Classification of spin Hall effect in two-dimensional systems

Abstract

Physical properties such as the conductivity are usually classified according to the symmetry of the underlying system using Neumann’s principle, which gives an upper bound for the number of independent components of the corresponding property tensor. However, for a given Hamiltonian, this global approach usually can not give a definite answer on whether a physical effect such as spin Hall effect (SHE) exists or not. It is found that the parity and types of spin-orbit interactions (SOIs) are good indicators that can further reduce the number of independent components of the spin Hall conductivity for a specific system. In terms of the parity as well as various Rashba-like and Dresselhaus-like SOIs, we propose a local approach to classify SHE in two-dimensional (2D) two-band models, where sufficient conditions for identifying the existence or absence of SHE in all 2D magnetic point groups are presented.

Graphical abstract

Keywords

spin Hall effect / symmetry / two-dimensional system

1 Introduction

Berry curvature related band geometric quantities are widely adopted to describe various Hall effects [1, 2]. For instance, nonzero Berry phase accounts for both quantum Hall effect [3] and quantum anomalous Hall effect [4], which are intrinsic responses and involve the breaking of time-reversal symmetry. In time-reversal invariant systems, extrinsic Hall effect can exist in the nonlinear response regime, such as the second-order Hall effect induced by Berry curvature dipole [5–7] and the third-order Hall effect induced by Berry-connection polarizability tensor [8], which can also be studied in multiterminal systems using the scattering matrix theory [9, 10]. Higher-order nonlinear anomalous Hall effects induced by Berry curvature multipoles such as quadrupole and hexapole have been discussed in certain materials with magnetic point group symmetry [11]. In particular, intrinsic second-order anomalous Hall effect has been discovered in \mathcal{P}\mathcal{T}-symmetric antiferromagnets [12–14] and intrinsic third-order anomalous Hall effect was discussed [15]. In addition to Hall currents, it was found that Berry curvature dipole and Berry curvature fluctuation can give rise to linear and nonlinear thermal Hall noises [16]. The existence of these nonlinear Hall effects are classified by symmetry and relevant constraints [5, 6, 11–13, 15], since band geometry is strongly affected by symmetry.

In fact, symmetry plays essential roles in classifying a variety of physical properties [17]. For example, the famous universal conductance fluctuation (UCF) [18–22] in mesoscopic transport depends on symmetry and dimensionality of the system. To describe UCF, the system Hamiltonian is categorized into three ensembles, i.e., Gaussian orthogonal/unirary/sympletic ensemble, based on the presence or absence of time-reversal and spin-rotation symmetries. When the particle-hole and chiral symmetries are further included, it has been extended to the ten-fold way [23, 24], which is widely used in classifying topological insulators (TI) and topological superconductors. Spin photovoltaic effect in antiferromagnetic materials can be classified in terms of the \mathcal{P}\mathcal{T} symmetry and spacial symmetries [25]. However, there are also exceptions using symmetry analysis. The remarkable example is the phase transition between TI and band insulator, which occurs without symmetry breaking but with the changing of global (topological) invariants [26, 27]. In this work, we find another example: for a system with a given symmetry, spin Hall effect (SHE) can be present or absent depending on the parity and spin-orbit interaction (SOI) of the Hamiltonian.

SHE is a relativistic phenomenon, where charge current drives transverse spin current in spin-orbit coupled systems [28]. Similar to charge Hall effects, SHE can also be intrinsic or extrinsic. Intrinsic SHE [29, 30] is not influenced by transport process, which has been proposed in Weyl semimetals such as TaAs [31]. The orientation of spin polarization in SHE can be either in-plane or out-of-plane. In different materials, the dominant SOI could be Dresselhaus-like or Rashba-like, or the combination of them [32–34]. The existence of SHE and inverse SHE [35] have been solidly confirmed by a series of experiments [36, 37], and theoretical description of SHE usually involves spin Berry curvature [31]. However, less attention has focused on classifying SHE, which could be carried out using Neumann’s principle. Neumann’s principle has been applied to analyze the conductivity tensor for nonlinear Hall effects [11–13], which can be stated as: if a crystal is invariant with respect to certain symmetry, any of its physical properties must also be invariant with respect to the same symmetry. As a consequence of this principle, symmetry imposes constraint on physical quantities, including the spin Hall conductivity.

In this work, we demonstrate that the symmetry alone is insufficient to determine the existence of SHE in 2D systems, since SHE can be switched on or off by varying the SOI while maintaining the symmetry of the system. Compared with Neumann’s principle, the parity and types of SOIs are better indicators for identifying SHE. Based on the parity as well as different orders of in-plane or out-of-plane, Dresselhaus-like or Rashba-like SOIs, we propose a local approach to classify SHE, which gives sufficient conditions for the existence of SHE in 2D two-band models. Complete analysis on all 2D magnetic point groups (MPGs) are carried out and possible materials for experimental verification are discussed.

The paper is organized as follows. In Section 2, the Kubo formula for the spin Hall conductivity of a 2D two-band model is introduced. In Section 3, classification of 2D SHE according to Neumann’s principle is discussed. For a Hamiltonian with certain symmetry, this general classification can only give an indefinite answer on the presence or absence of SHE. In Section 4, a different symmetry analysis based on parity and types of SOIs constituting the 2D Hamiltonian is given, which allows identifying the existence of SHE. Finally, discussion and conclusion are given in Section 5.

