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

Topological parafermion corner states
in clock-symmetric non-Hermitian second-order topological insulator

Motohiko Ezawa Department of Applied Physics, University of Tokyo, Hongo 7-3-1, 113-8656, Japan
Abstract

Parafermions are a natural generalization of Majorana fermions. We consider a breathing Kagome lattice with complex hoppings by imposing 3\mathbb{Z}_{3} clock symmetry in the complex energy plane. It is a non-Hermitian generalization of the second-order topological insulator characterized by the emergence of topological corner states. We demonstrate that the topological corner states are parafermions in the present 3\mathbb{Z}_{3} clock-symmetric model. It is also shown that the model is realized in electric circuits properly designed, where the parafermion corner states are observed by impedance resonance. We also construct 4\mathbb{Z}_{4} and 6\mathbb{Z}_{6} parafermions on breathing square and honeycomb lattices, respectively.

I Introduction

Topological quantum computation is a fault-tolerant quantum computationMoore ; Das ; Kitaev ; TQC ; SternA ; Stern ; NPJ . Majorana fermions provide us with a most studied platform of topological quantum computationBeen ; Stan ; Elli ; AA ; Ivanov ; Halperin . The Majorana operator γ\gamma satisfies γ2=1\gamma^{2}=1. They are realized as topological boundary states of topological superconductorsAliceaBraid ; Qi ; Lei ; Alicea ; Sato ; Tanaka and Kitaev spin liquidsKitaev ; Matsuda . However, it is impossible to perform universal quantum computation only by the braiding of Majorana fermions since they can reproduce only a part of Clifford gatesAhl .

Parafermions are straightforward generalization of Majorana fermionsFendJSM ; Fend ; AliceaPara ; Jerm ; Ebisu , where the parafermion operator γ\gamma satisfies γd=1\gamma^{d}=1 for d3d\geq 3. Braiding of parafermions with d=3d=3 are known to reproduce all the Clifford gatesHutter although universal quantum computation is not yet possible. In this sense, parafermions are more powerful than Majorana fermions in the context of quantum computation. Parafermions are realized in clock-spin modelsBaxter ; Fend , fractional quantum Hall effectsReadR ; Clarke , fractional topological superconductorsLaub1 and twisted bilayer grapheneLaub2 . Among them, the d\mathbb{Z}_{d} clock-spin modelBaxter is non-Hermitian and its energy spectrum is d\mathbb{Z}_{d} symmetric in the complex plane. It is an interesting problem if they also emerge as topological boundary states in certain lattice structures just as Majorana fermions do.

Higher-order topological insulators and superconductors are generalization of topological insulators and superconductorsFan ; Science ; APS ; Peng ; Lang ; Song ; Bena ; Schin ; FuRot ; Kagome ; EzawaPhos ; Gei ; Kha ; EzawaMajo . They are prominent by the emergence of zero-energy corner states instead of gapless edge states. These zero-energy corner states are topologically protected. A typical example is given by the breathing Kagome latticeKagome , where three topological corner states emerge. There are some generalization to non-Hermitian higher-order topological insulatorsLiuSOTI ; EzawaLCR ; EzawaSkin ; Berg .

In this paper, generalizing the breathing Kagome second-order topological insulator model by imposing 3\mathbb{Z}_{3} clock symmetry, we propose a new type of non-Hermitian higher-order topological insulator, where the topological corner states are parafermions. This model is non-Hermitian, where the energy spectrum is 3\mathbb{Z}_{3} symmetric in the complex energy plane as in the case of the 3\mathbb{Z}_{3} clock-spin model. We demonstrate how to implement the present model of parafermions in an electric circuit. We also construct 4\mathbb{Z}_{4} and 6\mathbb{Z}_{6} parafermions as topological corner states on breathing square and honeycomb lattices.

II Majorana fermion and parafermion

Majorana fermion operators γi\gamma_{i} satisfy the relations

(γj)2=1,γjγk=γkγj.\left(\gamma_{j}\right)^{2}=1,\qquad\gamma_{j}\gamma_{k}=-\gamma_{k}\gamma_{j}. (1)

Majorana fermions are realized as zero-energy states of a topological superconductor, where particle-hole symmetry (PHS) preserves. It is understood as follows. The PHS operator Ξ\Xi acts as Ξ1HΞ=H\Xi^{-1}H\Xi=-H with the eigen equation H|ψ=E|ψH\left|\psi\right\rangle=E\left|\psi\right\rangle. If a particle has an energy EE, its antiparticle has the energy E-E in the presence of PHS. Namely, the wave functions always appear in a particle-hole pair with a pair of energies (E,E)\left(E,-E\right). If the states satisfy the relation E=EE=-E, the particle is identical to its antiparticle, and a pair of Majorana fermions emerge. Hence, the zero-energy (E=0E=0) states respecting PHS are Majorana states.

