn,\beta})}{(E_n-E_m)^2+\eta^2}\right] \tag*{(8)}

and

χαβS,odd=χαβS,sea+χαβS,Drude=ℏ∫[dk][∑n≠m(fn−fm)Re⁡(σnm,αTmn,β)En−Em+∑nσnn,αTnn,β(∂fn∂E)].(9)\begin{aligned} \chi^{S,\mathrm{odd}}_{\alpha\beta}&=\chi^{S,\mathrm{sea}}_{\alpha\beta}+\chi^{S,\mathrm{Drude}}_{\alpha\beta}\\ &=\hbar\int[dk]\left[\sum_{n\ne m}(f_n-f_m)\frac{\operatorname{Re}(\sigma_{nm,\alpha}\mathcal{T}_{mn,\beta})}{E_n-E_m}\right.\\ &\qquad\left.+\sum_n\sigma_{nn,\alpha}\mathcal{T}_{nn,\beta}\left(\frac{\partial f_n}{\partial E}\right)\right]. \tag*{(9)} \end{aligned}

According to the Sokhotski-Plemelj formula, lim⁡η→0η(En−Em)2+η2=πδ(En−Em)∼τ\lim_{\eta\to0}\frac{\eta}{(E_n-E_m)^2+\eta^2}=\pi\delta(E_n-E_m)\sim\tau. Hence, the T\mathcal{T} odd contribution mainly arises when band nn and mm are both near Fermi level (with finite fn−fmf_n-f_m). We also denote it as Fermi surface (sf) contribution (χαβS,sf\chi^{S,\mathrm{sf}}_{\alpha\beta}) in the following discussion. This χαβS,sf∼τ−1\chi^{S,\mathrm{sf}}_{\alpha\beta}\sim\tau^{-1} shows a conductive-like feature, and is potentially tuned through extrinsic environmental factors such as temperature and disorder levels. We note that the presence of dissipative broadening does not invalidate Onsager reciprocity. Instead, the response tensor can be classified according to the time-reversal parities of the generalized forces and observables. The Fermi surface and Fermi sea components may exhibit different symmetry features under magnetic-order reversal (χαβ(m,n)=ϵαϵβχαβ(−m,−n)\chi_{\alpha\beta}(\mathbf{m},\mathbf{n})=\epsilon_\alpha\epsilon_\beta\chi_{\alpha\beta}(-\mathbf{m},-\mathbf{n}) with ϵα,ϵβ=±1\epsilon_\alpha,\epsilon_\beta=\pm1 denoting the time-reversal characters of their corresponding force and observable operators), but both remain compatible with the general Onsager framework.

Furthermore, we find that in the clean limit (η→0\eta\to0), the Fermi surface contributed χαβS,sf\chi^{S,\mathrm{sf}}_{\alpha\beta} is related to a “magnetic” Berry curvature χαβS,sf=−2η∫[dk]Im⁡∑nocc⟨∂unk∂ϕβ∣∂unk∂Jα⟩\chi^{S,\mathrm{sf}}_{\alpha\beta}=-2\eta\int[dk]\operatorname{Im}\sum_n^{\mathrm{occ}}\langle\frac{\partial u_{nk}}{\partial\phi_\beta}|\frac{\partial u_{nk}}{\partial\mathcal{J}_\alpha}\rangle, where we denote the exchange field strength and magnetic unit vector as a whole vector (J⃗=Jm\vec{\mathcal{J}}=\mathcal{J}\mathbf{m}). According to Eq. (1), H=H0+J⃗⋅σH=H_0+\vec{\mathcal{J}}\cdot\boldsymbol{\sigma}— (the site index is omitted for clarity). One can write

∂∣unk⟩∂Jα=∑m≠n∣umk⟩⟨umk∣∂H∂Jα∣unk⟩Enk−Emk+imnk∣unk⟩=∑m≠n∣umk⟩⟨umk∣σα∣unk⟩Enk−Emk+imnk∣unk⟩,(10)\frac{\partial|u_{nk}\rangle}{\partial\mathcal{J}_{\alpha}}=\sum_{m\ne n}\frac{|u_{mk}\rangle\langle u_{mk}|\frac{\partial H}{\partial\mathcal{J}_{\alpha}}|u_{nk}\rangle}{E_{nk}-E_{mk}}+i m_{nk}|u_{nk}\rangle =\sum_{m\ne n}\frac{|u_{mk}\rangle\langle u_{mk}|\sigma_{\alpha}|u_{nk}\rangle}{E_{nk}-E_{mk}}+i m_{nk}|u_{nk}\rangle, \tag*{(10)}

where mnkm_{nk} is a gauge-dependent real number. This measures the variation of Bloch wavefunction over exchange field components. One can artificially rotate m^\hat{m} by a small angle ϕ\phi. For instance, when we assume that the equilibrium magnetic axis is along zz and the rotation along the xx direction, then we have H(ϕx)=eiσxϕxHe−iσxϕxH(\phi_x)=e^{i\sigma_x\phi_x}He^{-i\sigma_x\phi_x}. Hence, the change of Hamiltonian with respect to this angle is

∂H∂ϕx∣ϕx=0=iσxeiσxϕxHe−iσxϕx−ieiσxϕxHσxe−iσxϕx=i[σx,H].(11)\left.\frac{\partial H}{\partial\phi_x}\right|_{\phi_x=0}=i\sigma_x e^{i\sigma_x\phi_x}He^{-i\sigma_x\phi_x}-ie^{i\sigma_x\phi_x}H\sigma_xe^{-i\sigma_x\phi_x} =i[\sigma_x,H]. \tag*{(11)}

Note that Tx=1ℏ[σx,H]=−1ℏ∂H∂ϕxT_x=\frac{1}{\hbar}[\sigma_x,H]=-\frac{1}{\hbar}\frac{\partial H}{\partial\phi_x} near the magnetic equilibrium. Similar performance can be conducted for yy rotation. Hence, in general one can write Tβ=−1ℏ∂H∂ϕβT_\beta=-\frac{1}{\hbar}\frac{\partial H}{\partial\phi_\beta} when we introduce a vector angle ϕ⃗\vec{\phi}. In this case,

∂∣unk⟩ℏ∂ϕβ=∑m≠n∣umk⟩⟨umk∣∂Hℏ∂ϕβ∣unk⟩Enk−Emk+itnk∣unk⟩=−∑m≠n∣umk⟩⟨umk∣Tβ∣unk⟩Enk−Emk+itnk∣unk⟩.(12)\frac{\partial|u_{nk}\rangle}{\hbar\partial\phi_\beta}=\sum_{m\ne n}\frac{|u_{mk}\rangle\langle u_{mk}|\frac{\partial H}{\hbar\partial\phi_\beta}|u_{nk}\rangle}{E_{nk}-E_{mk}}+it_{nk}|u_{nk}\rangle =-\sum_{m\ne n}\frac{|u_{mk}\rangle\langle u_{mk}|T_\beta|u_{nk}\rangle}{E_{nk}-E_{mk}}+it_{nk}|u_{nk}\rangle. \tag*{(12)}

Here, tnkt_{nk} is another gauge-dependent real number. We note that this equation also equals to (m^×∂∣unk⟩∂m^)β\left(\hat{m}\times\frac{\partial|u_{nk}\rangle}{\partial\hat{m}}\right)_\beta. Therefore,