2 Formalism for spin Hall effect

The spin Hall conductivity is [29, 31, 38] (\hslash =e=1)

(1){\sigma }_{xy}^{\alpha }={\displaystyle {\int }_k{\displaystyle \sum _n{f}_{n,\mathit{k}}{\mathrm{\Omega }}_{n,xy}^{\alpha }(\mathit{k}),}}

where {\int }_k={\int }_{BZ} {\mathrm{d}}^{2}k/(2\pi {)}^2 and \alpha labels the direction of spin polarization of SHE. The “spin” Berry curvature is defined as

(2){\mathrm{\Omega }}_{n,xy}^{\alpha }(\mathit{k})=- 2\mathrm{I}\mathrm{m}{\displaystyle \sum _{n^{\prime }\ne n}{\displaystyle \frac{⟨n|J_x^{\alpha }|n^{\prime }⟩⟨n^{\prime }|v_y|n⟩}{({ϵ}_n-{ϵ}_{n^{\prime }}{)}^2}},}

with spin current operator J_i^{\alpha }= \frac{1}{2}\{v_i,s_{\alpha }\}, where s_{\alpha } is the spin operator and v_i =\frac{\mathrm{\partial }H}{\mathrm{\partial }{k}_i} is the velocity operator. Note that the “spin” Berry curvature resembles usual Berry curvature only if the spin is a good quantum number. In the presence of spin−orbit interaction (SOI), it is very different from the usual Berry curvature. We focus on a two-band model defined as H=d_0{\sigma }_0+\mathit{d}\cdot \mathit{\sigma }, from which we have

(3)v_i={\mathrm{\partial }}_i{d}_0+({\mathrm{\partial }}_i{d}_j){\sigma }_j,

where summation over repeated indices is implied and

(4)J_i^{\alpha } ={\displaystyle \frac{\hslash }{2}}({\sigma }_{\alpha }{\mathrm{\partial }}_i{d}_0 +{\mathrm{\partial }}_i{d}_{\alpha } ).

After some algebra, we find the spin Berry curvature for the lower band

(5){\mathrm{\Omega }}_{xy}^{\alpha }(\mathit{k})={\displaystyle \frac{{\mathrm{\partial }}_x{d}_0({\mathrm{\partial }}_y\mathit{d}\times \mathit{d}{)}_{\alpha }}{4d^3}},

where d^2=\mathit{d}\cdot \mathit{d}. It is easy to show that for the linear Rashba SOI, Eq. (5) reproduces the spin Hall conductance of e/(8\pi ), which was obtained by Sinova et al. [29]. For comparison, we show the expression of Berry curvature

(6){\mathrm{\Omega }}_{\pm ,xy}(\mathit{k})=\pm {\displaystyle \frac{\hslash {\mathrm{\partial }}_x\mathit{d}\cdot ({\mathrm{\partial }}_y\mathit{d}\times \mathit{d})}{2d^3}}.

Obviously, if H_0=0 there is no linear SHE. From now on, we will work on systems with broken particle−hole symmetry (d_0=k^2) and focus on {\sigma }_{xy}^{\alpha }. The analysis of {\sigma }_{yx}^{\alpha } is similar.

In order to have spin Hall effect, SOI must be present which can be classified according to the symmetry as well as linear or nonlinear orders of momentum. For instance, two typical in-plane linear order SOIs (IP-SOI) in 2D systems are the Rashba SOI k_y{\sigma }_x-k_x{\sigma }_y and Dresselhaus SOI k_x{\sigma }_x-k_y{\sigma }_y. For the classification reason, we define two types of SOI, Dresselhaus-like (H_D) and Rashba-like (H_R) SOI, as follows. For IP-SOI, we define A= \sum _{n=1}^N({\alpha }_n{k}_{+}^{N-n}{k}_{-}^n{\sigma }_{+} +{\beta }_n k_{-}^{N-n}{k}_{+}^n{\sigma }_{-}) with k_{\pm }=k_x \pm \mathrm{i}{k}_y, then H_D=A+A^{†} and H_R=-\mathrm{i}(A-A^{†}). For the out-of-plane SOI (OP-SOI), we have H_D=B+ B^{†} and H_R=-\mathrm{i}(B-B^{†}) where B= \sum _{n=1}^N{\gamma }_n k_{+}^{N-n} k_{-}^n {\sigma }_z. In Tab.1, we list a few IP-SOIs and OP-SOIs which are basic building blocks of the SOI Hamiltonian. In particular, the SOI with N=3 is called cubic Dresselhaus and Rashba SOI, respectively in the literatures [32, 38, 46–48]. According to our notation, k_y^3{\sigma }_x-k_x^3{\sigma }_y\sim \mathrm{I}\mathrm{m}[( k_{+}^3+3k_{+} k_{-}^2 ){\sigma }_{+}], ( k_x^2-k_y^2)(k_y{\sigma }_x+k_x{\sigma }_y)\sim \mathrm{I}\mathrm{m}[(k_{+}^2+k_{-}^2)k_{+} {\sigma }_{+}], k_x^2{k}_y^2(k_y{\sigma }_x-k_x{\sigma }_y)\sim \mathrm{I}\mathrm{m}[(k_{+}^2-k_{-}^2{)}^2{k}_{+}{\sigma }_{-}], and k_x{k}_y(k_x {\sigma }_x- k_y{\sigma }_y)\sim \mathrm{I}\mathrm{m}[(k_{+}^2-k_{-}^2)k_{+} {\sigma }_{-}] belong to Rashba-like SOI with C_{4v} symmetry [49–51] while k_x{k}_y(k_y {\sigma }_x- k_x{\sigma }_y)\sim \mathrm{R}\mathrm{e}[(k_{+}^2-k_{-}^2)k_{+} {\sigma }_{-}] and k_x^3{\sigma }_x+k_y^3{\sigma }_y \sim \mathrm{I}\mathrm{m}[( k_{+}^3+3k_{+} k_{-}^2 ){\sigma }_{+}] belongs to Dresselhaus-like SOI with C_4 symmetry [42, 52].

3 Classification of SHE from Neumann’s principle

SHE can be classified using Neumann’s principle. Note that the spin current is obtained from spin (pseudo-vector) and velocity (vector) operators, it forms a second rank pseudo-tensor while the electric field is a vector. As a result, the spin conducticity tensor {\sigma }_{\beta \gamma }^{\alpha } is a third rank pseudo-tensor (\beta ,\gamma =x,y) [53, 54]. Different from the Hall effect, generally speaking, spin conductance does not enjoy antisymmetric property with respective to \beta and \gamma. According to Neumann’s principle, the spin conductivity tensor is expressed in terms of a rotation matrix R as

(7){\sigma }_{\beta \gamma }^{\alpha }= \eta \mathrm{d}\mathrm{e}\mathrm{t}( R)R_{\alpha {\alpha }^{\prime }}{R}_{\beta {\beta }^{\prime }}{R}_{\gamma {\gamma }^{\prime }}{\sigma }_{{\beta }^{\prime }{\gamma }^{\prime }}^{{\alpha }^{\prime }}.

