This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

Anomalous photon thermal Hall effect

A. Ott Institut für Physik, Carl von Ossietzky Universität, D-26111 Oldenburg, Germany.    S.-A. Biehs [email protected] Institut für Physik, Carl von Ossietzky Universität, D-26111 Oldenburg, Germany.    P. Ben-Abdallah [email protected] Laboratoire Charles Fabry, UMR 8501, Institut d’Optique, CNRS, Université Paris-Saclay, 2 Avenue Augustin Fresnel, 91127 Palaiseau Cedex, France.
(March 3, 2025)
Abstract

We predict an anomalous thermal Hall effect (ATHE) mediated by photons in networks of Weyl semi-metals. Contrary to the photon thermal Hall effect in magneto-optical systems which requires the application of an external magnetic field the ATHE in a Weyl semi-metals network is an intrinsic property of these systems. Since the Weyl semi-metals can exhibit a strong nonreciprocal response in the infrared over a broad spectral range the magnitude of thermal Hall flux in these systems can be relatively large compared to the primary flux. This ATHE paves the way for a directional control of heat flux by localy tuning the magnitude of temperature field without changing the direction of temperature gradient.

pacs:
44.40.+a, 78.20.N-, 03.50.De, 66.70.-f

The classical Hall effect Hall results in the appearance of a transverse electric current inside a conductor under the action of an external magnetic field applied in the direction orthogonal to the primary bias voltage. This effect stems from the Lorentz force which acts transversally on the electric charges in motion through the magnetic field curving so their trajectories. Very shortly after this discovery, a thermal analog of this effect has been observed by Righi and Leduc Leduc when a temperature gradient is applied throughout an electric conductor. As for the classical Hall effect, this effect is intrinsically related to the presence of free electric charges. Hence, one would not expect a thermal Hall effect with neutral particles or quasiparticles. Nevertheless, during the last decade researchers have highlighted such an effect in non-conducting materials due to phonons Strohm ; Inyushkin , magnons (spin waves) Fujimoto ; Katsura ; Onose and even photons PBA_PRL2016 ; Ott_JPE in non-reciprocal many-body systems due to different mechanisms of local broken symmetry induced by application of an external magnetic field. Beside these ’normal’ Hall effects, anomalous effects AHE_RMP ; Karplus have been predicted in solids such as ferromagnets without external field application. In these media an intrinsic mechanism (a Berry curvature acting as a fictitious field on electrons, a skew scattering that is an asymmetric impurity scattering or a spin-orbit coupling) is responsible for the local symmetry breaking which gives rise to a Hall current. More recently thermal analogs of this effect, also called anomalous thermal Hall effect Ferreiros ; Sugii ; Huang2 have been measured in these media.

Refer to caption
Figure 1: Four terminal junction with particles made of WSM forming a square with C4C_{4} symmetry. To measure the photon thermal Hall flux a temperature gradient ΔT\Delta T is applied between particles 11 and 22 along the 𝐱\mathbf{x}-axis and the Hall flux is evaluated by measuring the temperature difference T3(st)T4(st)T_{3}^{({\rm st})}-T_{4}^{({\rm st})} in the transveral direction along the 𝐲\mathbf{y}-axis in steady state. In this figure, the heat flux φ13\varphi_{13} exchanged from the first to the third particle is different than the flux φ31\varphi_{31} transferred in the opposite direction.

In this paper we predict that many-body interactions mediated by thermal photons in Weyl semi-metal (WSM) networks as depicted in Fig. 1 can lead to an anomalous Hall flux. WSMs Huang ; Xu ; Lv are materials where valence and conduction bands cross in single points. Some of these media can exhibit, because of their unique topologically nontrivial electronic states Armitage ; Yan , a strong nonreciprocal optical response Kotov1 ; Kotov2 .

