Planar Hall Plateau in Magnetic Weyl Semimetals

Abstract

Despite the rapid progress in the study of planar Hall effect (PHE) in recent years, all the previous works only showed that the PHE is connected to local geometric quantities, such as Berry curvature. Here, for the first time, we point out that the PHE in magnetic Weyl semimetals is directly related to a global quantity, namely, the Chern number of the Weyl point. This leads to a remarkable consequence that the PHE observation predicted here is robust against many system details, including the Fermi energy. The main difference between non-magnetic and magnetic Weyl points is that the latter breaks time-reversal symmetry ${\cal{T}}$, thus generally possessing an energy tilt. Via semiclassical Boltzmann theory, we investigate the PHE in generic magnetic Weyl models with energy tilt and arbitrary Chern number. We find that by aligning the magnetic and electric fields in the same direction, the trace of the PHE conductivity contributed from Berry curvature and orbital moment is proportional to the Chern number and the energy tilt of the Weyl points, resulting in previously undiscovered quantized PHE plateau by varying Fermi energy. We further confirm the existence of PHE plateaus in a more realistic lattice model without ${\cal{T}}$ symmetry. By proposing a new quantized physical quantity, our work not only provides a new tool for extracting the topological character of the Weyl points but also suggests that the interplay between topology and magnetism can give rise to intriguing physics.

Introduction.— The band topology, referring to the topological properties of the electronic band structure of materials, have attracted considerable attention in the field of condensed matter physics [1, 2]. Generally, one can utilize the geometric quantities, which are locally defined at each momentum point in the Brillouin zone (BZ), to describe it. Some common geometric quantities include Berry curvature and its dipole [1, 3, 4, 5, 6, 7], Berry-connection polarizability and its dipole [8, 9, 10, 11, 12, 13], and quantum metric [14, 15, 16, 17, 18]. However, the essential aspect of topological states is that they host (at least) a global quantity, i.e. topological charge [19, 20, 21, 22, 23], which also is used to characterize and distinguish topological states. For example, the integral of the Berry curvature over a 2D closed surface in momentum space may result in Chern number [24]. More importantly, in stark contrast to local geometric quantities and general physical observations, global topological quantities are integers and are robust to many material details [25, 26].

Recently, various Hall effects, including nonlinear Hall effects [27, 8, 9, 28, 29, 30, 10, 31, 11, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44] and many kinds of planar Hall effect (PHE) [45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56], have been proposed and demonstrated to have a close relationship with geometric quantities. Since the geometric quantities are significantly enhanced around band degeneracies, these unconventional Hall effects are widely investigated in topological semimetals [57, 58, 59, 60, 61, 62, 63, 64, 65, 66]. As aforementioned, the existence of a global quantity rather than local geometric quantities is the most remarkable feature of topological states. Unfortunately, the PHEs have not yet been reported to be associated with a global quantity.

Figure 1: Sketch of PHE in topological Weyl semimetals with magnetic and electric fields aligned in the same direction. In this case, the trace of the PHE conductivity ${\rm Tr}[\chi_{i}]=\sum_{j}\chi_{i;jj}$ is in direct proportion to the Chern number of the WP, and is robust against many system details, leading to a flat plateau at low energy.

In this work, we show that the PHE in magnetic Weyl semimetals is directly linked to the topological charge (Chern number) of the Weyl point. In PHE, the magnetic field, the driving electric field, and the transverse Hall current are in the same plane [58, 67], as illustrated in Fig. 1, which is completely different from the setup in ordinary Hall effect. Initially, PHE was attributed to system anisotropy [45]. Subsequent findings unveiled that the Berry curvature also affects PHE, leading to a surge in research on PHE in topological Weyl semimetals [68, 69, 70, 71, 72, 73, 74, 75]. Recently, Li et al. [67] found that the PHE resulting from orbital moment is comparable with that from Berry curvature. However, a common observation is that the PHE mechanisms in these studies are independent of any global quantity.