where \eta =-1 for prime operation and the presence of \mathrm{d}\mathrm{e}\mathrm{t}(R) is needed for a pseudo-tensor. From Eq. (1), it can also be shown as

(8){\sigma }_{\beta \gamma }^{\alpha }={\displaystyle {\int }_k{\mathrm{\Omega }}_{\beta \gamma }^{\alpha }(\mathit{k})=\eta \mathrm{d}\mathrm{e}\mathrm{t}(\mathrm{R}){\displaystyle {\int }_k{\mathrm{\Omega }}_{\beta \gamma }^{\alpha }(R\mathit{k}),}}

Since 2D MPG is imbedded in 3D MPG, we can use Bilbao Crystallographic Server [55] to find nonzero components of {\sigma }_{\beta \gamma }^{\alpha }, which is similar to the discussion of higher-order anomalous Hall effects [11–13].

The results obtained from Bilbao Crystallographic Server, using Jahn Symbol {\mathrm{e}\mathrm{V}}^3, can be summarized as follows. For in-plane SHE (IP-SHE), we find: (i) for MPG m, m1', {\sigma }_{xy}^y={\sigma }_{yx}^y=0 and {\sigma }_{xy}^x and {\sigma }_{yx}^x are indefinite; (ii) for MPG 2', m', \mathcal{T}, both {\sigma }_{xy}^{x/y} and {\sigma }_{yx}^{x/y} are indefinite; (iii) for MPG 3, 31', 3m', 6', {\sigma }_{xy}^{x/y}={\sigma }_{yx}^{x/y}; (iv) for MPG 3m, 3m1', 6'mm', {\sigma }_{xy}^x={\sigma }_{yx}^x =0 and {\sigma }_{xy}^y ={\sigma }_{yx}^y. For other 2D MPG elements, both {\sigma }_{xy}^{x/y} and {\sigma }_{yx}^{x/y} are zero. For out-of-plane SHE (OP-SHE), we have: (i) {\sigma }_{xy}^z=-{\sigma }_{yx}^z for most of the high symmetry rotations: 4, 41', 4mm, 4mm1', 4'm'm, 4m'm', 3, 31', 3m, 3m1', 3m', 6, 61', 6', 6mm, 6mm1', 6'mm', 6m'm'; (ii) {\sigma }_{xy}^z and {\sigma }_{yx}^z are indefinite for the rest of the MPGs. The in-plane and out-of-plane SHEs are schematically shown in Fig.1.

These results are summarized in Tab.2, from which we find that for many systems the existence of SHE can not be determined solely by the symmetry. Moreover, we observe that {\sigma }_{xy}^{\alpha } can be switched on and off while maintaining the symmetry of the system. For instance, H_{\mathrm{S}\mathrm{O}\mathrm{I}}=k_x^2{\sigma }_x+k_y^2{\sigma }_y has C_2\mathcal{T} symmetry and it is easy to verify that {\sigma }_{xy}^z is zero for H_{\mathrm{S}\mathrm{O}\mathrm{I}} while {\sigma }_{xy}^z\ne 0 for H_1+H_{\mathrm{S}\mathrm{O}\mathrm{I}} where H_1 is the linear Rashba SOI listed in Tab.1. This shows that for a given Hamiltonian, symmetry alone cannot characterize the OP-SHE. This conclusion is also valid for IP-SHE. Note that Eq. (7) relates different components of SHE and imposes constraint on them (global constraint). It gives an upper bound of the number of independent components solely from the symmetry of the third rank pseudo-tensor, regardless of the physical system. Once the expression of SHE is given, we can use Eq. (8) to find further constraint on a particular component of SHE (local constraint). It turns out that the parity and types (H_D or H_R) of SOI are good indicators to classify SHE. We recognize that parity is also a kind of symmetry, but in this work the word “symmetry” refers solely to spatial symmetries. In the following, we will give the sufficient condition under which SHE may vanish.

4 Classification of SHE based on the parity and types of SOIs

The order of SOI, N, is determined by the power of k (regardless of k_x and k_y). In general, the parities of H_D and H_R are discussed in Appendix A. We find that the parity (-1{)}^N plays a critical role in determining the existence of IP-SHE ({\sigma }_{xy}^{x/y}). Specifically, we demonstrate that, for a system with several SOI components, the IP-SHE can be switched on and off by tuning these SOIs while maintaining the symmetry of the system. We further show that SHE can be classified by the parity of SOI.

For IP-SHE, there are 12 MPGs in Tab.3. The results are summarized as: (i) for MPG m, m1', 3m1, we find {\sigma }_{xy}^x=0 and {\sigma }_{xy}^y= 1 (“1” stands for nonzero) when m=m_x; (ii) for MPG m', 3m, 3m', we find {\sigma }_{xy}^x=0 while {\sigma }_{xy}^y=0/1 (“0/1” stands for zero or nonzero depending on the parity of individual SOI component in the Hamiltonian); (iii) for MPG \mathcal{T}, 3, 31', {\sigma }_{xy}^{x/y}=0/1; (iv) for MPG 2', 21', 2'mm', 2mm1', 41', 4mm1', 6', 6'm'm, 61', and 6mm1', no SOI Hamiltonians are available for 2D two-band models with nonzero d_z.

For OP-SHE, {\sigma }_{xy}^z is nonzero for most of MPGs. The condition for {\sigma }_{xy}^z=0 is that d_x{d}_y is an odd function in momentum space. We find that {\sigma }_{xy}^z is nonzero for all 2D MPGs except three (m, 1, 2') with the following Hamiltonians: (i) the Hamiltonian with MPG m (mirror symmetry), e.g., H= k_y^2 {\sigma }_x+ k_x{\sigma }_y. Other Hamiltonian with the same symmetry may not have vanishing OP-SHE. (ii) the Hamiltonian with no symmetry at all, e.g., H=k_x{\sigma }_x+k_y^2{\sigma }_y. (iii) the Hamiltonian with MPG 2', e.g., H= k_y^2 {\sigma }_x- k_x^2 {\sigma }_y. Another straightforward condition is that the system either has the chiral symmetry {\sigma }_x [56, 57] or chiral symmetry {\sigma }_y [58]. In the following, we give an example to demonstrate our findings and then verify it in a general account.