Parafermions are natural generalization of Majorana fermions. d\mathbb{Z}_{d} parafermions are defined through the relations

(γj)d=1,γjγk=ωγkγj,\left(\gamma_{j}\right)^{d}=1,\qquad\gamma_{j}\gamma_{k}=\omega\gamma_{k}\gamma_{j}, (2)

where ω=e2πi/d\omega=e^{2\pi i/d}. The minimal model consists of two elements γ1\gamma_{1} and γ2\gamma_{2}.

III 3\mathbb{Z}_{3} Parafermion

We start with a minimal model by setting d=3d=3 in Eq.(2). 3\mathbb{Z}_{3} parafermions are represented by the shift operatorZohar ; Fend ; AliceaPara

γ1τ=(001100010),\gamma_{1}\equiv\tau=\left(\begin{array}[]{ccc}0&0&1\\ 1&0&0\\ 0&1&0\end{array}\right), (3)

and the clock operatorZohar ; Fend ; AliceaPara

γ2σ=diag.(1,ω,ω2),\gamma_{2}\equiv\sigma=\text{diag.}\left(1,\omega,\omega^{2}\right), (4)

where ω=e2πi/3\omega=e^{2\pi i/3}. Here, τ\tau and σ\sigma satisfy the 3\mathbb{Z}_{3} parafermion relations,

τ3=σ3=1,στ=ωτσ.\tau^{3}=\sigma^{3}=1,\qquad\sigma\tau=\omega\tau\sigma. (5)

In the 3\mathbb{Z}_{3} clock-symmetric model, the energy spectrum is composed of tripletsBaxter En(0,1,2)E_{n}^{\left(0,1,2\right)}, n=0,1,2,n=0,1,2,\cdots,

En(0,1,2)=εn,ωεn,ω2εn,E_{n}^{\left(0,1,2\right)}=\varepsilon_{n},\quad\omega\varepsilon_{n},\quad\omega^{2}\varepsilon_{n}, (6)

satisfying εn+ωεn+ω2εn=0\varepsilon_{n}+\omega\varepsilon_{n}+\omega^{2}\varepsilon_{n}=0. The system is necessarily non-Hermitian because the eigen energies are complex except for zero-energy states.

Refer to caption

Figure 1: (a) Breathing Kagome lattice, (b) breathing square lattice and (c) breathing honeycomb lattice. The hopping parameters are ta=0t_{a}=0 for (a1)\sim(c1), tatb0t_{a}t_{b}\neq 0 for (a2)\sim(c2), and tb=0t_{b}=0 for (a3)\sim(c3). There emerge three, four and six parafermion corner states in the topological phase. When ta=0t_{a}=0, they are isolated as marked by blue disks in (a1), (b1) and (c1), respectively. The unit cells are given in (a4), (b4) and (c4) with the hopping parameters indicated.

III.1 Zero-energy parafermion states

It follows from Eq.(6) that only the zero-energy states form a set of degenerate states respecting 3\mathbb{Z}_{3} clock symmetry. They are 3\mathbb{Z}_{3} parafermion states. We denote them as |ψ0\left|\psi_{0}\right\rangle, |ψ1\left|\psi_{1}\right\rangle and |ψ2\left|\psi_{2}\right\rangle. They are characterized by the properties

τ|ψ0\displaystyle\tau\left|\psi_{0}\right\rangle =|ψ1,τ|ψ1=|ψ2,τ|ψ2=|ψ0,\displaystyle=\left|\psi_{1}\right\rangle,\quad\tau\left|\psi_{1}\right\rangle=\left|\psi_{2}\right\rangle,\quad\tau\left|\psi_{2}\right\rangle=\left|\psi_{0}\right\rangle, (7)
σ|ψ0\displaystyle\sigma\left|\psi_{0}\right\rangle =|ψ0,σ|ψ1=ω|ψ1,σ|ψ2=ω2|ψ2,\displaystyle=\left|\psi_{0}\right\rangle,\quad\sigma\left|\psi_{1}\right\rangle=\omega\left|\psi_{1}\right\rangle,\quad\sigma\left|\psi_{2}\right\rangle=\omega^{2}\left|\psi_{2}\right\rangle, (8)

from which the matrix representations (3) and (4) follow. Then, the 3\mathbb{Z}_{3} parafermion relations (5) are verifed. Namely, it is necessary and sufficient to examine Eqs.(7) and (8) for a triplet set of zero-energy states in order to show that they are 3\mathbb{Z}_{3} parafermions.