Refer to caption
Refer to caption
Figure 2: (a) The non-diagonal element of the permittivity tensor of the WSM and InSb (parameters are given in Ref. SM ) with an applied magnetic field in z-direction with amplitude B=0.1,10TB=0.1,10\,{\rm T} in the infrared range at T=300 K. Inset: |ϵ1|/|ϵ2||\epsilon_{1}|/|\epsilon_{2}| for the same materials. (b) The non-diagonal element Re(αxy){\rm Re}({\alpha_{xy}}) of (𝜶i𝜶i)/2i(\boldsymbol{\alpha}_{i}-\boldsymbol{\alpha}_{i}^{\dagger})/2{\rm i} entering the transmission coefficient in Eq. (7) for WSM with 2b=0.37,0.47,0.5Å12b=0.37,0.47,0.5\,\mathring{A}^{-1}. Inset: Re(αxy){\rm Re}({\alpha_{xy}}) for InSb (parameters are given in Ref. SM ) with an applied magnetic field in z-direction with amplitude B=0.1,1,10TB=0.1,1,10\,{\rm T} in the infrared range at T=300 K.

To investigate this effect we consider the system sketched in Fig. 1. Identical spherical particles of radius rr of WSMs are arranged in a four-terminal junction. The permittivity tensor of these particles takes the following form Kotov1 ; Kotov2

𝜺=(ε1iε20iε2ε1000ε1)\boldsymbol{\varepsilon}=\left(\begin{array}[]{ccc}\varepsilon_{1}&-i\varepsilon_{2}&0\\ i\varepsilon_{2}&\varepsilon_{1}&0\\ 0&0&\varepsilon_{1}\end{array}\right) (1)

with

ε1=εb+iσΩ\varepsilon_{1}=\varepsilon_{b}+i\frac{\sigma}{\Omega} (2)

and

ε2=be22π2ω.\varepsilon_{2}=\frac{be^{2}}{2\pi^{2}\hbar\omega}. (3)

Here

σ=rsg6ΩG(Ω/2)+irsgΩ6π[4Ω2(1+π23(kBTEF)2)+80ξcG(ξ)G(Ω/2)Ω24ξ2ξdξ]\begin{split}\sigma&=\frac{r_{s}g}{6}\Omega G(\Omega/2)+i\frac{r_{s}g\Omega}{6\pi}\bigg{[}\frac{4}{\Omega^{2}}\biggl{(}1+\frac{\pi^{2}}{3}\biggl{(}\frac{k_{B}T}{E_{F}}\biggr{)}^{2}\biggr{)}\\ &\qquad+8\int_{0}^{\xi_{c}}\frac{G(\xi)-G(\Omega/2)}{\Omega^{2}-4\xi^{2}}\xi d\xi\biggr{]}\end{split} (4)

where vFv_{F} is the Fermi velocity, Ω\Omega is the complex frequency normalized by the chemical potential, gg is the number of Weyl points, rs=e2/4πε0vFr_{s}=e^{2}/4\pi\varepsilon_{0}\hbar v_{F} is the effective fine structure constant, ee being the electron charge, and ξc=Ec/EF\xi_{c}=E_{c}/E_{F} where Ec is the cutoff energy (see details in Refs. Kotov1 ; Kotov2 ; SM ). If not explicitely stated we use for the WSM the parameter set vF=106m/s,g=2,2b=0.47Å1v_{F}=10^{6}\,{\rm m/s},g=2,2b=0.47\,\mathring{A}^{-1}, ξc=3\xi_{c}=3, and τ=1012s\tau=10^{-12}\,{\rm s}. Note that the permittivity tensor is non-reciprocal (i.e. 𝜺𝜺t\boldsymbol{\varepsilon}\neq\boldsymbol{\varepsilon}^{t}) so that for WSMs we intrinsically have such effects like a persistent heat currents OttEtAl2018 ; zhufan ; zhufan2 ; Silveirinha , persistent angular momentum and spin OttEtAl2018 ; Silveirinha ; Zubin2019 as found for non-reciprocal magneto-optical materials and as we will show the presence of an ATHE. Contrary to usual magneto-optical materials the non-reciprocity of WSMs is strong because the separations of the Weyl nodes in momentum space 2b2b can have relatively large values with compounds like Eu2IrO7, Co3S2Sn2, and Co3S2Se2 Kotov1 . This can be seen by comparison, for instance, with the non-diagonal term of permittivity tensor of usual magneto-optical material such as Indium Antimonide (InSb) under a strong magnetic field 𝐁=B𝐞z\mathbf{B}=B\mathbf{e}_{z} applied in z direction (see Fig. 2(a)). In this case the off-diagonal term is given by SM