We consider the following Hamiltonian

(9)\begin{array}{rl}H=& k^2{\sigma }_0+{\lambda }_1(k_y{\sigma }_x-k_x{\sigma }_y)+{\displaystyle \frac{{\lambda }_2}{2}}(k_{+}^2{\sigma }_{+}+k_{-}^2{\sigma }_{-})\\ & -\mathrm{i}{\displaystyle \frac{{\lambda }_3}{2}}(k_{+}^5{\sigma }_{+}-k_{-}^5{\sigma }_{-})+{\displaystyle \frac{{\lambda }_4}{2}}(k_{+}^3+k_{-}^3){\sigma }_z-{\displaystyle \frac{\mathrm{i}{\lambda }_5}{2}}(k_{+}^3\\ & -k_{-}^3){\sigma }_z+{\displaystyle \frac{{\lambda }_6}{2}}(k_{+}^6+k_{-}^6){\sigma }_z-{\displaystyle \frac{\mathrm{i}{\lambda }_7}{2}}(k_{+}^6-k_{-}^6){\sigma }_z.\end{array}

Here {\sigma }_{\pm }={\sigma }_x\pm \mathrm{i}{\sigma }_y. In this Hamiltonian, different parts of SOI have been used before. For instance, the {\lambda }_2 term was used in Ref. [39] to address 2D dual topological insulator of Na{}_3Bi. By fitting the experimental data, the Dresselhaus-like SOI ({\lambda }_3 term) can account for the strong out-of-plane spin component at the Fermi surface of 2D Au/Ge(111) surface [34]. The {\lambda }_4 term was proposed [41] to explain the experimentally observed warping effect of Fermi surface for topological insulator Bi{}_2Te{}_3 and the {\lambda }_5 term is crucial in determining the origin of experimental finding, the giant Zeeman-type spin polarization in WSe{}_2 [59]. Note that the linear Rashba SOI ({\lambda }_1 term) has the highest symmetry while the {\lambda }_2 term (6'm'm), {\lambda }_3 term (6mm1'), {\lambda }_6 term (6), and {\lambda }_7 term (6mm) are hexagonal SOI. The {\lambda }_4 and {\lambda }_5 terms are trigonal SOI having m_x and m_y symmetries, respectively, making Eq. (9) a good testing platform for symmetry analysis. For instance, when {\lambda }_2 and {\lambda }_5 are nonzero, the system has (C_3, m_y\mathcal{T}) symmetry (see Appendix B for more details). Turning on and off any of {\lambda }_1 [60], {\lambda }_3, or {\lambda }_6 does not affect the symmetry of the system. Fixing {\lambda }_2 and {\lambda }_4 to be nonzero while switching off {\lambda }_5 and {\lambda }_6 changes the system symmetry to (C_3, m_x). Therefore IP-SHE of this trigonal symmetry can be studied by tuning any of parameters of {\lambda }_1, {\lambda }_3, and {\lambda }_7. For nonzero ( {\lambda }_2, {\lambda }_4, {\lambda }_5), the system has C_3 symmetry. Its IP-SHE can be studied by varying {\lambda }_1. In addition, IP-SHE of hexagonal MPG 6 can be examined by setting ({\lambda }_2, {\lambda }_4, {\lambda }_5)=0 and({\lambda }_3, {\lambda }_6, {\lambda }_7)\ne 0 while varying {\lambda }_1. IP-SHE of MPG 6mm (6m'm') can be studied by switching on {\lambda }_3 and {\lambda }_7 ({\lambda }_6) and turning on and off {\lambda }_1.

In the following, we focus on c_{\alpha }\equiv {\mathrm{\partial }}_x d_0({\mathrm{\partial }}_y\mathbf{d}\times \mathbf{d}{)}_{\alpha } which differs from {\mathrm{\Omega }}_{xy}^{\alpha } by 1/ (4d^3). For convenience, we define a_N=k_{+}^N+ k_{-}^N and b_N=\mathrm{i}(k_{+}^N-k_{-}^N) so that all SOI Hamiltonians in Tab.1 can be expressed in terms of a_N, b_N, and {\sigma }_{\alpha }. The parities of a_N and b_N with respect to k_x and k_y are found to be a_N\sim k_x^N and b_N\sim k_x{k}_y{a}_N (see Appendix C). For Eq. (9), in terms of a_N and b_N, we have d_x={\lambda }_1{k}_y+{\lambda }_2{a}_2+{\lambda }_3{b}_5, d_y=-{\lambda }_1{k}_x +{\lambda }_2 b_2- {\lambda }_3{a}_5, and d_z={\lambda }_4{a}_3+ {\lambda }_5{b}_3+{\lambda }_6{a}_6 +{\lambda }_7 b_6. Using Eq. (5), the expression of c_{\alpha } is shown in Appendix D.

Now we examine the behaviors of IP-SHE for various symmetries 6mm (6m'm'), 6, (C_3, m_y\mathcal{T}), (C_3, m_x) and C_3. 1) For MPG 6mm or 6m'm', we find {\sigma }_{xy}^{x/y}=0. The reason behind this can be understood by simple parity analysis. For a 2D system having two mirror symmetries or both m_{x/y}\mathcal{T} symmetries, d^2 is an even function of both k_x and k_y, and hence we can focus on the parity of c_{\alpha }. For {\sigma }_{xy}^{\alpha } to be nonzero, c_{\alpha } must be an even function of k_0 if we set k_x=k_y =k_0. In addition, the in-plane (out-of-plane) SOIs must be odd (even) functions of k_0. This is because {\sigma }_{\pm } rotates in the same way as k_{\pm } while {\sigma }_z remains unchanged under rotation. Since c_{x/y} scales like d_{y/x}{d}_z{k}_x{k}_y according to Eq. (5), it is impossible for c_{x/y} to be an even function of k_0. Thus, we find that {\sigma }_{xy}^{x/y}=0 for systems with two mirror symmetries, which include symmetry groups: 2mm, 4mm, 6mm, 2mm1', 4mm1', 6mm1', 2m'm', 4m'm', and 6m'm'.