Im⁡⟨∂unk∂Jα|∂unk∂ϕβ⟩=−ℏIm⁡⟨unk∣σα∣umk⟩⟨umk∣Tβ∣unk⟩(Enk−Emk)2.(13)\operatorname{Im}\left\langle\frac{\partial u_{nk}}{\partial\mathcal{J}_{\alpha}}\middle|\frac{\partial u_{nk}}{\partial\phi_\beta}\right\rangle=-\hbar\operatorname{Im}\frac{\langle u_{nk}|\sigma_{\alpha}|u_{mk}\rangle\langle u_{mk}|T_\beta|u_{nk}\rangle}{(E_{nk}-E_{mk})^2}. \tag*{(13)}

The gauge-dependent factors would vanish as they do not contribute to the imaginary parts. Therefore, the clean limit Fermi surface contribution takes the form

χαβS,sf=−2η∫[dk]∑nfnkIm⁡⟨∂unk∂Jα|∂unk∂ϕβ⟩.(14)\chi_{\alpha\beta}^{S,\mathrm{sf}}=-2\eta\int[dk]\sum_n f_{nk}\operatorname{Im}\left\langle\frac{\partial u_{nk}}{\partial\mathcal{J}_{\alpha}}\middle|\frac{\partial u_{nk}}{\partial\phi_\beta}\right\rangle. \tag*{(14)}

This clearly indicates that it measures a mixed (J−ϕ)(\mathcal{J}-\phi) Berry curvature that is defined in the exchange field space, over the magnetic exchange field strength (J\mathcal{J}) and direction (ϕ\phi). The magnetic sublattice dependent responses can be brought back to describe the local mixed Berry curvature.

On the other hand, for the T\mathcal{T} odd susceptibility Eq. (9), we find that the Drude-like intraband Fermi-surface contribution χαβS,Drude=−∫[dk]∑nσnn,αTnn,βδ(En−EF)\chi_{\alpha\beta}^{S,\mathrm{Drude}}=-\int[dk]\sum_n \sigma_{nn,\alpha}T_{nn,\beta}\delta(E_n-E_F) is order of magnitude smaller than the rest part and can be safely omitted. This is because both α\alpha and β\beta should along the lateral directions that is normal to the equilibrium magnetization axis, which is extremely small in collinear AFMs. Hence, this T\mathcal{T} odd contribution is dominated by the Fermi sea contribution, which is intrinsic as χαβS,sea∼τ0\chi_{\alpha\beta}^{S,\mathrm{sea}}\sim\tau^0.

FunctionsP\mathcal{P}T\mathcal{T}PT\mathcal{PT}
χsf\chi^{\mathrm{sf}}✓✓✓
χsea\chi^{\mathrm{sea}}✓✗✗
ζsf\zeta^{\mathrm{sf}}✗✓✗
ζsea\zeta^{\mathrm{sea}}✗✗✓

Table 1. Symmetry constraints for total and staggered spin polarization susceptibility functions in PT\mathcal{PT} AFMs. Here, the ✓ and ✗ represent invariant and sign flip under transformation.

One can breakdown the susceptibility functions Eqs. (8) and (9) into different magnetic sublattice contributions, χαβS,sf/sea=χαβSA,sf/sea+χαβSB,sf/sea\chi_{\alpha\beta}^{S,\mathrm{sf/sea}}=\chi_{\alpha\beta}^{S_A,\mathrm{sf/sea}}+\chi_{\alpha\beta}^{S_B,\mathrm{sf/sea}} for an AFM bipartite lattice. Under strong exchange condition, the uniform (staggered) local spin sm=(sA+sB)/2s_m=(s^A+s^B)/2 (sn=(sA−sB)/2s_n=(s^A-s^B)/2) aligns with the total magnetic moment m^\hat{m} (Néel vector n^\hat{n}). In this regards, we define the spin difference between site-AA and BB that describes the staggered spin accumulation under magnetization precession, which are denoted as ζαβS,sf=χαβSA,sf−χαβSB,sf\zeta_{\alpha\beta}^{S,\mathrm{sf}}=\chi_{\alpha\beta}^{S_A,\mathrm{sf}}-\chi_{\alpha\beta}^{S_B,\mathrm{sf}} and ζαβS,sea=χαβSA,sea−χαβSB,sea\zeta_{\alpha\beta}^{S,\mathrm{sea}}=\chi_{\alpha\beta}^{S_A,\mathrm{sea}}-\chi_{\alpha\beta}^{S_B,\mathrm{sea}}. Both variations of total and staggered spin δ⟨sm⟩=τXD⃗\delta\langle s_m\rangle=\tau\mathcal{X}\vec{\mathcal{D}} and δ⟨sn⟩=τSD⃗\delta\langle s_n\rangle=\tau\mathcal{S}\vec{\mathcal{D}} may be separately measured in AFM materials.

Symmetry arguments

In order to elucidate the symmetry constraints of both χ\chi and ζ\zeta, we perform group-theoretical analysis according to Neumann’s principle. We focus on collinear AFMs with symmetrically constrained zero net magnetization. For the PT\mathcal{PT} AFMs, they conceive Kramers degeneracy in the band dispersion, as PTEn,σ(k)=En,−σ(k)\mathcal{PT}E_{n,\sigma}(\boldsymbol{k})=E_{n,-\sigma}(\boldsymbol{k}) [41]. While the spin operator is P\mathcal{P} invariant (Pσnm=σnm\mathcal{P}\sigma_{nm}=\sigma_{nm}) and reverses under T\mathcal{T} (Tσnm=−σnm\mathcal{T}\sigma_{nm}=-\sigma_{nm}), the Néel vector flips its sign under both P\mathcal{P} and T\mathcal{T}. Hence, one can show that χsf/sea\chi^{\mathrm{sf/sea}} is unchanged under P\mathcal{P}, while ζsf/sea\zeta^{\mathrm{sf/sea}} is forbidden. We tabulate the basic symmetry constraints in Table I for PT\mathcal{PT} AFMs. Note that the staggered spin polarization pumping depends on specific symmetry of the lattice. For example, if altermagnetic (AM, with collinear magnetization [43, 45] [44]) is considered, in general the space inversion does not flip the Néel vector. We discuss their symmetry arguments in SM (including TL\mathcal{T}_L-AFMs with tt representing half lattice translation), and focus on the PT\mathcal{PT} AFMs in the main text.

According to Table I, one observes that the χsea\chi^{\mathrm{sea}} and ζsf\zeta^{\mathrm{sf}} are always symmetrically forbidden. The former indi...cates that for an intrinsic AFM semiconductor without a finite Fermi surface, this magnetization dynamics induced spin polarization pumping does not appear. This is the reason that the spin pumping is usually discussed in ferromagnets rather than AFMs previously. However, when one looks at the staggered spin pumping, it only arises from Fermi sea contribution. This clearly indicates a hidden mode for the spin polarization driven by D⃗\vec{\mathcal{D}}, which requires further investigation for its emergence and tuning. The implication for deterministic Néel vector switching is not determined by the perturbative theory and is beyond the present linear response framework.