ϵ2InSb=ϵωp2ωcω((ω+iγ)2ωc2)\epsilon_{2}^{\rm InSb}=\frac{\epsilon_{\infty}\omega_{p}^{2}\omega_{c}}{\omega\bigl{(}(\omega+{\rm i}\gamma)^{2}-\omega_{c}^{2}\bigr{)}} (5)

where ωp\omega_{p} and ωc=eB/m\omega_{c}=eB/m^{*} are the plasma and cyclotron frequency, γ\gamma is a phenomenological damping constant, mm^{*} the effective electron mass, and ϵ\epsilon_{\infty} is a material specific constant factor. This term clearly depends on the magnetic field strength BB via ωc\omega_{c} and as can be seen in Fig. 2(a) even when applying a magnetic field of 10T10\,{\rm T} we see that the non-reciprocity quantified by the ratio |ϵ2|/|ϵ1||\epsilon_{2}|/|\epsilon_{1}| is much smaller for InSb than for WSMs over a broad spectral range in the infrared. Hence, the WSMs can be seen as a magneto-optical material under the action of an extremely large magnetic field.

Using the Landauer formalism for N-body systems PBA_PRL2016 ; PBAEtAl2011 ; Riccardo ; Ivan_PRL2017 ; zhufan ; zhufan2 ; Cuevas ; Cuevas2 the non-equilibrium heat flux exchanged from the ithi^{th} to the jthj^{th} particle in the network reads

φij=0dω2π[Θ(ω,Ti)Θ(ω,Tj)]𝒯ij,\varphi_{ij}=\int_{0}^{\infty}\frac{{\rm d}\omega}{2\pi}\,[\Theta(\omega,T_{i})-\Theta(\omega,T_{j})]\mathcal{T}_{ij}, (6)

where Θ(ω,T)=ω/[eωkBT1]\Theta(\omega,T)={\hbar\omega}/[{e^{\frac{\hbar\omega}{k_{B}T}}-1}] is the mean energy of a harmonic oscillator in thermal equilibrium at temperature TT and 𝒯ij(ω)\mathcal{T}_{ij}(\omega) denotes the transmission coefficient, at the frequency ω\omega, between the particles ii and jj. In the dipolar approximation the transmission coefficient reads Cuevas ; Cuevas2

𝒯ij(ω)=43k04ImTr[𝜶j𝔾ji𝜶i𝜶i2i𝔾ji],\mathcal{T}_{ij}(\omega)=\frac{4}{3}k_{0}^{4}{\rm Im}{\rm Tr}\biggl{[}\boldsymbol{\alpha}_{j}\mathds{G}_{ji}\frac{\boldsymbol{\alpha}_{i}-\boldsymbol{\alpha}_{i}^{\dagger}}{2{\rm i}}\mathds{G}_{ji}^{\dagger}\biggr{]}, (7)

where k0=ω/ck_{0}=\omega/c with the light velocity cc and 𝔾ij\mathds{G}_{ij} denotes the dyadic Green tensor between the ithi^{th} and the jthj^{th} particle in the N-dipole system Purcell and 𝜶i\boldsymbol{\alpha}_{i} is the polarizability tensor of the ithi^{th} particle which reads for anisotropic particles in vacuum Albaladejo

𝜶i(ω)=(𝟏ik036π𝜶𝟎i)1𝜶𝟎i,\boldsymbol{\alpha}_{i}(\omega)=\biggl{(}\boldsymbol{1}-i\frac{k^{3}_{0}}{6\pi}\boldsymbol{\alpha_{0}}_{i}\biggr{)}^{-1}\boldsymbol{\alpha_{0}}_{i}, (8)

where 𝜶𝟎i\boldsymbol{\alpha_{0}}_{i} denotes the quasi-static or undressed polarizability of the ithi^{th} particle. For anisotropic spheres embedded in vacuum it reads LakhtakiaEtAl1991

𝜶𝟎i(ω)=4πr3(𝜺𝟙)(𝜺+2𝟙)1.\boldsymbol{\alpha_{0}}_{i}(\omega)=4\pi r^{3}(\boldsymbol{\varepsilon}-\mathds{1})(\boldsymbol{\varepsilon}+2\mathds{1})^{-1}. (9)