Here, based on semiclassical Boltzmann theory, we investigate PHE in the charge-$n$ (with $n$ an integer) Weyl points (C-$n$ WPs) with energy tilt, which is a common feature for magnetic WPs. For magnetic WPs, the leading order of PHE generally is linear due to the breaking of time reversal symmetry (${\cal{T}}$) [67]. The linear PHE conductivity contains six terms, with one term contributed by Lorentz force and the remaining terms originating from geometric quantities, i.e. Berry curvature and orbital moment. Remarkably, by aligning the magnetic and electric fields in the same direction (see Fig. 1), we demonstrate that the trace of the PHE conductivity contributed by Berry curvature and orbital moment (Tr[$\chi_{i}$] with $i=x,y,z$) is quantized for all the C-$n$ WPs resulting from the global topology of the WP, and then is robust against many material details. The analytical derivation of Tr[$\chi_{i}$] shows that the Tr[$\chi_{i}$] is fully determined by the Chern number and the energy tilt of the WPs. This feature becomes increasingly more precise when the Fermi energy approaches the WPs. By unveiling a novel magnetotransport phenomenon directly related to the global quantities of WP in magnetic Weyl semimetals, our work deepens the understanding on the PHE, and provides a promising method to detect the topological charge and the energy tilt of the WPs.

Linear PHE conductivity.— Generally, the PHE conductivity under weak fields can be obtained from semiclassical equations of motion [1] and the Boltzmann transport equation [76]. Here, we focus on the setup that the driving electric field $\bm{E}$ and the magnetic field $\bm{B}$ are aligned in the same direction, i.e., ${\bm{E}}\|{\bm{B}}$ (see Fig. 1). For the three-dimensional magnetic systems, the leading terms of the PHE that have topological origin are linearly dependent on both $\bm{E}$ and $\bm{B}$, expressed as [67]

$$J_{i}=\chi_{i;jj}B_{j}E_{j},$$ (1)

where $\bm{J}$ is the current density and $\chi_{i;jj}$ denotes the PHE conductivity. The PHE conductivity $\chi_{i;jk}$ is a third-order tensor, and the $\chi_{i;jj}$ in Eq. (1) denotes a diagonal element of the matrix $\chi_{i}$. $\chi_{i;jj}$ can be further decomposed into the following five parts [67]:

$$\chi_{i;jj}=\chi_{i;jj}^{(1)}+\chi_{i;jj}^{(2)}+\chi_{i;jj}^{(3)}+\chi_{i;jj}^{(4)}+\chi_{i;jj}^{(5)},$$ (2)

with

$$\chi_{i;jj}^{(1)} = -e^{3}\tau\int[d\bm{k}]v_{i}v_{j}\Omega_{j}\partial_{\varepsilon}f_{k}^{0},$$

$$\chi_{i;jj}^{(2)} = e^{3}\tau\int[d\bm{k}]v_{j}\left(\bm{v}\cdot\bm{\Omega}\right)\partial_{\varepsilon}f_{k}^{0},$$

$$\chi_{i;jj}^{(3)} = e^{3}\tau\int[d\bm{k}]v_{i}\left(\bm{v}\cdot\bm{\Omega}\right)\partial_{\varepsilon}f_{k}^{0},$$

$$\chi_{i;jj}^{(4)} = -e^{2}\tau\int[d\bm{k}]v_{i}\partial_{k_{j}}m_{j}\partial_{\varepsilon}f_{k}^{0},$$

$$\chi_{i;jj}^{(5)} = -e^{2}\tau\int[d\bm{k}]v_{j}\partial_{k_{i}}m_{j}\partial_{\varepsilon}f_{k}^{0}.$$ (3)

Here, $\int[d\bm{k}]\equiv-\frac{1}{(2\pi)^{3}\hbar}\int d^{3}k$, $\varepsilon$ is the band energy, $\bm{v}=\partial_{\bm{k}}\varepsilon/\hbar$ denotes the velocity, $\bm{\Omega}$ and $\bm{m}$ respectively are the Berry curvature and the orbital moment, $\tau$ is the relaxation time, and $f_{k}^{0}$ is the Fermi-Dirac distribution. The electron charge is taken as $-e$ (i.e., , $e>0$).