Next, we explore the symmetry constraints of both χsf\chi^{\mathrm{sf}} and χsea\chi^{\mathrm{sea}} (the other two are ignored here as they are always forbidden), we carry out magnetic group theory analysis of all the PT\mathcal{PT} invariant systems. Note that as the torque TT includes both the exchange and SOC contributions, in general the SU(2)\mathrm{SU}(2) rotational invariant does not exist. Hence, we use magnetic group theory to conduct the analysis. In order to specifically determine how the spin and Néel polarization are generated under magnetic precession, we take the 2/m′2/m' group as an example. In this case, the equilibrium magnetization can be along the xx, yy, or zz direction. We assume that the two-fold axis is along the zz axis, then it contains four operators, namely, {E,C2z,PT,MzT}\{E,C_{2z},\mathcal{PT},M_z\mathcal{T}\}. As TT applies on the inversion and mirror reflection, one can use ΓT=Au\Gamma_T=A_u to represent it, according to isomorphic group method [45, 46, 47]. The character table of 2/m2/m can be seen in SM. When the equilibrium Néel vector is along zz, the in-plane spin polarization transforms according to Γσx,y=Bg\Gamma_{\sigma_{x,y}}=B_g. Hence, the in-plane torque Tx,yT_{x,y} belongs to ΓTx,y=Bu\Gamma_{T_{x,y}}=B_u that flips its sign under C2zC_{2z} and inversion PP, while keeps invariant under mirror reflection MzM_z. Therefore, the χαβS,sf\chi_{\alpha\beta}^{S,\mathrm{sf}} follows ΓT⊗Γσα⊗ΓTβ=Ag\Gamma_T\otimes\Gamma_{\sigma_\alpha}\otimes\Gamma_{T_\beta}=A_g which is invariant and symmetrically allowed.

The MPGs are used to describe the magnetic order parameter m^\hat{m}, which is PP even and T\mathcal{T} odd. Unfortunately, it is not straightforward to describe the transformation of Néel vector n^\hat{n}, which is both PP and T\mathcal{T} odd. Hence, it is necessary to identify the groups that describe the Néel vector in collinear AFMs, which could then be used to determine the emergence of staggered spin pumping χαβS,sea\chi_{\alpha\beta}^{S,\mathrm{sea}}. For instance, one can divide the operators in 2/m2/m into two types, namely, O1O_1 that maps the same magnetic sublattices and O2O_2 between the different magnetic sublattices. For PT\mathcal{PT} symmetric AFMs, it is evident that E∈{O1}E\in\{O_1\} and P∈{O2}P\in\{O_2\}. We denote the irreducible representation Γn\Gamma_n that changes the sign of {O2}\{O_2\}, yet keeping O2TO_2\mathcal{T} invariant. It depends on the specific Néel vector n^\hat{n} direction. When n^∥z\hat{n}\parallel z, we find Γn=Au\Gamma_n=A_u (see SM). In this case, Néel vector follows the same MPG with magnetization, 2/m′2/m'. Following the same route, we tabulate all the MPGs that describe the Néel vector (denoted as N\mathcal{N}) for all 18 PT\mathcal{PT} invariant MPGs with collinear magnets in Table II. It is evident that the N\mathcal{N}'s are also black-white groups with PT\mathcal{PT} symmetry.

MPG M\mathcal{M}IRN\mathrm{IR}_{\mathcal{N}}N\mathcal{N}χαβsea\chi_{\alpha\beta}^{\mathrm{sea}}
1′1'AuA_u1′1'all allowed
2/m′  (n^∥z)2/m'\;(\hat{n}\parallel z)AuA_u2/m′2/m'all allowed
2/m′  (n^∥x)2/m'\;(\hat{n}\parallel x)BuB_u2′/m2'/mχyzsea,χzysea\chi_{yz}^{\mathrm{sea}},\chi_{zy}^{\mathrm{sea}}
2/m′  (n^∥y)2/m'\;(\hat{n}\parallel y)BuB_u2′/m2'/mχxzsea,χzxsea\chi_{xz}^{\mathrm{sea}},\chi_{zx}^{\mathrm{sea}}
2′/m  (n^∥z)2'/m\;(\hat{n}\parallel z)BuB_u2′/m2'/mall allowed
2′/m  (n^∥x)2'/m\;(\hat{n}\parallel x)AuA_u2/m′2/m'χyzsea,χzysea\chi_{yz}^{\mathrm{sea}},\chi_{zy}^{\mathrm{sea}}
2′/m  (n^∥y)2'/m\;(\hat{n}\parallel y)AuA_u2/m′2/m'χxzsea,χzxsea\chi_{xz}^{\mathrm{sea}},\chi_{zx}^{\mathrm{sea}}
m′mm  (n^∥x)m'mm\;(\hat{n}\parallel x)AuA_um′m′m′m'm'm'χyzsea,χzysea\chi_{yz}^{\mathrm{sea}},\chi_{zy}^{\mathrm{sea}}
m′mm  (n^∥y)m'mm\;(\hat{n}\parallel y)B2uB_{2u}mmm′mmm'χxzsea,χzxsea\chi_{xz}^{\mathrm{sea}},\chi_{zx}^{\mathrm{sea}}
m′mm  (n^∥z)m'mm\;(\hat{n}\parallel z)B3uB_{3u}m′mmm'mmχxysea,χyxsea\chi_{xy}^{\mathrm{sea}},\chi_{yx}^{\mathrm{sea}}
m′m′m′  (n^∥x)m'm'm'\;(\hat{n}\parallel x)B3uB_{3u}m′mmm'mmχyzsea,χzysea\chi_{yz}^{\mathrm{sea}},\chi_{zy}^{\mathrm{sea}}
m′m′m′  (n^∥y)m'm'm'\;(\hat{n}\parallel y)B2uB_{2u}mm′mmm'mχxzsea,χzxsea\chi_{xz}^{\mathrm{sea}},\chi_{zx}^{\mathrm{sea}}
m′m′m′  (n^∥z)m'm'm'\;(\hat{n}\parallel z)B1uB_{1u}mmm′mmm'χxysea,χyxsea\chi_{xy}^{\mathrm{sea}},\chi_{yx}^{\mathrm{sea}}
4/m′  (n^∥z)4/m'\;(\hat{n}\parallel z)AuA_u4/m′4/m'χxxsea=χyysea=−χyxsea\chi_{xx}^{\mathrm{sea}}=\chi_{yy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
4′/m′  (n^∥z)4'/m'\;(\hat{n}\parallel z)BuB_u4′/m′4'/m'χxxsea=χyysea=−χxysea\chi_{xx}^{\mathrm{sea}}=\chi_{yy}^{\mathrm{sea}}=-\chi_{xy}^{\mathrm{sea}}
4/m′mm  (n^∥z)4/m'mm\;(\hat{n}\parallel z)A1uA_{1u}4/m′m′m′4/m'm'm'χxysea=−χyxsea\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
4′/m′mm  (n^∥z)4'/m'mm\;(\hat{n}\parallel z)B2uB_{2u}4′/m′m′m4'/m'm'mχxysea=−χyxsea\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
4/m′m′m′  (n^∥z)4/m'm'm'\;(\hat{n}\parallel z)A2uA_{2u}4/m′mm4/m'mmχxysea=−χyxsea\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
3′  (n^∥z)3'\;(\hat{n}\parallel z)AuA_u3′3'χxxsea=χyysea,χxysea=−χyxsea\chi_{xx}^{\mathrm{sea}}=\chi_{yy}^{\mathrm{sea}},\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
3ˉ′m  (n^∥z)\bar{3}'m\;(\hat{n}\parallel z)A1uA_{1u}3ˉ′m\bar{3}'mχxxsea=χyysea\chi_{xx}^{\mathrm{sea}}=\chi_{yy}^{\mathrm{sea}}
3ˉ′m′  (n^∥z)\bar{3}'m'\;(\hat{n}\parallel z)A2uA_{2u}3ˉ′m\bar{3}'mχxysea=−χyxsea\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
6′/m  (n^∥z)6'/m\;(\hat{n}\parallel z)BuB_u6′/m6'/mχxxsea=χyysea=−χyxsea\chi_{xx}^{\mathrm{sea}}=\chi_{yy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
6′/m′  (n^∥z)6'/m'\;(\hat{n}\parallel z)AuA_u6′/m′6'/m'χxxsea=χyysea=−χxysea\chi_{xx}^{\mathrm{sea}}=\chi_{yy}^{\mathrm{sea}}=-\chi_{xy}^{\mathrm{sea}}
6/m′mm  (n^∥z)6/m'mm\;(\hat{n}\parallel z)A1uA_{1u}6/m′m′m′6/m'm'm'χxysea=−χyxsea\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
6/m′m′m′  (n^∥z)6/m'm'm'\;(\hat{n}\parallel z)B2uB_{2u}6′/m′mm6'/m'mmχxysea=−χyxsea\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}
6/m′m′m′  (n^∥z)6/m'm'm'\;(\hat{n}\parallel z)A2uA_{2u}6′/m′m′m′6'/m'm'm'χxysea=−χyxsea\chi_{xy}^{\mathrm{sea}}=-\chi_{yx}^{\mathrm{sea}}