Due to the non-reciprocity of the permittivity the polarizability tensor is non-reciprocal as well, i.e. 𝜶0i𝜶0it\boldsymbol{\alpha}_{0i}\neq\boldsymbol{\alpha}_{0i}^{t} and in particular α0ixy=α0iyx{\alpha_{0i}}_{xy}=-{\alpha_{0i}}_{yx}. Hence, the skew hermitian part of the polarizability (𝜶i𝜶i)/2i(\boldsymbol{\alpha}_{i}-\boldsymbol{\alpha}_{i}^{\dagger})/2{\rm i} entering in the transmission coefficient has the diagonal elements Im(αiνν){\rm Im}({\alpha_{i}}_{\nu\nu}) (ν=x,y,z\nu=x,y,z) and the off-diagonal elements Re(αixy){\rm Re}({{\alpha_{i}}_{xy}}) and Re(αiyx)=Re(αixy){\rm Re}({{\alpha_{i}}_{yx}})=-{\rm Re}({{\alpha_{i}}_{xy}}), respectively. In Fig. 2(b) we show Re(αixy){\rm Re}({\alpha_{i}}_{xy}) which is responsible for the non-reciprocity and the thermal Hall effect. The two resonances in Re(αixy){\rm Re}({\alpha_{i}}_{xy}) as seen for InSb in Fig. 2(b) correspond to the dipolar resonances ωm=±1\omega_{m=\pm 1} of the nanoparticle, whereas the resonance ωm=0\omega_{m=0} can also be observed for the diagonal elements of the polarizability. Here mm represents the topological charge of particles OttEtAl2018 . For the WSM the high-frequency resonances at ωm=1\omega_{m=-1} dominate Re(αixy){\rm Re}({\alpha_{i}}_{xy}) for our choice of parameters so that only these resonances can be seen in Fig. 2(b). The magnitude of this resonance of the WSM is for the different values of 2b2b comparable to the corresponding resonances of InSb with a magnetic field between 0.1 and 1T. This might be astonishing because we have seen before that the non-diagonal elements of the permittivity tensor of the WSM are much stronger than those of InSb with a magnetic field of 1T to 10T. But the difference is due to the fact that for the WSM the resonance of ε2\varepsilon_{2} is at ω=0\omega=0 and that of InSb is at ωc\omega_{c}. By increasing the magnetic field the m=±1m=\pm 1 resonances frequency split and are separated by ωm=1ωm=+1ωc\omega_{m=-1}-\omega_{m=+1}\approx\omega_{c} while their magnitude increases as can be observed in Fig. 2(b). For the WSM the impact of the large non-diagonal elements of the permittivity tensor is that the splitting of the two resonances is much larger than for InSb, but the magnitude of the resonances does not get so large.

As a consequence of the non-reciprocity, even for four identical nano-particles we have 𝒯ij𝒯ji\mathcal{T}_{ij}\neq\mathcal{T}_{ji}. This asymmetry is a key condition for the existence of an ATHE. Note that this condition is not restricted to dipolar systems. For bigger particles the higher order modes (multipoles) should be taken into account to derive the optical responses (generalized susceptibility) of particles and transmission coefficients. As can be seen in the inset of Fig. 3(a), in the dipolar case, the non-reciprocity is for the used parameter set especially large for the m=1m=-1 resonance. For other choices of parameters also the m=+1m=+1 resonance can significantly contribute to the non-reciprocity. When 𝒯ij𝒯ji\mathcal{T}_{ij}\neq\mathcal{T}_{ji} we also have φijφji\varphi_{ij}\neq\varphi_{ji}. In particular, in the C4C_{4} symmetric configuration shown Fig. 1 we find

φ13=φ32=φ24=φ41φ14=φ42=φ23=φ31.\varphi_{13}=\varphi_{32}=\varphi_{24}=\varphi_{41}\neq\varphi_{14}=\varphi_{42}=\varphi_{23}=\varphi_{31}. (10)

In other words, due to the non-reciprocity the heat flow clockwise and counterclockwise is different as observed for the persistent heat current zhufan ; zhufan2 ; Ott_JPE leading to a circular heat flux OttEtAl2018 which is at the heart of the thermal Hall effect PBA_PRL2016 ; Ott_JPE . The circular heat flux around the WSM particles is shown in Fig. 3(b).