2) For the system with C_6 symmetry, ({\lambda }_3, {\lambda }_6, {\lambda }_7) is nonzero and varying {\lambda }_1 does not affect the symmetry. We find {\sigma }_{xy}^{x/y}=0 because d^2 has a mirror symmetry M_{x+y} with or without {\lambda }_1 so that terms involving c_{x/y}^{(2)} and c_{x/y}^{(3)} do not contribute to {\sigma }_{xy}^{x/y}. Discussion on the symmetry of energy spectrum or d^2 is shown in Appendix E. This result agrees with that obtained from Neumann’s principle.

3) For the system with (C_3, m_y\mathcal{T}) symmetry, we study the following cases. (a) The case when {\lambda }_2 and {\lambda }_5 are the only two nonzero parameters. Since there is only one SOI in either IP-SOI or OP-SOI, d^2 has two mirror symmetries. From Appendix D, we keep only c_x^{(1)} and c_y^{(1)} and find {\sigma }_{xy}^{x/y}=0 (no IP-SHE). As will be discussed later, if there is one IP-SOI and one OP-SOI with different parities, there is no IP-SHE. Case (3a) is just a special case. (b) Now we turn on {\lambda }_1 which respects m_y symmetry. Since the {\lambda }_2 term does have m_y symmetry, hence d_x^2+d_y^2 has only m_x symmetry which is the only mirror symmetry possessed by both IP-SOIs. Note that the symmetry of d_z^2 remains the same as in (3a) which makes d^2 asymmetric about k_y. Including both c_x^{(1)} and c_x^{(3)}, we obtain c_x=3{\lambda }_1{\lambda }_5 a_2{k}_x^2+{\lambda }_2{\lambda }_5(2 a_1{b}_3-3a_2{b}_2)k_x and c_y=0, which gives {\sigma }_{xy}^x\ne 0 and {\sigma }_{xy}^y= 0 [shown in Appendix F(a)]. (c) If we replace the {\lambda }_1 term by the {\lambda }_6 term, d_x^2+d_y^2 has two mirror symmetries as explained above. Note that d_z^2 has m_x symmetry. We have c_x=-4{\lambda }_2{\lambda }_6(a_1{a}_6+b_2{b}_5 )k_x+2{\lambda }_2{\lambda }_5 (3a_2{b}_2 -2a_1{b}_3)k_x and c_y=0, leading to nonvanishing {\sigma }_{xy}^x [shown in Appendix F(b)]. We see that {\sigma }_{xy}^x and {\sigma }_{xy}^y can be switched on and off while maintaining the system symmetry.

4) For the system with (C_3, m_x) symmetry, we require {\lambda }_2 and {\lambda }_4 to be nonzero and set {\lambda }_5={\lambda }_6=0. (a) When {\lambda }_1= {\lambda }_7=0, d^2 has two mirror symmetries and we find {\sigma }_{xy}^{x/y}=0 from the symmetry argument which is the same as case (3a). (b) When turning on {\lambda }_1, d^2 becomes symmetric about k_x only. Hence neglecting terms odd in k_x we find {\sigma }_{xy}^x=0 and c_y={\lambda }_1{\lambda }_4(a_3-3b_2{k}_y)k_x+{\lambda }_2{\lambda }_4(2a_3{b}_1-3a_2{b}_2)k_x making {\sigma }_{xy}^y\ne 0 [shown in Appendix F(c)]. (c) When replacing the {\lambda }_1 term by the {\lambda }_7 term, the symmetry of d^2 remains the same. We obtain {\sigma }_{xy}^x =0 and c_y={\lambda }_2{\lambda }_4(2a_3{b}_1-3a_2{b}_2)k_x which has nonzero contribution. The result is the same as (3b).

5) When ({\lambda }_2, {\lambda }_4, {\lambda }_5) are nonzero, the system has C_3 symmetry. It is easy to show that d^2 has only mirror symmetry m_{x+y} and therefore only terms involving c_{x/y}^{(1)} and c_{x/y}^{(4)} contribute. We find {\sigma }_{xy}^{x/y}=0. When {\lambda }_1 is present, the system is still maintained at C_3 while m_{x+y} symmetry is broken for d^2. We find c_x=- 6{\lambda }_1 {\lambda }_5{a}_2{k}_x^2+6{\lambda }_1{\lambda }_4{b}_2 k_x^2 -2{\lambda }_2{\lambda }_4 (2a_1{a}_3 +3b_2^2) k_x+2 {\lambda }_2{\lambda }_5( 3a_2{b}_2 -2a_1{b}_3)k_x and c_y=2{\lambda }_1{\lambda }_4 (a_3-3b_2{k}_y)k_x +2{\lambda }_1{\lambda }_5(b_3 +3a_2{k}_y)k_x +2{\lambda }_2{\lambda }_5 (2b_1{b}_3 +3a_2^2)k_x+2{\lambda }_2{\lambda }_4(2 a_3{b}_1-3a_2{b}_2)k_x. Hence both {\sigma }_{xy}^{x/y}\ne 0.

For OP-SHE, it is easy to see that: if OP-SHE vanishes for a particular H_{SOI} then H_1+H_{SOI} makes {\sigma }_{xy}^z\ne 0. Therefore, for a given symmetry, OP-SHE can also be tuned from zero to nonzero. One IP-SOI means a basic building block with a defnite parity and SOI type, such as the one in Tab.1.