Table II. Symmetry analysis for all PT\mathcal{PT} invariant magnetic point groups (MPGs). IRN\mathrm{IR}_{\mathcal{N}} is the irreducible representation that swaps two magnetic sublattices. The Néel vector MPGs N\mathcal{N} are listed. Here, we assume the local coordinate as the principal rotation axis lies along zz, and for the m′mmm'mm it is the Mx\mathcal{M}_x that multiplies with T\mathcal{T}. Note in 4′/m′mm′4'/m'mm' and 6′/m′m′m′6'/m'm'm', the xx and yy directions of N\mathcal{N} are different with its parental MPG.

With all the N\mathcal{N} elucidated, one can further analyze the emergence of χαβS,sea\chi_{\alpha\beta}^{S,\mathrm{sea}}. Being a T\mathcal{T}-odd term, it simply follows Γnα⊗ΓTβ\Gamma_{n_\alpha}\otimes\Gamma_{T_\beta}. The results are also listed in Table II, revealing a universal staggered spin accumulation in PT\mathcal{PT} AFMs. If the system contains a high order rotation axis, then there will be only one independent component, indicating an isotropic staggered spin accumulation under adiabatic magnetic precession. We do not include m′3′m'3', 3ˉ′m\bar{3}'m, and m′3′m′m'3'm' in Table II as they do not allow unidirectional magnetic pattern. With the discussions in SM [40], we have scrutinized all 91 MPGs that symmetrically describe both AFM and AM systems with vanishing net magnetic moments.

Low energy model simulation

We perform numerical simulations to elucidate the magnetic precession induced dynamical local spin generation using a low energy k⋅pk\cdot p model. To set the stage, we take a layered AFM model as an exemplary platform [Fig. 2(a)]. The Hamiltonian in 2D Brillouin zone is [48, 49, 50, 51]

H(k)=∑i,j[vF(z^×σ)⋅kτz+Jexiσnτ0+(ΔS+Bk2)σ0τx+λwarp(kx3−3kxky2)σ0τz]δi,j+(12ΔDσ0τ+δj,i+1+12ΔDσ0τ−δj,i−1)ck,i†ck,j.(15)H(\mathbf{k})=\sum_{i,j}\left[v_F(\hat{\mathbf{z}}\times\boldsymbol{\sigma})\cdot\mathbf{k}\tau_z+J_{\mathrm{ex}}^i\sigma_n\tau_0 +\left(\Delta_S+B k^2\right)\sigma_0\tau_x+\lambda_{\mathrm{warp}}\left(k_x^3-3k_xk_y^2\right)\sigma_0\tau_z\right]\delta_{i,j} +\left(\frac{1}{2}\Delta_D\sigma_0\tau_+\delta_{j,i+1}+\frac{1}{2}\Delta_D\sigma_0\tau_-\delta_{j,i-1}\right)c_{k,i}^\dagger c_{k,j}. \tag*{(15)}

Here, vFv_F is the Fermi velocity in the Fermi velocity for both top and bottom surfaces of each layer, indexed as ii (jj). Jexi=(−1)i∣Jex∣J_{\mathrm{ex}}^i=(-1)^i|J_{\mathrm{ex}}| represents an alternative magnetic exchange field that points along the n\mathbf{n}-direction. σ\sigma and τ\tau (τ±=τx±iτy\tau_{\pm}=\tau_x\pm i\tau_y) are Pauli matrices for spin and two surfaces of each layer, respectively (σn=σ⋅n^\sigma_n=\boldsymbol{\sigma}\cdot\hat{\mathbf{n}}). ΔD\Delta_D and ΔS\Delta_S measure the interlayer and intralayer interactions, respectively, and only the nearest neighbor surface interactions are accounted. λwarp\lambda_{\mathrm{warp}} is the hexagonal warping SOC strength. For a bilayer system, the calculated band dispersion is plotted in Figure 2(b) with a band gap of Eg=0.67vFE_g=0.67v_F, in which different color represents expectation value of Néel vector ⟨σzρz⟩\langle\sigma_z\rho_z\rangle with ρz\rho_z representing a Pauli-z matrix for layer pseudospin index. With finite SOC, the Néel vector value is continuously changed, being not a good quantum number.

Low energy k·p model results

Figure 2. Low energy k⋅p\mathbf{k}\cdot\mathbf{p} model results. (a) Schematic plot of a layered AFM model with magnetization in each layer of polar angle of θ\theta. ΔS\Delta_S and ΔD\Delta_D represent interactions within each layer (top and bottom sides) and between adjacent layers. In the simulation, we assume θ=0\theta=0 and adopt typical parameters with ∣Jex∣=1.5vF|J_{\mathrm{ex}}|=1.5v_F, ΔD=0.1vF\Delta_D=0.1v_F, ΔS=0.4vF\Delta_S=0.4v_F, B=0B=0, and λwarp=0.1vF\lambda_{\mathrm{warp}}=0.1v_F. (b) Band dispersion of a minimum bilayer AFM model with color-coding the Néel expectation value of each state. (c) and (d) are calculated χS,sf\chi^{S,\mathrm{sf}} and ςS,sea\varsigma^{S,\mathrm{sea}} tensor components with respect to chemical potential μ\mu, respectively. (e) shows field-like ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} with respect to magnetization angle θ\theta (in the xzxz-plane) while keeping other parameters unchanged. (f) Gate voltage induced variation of band gap EgE_g and χFLS,sea\chi^{S,\mathrm{sea}}_{\mathrm{FL}} at μ=−1.3vF\mu=-1.3v_F.