An analysis of Eq. (Planar Hall Plateau in Magnetic Weyl Semimetals) reveals two crucial features of the topological PHE conductivities $\chi_{i;jj}$. First, all the five PHE conductivities linearly depend on $\tau$. Thus, they can be easily extracted in experiments [77, 78], as the contributions from intrinsic Hall effect and Lorentz force are even functions of $\tau$ [67]. This linear dependence also indicates that the topological PHE conductivities only appear in the magnetic systems, as ${\cal{T}}$ reverses ${\bm{J}}$, ${\bm{B}}$ and $\tau$, but keeps ${\bm{E}}$ invariant.

Second, one observes that the trace of the matrix $\chi_{i}^{(1)}$ equals to $-\chi_{i;jj}^{(3)}$, i.e., ${\rm Tr}[\chi_{i}^{(1)}]=\sum_{j}\chi_{i;jj}^{(1)}=-\chi_{i;jj}^{(3)}$, and ${\rm Tr}[\chi_{i}^{(3)}]=3\chi_{i;jj}^{(3)}$. Thus, the trace of $\chi_{i}$ is

$${\rm Tr}[\chi_{i}]\equiv\sum_{j}\chi_{i;jj}={\rm Tr}[\chi_{i}^{(2)}+\frac{2}{3}\chi_{i}^{(3)}+\chi_{i}^{(4)}+\chi_{i}^{(5)}],$$ (4)

in which the contribution from Berry curvature and orbital moment are ${\rm Tr}[\chi_{i}^{\Omega}]={\rm Tr}[\chi_{i}^{(2)}+\frac{2}{3}\chi_{i}^{(3)}]$ and ${\rm Tr}[\chi_{i}^{m}]={\rm Tr}[\chi_{i}^{(4)}+\chi_{i}^{(5)}]$, respectively.

Third, the topological PHE conductivities $\chi_{i;jj}$ are determined by the properties of the Fermi surface, due to the presence of $\partial_{\varepsilon}f_{k}^{0}$. It has been demonstrated that the geometry quantities at the Fermi surface have a significant influence on the PHE conductivities [67, 68, 69]. However, the Fermi surface may also possess global quantities. The influence of the global quantities on the PHE conductivities has not been investigated. Since the most famous global quantity that may be hosted by a 2D Fermi surface is the Chern number (${\cal{C}}$), here we focus on the topological PHE conductivities $\chi_{i;jj}$ of generic magnetic WPs with different Chern numbers.

Linear WP and scaling analysis.— Before studying a generic Weyl model with arbitrary Chern number, it is instructive to begin with a conventional (linear) WP with $|{\cal{C}}|=1$, which has a linear dispersion along all the directions in momentum space [79]. Due to the absence of ${\cal{T}}$ symmetry, the magnetic WP generally has an energy tilt. For linear WP, the leading term for the energy tilt is also linear. Thus, a general Hamiltonian for the magnetic linear WP can be written as [79]

$$\mathcal{H}_{1}=w_{x}k_{x}+v_{1}\bm{k}\cdot\bm{\sigma},$$ (5)

with $w_{x}$ and $v_{1}$ the real parameters respectively representing the energy tilt and Fermi velocity of the WP, and $\sigma$ being the Pauli matrix. Here, without loss of generality, we have set the energy tilt to be along $k_{x}$ direction. We also assume $|w_{x}|<|v_{1}|$ to guarantee the Fermi surface of the WP close [80, 81]. It is easy to know that this model (5) exhibits a linear dispersion and exhibits a Chern number ${\cal{C}}={\rm Sign}(v_{1})$.

Figure 2: Variation of PHE conductivity with (a) Fermi level $E_{F}$ and (b) Fermi velocity $v_{1}$. In panel (a), the red and blue lines correspond to the $\chi_{x;yy}^{(3)}$ and $\chi_{x;xx}^{(4)}$, respectively. Solid lines represent results for $w_{x}>0$ and dashed lines for $w_{x}<0$. We set $|w_{x}|=0.2\ \text{eV}\cdot\text{\AA}$, and $v_{1}=1.0\ \text{eV}\cdot\text{\AA}$ in (a), and set $w_{x}=0.2\ \text{eV}\cdot\text{\AA}$ and $E_{F}=200~{}\rm{meV}$ in (b). We also set $\tau=0.1~{}\rm{ps}$ in all calculations.