Now we give the sufficient conditions for vanishing IP-SHE in 2D systems in the following, and present the proof in Appendix I. (i) If the SOI Hamiltonian has chiral symmetry {\sigma }_z, then {\sigma }_{xy}^{x/y}=0. (ii) If the SOI Hamiltonian contains only one in-plane and one out-of-plane components, there is no IP-SHE when the two componants have different parities [see the examples, cases (3a) and (4a)]. Supposing they have the same parity, we find that {\sigma }_{xy}^y=0 if both components are the same type of SOI while {\sigma }_{xy}^x=0 if they are different types of SOI. (iii) The case that d^2 has one mirror symmetry m_x. Supposing the Hamiltonian is given by H=H_0+{\lambda }_1{R}_{IP} +{\lambda }_2 D_{IP}+{\lambda }_3{D}_{OP} +{\lambda }_4 R_{OP} where R_{IP} and D_{OP} stand for in-plane Rashba-like SOI and out-of-plane Dresselhaus-like SOI, respectively. Each SOI A_B with A=R,D and B=IP,OP may have several terms but must have the same parity. For instance, D_{IP}=D_{IP}^a+D_{IP}^b, IP-SHE vanishes only if a and b have the same parity. We find that the condition for {\sigma }_{xy}^x =0 is: the parity of ( D_{IP},R_{IP},D_{OP} ,R_{OP})=(\pm ,\mp ,\mp ,\pm ) where + stands for even parity while the condition for {\sigma }_{xy}^y=0 is: the parity of (D_{IP} ,R_{IP},D_{OP},R_{OP} )=(\pm ,\mp ,\pm ,\mp ) [see the examples, cases (3b) and (4b-d)]. (iv) Assuming the mirror symmetry of d^2 is m_y, we find that the condition for {\sigma }_{xy}^x= 0 is H=R_{IP}^{-}+R_{IP}^{+}+D_{OP}^{-}+D_{OP}^{+} or H= D_{IP}^{-}+ D_{IP}^{+}+ R_{OP}^{-}+ R_{OP}^{+}; while for {\sigma }_{xy}^y=0, we require H= R_{IP}^{-}+ R_{IP}^{+}+ R_{OP}^{-}+ R_{OP}^{+} or H= D_{IP}^{-}+ D_{IP}^{+}+ D_{OP}^{-}+ D_{OP}^{+}. (v) If all IP-SOIs have one parity and all OP-SOIs have another parity, {\sigma }_{xy}^{x/y}=0.

We present examples of SOI Hamiltonians for all 2D MPGs in Table A2 of Appendix J. As shown in Appendix K, the vanishing of IP-SHE for all 2D MPGs with d_z\ne 0 in Tab.3 can be predicted using conditions (2)−(5) without involving Neumann’s principle.

5 Discussion and conclusion

In summary, we have demonstrated that the symmetry of the system along is not enough to characterize the existence of spin Hall effect, while the parity and symmetry types of constituent SOI play a crucial role. It is found that 2D in-plane SHE can be switched on and off by combining different SOI components, including in-plane and out-of-plane Rashba-like and Dresselhaus-like SOIs of different orders, while maintaining the system symmetry. Sufficient conditions for the existence of SHE are presented, accompanied by complete analysis on all 2D magnetic point groups.

The verification of our findings would be straightforward. SHE has been experimentally realized over fifteen years [36, 64]. In Ref. [65], H=H_1+ H_9+H_{13x} was used to model the surface states in the Bi{}_2Te{}_3 family of 3D TI using first-principles calculation, which has been experimentally verified [34]. However, we need a system with both types of SOI and different parities. In Ref. [66], a new class of atom-laser coupling schemes were introduced to describe the spin−orbit coupled Hamiltonian of ultracold neutral atoms, which has the form H=H_1+H_{3x} +H_{3y}+H_5+H_6 and could be adopted to the switching on and off of SHE.

References

[1] D. Xiao, M. C. Chang, Q. Niu. Berry phase effects on electronic properties. Rev. Mod. Phys., 2010, 82(3): 1959

[2] Y. Gao. Semiclassical dynamics and nonlinear charge current. Front. Phys., 2019, 14(3): 33404

[3] K. von Klitzing. The quantized Hall effect. Rev. Mod. Phys., 1986, 58(3): 519

[4] N. Nagaosa, J. Sinova, S. Onoda, A. H. MacDonald, N. P. Ong. Anomalous Hall effect. Rev. Mod. Phys., 2010, 82(2): 1539

[5] I. Sodemann, L. Fu. Quantum nonlinear Hall effect induced by Berry curvature dipole in time-reversal invariant materials. Phys. Rev. Lett., 2015, 115(21): 216806

[6] T. Low, Y. Jiang, F. Guinea. Topological currents in black phosphorus with broken inversion symmetry. Phys. Rev. B, 2015, 92(23): 235447

[7] Z. Z. Du, H. Z. Lu, X. C. Xie. Nonlinear Hall effects. Nat. Rev. Phys., 2021, 3(11): 744

[8] S. Lai, H. Liu, Z. Zhang, J. Zhao, X. Feng, N. Wang, C. Tang, Y. Liu, K. S. Novoselov, S. A. Yang, W. Gao. Third-order nonlinear Hall effect induced by the Berry-connection polarizability tensor. Nat. Nanotechnol., 2021, 16(8): 869

[9] M. Wei, B. Wang, Y. Yu, F. Xu, J. Wang. Nonlinear Hall effect induced by internal Coulomb interaction and phase relaxation process in a four-terminal system with time-reversal symmetry. Phys. Rev. B, 2022, 105(11): 115411

[10] M. Wei, L. Xiang, L. Wang, F. Xu, J. Wang. Quantum third-order nonlinear Hall effect of a four-terminal device with time-reversal symmetry. Phys. Rev. B, 2022, 106(3): 035307

[11] C. P. Zhang, X. J. Gao, Y. M. Xie, H. C. Po, K. T. Law. Higher-order nonlinear anomalous Hall effects induced by Berry curvature multipoles. Phys. Rev. B, 2023, 107(11): 115142

[12] C. Wang, Y. Gao, D. Xiao. Intrinsic nonlinear Hall effect in antiferromagnetic tetragonal CuMnAs. Phys. Rev. Lett., 2021, 127(27): 277201

[13] H. Y. Liu, J. Z. Zhao, Y. X. Huang, W. K. Wu, X. L. Sheng, C. Xiao, S. Y. A. Yang. Intrinsic second-order anomalous Hall effect and its application in compensated antiferromagnets. Phys. Rev. Lett., 2021, 127(27): 277202