Our simulated interband and intraband uniform and staggered spin accumulations are in fully agreement with previous symmetry constraints. In detail, one sees that both χS,sea\chi^{S,\mathrm{sea}} and ςS,sf\varsigma^{S,\mathrm{sf}} totally vanish, and the χS,sf\chi^{S,\mathrm{sf}} and ςS,sea\varsigma^{S,\mathrm{sea}} with respect to chemical potential μ\mu are depicted in Figure 2(c) and Figure 2(d). The diagonal terms of χS,sf\chi^{S,\mathrm{sf}} becomes nonzero when the chemical potential crossing either valence or conduction bands, while the off-diagonal terms remain to be vanishing throughout the energy window. As for the ςS,sea\varsigma^{S,\mathrm{sea}}, its diagonal components are completely zero, and the off-diagonal terms exhibit antisymmetric behaviors (ςxyS,sea=−ςyxS,sea\varsigma^{S,\mathrm{sea}}_{xy}=-\varsigma^{S,\mathrm{sea}}_{yx}). These facts are consistent with the symmetry constraints. In the following, we refer to the diagonal and off-diagonal terms the dampinglike (DL, ∝m^×(m^×σ)\propto\hat{\mathbf{m}}\times(\hat{\mathbf{m}}\times\boldsymbol{\sigma})) and fieldlike (FL, ∝m^×σ\propto\hat{\mathbf{m}}\times\boldsymbol{\sigma}) responses, respectively [34].

We find that the equilibrium Néel direction n^\hat{\mathbf{n}} would affect the responses. In Figure 2(e), we rotate n^\hat{\mathbf{n}} in the xzxz-plane (keeping the azimuthal angle φn=0\varphi_n=0) and plot the ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} variation. Other φn\varphi_n values do not significantly affect the results. We observe that the ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} reduces from positive values to negative values when the equilibrium θn\theta_n rotates from 0 (out-of-plane) to π/2\pi/2 (in-plane), and then increases back at π\pi.

We further propose that one can apply a gate voltage to break PT\mathcal{PT} and trigger the emergence of χS,sea\chi^{S,\mathrm{sea}} and ςS,sf\varsigma^{S,\mathrm{sf}}. Here, we introduce HV=Vgσ0τ0ρzH_V=V_g\sigma_0\tau_0\rho_z that generates an potential gradient at different magnetic sublattices. Figure 2(f) shows the variation of band gap EgE_g and χFLS,sea\chi^{S,\mathrm{sea}}_{\mathrm{FL}}. Under such an electric Stark effect, the former gradually reduces, while the latter increases monotonously. These provide efficient approaches to modulate the spin and Néel dynamics under magnetic precession.

We would like to remark that the above discussions are not constrained in layered-AFMs. One can refer to results of other type AFMs in SM [40]. This is different from usual electric controlling magnetic responses in multiferroic systems [52, 53, 54, 55, 56, 57, 58, 59], as here we only adopt conventional AFMs. Thus, breaking PT\mathcal{PT} symmetry restores previously forbidden uniform and staggered channels, enabling tunable dynamical responses.

Ab initio calculations on realistic even layer MnBi2_2Te4_4

Armed with low energy model results, we perform ab initio calculations for a realistic material example, MnBi2_2Te4_4 with even septuple layers. We consider the magnetic axis along the zz direction, giving a MPG of 3ˉ′m′\bar{3}^{\prime}m^{\prime}. With the strong correlation of Mn-d orbitals treated by the Hubbard UU corrections, the calculated bandgap of bilayer MnBi2_2Te4_4 is Eg=0.076E_g = 0.076 eV. Figure 3 shows our simulation results. The dampinglike total spin generation χDLS,sf\chi^{S,\mathrm{sf}}_{\mathrm{DL}} reaches −63.9ℏ/nm2-63.9 \hbar/\mathrm{nm}^2 when μ=−1\mu=-1 eV below the intrinsic Fermi level, while the fieldlike χFLS,sf\chi^{S,\mathrm{sf}}_{\mathrm{FL}} is completely silent according to symmetry constraints [Figure 3(a)]. The susceptibility function is measured in unit of ℏ/nm2\hbar/\mathrm{nm}^2, indicating a steady-state angular momentum accumulation rate per 2D area. From Table II, its Néel vector follows N=3ˉ′m\mathcal{N}=\bar{3}^{\prime}m, giving only one independent fieldlike ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} (=ςxyS,sea=−ςyxS,sea=\varsigma^{S,\mathrm{sea}}_{xy}=-\varsigma^{S,\mathrm{sea}}_{yx}). This Fermi sea contribution is notably large compared to the Fermi sea part in conventional spin-orbit torque setups, suggesting it is potentially within reach of experimental detection conditions. As shown in Figure 3(b), in the intrinsic bandgap one has ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} reaches 113.2ℏ/nm2113.2\hbar/\mathrm{nm}^2. Its BZ-resolved contribution is plotted in the inset, giving a clear hexagonal pattern and the main contributions from the states with large interlayer mixing near MM point. We provide an order of magnitude estimate of the Néel vector accumulation. Taking a typical ferromagnetic resonance precession frequency of ω0/2π=10\omega_0/2\pi=10 GHz, and a small cone angle of θ∼1∘\theta\sim1^\circ, the adiabatic damping vector magnitude is then ∣D∣=ω0sin⁡θ≃1×109|\mathcal{D}|=\omega_0\sin\theta\simeq1\times10^9 rad/s. This yields a steady-state sublattice δ⟨sn⟩\delta\langle s_n\rangle on the order of 10−3ℏ/nm210^{-3}\hbar/\mathrm{nm}^2 (taking a typical relaxation lifetime of 10 fs). We emphasize that this is only an order estimate, but it could be within the sensitivity range of current magneto-optical or spin-resolved probes.

Results for a MnBi2Te4 bilayer with equilibrium out-of-plane magnetization

Figure 3. Results for a MnBi2_2Te4_4 bilayer with equilibrium out-of-plane magnetization. (a) χS,sf\chi^{S,\mathrm{sf}} and (b) ςS,sea\varsigma^{S,\mathrm{sea}} for different chemical potential. The BZ-resolved ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} at the Fermi level is shown as inset plot.

To further illustrate the universality and robustness of this Néel accumulation in such an intrinsic semiconductors, we focus on the in-gap responses under external fields and disorder scattering on the bilayer MnBi2_2Te4_4. We first examine the dependence of the response on the scattering rate η\eta, which semiempirically incorporates the disorder level and temperature effect. Figure 4(a) shows that the in-gap ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} remains constant over a wide range of η\eta (from 0.01 to 0.2 eV). It is noteworthy that this robustness persists even when the scattering energy scale exceeds the intrinsic band gap (Eg=0.076E_g=0.076 eV). This persistent stability against phenomenological broadening further corroborates the intrinsic deep sea nature of the generated staggered spin polarization.