A striking feature of the system that exhibits same order of dispersion along all the directions is that its various important properties can be obtained by applying a simple scaling analysis [82]. For this model (5), we consider a scaling transformation in momentum and energy: $\bm{k}\rightarrow\lambda\bm{k}$ and $E_{F}\rightarrow\lambda E_{F}$, where $\lambda$ is a real number. Due to the linear feature of Eq. (5), one has ${\cal{H}}_{1}(\lambda\bm{k})=\lambda{\cal{H}}_{1}(\bm{k})$ and $\varepsilon(\lambda\bm{k})=\lambda\varepsilon(\bm{k})$. Note that the velocity and the eigenstates of the system are invariant under the scaling, i.e., $v(\lambda\bm{k})=v(\bm{k})$ and $|u_{n,\lambda\bm{k}}\rangle=|u_{n,\bm{k}}\rangle$. Then, the Berry curvature and orbital moment transform as ${\bm{\Omega}_{\lambda\bm{k}}}=\lambda^{-2}{\bm{\Omega}_{\bm{k}}}$ and ${\bm{m}}_{\lambda\bm{k}}=\lambda^{-1}{\bm{m}}_{\bm{k}}$, respectively.

Figure 3: Variation of PHE conductivity (a) $\mathrm{Tr}[\chi^{\Omega}_{x}]$ and (b) $\mathrm{Tr}[\chi^{m}_{x}]$ with Fermi level for $n=1,2,3$ when $v_{1}>0$. The corresponding value of $\eta$ and $\beta$ in Table 1 are $\eta=0.474~{}\rm{S\cdot\AA\cdot cm^{-1}\cdot eV^{-1}\cdot T^{-1}}$ and $\beta=0.112$. (c) Variation of $\mathrm{Tr}[\chi^{\Omega}_{x}]$ and $\mathrm{Tr}[\chi^{m}_{x}]$ with Fermi level for $n=2$ when $v_{1}<0$. The corresponding $\beta=0.143$. (d) Variation of $\mathrm{Tr}[\chi^{\Omega}_{x}]$ and $\mathrm{Tr}[\chi^{m}_{x}]$ with energy tilt $w_{x}$ for $n=3$ when $v_{1}>0$. We set $w_{x}=0.5\ \text{eV}\cdot\text{\AA}$ in (a-c), $v_{1}=1.0\ \text{eV}\cdot\text{\AA}$, $c_{+}=0.3\ \text{eV}\cdot\text{\AA}^{n}$, and $c_{-}=1.2\ \text{eV}\cdot\text{\AA}^{n}$ in (a), (b) and (d); $v_{1}=-0.9\ \text{eV}\cdot\text{\AA}$, $c_{+}=0.4\ \text{eV}\cdot\text{\AA}^{n}$ and $c_{-}=1.0\ \text{eV}\cdot\text{\AA}^{n}$ in (c). We also set $E_{F}=200~{}\text{meV}$ in (d) and $\tau=0.1~{}\rm{ps}$ in all calculations.

As a consequence, for all the five PHE conductivities in Eq. (Planar Hall Plateau in Magnetic Weyl Semimetals), we have

$$\chi_{i;jj}^{(l)}(\lambda E_{F})=\chi_{i;jj}^{(l)}(E_{F}),$$ (6)