[14] A. Gao, Y. F. Liu, J. X. Qiu, B. Ghosh, T. V. Trevisan, Y. Onishi, C. Hu, T. Qian, H. J. Tien, S. W. Chen, M. Huang, D. Bérubé, H. Li, C. Tzschaschel, T. Dinh, Z. Sun, S. C. Ho, S. W. Lien, B. Singh, K. Watanabe, T. Taniguchi, D. C. Bell, H. Lin, T. R. Chang, C. R. Du, A. Bansil, L. Fu, N. Ni, P. P. Orth, Q. Ma, S. Y. Xu. Quantum metric nonlinear Hall effect in a topological antiferromagnetic heterostructure. Science, 2023, 381(6654): 181

[15] L. Xiang, C. Zhang, L. Wang, J. Wang. Third-order intrinsic anomalous Hall effect with generalized semiclassical theory. Phys. Rev. B, 2023, 107(7): 075411

[16] M. Wei, L. Wang, B. Wang, L. Xiang, F. Xu, B. Wang, J. Wang. Quantum fluctuation of the quantum geometric tensor and its manifestation as intrinsic Hall signatures in time-reversal invariant systems. Phys. Rev. Lett., 2023, 130(3): 036202

[17] L. Shi, H. Z. Lu. Quantum transport in topological semimetals under magnetic fields (III). Front. Phys., 2023, 18(2): 21307

[18] L. B. Altshuler. Fluctuations in the extrinsic conductivity of disordered conductors. JETP Lett., 1985, 41: 648

[19] P. A. Lee, A. D. Stone. Universal conductance fluctuations in metals. Phys. Rev. Lett., 1985, 55(15): 1622

[20] P. A. Lee, A. D. Stone, H. Fukuyama. Universal conductance fluctuations in metals: Effects of finite temperature, interactions, and magnetic field. Phys. Rev. B, 1987, 35(3): 1039

[21] C. W. J. Beenakker. Random-matrix theory of quantum transport. Rev. Mod. Phys., 1997, 69(3): 731

[22] Y. L. Han, Z. H. Qiao. Universal conductance fluctuations in Sierpinski carpets. Front. Phys., 2019, 14(6): 63603

[23] S. Ryu, A. Schnyder, A. Furusaki, A. Ludwig. Topological insulators and superconductors: Tenfold way and dimensional hierarchy. New J. Phys., 2010, 12(6): 065010

[24] C. K. Chiu, J. C. Y. Teo, A. P. Schnyder, S. Ryu. Classification of topological quantum matter with symmetries. Rev. Mod. Phys., 2016, 88(3): 035005

[25] R. C. Xiao, Y. J. Jin, H. Jiang. Spin photovoltaic effect in antiferromagnetic materials: Mechanisms, symmetry constraints, and recent progress. APL Mater., 2023, 11(7): 070903

[26] X. L. Qi, S. C. Zhang. Topological insulators and superconductors. Rev. Mod. Phys., 2011, 83(4): 1057

[27] B.A. Bernevig, Topological Insulators and Topological Superconductors, Princeton University Press, New Jersey, 2013

[28] J. Sinova, S. O. Valenzuela, J. Wunderlich, C. H. Back, T. Jungwirth. Spin Hall effects. Rev. Mod. Phys., 2015, 87(4): 1213

[29] J. Sinova, D. Culcer, Q. Niu, N. A. Sinitsyn, T. Jungwirth, A. H. MacDonald. Universal intrinsic spin Hall effect. Phys. Rev. Lett., 2004, 92(12): 126603

[30] Y. Yang, Z. Xu, L. Sheng, B. Wang, D. Y. Xing, D. N. Sheng. Time-reversal-symmetry-broken quantum spin Hall effect. Phys. Rev. Lett., 2011, 107(6): 066602

[31] Y. Sun, Y. Zhang, C. Felser, B. Yan. Strong intrinsic spin Hall effect in the TaAs family of Weyl semimetals. Phys. Rev. Lett., 2016, 117(14): 146403

[32] H. J. Zhao, H. Nakamura, R. Arras, C. Paillard, P. Chen, J. Gosteau, X. Li, Y. R. Yang, L. Bellaiche. Purely cubic spin splittings with persistent spin textures. Phys. Rev. Lett., 2020, 125(21): 216405

[33] I. A. Nechaev, E. E. Krasovskii. Spin polarization by first-principles relativistic k⋅p theory: Application to the surface alloys PbAg₂ and BiAg₂. Phys. Rev. B, 2019, 100(11): 115432

[34] P. Höpfner, J. Schafer, A. Fleszar, J. H. Dil, B. Slomski, F. Meier, C. Loho, C. Blumenstein, L. Patthey, W. Hanke, R. Claessen. Three-dimensional spin rotations at the Fermi surface of a strongly spin‒orbit coupled surface system. Phys. Rev. Lett., 2012, 108(18): 186801

[35] J. Sinova, S. O. Valenzuela, J. Wunderlich, C. H. Back, T. Jungwirth. Spin Hall effects. Rev. Mod. Phys., 2015, 87(4): 1213

[36] J. Wunderlich, B. Kaestner, J. Sinova, T. Jungwirth. Experimental observation of the spin-Hall effect in a two-dimensional spin‒orbit coupled semiconductor system. Phys. Rev. Lett., 2005, 94(4): 047204

[37] E. Saitoh, M. Ueda, H. Miyajima, G. Tatara. Conversion of spin current into charge current at room temperature: Inverse spin-Hall effect. Appl. Phys. Lett., 2006, 88(18): 182509

[38] J. Schliemann, D. Loss. Spin-Hall transport of heavy holes in III‒V semiconductor quantum wells. Phys. Rev. B, 2005, 71(8): 085308

[39] C. M. Acosta, A. Fazzio. Spin-polarization control driven by a Rashba-type effect breaking the mirror symmetry in two-dimensional dual topological insulators. Phys. Rev. Lett., 2019, 122(3): 036401

[40] S. D. Stolwijk, K. Sakamoto, A. B. Schmidt, P. Kruger, M. Donath. Spin texture with a twist in momentum space for Tl/Si(111). Phys. Rev. B, 2015, 91(24): 245420

[41] L. Fu. Hexagonal warping effects in the surface states of the topological insulator Bi₂Te₃. Phys. Rev. Lett., 2009, 103(26): 266801