Refer to caption
Refer to caption
Figure 3: (a) Plot of the transmission coefficients 𝒯12\mathcal{T}_{12}, 𝒯13\mathcal{T}_{13} and 𝒯31\mathcal{T}_{31} for the configuration in Fig. 1 with d=300nmd=300\,{\rm nm} and radius r=100nmr=100\,{\rm nm} for four nanoparticles made of a WSM. The three dipolar resonances are at ωm=+1=1.486×1013rad/s\omega_{m=+1}=1.486\times 10^{13}\,{\rm rad/s}, ωm=0=7.714×1013rad/s\omega_{m=0}=7.714\times 10^{13}\,{\rm rad/s}, and ωm=1=3.856×1014rad/s\omega_{m=-1}=3.856\times 10^{14}\,{\rm rad/s}. The inset shows the ratio 𝒯31/𝒯13\mathcal{T}_{31}/\mathcal{T}_{13} illustrating the asymmetry of clockwise and counterclockise heat flux due to the non-reciprocity. (b) Normalized mean Poynting vector (arrows) and its magnitude (colour scale) for the Hall configuration in Fig. 1 with WSM nanoparticles having the temperatures as indicated in the plot. For the calculation of the mean Poynting vector we have used the method described in Ref. Ott_JPE .

Now let us assume that the two particles along the 𝐱\mathbf{x}-axis are connected to two thermostats fixing their temperatures so that a heat flux flows through the system between these two particles. While in a reciprocal system, no heat flux exists between the two other unthermostated particles for symmetry reasons, in non-reciprocal systems a Hall flux appears giving rise to a transversal temperature gradient in steady state. The magnitude of this Hall effect can be evaluated using the relative Hall temperature difference

RH=T3T4T1T2R_{H}=\frac{T_{3}-T_{4}}{T_{1}-T_{2}} (11)

where all temperatures are the steady states temperatures. In linear response regime, this expression can also be rewritten in term of the thermal conductance

Gij=30dω2πΘ(ω,T)T|T=Tj𝒯ij(ω)G_{ij}=3\int_{0}^{\infty}\!\!\frac{{\rm d}\omega}{2\pi}\,\frac{\partial\Theta(\omega,T)}{\partial T}\biggr{|}_{T=T_{j}}\mathcal{T}_{ij}(\omega) (12)

between the ithi^{th} and the jthj^{th} particle as PBA_PRL2016

RH=G13G24G14G23j3Gj3j4Gj4G43G34.R_{H}=\frac{G_{13}G_{24}-G_{14}G_{23}}{\underset{j\neq 3}{\sum}G_{j3}\underset{j\neq 4}{\sum}G_{j4}-G_{43}G_{34}}. (13)

Taking the symmetries in Eq. (10) due to the C4 configuration into account we can rewrite RHR_{H} as

RH=G13G31G13+G31+2G34.R_{H}=\frac{G_{13}-G_{31}}{G_{13}+G_{31}+2G_{34}}. (14)

This expression clearly shows that the assymetry in clockwise and conter-clockwise heat flow expressed by G13G31G_{13}-G_{31} introduced by the non-reciprocal Weyl metal is at the heart of the ATHE. Alternatively, to determine RHR_{H} the temperatures T1T_{1} and T2T_{2} are fixed and the equilibrium temperatures of particles 33 and 44 are calculated numerically by solving the system of overall power flow  Ott_JPE

ϕ3(T3,T4)=𝑗φj3=0,\displaystyle\phi_{3}(T_{3},T_{4})=\underset{j}{\sum}\varphi_{j3}=0, (15)
ϕ4(T3,T4)=𝑗φj4=0.\displaystyle\phi_{4}(T_{3},T_{4})=\underset{j}{\sum}\varphi_{j4}=0. (16)

These two methods are equivalent and give the same results for RHR_{H} provided that |T1T2|min(T1,T2)|T_{1}-T_{2}|\ll\min(T_{1},T_{2}).