indicating that the $\chi_{i;jj}^{(l)}$ ($l=1,\cdots,5$) is a quantity that surprisingly has no dependence on Fermi energy, leading to a PHE plateau in energy space, as shown in Fig. 2(a). Further numerical calculations show that the trace of the PHE conductivity ${\rm Tr}[\chi_{i}^{(l)}]$ is also robust to the absolute value of $v_{1}$, and is only dependent on its sign [see Fig. 2(b)], which is reminiscent of the Chern number of the WP ${\cal{C}}={\rm Sign}(v_{1})$. Notice that in the field of condensed matter physics, it is quite rare for an observable to have such a robust feature. All these results indicate that the PHE conductivity ${\rm Tr}[\chi_{i}^{(l)}]$ in WP should be more nontrivial than previously understood. Besides, from symmetry analysis, one knows that the value of the PHE plateau would be reversed when the sign of $w_{x}$ is changed, as both $w_{x}$ and $\chi_{i,jj}^{(l)}$ are anti-symmetric under ${\cal{T}}$ symmetry. This is confirmed by our numerical calculations, as shown in Fig. 2(a).

Beyond linear model.— We then discuss the trace of PHE conductivity ${\rm Tr}[\chi_{i}]$ in a generic magnetic WP with arbitrary Chern number, for which the Hamiltonian reads [83, 84],

$$\mathcal{H}_{2}=w_{x}k_{x}+\left[\begin{array}[]{cc}v_{1}k_{x}&c_{+}k_{+}^{n}+c_{-}k_{-}^{n}\\ c_{+}k_{-}^{n}+c_{-}k_{+}^{n}&-v_{1}k_{x}\end{array}\right],$$ (9)

with $k_{\pm}=k_{y}\pm ik_{z}$, $n$ an integer (which can be 1, 2, and 3 in crystals), and $c_{\pm}$ a real parameter. The Chern number of this WP is ${\cal{C}}=n\times{\rm Sign}(v_{1})$ for $|c_{+}|>|c_{-}|$ and is ${\cal{C}}=-n\times{\rm Sign}(v_{1})$ for $|c_{+}|<|c_{-}|$. Again, we require the WP to be a type-I point with $|w_{x}|<|v_{1}|$. For $n\neq 1$, the WP has a linear dispersion along $k_{x}$, but a nonlinear dispersion along other directions. Thus, it can not be investigated by scaling analysis.

The numerical results of ${\rm Tr}[\chi_{i}^{\Omega(m)}]$ based on Hamiltonian $\mathcal{H}_{2}$ (9) are presented in Fig. 3. A remarkable observation is that while $\chi_{i,jj}^{\Omega(m)}$ is no longer independent of Fermi energy, due to the loss of linear dispersion, the trace, i.e. ${\rm Tr}[\chi_{i}^{\Omega(m)}]$ remain robust against $E_{F}$, as shown in Fig. 3(a-c). More than that, the plateau of ${\rm Tr}[\chi_{i}^{\Omega(m)}]$ is completely not effected by the nonlinear dispersion and the band anisotropy induced by the parameters $c_{\pm}$ (see Fig. 3). Again, we find that the PHE plateau is sensitive to the value of $n$ and the sign of $v_{1}$, i.e., the Chern number of the WP. Moreover, we find that ${\rm Tr}[\chi_{i}^{\Omega}]$ linearly depend on the energy tilt $w_{x}$, while ${\rm Tr}[\chi_{i}^{m}]$ shows a linear dependence on $w_{x}$ when $w_{x}$ is small, but increasingly deviates it when $w_{x}$ increases, as shown in Fig. 3(d). All these results strongly suggest that the trace of the PHE conductivity ${\rm Tr}[\chi_{i}^{\Omega(m)}]$ has a deeper physical origin beyond local geometric quantities.

Table 1: The PHE conductivity ${\rm Tr}[\chi_{i}^{(l)}]$ of a generic WP model (9) where the energy tilt is along $k_{x}$ direction. Here, $\mathcal{C}$ is the Chern number of the WP, $\eta=e^{3}\tau/(12\pi^{2}\hbar^{3})$, and $\beta=[-2\alpha^{3}+3\alpha+3(\alpha^{2}-1)\operatorname{arctanh}\alpha]/\alpha^{3}$. $w_{x}$ and $\alpha=|w_{x}/v_{1}|<1$ denote the absolute and relative energy tilt, respectively.