Three panels showing field effects on in-gap spin accumulation and bandgap

Figure 4. Field effect to modulate the SFLS,sea\mathcal{S}_{\mathrm{FL}}^{S,\mathrm{sea}} in bilayer MnBi2_2Te4_4. (a) The dependence of the in-gap SFLS,sea\mathcal{S}_{\mathrm{FL}}^{S,\mathrm{sea}} on the phenomenological scattering rate η\eta. (b) Gate electric field induced variation of bandgap EgE_g (red squares, right axis) and in-gap SFLS,sea\mathcal{S}_{\mathrm{FL}}^{S,\mathrm{sea}} (black circles, left axis). (c) Floquet light-dressed variation of EgE_g and in-gap SFLS,sea\mathcal{S}_{\mathrm{FL}}^{S,\mathrm{sea}} as a function of light field intensity eA0/ℏeA_0/\hbar under a left-handed circularly polarized light with pump frequency ω=5 eV\omega= 5\,\mathrm{eV}.

Secondly, similar as in the model calculations, we consider a gate electric field effect VgV_g along the out-of-plane direction, added through the saw-tooth electric potential method. As illustrated in Figure 4(b), the bandgap EgE_g decreases drastically as VgV_g gradually applied. Interestingly, the fieldlike staggered spin generation ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} within the bandgap also does not change too much as the voltage increases. It maintains a sizable value with only marginal fluctuations even when the gap nearly closes. This distinction implies that while the electric field effectively modulates the band-edge spectrum, the induced Néel polarization ς\varsigma arises primarily from the global topology of the deep Fermi sea. Consequently, it is topologically protected and insensitive to the local bandgap modulation, unlike the surface-sensitive or symmetry-breaking dependent responses.

Furthermore, we propose that a light field with ultrafast periodic time dressing could serve as an effective knob to manipulate the band dispersion and staggered spin generations. Here, we simulate such an effect via the Floquet band engineering [60, 61, 62, 63, 64, 65] under a left-handed circularly polarized light field, A(t)=A0(cos⁡ωt,sin⁡ωt,0)A(t)=A_0(\cos\omega t,\sin\omega t,0). The pump frequency ω=5\omega=5 eV is chosen to be significantly higher than the characteristic time scales of the relevant band energy and magnetic precession, allowing the application of the van Vleck-Magnus expansion to approximate the light dressed Hamiltonian,

HF=H0+∑n≥1[H−n,Hn]nℏω+O(1ω2).H^{F}=H_0+\sum_{n\geq1}\frac{[H_{-n},H_n]}{n\hbar\omega}+\mathcal{O}\left(\frac{1}{\omega^2}\right).

Here, Hl=12π∫02π/ωH(k+eA(t)ℏ)eilωt dtH_l=\frac{1}{2\pi}\int_0^{2\pi/\omega}H\left(\mathbf{k}+\frac{eA(t)}{\hbar}\right)e^{il\omega t}\,dt denotes the nn-th Fourier component of the time-periodic Hamiltonian, and we retain terms up to l=2l=2 to ensure convergence. As shown in Figure 4(c), as the light intensity enhances, the bandgap of bilayer MnBi2_2Te4_4 gradually reduces and closes at eA0/ℏ≃0.4eA_0/\hbar\simeq0.4 Å−1^{-1}. Further increasing light intensity would drive a topological phase transition [65], yet is challenging to be achieved experimentally. Concomitantly, the intrinsic staggered spin pumping ςFLS,sea\varsigma^{S,\mathrm{sea}}_{\mathrm{FL}} within the bandgap shows a monotonic increase. This indicates that light field could serve as an effective scheme to tune the staggered spin polarization pumping in intrinsic semiconducting AFMs.

Discussion and conclusion

In addition to spin pumping, recent years have witnessed a lot of researches on the orbital accumulation and generation, dubbed orbital pumping (and orbital transport) in orbitronic field [66, 67, 68, 69, 70, 71, 72]. This site-dependent hidden spin pumping can be extended to orbital contributions, especially when localized orbitals are within consideration (such as van der Waals layered structures). Because the torque operator couples equally to spin and orbital angular momentum, the same formalism yields a staggered orbital pumping tensor upon replacing local σ^\hat{\sigma} by L^\hat{L}.

In this work we are focusing on the perturbative response theory and symmetry constraint analysis. Here, the magnetic torque is considered as a source to generate staggered spin polarization, which does not correspond to the Landau-Lifshitz-Gilbert equation based magnetic dynamics for Néel control. Experimentally, the dynamically generated staggered spin accumulation may be detected by measuring spin changes within the AFM sample, or by attaching a heavy metal layer, where the resulting spin signal could be converted into a measurable charge current or voltage through spin-charge conversion processes such as inversion spin Hall and inverse spin Edelstein effect. While the present theoretical framework and calculations establish the symmetry conditions and microscopic mechanisms for the proposed pumping response, a quantitative estimate of experimentally detectable signal magnitudes would additionally require realistic modeling of dynamical precession amplitudes, damping, and device-specific conversion efficiencies. Such quantitative analysis is beyond the scope of the present perturbative treatment and will be investigated in future work.

In summary, we propose the hidden staggered spin accumulation under magnetic precession in AFM systems. We show how the Fermi surface and Fermi sea contributions are symmetrically constrained. We develop a magnetic group-theoretical method to elucidate the MPG that n^\hat{n} follows. The Fermi surface contributions arises from multiple bands, and scales with a mixed Berry curvature in the magnetization space. Furthermore, our theory are illustrated using low energy k⋅p\boldsymbol{k}\cdot\boldsymbol{p} model and ab initio simulations for PT\mathcal{PT}-symmetric AFMs. The generation of Néel polarizations can be effectively controlled by light field, yet is robust against extrinsic scattering factors.

Acknowledgments

Acknowledgments. J.Z. acknowledges useful discussions with C. Niu. This work is supported by the National Natural Science Foundation of China (NSFC) under Grant No. 12374065.

Data availability. The data that support the findings of this article are not publicly available yet. The data are available from the authors upon reasonable request.