Refer to caption
Refer to caption
Figure 4: (RHR_{H} from Eq. (14) as function of temperature. (a) RHR_{H} for different WSM parameters g=2,24g=2,24 and 2b=0.37Å1,0.47Å1,0.5Å12b=0.37\mathring{A}^{-1},0.47\mathring{A}^{-1},0.5\mathring{A}^{-1} with r=100nmr=100\,{\rm nm} and d=300nmd=300\,{\rm nm}. The horizontal dashed black (pink) line represents the value of RHR_{H} for InSb with a magnetic field of 0.1T0.1\,{\rm T} (1T1\,{\rm T}) in z-direction and T=300KT=300\,{\rm K}. (b) RHR_{H} for different distances using g=2g=2 and 2b=0.47Å12b=0.47\mathring{A}^{-1}.

In Fig. 4(a) we show the temperature dependence of the relative Hall temperature difference RHR_{H} for different WSM parameters. It is apparent that the magnitude and the directionality highly depend on these parameters. For a number of Weyl points g=2g=2 we find that RHR_{H} is mainly positive, i.e. particle 3 is heated up more efficiently than particle 4. This tendency could be anticipated from Fig. 3(a) which clearly shows that 𝒯13<𝒯31\mathcal{T}_{13}<\mathcal{T}_{31} and from Fig. 3(b) where the Poynting vector in the inner circle of the particles is bend towards particle 3. Furthermore, the ATHE becomes stronger for higher temperatures or smaller values of momentum-separation 2b2b of the Weyl nodes. The magnitude of Hall effect is comparable to the case of InSb particles with B=0.11TB=0.1-1\,{\rm T}. On the other hand, with g=24g=24 the Hall effect is the strongest for low temperatures and the direction of Hall flux is reversed, i.e. particle 4 is heated up more efficiently than particle 3. This can be traced back to the fact that for g=24g=24 the dipole resonance at ωm=+1\omega_{m=+1} dominates the heat flux so that the circularity is inverted and 𝒯13>𝒯31\mathcal{T}_{13}>\mathcal{T}_{31}. Finally, in Fig. 4 it can be seen that the strength of the ATHE becomes smaller in the far-field regime and even changes its directionality as also observed for the circular heat flux Ott_JPE .

In conclusion, we have demonstrated that the optical non-reciprocity in WSM nanoparticles networks induces circular heat flux around the particles resulting in an ATHE without applied external magnetic field. We have shown that the intrinsic time-reversal symmetry breaking in these systems gives rise to a Hall effect which is comparable to those observed in magneto-optical networks with magnetic fields of B=0.1B=0.1 to 1T1\,{\rm T}. It is worthwile to note that the non-reciprocity in WSM should also be responsible of persistent heat flux, persistent angular momentum and spin OttEtAl2018 ; zhufan ; zhufan2 ; Silveirinha ; Zubin2019 of thermal radiation. Besides its fundamental interest the ATHE for WSM opens up the possibility to control the direcionality of the radiative heat flux in nanoscale systems without the necessity of applying strong magnetic fields. This opens the route for an alternative way for thermal management and heat flux guiding in nanoscale systems.

Acknowledgements.
S.-A. B. acknowledges support from Heisenberg Programme of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under the project No. 404073166 and discussions with R. Messina.