	Berry curvature			orbital moment
	$l$ = 1	$l$ = 2	$l$ = 3	$l$ = 4	$l$ = 5
${\rm Tr}[\chi_{x}^{(l)}]/\eta$	$-3w_{x}\mathcal{C}$	$3w_{x}\mathcal{C}$	$9w_{x}\mathcal{C}$	$(-2-\beta)w_{x}\mathcal{C}$	$(-2-\beta)w_{x}\mathcal{C}$
${\rm Tr}[\chi_{y}^{(l)}]/\eta$	0	$3w_{x}\mathcal{C}$	0	0	0
${\rm Tr}[\chi_{z}^{(l)}]/\eta$	0	$3w_{x}\mathcal{C}$	0	0	0

Analytical expression of ${\rm Tr}[\chi_{i}]$.—Next, we provide a rigorous explanation for the novel and intriguing results obtained above. Due to the presence $\partial_{\varepsilon}f_{k}^{0}$, at zero temperature the integral in Eq. (Planar Hall Plateau in Magnetic Weyl Semimetals) actually is performed on the Fermi surface. Since ${\rm Tr}[\chi_{i}^{(1)}]=-\chi_{i;jj}^{(3)}=-{\rm Tr}[\chi_{i}^{(3)}]/3$, we start our analysis from ${\rm Tr}[\chi_{i}^{(3)}]$, which can be rewritten as

$${\rm Tr}[\chi_{i}^{(3)}]=3\frac{e^{3}\tau}{(2\pi)^{3}\hbar^{2}}\oint_{\varepsilon=E_{F}}v_{i}\Omega_{\perp}dS,$$ (10)

where $S$ denotes the Fermi surface, the subscript $\perp$ represents the normal component of the corresponding quantities with respect to $S$, and the equations $\bm{v_{k}}\cdot\bm{\Omega_{k}}=v_{\perp}\Omega_{\perp}$ and $v_{\perp}=\partial_{k_{\perp}}\varepsilon/\hbar$ are used. For model (9), its dispersion is $\varepsilon=w_{x}k_{x}+\varepsilon_{0}$ with $\varepsilon_{0}$ denotes the dispersion without energy tilt. Then, one has $v_{x}=w_{x}/\hbar+v_{x,0}$ with $v_{x,0}=\partial_{k_{x}}\varepsilon_{0}/\hbar$. Interestingly, the $x$-component of ${\rm Tr}[\chi^{(3)}]$ becomes [see the Supplementary Material [85] for details]

$${\rm Tr}[\chi_{x}^{(3)}] =\frac{3e^{3}\tau}{(2\pi)^{3}\hbar^{3}}(w_{x}\oint\Omega_{\perp}dS+\oint\hbar v_{x,0}\Omega_{\perp}dS)$$

$$=9\eta w_{x}\mathcal{C},$$ (11)

with $\eta=\frac{e^{3}\tau}{12\pi^{2}\hbar^{3}}$ and $\mathcal{C}$ the Chern number of the 2D Fermi surface, which is exactly the Chern number of the WP. Since the energy tilt along $k_{y(z)}$ is zero, one has ${\rm Tr}[\chi_{y}^{(3)}]={\rm Tr}[\chi_{z}^{(3)}]=0$ [85]. Besides, from Eq. Planar Hall Plateau in Magnetic Weyl Semimetals, we have ${\rm Tr}[\chi^{(2)}_{i}]=({\rm Tr}[\chi_{x}^{(3)}]+{\rm Tr}[\chi_{y}^{(3)}]+{\rm Tr}[\chi_{z}^{(3)}])/3$, indicating that ${\rm Tr}[\chi_{x/y/z}^{(2)}]=3\eta w_{x}\mathcal{C}$. Therefore, the PHE conductivity contributed by Berry curvature is

$${\rm Tr}[\chi_{x}^{\Omega}] = 9\eta w_{x}\mathcal{C},$$ (12)

$${\rm Tr}[\chi_{y}^{\Omega}] = {\rm Tr}[\chi_{z}^{\Omega}]=3\eta w_{x}\mathcal{C}.$$ (13)