References

  1. [1]P. Němec, M. Fiebig, T. Kampfrath, and A. V. Kimel, “Antiferromagnetic opto-spintronics,” Nat. Phys. 14, 229 (2018).DOI
  2. [2]V. Baltz, A. Manchon, M. Tsoi, T. Moriyama, T. Ono, and Y. Tserkovnyak, “Antiferromagnetic spintronics,” Rev. Mod. Phys. 90, 015005 (2018).DOI
  3. [3]T. Jungwirth, X. Marti, P. Wadley, and J. Wunderlich, “Antiferromagnetic spintronics,” Nat. Nanotech. 11, 231 (2016).DOI
  4. [4]J. Han, R. Cheng, L. Liu, H. Ohno, and S. Fukami, “Coherent antiferromagnetic spintronics,” Nat. Mater. 22, 684 (2023).DOI
  5. [5]R. Cheng, J. Xiao, Q. Niu, and A. Brataas, “Spin pump- ing and spin-transfer torques in antiferromagnets,” Phys. Rev. Lett. 113, 057601 (2014).
  6. [6]H. Wang, Y. Xiao, M. Guo, E. Lee-Wong, G. Q. Yan, R. Cheng, and C. R. Du, “Spin pumping from an easy-plane antiferromagnet enhanced by dzyaloshinskii–moriya interaction,” Phys. Rev. Lett. 127, 117202 (2021).DOI
  7. [7]Ø. Johansen and A. Brataas, “Spin pumping and inverse spin hall voltages from dynamical antiferromagnets,” Phys. Rev. B 95, 220408 (2017).DOI
  8. [8]R. Islam, S. Ahmad, and F. Xue, “Spin-orbit torque control of topology in intrinsic antiferromagnetic insulators,” (2025), arXiv:2509.01810 [cond-mat.mes-hall].
  9. [9]L. Han, X. Fu, R. Peng, X. Cheng, J. Dai, L. Liu, Y. Li, Y. Zhang, W. Zhu, H. Bai, et al., “Electrical 180° switching of néel vector in spin-splitting antiferromagnet,” Sci. Adv. 10, eadn0479 (2024).
  10. [10]Y. Chen, X. Liu, H.-Z. Lu, and X. C. Xie, “Electrical switching of altermagnetism,” Phys. Rev. Lett. 135, 016701 (2025).DOI
  11. [11]V. Lopez-Dominguez, H. Almasi, and P. K. Amiri, “Pi-cosecond electric-field-induced switching of antiferromagnets,” Phys. Rev. Appl. 11, 024019 (2019).
  12. [12]I. J. Park, S. Kwon, and R. K. Lake, “Effects of filling, strain, and electric field on the néel vector in antiferromagnetic crsb,” Phys. Rev. B 102, 224426 (2020).DOI
  13. [13]M. Ezawa, “Out-of-plane edelstein effects: Electric field induced magnetization in p-wave magnets,” Phys. Rev. B 111, L161301 (2025).DOI
  14. [14]R. Cheng, J. Xiao, Q. Niu, and A. Brataas, “Spin pumping and spin-transfer torques in antiferromagnets,” Phys. Rev. Lett. 113, 057601 (2014).DOI
  15. [15]P. Vaidya, S. Morley, J. Tol, Y. Liu, R. Cheng, A. Brataas, D. Lederman, and E. Del Barco, “Subterahertz spin pumping from an insulating antiferromagnet,” Science 368, 160–165 (2020).DOI
  16. [16]T. Zhu, H. Wang, H. Zhang, and D. Xing, “Tunable dynamical magnetoelectric effect in antiferromagnetic topological insulator mnbi2te4 films,” npj Comput. Mater. 7 (2021).DOI
  17. [17]J. Zhou, X. Wang, Y. Liu, J. Yu, H. Fu, L. Liu, S. Chen, J. Deng, W. Lin, X. Shu, et al., “Large spin-orbit torque efficiency enhanced by magnetic structure of collinear antiferromagnet irmn,” Sci. Adv. 5, eaau6696 (2019).DOI
  18. [18]E. Cogulu, H. Zhang, N. N. Statuto, Y. Cheng, F. Yang, R. Cheng, and A. D. Kent, “Quantifying spin-orbit torques in antiferromagnet-heavy-metal heterostructures,” Phys. Rev. Lett. 128, 247204 (2022).DOI
  19. [19]G. Kresse and J. Hafner, “Ab initio molecular dynamics for liquid metals,” Phys. Rev. B 47, 558–561 (1993).pubmed.ncbi.nlm.nih.gov/10004490
  20. [20]G. Kresse and J. Furthmüller, “Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set,” Phys. Rev. B 54, 11169–11186 (1996).pubmed.ncbi.nlm.nih.gov/9984901
  21. [21]P. E. Blöchl, “Projector augmented-wave method,” Phys. Rev. B 50, 17953–17979 (1994).pubmed.ncbi.nlm.nih.gov/9976227
  22. [22]J. P. Perdew, K. Burke, and M. Ernzerhof, “Generalized gradient approximation made simple,” Phys. Rev. Lett. 77, 3865–3868 (1996).pubmed.ncbi.nlm.nih.gov/10062328
  23. [23]J. P. Perdew, A. Ruzsinszky, G. I. Csonka, O. A. Vydrov, G. E. Scuseria, L. A. Constantin, X. Zhou, and K. Burke, “Restoring the density-gradient expansion for exchange in solids and surfaces,” Phys. Rev. Lett. 100, 136406 (2008).DOI
  24. [24]H. J. Monkhorst and J. D. Pack, “Special points for brillouin-zone integrations,” Phys. Rev. B 13, 5188–5192 (1976).
  25. [25]A. A. Mostofi, J. R. Yates, G. Pizzi, Y.-S. Lee, I. Souza, D. Vanderbilt, and N. Marzari, “An updated version of wannier90: A tool for obtaining maximally-localised wannier functions,” Comput. Phys. Commun. 185, 2309–2310 (2014).
  26. [26]P. M. Haney and M. D. Stiles, “Current-induced torques in the presence of spin-orbit coupling,” Phys. Rev. Lett. 105, 126602 (2010).DOI
  27. [27]F. Mahfouzi and N. Kioussis, “First-principles study of the angular dependence of the spin-orbit torque in pt/co and pd/co bilayers,” Phys. Rev. B 97, 224426 (2018).DOI
  28. [28]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).DOI
  29. [29]S. Wimmer, K. Chadova, M. Seemann, D. Ködderitzsch, and H. Ebert, “Fully relativistic description of spin-orbit torques by means of linear response theory,” Phys. Rev. B 94, 054415 (2016).DOI
  30. [30]K. D. Belaschenko, A. A. Kovalev, and M. van Schilfgaarde, “First-principles calculation of spin-orbit torque in a co/pt bilayer,” Phys. Rev. Mater. 3, 011401 (2019).DOI
  31. [31]G. Géranton, F. Freimuth, S. Blügel, and Y. Mokrousov, “Spin-orbit torques in l10− FePt/Pt thin films driven by electrical and thermal currents,” Phys. Rev. B 91, 014417 (2015).DOI
  32. [32]E. Šimánek and B. Heinrich, “Gilbert damping in magnetic multilayers,” Phys. Rev. B 67, 144418 (2003).DOI
  33. [33]D. L. Mills, “Ferromagnetic resonance relaxation in ultrathin metal films: The role of the conduction electrons,” Phys. Rev. B 68, 014419 (2003).DOI
  34. [34]D. Go, K. Ando, A. Pezo, S. Blügel, A. Manchon, and Y. Mokrousov, “Orbital pumping by magnetization dynamics in ferromagnets,” Phys. Rev. B 111, L140409 (2025).arxiv.org/abs/2309.14817
  35. [35]B. I. Sturman, “Ballistic and shift currents in the bulk photovoltaic effect theory,” Phys.Usp. 63, 407–411 (2020).DOI
  36. [36]R. von Baltz and W. Krout, “Theory of the bulk photovoltaic effect in pure crystals,” Phys. Rev. B 23, 5590–5596 (1981).DOI
  37. [37]Z. Dai and A. M. Rappe, “Recent progress in the theory of bulk photovoltaic effect,” Chem. Phys. Rev. 4, 011303 (2023).arxiv.org/abs/2206.00602