References

  • (1) E. Hall, Am. J. Math. 2, 287 (1879).
  • (2) M.A. Leduc, J. Phys. 2e série 6, 378 (1887).
  • (3) C. Strohm, G. L. J. A. Rikken, and P. Wyder, Phys. Rev. Lett. 95, 155901 (2005).
  • (4) A. V. Inyushkin and A. N. Taldenkov, JETP Lett. 86, 379 (2007).
  • (5) S. Fujimoto, Phys. Rev. Lett. 103, 047203 (2009).
  • (6) H. Katsura, N. Nagaosa, and P. A. Lee, Phys. Rev. Lett. 104, 066403 (2010).
  • (7) Y. Onose, T. Ideue, H. Katsura, Y. Shiomi, N. Nagaosa, and Y. Tokura, Science 329, 297 (2010).
  • (8) P. Ben-Abdallah, Phys. Rev. Lett. 116, 084301 (2016).
  • (9) A. Ott, R. Messina, P. Ben-Abdallah, S.-A. Biehs,J. Photon. Energy 9, 032711 (2019).
  • (10) N. Nagaosa, J. Sinova, S. Onoda, A. H. MacDonald and N. P. Ong, Rev. Mod. Phys. 82, 1539 (2010)
  • (11) R. Karplus and J. M. Luttinger, Phys. Rev. 95, 1154 (1954).
  • (12) Y. Ferreiros, A. A. Zyuzin and J. H. Bardarson, Phys. Rev. B 96, 115202 (2017).
  • (13) K.Sugii et al., Anomalous thermal Hall effect in the topological antiferromagnetic state, ArxiV, 1902.06601.
  • (14) Z.-M. Huang, B. Han, M. Stone, Phys. Rev. B 101, 125201 (2020).
  • (15) S.-M. Huang, S.-Y. Xu, I. Belopolski, C.-C. Lee, G. Chang, B. Wang, N. Alidoust, G. Bian, M. Neupane, C. Zhang, S. Jia, A. Bansil, H. Lin, and M. Z. Hasan, Nat. Commun. 6, 7373 (2015).
  • (16) S.-Y. Xu, I. Belopolski, N. Alidoust, M. Neupane, G. Bian, C. Zhang, R. Sankar, G. Chang, Z. Yuan, C.-C. Lee, S.-M. Huang, H. Zheng, J. Ma, D. S. Sanchez, B. Wang, A. Bansil, F. Chou, P. P. Shibayev, H. Lin, S. Jia, and M. Z. Hasan, Science 349, 613 (2015).
  • (17) B. Q. Lv, N. Xu, H. M. Weng, J. Z. Ma, P. Richard, X. C. Huang, L. X. Zhao, G. F. Chen, C. E. Matt, F. Bisti, V. N. Strocov, J. Mesot, Z. Fang, X. Dai, T. Qian, M. Shi, and H. Ding, Nat. Phys. 11, 724 (2015).
  • (18) N. P. Armitage, E. J. Mele, A. Vishwanath, Rev. Mod. Phys. 90, 015001 (2018).
  • (19) B. Yan, C. Felser, Annual Rev. of Cond. Mat. Phys. 8, 337-354 (2017).
  • (20) O. V. Kotov, Y. E. Lozovik, Phys. Rev. B, 98, 195446 (2018).
  • (21) O. V. Kotov and Y. E. Lozovik, Phys. Rev. B 93, 235417 (2016).
  • (22) Supplemental Material.
  • (23) A. Ott, P. Ben-Abdallah, and S.-A. Biehs, Phys. Rev. B 97, 205414 (2018).
  • (24) L. Zhu and S. Fan, Phys. Rev. Lett. 117, 134303 (2016).
  • (25) L. Zhu, Y. Guo and S. Fan, Phys. Rev. B 97, 094302 (2018).
  • (26) M. G. Silveirinha, Phys. Rev. B 95, 115103 (2017).
  • (27) C. Khandekar, Z. Jacob, New J. Phys. 21, 103030 (2019).
  • (28) P. Ben-Abdallah, S.-A. Biehs, and K. Joulain, Phys. Rev. Lett. 107, 114301 (2011).
  • (29) R. Messina, M. Tschikin, S.-A. Biehs, and P. Ben-Abdallah, Phys. Rev. B 88, 104307 (2013).
  • (30) I. Latella and P. Ben-Abdallah, Phys. Rev. Lett. 118, 173902 (2017).
  • (31) R. M. Abraham Ekeroth, A. García-Martín, and J. C. Cuevas, Phys. Rev. B 95, 235428 (2017).
  • (32) R. M. Abraham Ekeroth, P. Ben-Abdallah, J. C. Cuevas and A. García-Martín, ACS Photonics 5 705 (2018).
  • (33) A. Lakhtakia, V. K. Varadan, and V. V. Varadan, International Journal of Infrarared and Millimeter Waves 12, pp. 1253-1264 (1991).
  • (34) E. M. Purcell and C. R. Pennypacker, Astrophys. J. 186, 705 (1973).
  • (35) S. Albaladejo, R. Gómez-Medina, L. S. Froufe-Pérez, H. Marinchio, R. Carminati, J. F. Torrado, G. Armelles, A. García-Martín, and J. J. Sáenz, Optics Express, 18, 4, pp. 3556-3567 (2010).