A more complicated calculation demonstrates that ${\rm Tr}[\chi_{i}^{(4)}]={\rm Tr}[\chi_{i}^{(5)}]$, and the PHE conductivity contributed by orbital moment is [85]

$${\rm Tr}[\chi_{x}^{m}] = -(4+2\beta)\eta w_{x}\mathcal{C},$$ (14)

$${\rm Tr}[\chi_{y}^{m}] = {\rm Tr}[\chi_{z}^{m}]=0,$$ (15)

with $\beta=[-2\alpha^{3}+3\alpha+3(\alpha^{2}-1)\operatorname{arctanh}\alpha]/\alpha^{3}$ and $\alpha=|w_{x}/v_{1}|<1$ denoting the relative energy tilt. Obviously, $\beta$ represents the deviation of ${\rm Tr}[\chi^{(m)}_{x}]$ from a simple linear relationship with $w_{x}{\cal{C}}$. It rapidly approaches zero as $w_{x}$ decreases, which can be inferred from the alternative expression of $\beta$: $\beta=\sum_{n=1}^{\infty}\frac{6\alpha^{2n}}{4n^{2}+8n+3}$. The Eqs. (12-15) are the main results of this work, and is summarized in Table 1. These results directly show that the PHE is highly nontrivial, and originates from the global topological quantities of the Fermi surface. Consequently, in contrast to various Hall effects, the PHE conductivity in magnetic Weyl semimetals are robust against many kinds of small perturbations, as shown in Fig. 3.

Figure 4: (a) Band structure of the Weyl model (Planar Hall Plateau in Magnetic Weyl Semimetals). (b) Variation of the PHE conductivity with Fermi energy where the corresponding $\beta=0.069$. We set $\tau=0.1~{}\rm{ps}$, $m=t=1.5\ \text{eV}\cdot\text{\AA}$, $w_{x}=0.6\ \text{eV}\cdot\text{\AA}$, $\Delta=0.06\ \text{eV}\cdot\text{\AA}$ and $k_{0}=\pi/2\ \text{\AA}^{-1}$.

Lattice model.—Our results so far are obtained on the effective model around a WP. We point out that the PHE plateau can persist in a realistic lattice model, and is almost precise when the Fermi energy is around WPs. Consider a minimal lattice models for topological magnetic Weyl semimetal, expressed as [86]

$$\mathcal{H}_{3}= \left[m\left(\cos k_{y}+\cos k_{z}-2\right)-t\left(\cos k_{x}-\cos k_{0}\right)\right]\sigma_{x}$$

$$-t\sin k_{y}\sigma_{y}-t\sin k_{z}\sigma_{z}+w_{x}\left(\cos k_{0}-\cos k_{x}\right)$$

$$+\Delta\sin k_{x},$$ (16)

where $m$, $t$ and $k_{0}$ are model parameters, and $w_{x}$ denotes the energy tilt. The band structure of the Weyl model (Planar Hall Plateau in Magnetic Weyl Semimetals) is plotted in Fig. 4(a), in which two tilted linear WPs with $|{\cal C}=1|$ at $(\pm k_{0},0,0)$ position can be observed when $w_{x}$ is finite. Since the two WPs have both opposite energy tilt and Chern number, one can expect that the total PHE conductivity ${\rm Tr}[\chi_{i}](\propto w_{x}{\cal C})$ from the two WPs would not be canceled but is added.

The numerically calculated PHE conductivities of model (Planar Hall Plateau in Magnetic Weyl Semimetals) as a function of $E_{F}$ are plotted in Fig. 4(b). These conductivities continue to display a flat plateau and preserve their mutual proportional relationships, as revealed in Eqs. (12-15) and Table 1. Moreover, the plateau values have now doubled due to the presence of two WPs.

Conclusions.— In conclusion, we demonstrate that the PHE conductivity can be directly related to the topological charge of WPs, and then features quantized plateau by varying many material parameters. Our work significantly advances the understanding of magnetotransport in magnetic Weyl semimetals, highlighting the crucial role of Chern number and energy tilt of the WPs, and providing an attractive prospect for future explorations on quantized Hall transport phenomena in topological materials.