  38. [38]Q. Xue, X. Mu, Y. Sun, and J. Zhou, “Valley contrasting bulk photovoltaic effect in a PT-symmetric MnPSe3 monolayer,” Phys. Rev. B 107, 245404 (2023).
  39. [39]F. Freimuth, S. Blügel, and Y. Mokrousov, “Direct and inverse spin-orbit torques,” Phys. Rev. B 92, 064415 (2015).DOI
  40. [40]See Supplemental Material for details on the symmetry consideration, and additional simulation results that includes Refs. [41, 42].
  41. [41]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,” Phys. Rev. Mater. 5, 014409 (2021).DOI
  42. [43]L. Šmejkal, J. Sinova, and T. Jungwirth, “Emerging research landscape of altermagnetism,” Phys. Rev. X 12, [44] L. Šmejkal, J. Sinova, and T. Jungwirth, “Beyond conventional ferromagnetism and antiferromagnetism: A phase with nonrelativistic spin and crystal rotation symmetry,” Phys. Rev. X 12, 031042 (2022).
  43. [45]J. Zhou and C. Zhang, “Contrasting light-induced spin torque in antiferromagnetic and altermagnetic systems,” Phys. Rev. Lett. 134, 176902 (2025).DOI
  44. [46]R. R. Birss, ed., Symmetry and Magnetism (North-Holland Publishing Company, 1964).
  45. [47]H. Wang and X. Qian, “Electrically and magnetically switchable nonlinear photocurrent in PT-symmetric magnetic topological quantum materials,” npj Comput. Mater. 6, 199 (2020).DOI
  46. [48]A. A. Burkov and L. Balents, “Weyl semimetal in a topological insulator multilayer,” Phys. Rev. Lett. 107, 127205 (2011).DOI
  47. [49]L. Fu, “Hexagonal warping effects in the surface states of the topological insulator bi₂te₃,” Phys. Rev. Lett. 103, 266801 (2009).pubmed.ncbi.nlm.nih.gov/20366328
  48. [50]C. Lei, S. Chen, and A. H. MacDonald, “Magnetized topological insulator multilayers,” Proceedings of the National Academy of Sciences 117, 27224–27230 (2020).DOI
  49. [51]J. Li, Q. Wu, and Weng H., “Stacking-dependent electronic and topological properties in van der waals antiferromagnet mmbi₂te₄ films,” npj Comput. Mater. 11, 44 (2025).DOI
  50. [52]N. Hur, S. Park, P. A. Sharma, J. S. Ahn, S. Guha, and S.-W. Cheong, “Electric polarization reversal and memory in a multiferroic material induced by magnetic fields,” Nature 429, 392 (2004).pubmed.ncbi.nlm.nih.gov/15164057
  51. [53]H. J. Xiang, S.-H. Wei, M.-H. Whangbo, and J. L. F. Da Silva, “Spin-orbit coupling and orbital displacements in multiferroic tbmno₃,” Phys. Rev. Lett. 101, 037209 (2008).DOI
  52. [54]J. Íñiguez, “First-principles approach to lattice-mediated magnetoelectric effects,” Phys. Rev. Lett. 101, 117201 (2008).DOI
  53. [55]E. Bousquet and N. Spaldin, “Induced magnetoelectric response in p⁢nma perovskites,” Phys. Rev. Lett. 107, 197603 (2011).pubmed.ncbi.nlm.nih.gov/22181646
  54. [56]M. Ye and D. Vanderbilt, “Dynamical magnetic charges and linear magnetoelectricity,” Phys. Rev. B 89, 064301 (2014).DOI
  55. [57]K. Zhai, Y. Wu, S. Shen, W. Tian, H. Cao, Y. Chai, B. C. Chakoumakos, D. Shang, L. Yan, F. Wang, and Y. Sun, “Giant magnetoelectric effects achieved by tuning spin cone symmetry in y-type hexaferrites,” Nat. Commun. 8, 159 (2017).DOI
  56. [58]K. Yamauchi, P. Barone, and S. Picozzi, “Bulk rashba effect in multiferroics: A theoretical prediction for bicoo₃,” Phys. Rev. X 9, 041046 (2019).DOI
  57. [59]J. Jiang, Z. L. Bai, Z. H. Chen, L. He, D. W. Zhang, Q. H. Zhang, J. A. Shi, M. H. Park, J. F. Scott, C. S. Hwang, and A. Q. Jiang, “Temporary formation of highly conducting domain walls for non-destructive read-out of ferroelectric domain-wall resistance switching memories,” Nat. Mater. 17, 49 (2018).DOI
  58. [60]A. Eckardt, “Colloquium: Atomic quantum gases in periodically driven optical lattices,” Rev. Mod. Phys. 89, 011004 (2017).DOI
  59. [61]J. H. Shirley, “Solution of the schrödinger equation with a hamiltonian periodic in time,” Phys. Rev. 138, B979–B987 (1965).DOI
  60. [62]H. Sambe, “Steady states and quasienergies of a quantum-mechanical system in an oscillating field,” Phys. Rev. A 7, 2203–2213 (1973).DOI
  61. [63]D. H. Dunlap and V. M. Kenkre, “Dynamic localization of a charged particle moving under the influence of an electric field,” Phys. Rev. B 34, 3625–3633 (1986).pubmed.ncbi.nlm.nih.gov/9940124
  62. [64]F. Zhan, R. Chen, Z. Ning, D.-S. Ma, Z. Wang, D.-H. Xu, and R. Wang, “Perspective: Floquet engineering topological states from effective models towards realistic materials,” Quantum Front. 3, 21 (2024).arxiv.org/abs/2409.02774
  63. [65]C. Zhou and J. Zhou, “Light-induced topological phase transition with tunable layer Hall effect in axion antiferromagnets,” Nano Lett. 24, 7311 (2024).DOI
  64. [66]B. A. Bernevig, T. L. Hughes, and S.-C. Zhang, “Orbitronics: The intrinsic orbital current in p-doped silicon,” Phys. Rev. Lett. 95, 066601 (2005).pubmed.ncbi.nlm.nih.gov/16090968
  65. [67]D. Jo, D. Go, G.-M. Choi, and H.-W. Lee, “Spintronics meets orbitronics: Emergence of orbital angular momentum in solids,” npj Spintronics 2, 19 (2024).DOI
  66. [68]T. P. Cysne, M. Costa, L. M. Canonico, M. B. Nardelli, R. B. Muniz, and T. G. Rappoport, “Disentangling orbital and valley hall effects in bilayers of transition metal dichalcogenides,” Phys. Rev. Lett. 126, 056601 (2021).DOI
  67. [69]D. Go, D. Jo, K.-W. Kim, S. Lee, M.-G. Choi, B.-G. Park, S. Blügel, H.-W. Lee, and Y. Mokrousov, “Long-range orbital torque by momentum-space hotspots,” Phys. Rev. Lett. 130, 246701 (2023).DOI
  68. [70]S. Han, H.-W. Ko, J. H. Oh, H.-W. Lee, K.-J. Lee, and K.-W. Kim, “Orbital pumping incorporating both orbital angular momentum and position,” Phys. Rev. Lett. 134, 036305 (2025).DOI
  69. [71]D. Go, F. Freimuth, J.-P. Hanke, F. Xue, O. Gomonay, K.-J. Lee, S. Blügel, P. M. Haney, H.-W. Lee, and Y. Mokrousov, “Theory of current-induced angular momentum transfer dynamics in spin-orbit coupled systems,” Phys. Rev. Research 2, 033401 (2020).DOI
  70. [72]X. Mu and J. Zhou, “Valley-dependent giant orbital moments and transport features in rhombohedral graphene multilayers,” Phys. Rev. B 111, 165102 (2025).DOI

Paper details

vlnc:W_kgv6qctp
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