[42] S. Vajna, E. Simon, A. Szilva, K. Palotas, B. Ujfalussy, L. Szunyogh. Higher-order contributions to the Rashba‒Bychkov effect with application to the Bi/Ag(111) surface alloy. Phys. Rev. B, 2012, 85(7): 075404

[43] M. Michiardi, M. Bianchi, M. Dendzik, J. A. Miwa, M. Hoesch, T. K. Kim, P. Matzen, J. L. Mi, M. Bremholm, B. B. Iversen, P. Hofmann. Strongly anisotropic spin‒orbit splitting in a two-dimensional electron gas. Phys. Rev. B, 2015, 91(3): 035445

[44] S. Bandyopadhyay, A. Paul, I. Dasgupta. Origin of Rashba‒Dresselhaus effect in the ferroelectric nitride perovskite LaWN₃. Phys. Rev. B, 2020, 101(1): 014109

[45] M. S. Bahramy, B. J. Yang, R. Arita, N. Nagaosa. Emergence of non-centrosymmetric topological insulating phase in BiTeI under pressure. Nat. Commun., 2012, 3(1): 679

[46] R. Moriya, K. Sawano, Y. Hoshi, S. Masubuchi, Y. Shiraki, A. Wild, C. Neumann, G. Abstreiter, D. Bougeard, T. Koga, T. Machida. Cubic Rashba spin‒orbit interaction of a two-dimensional hole gas in a strained-Ge/SiGe quantum well. Phys. Rev. Lett., 2014, 113(8): 086601

[47] L. G. Gerchikov, A. V. Subashiev. Spin splitting of size-quantization subbands in asymmetric heterostructures. Sov. Phys. Semicond., 1992, 26: 73

[48] O. Bleibaum, S. Wachsmuth. Spin Hall effect in semiconductor heterostructures with cubic Rashba spin‒orbit interaction. Phys. Rev. B, 2006, 74(19): 195330

[49] K. V. Shanavas. Theoretical study of the cubic Rashba effect at the SrTiO₃ (001) surfaces. Phys. Rev. B, 2016, 93(4): 045108

[50] R. Arras, J. Gosteau, H. J. Zhao, C. Paillard, Y. Yang, L. Bellaiche. Rashba-like spin‒orbit and strain effects in tetragonal PbTiO₃. Phys. Rev. B, 2019, 100(17): 174415

[51] L. G. D. da Silveira, P. Barone, S. Picozzi. Rashba‒Dresselhaus spin-splitting in the bulk ferroelectric oxide BiAlO₃. Phys. Rev. B, 2016, 93(24): 245159

[52] D. C. Marinescu. Cubic Dresselhaus interaction parameter from quantum corrections to the conductivity in the presence of an in-plane magnetic field. Phys. Rev. B, 2017, 96(11): 115109

[53] M. Glazov, A. Kavokin. Spin Hall effect for electrons and excitons. J. Lumin., 2007, 125(1−2): 118

[54] M. Seemann, D. Kodderitzsch, S. Wimmer, H. Ebert. Symmetry-imposed shape of linear response tensors. Phys. Rev. B, 2015, 92(15): 155138

[55] S. V. Gallego, J. Etxebarria, L. Elcoro, E. S. Tasci, J. M. Perez-Mato. Automatic calculation of symmetry-adapted tensors in magnetic and non-magnetic materials: A new tool of the Bilbao crystallographic server. Acta Crystallogr. A Found. Adv., 2019, 75(3): 438

[56] Y. J. Lin, K. Jiménez-García, I. B. Spielman. Spin–orbit-coupled Bose–Einstein condensates. Nature, 2011, 471(7336): 83

[57] G. Orso. Anderson transition of cold atoms with synthetic spin‒orbit coupling in two-dimensional speckle potentials. Phys. Rev. Lett., 2017, 118(10): 105301

[58] H. Zhai. Spin‒orbit coupled quantum gases. Int. J. Mod. Phys. B, 2012, 26(1): 1230001

[59] H. Yuan, M. S. Bahramy, K. Morimoto, S. Wu, K. Nomura, B. J. Yang, H. Shimotani, R. Suzuki, M. Toh, C. Kloc, X. Xu, R. Arita, N. Nagaosa, Y. Iwasa. Zeeman-type spin splitting controlled by an electric field. Nat. Phys., 2013, 9(9): 563

[60] The suppression of linear Rashba SOI was discussed in Ref. [49]. A transition from the cubic Rashba effect to the coexistence of linear and cubic Rashba effects was observed experimentally in oxide heterostructures [61]. For heavy holes in III−V semiconductor quantum wells, the linear Rashba SOI can be absent making the cubic SOI as the leading order [38, 62, 63].

[61] W. Lin, L. Li, F. Dŏgan, C. Li, H. Rotella, X. Yu, B. Zhang, Y. Li, W. S. Lew, S. Wang, W. Prellier, S. J. Pennycook, J. Chen, Z. Zhong, A. Manchon, T. Wu. Interface-based tuning of Rashba spin‒orbit interaction in asymmetric oxide heterostructures with 3d electrons. Nat. Commun., 2019, 10(1): 3052

[62] R. Winkler, H. Noh, E. Tutuc, M. Shayegan. Anomalous Rashba spin splitting in two-dimensional hole systems. Phys. Rev. B, 2002, 65(15): 155303

[63] K. Nomura, J. Wunderlich, J. Sinova, B. Kaestner, A. H. Mac-Donald, T. Jungwirth. Edge-spin accumulation in semiconductor two-dimensional hole gases. Phys. Rev. B, 2005, 72(24): 245330

[64] Y. K. Kato, R. C. Myers, A. C. Gossard, D. D. Awschalom. Observation of the spin Hall effect in semiconductors. Science, 2004, 306(5703): 1910

[65] S. Basak, H. Lin, L. A. Wray, S. Y. Xu, L. Fu, M. Z. Hasan, A. Bansil. Spin texture on the warped Dirac-cone surface states in topological insulators. Phys. Rev. B, 2011, 84: 121401(R)

[66] D. L. Campbell, G. Juzeliunas, I. B. Spielman. Realistic Rashba and Dresselhaus spin‒orbit coupling for neutral atoms. Phys. Rev. A, 2011, 84(2): 025602

RIGHTS & PERMISSIONS

Higher Education Press