III.2 Breathing Kagome lattice

We propose a model possessing parafermions on the breathing Kagome lattice. The bulk Hamiltonian is given by

H=(0h12ωh13h120ω2h23ωh13ω2h230),H=\left(\begin{array}[]{ccc}0&h_{12}&\omega h_{13}\\ h_{12}^{\ast}&0&\omega^{2}h_{23}\\ \omega h_{13}^{\ast}&\omega^{2}h_{23}^{\ast}&0\end{array}\right), (9)

with

h12\displaystyle h_{12} =ta+tbeikx,\displaystyle=t_{a}+t_{b}e^{ik_{x}}, (10)
h23\displaystyle h_{23} =ta+tbei(kx/2+3ky/2),\displaystyle=t_{a}+t_{b}e^{-i\left(k_{x}/2+\sqrt{3}k_{y}/2\right)}, (11)
h13\displaystyle h_{13} =ta+tbei(kx/23ky/2),\displaystyle=t_{a}+t_{b}e^{-i\left(k_{x}/2-\sqrt{3}k_{y}/2\right)}, (12)

where we have introduced two hopping parameters tat_{a} and tbt_{b}, corresponding to the magenta link and the cyan link along the horizontal axis in Fig.1(a4). The hopping parameters along the other two triangle sides are given by ωta\omega t_{a} and ω2ta\omega^{2}t_{a} for a magenta triangle, and ωtb\omega t_{b} and ω2tb\omega^{2}t_{b} for a cyan triangle. This model is non-Hermitian due to the presence of ω\omega.

A comment is in order with respect to the breathing Kagome model. It is a typical model for the conventional second-order topological insulatorKagome , where the factor ω\omega is absent and it is Hermitian. The present generalization of the breathing Kagome lattice model provides us with a new type of non-Hermitian second-order topological insulators.

Refer to caption

Figure 2: (a) Topological number (21) as a function of ta/tbt_{a}/t_{b}. It is quantized in the insulators. (b)\sim(c) Energy spectrum as a function of ta/tbt_{a}/t_{b} for (b) the bulk, (c) a nanoribbon with width 64, and (d) a triangle with size 20. In (b), the blue curves represent the bulk energy (20) at the KK and KK^{\prime} points, and the cyan lines represent the bulk energy (19) at the Γ\Gamma point, which are analytically obtained. In the topological phase, gapless edge states are absent in nanoribbon geometry as in (c), but zero-energy corner states emerge in triangle geometry as indicated by the magenta flat-line segment in (d). The segment slightly deviates from the region (1,1/2)(-1,1/2) due to the finite size effect.

III.3 Clock symmetry

The Hamiltonian (9) has 3\mathbb{Z}_{3} clock symmetryFend ,

τH(𝐤)τ=ωH(R𝐤),\tau H\left(\mathbf{k}\right)\tau^{\dagger}=\omega H\left(R\mathbf{k}\right), (13)

where RR rotates the momentum by 120 degrees as

R(kx,0)\displaystyle R\left(k_{x},0\right) =(kx2,3ky2),\displaystyle=\left(-\frac{k_{x}}{2},\frac{\sqrt{3}k_{y}}{2}\right), (14)
R(kx2,3ky2)\displaystyle R\left(-\frac{k_{x}}{2},\frac{\sqrt{3}k_{y}}{2}\right) =(kx2,3ky2),\displaystyle=\left(-\frac{k_{x}}{2},-\frac{\sqrt{3}k_{y}}{2}\right), (15)
R(kx2,3ky2)\displaystyle R\left(-\frac{k_{x}}{2},-\frac{\sqrt{3}k_{y}}{2}\right) =(kx,0),\displaystyle=\left(k_{x},0\right), (16)

making the energy spectrum have 3\mathbb{Z}_{3} symmetry in the complex plane as in Eq.(6). In addition, there is an anti-unitary symmetry

KH(𝐤)K=H(𝐤),KH\left(\mathbf{k}\right)K=H^{\ast}\left(\mathbf{k}\right), (17)

where KK implies complex conjugate. It leads to reflection symmetry between EE and EE^{\ast}. As a result, the energy spectrum has C3vC_{3v} symmetry in the complex plane, which consists of the three-fold rotational symmetry and three reflection symmetries.

The system also has a generalized chiral symmetry for a three-band modelNi ,

σHσ1\displaystyle\sigma H\sigma^{-1} =\displaystyle= H1,σH1σ1=H2,\displaystyle H_{1},\qquad\sigma H_{1}\sigma^{-1}=H_{2},
H+H1+H2\displaystyle H+H_{1}+H_{2} =\displaystyle= 0.\displaystyle 0. (18)

We show the bulk energy spectrum in Fig.3(a) and Fig.4(a), where C3vC_{3v} symmetry is manifest for all parameters.

III.4 Phase diagram

The notion of insulator and metal is generalized to the non-Hermitian Hamiltonian in two ways. On is a point-gap insulatorGong ; Kawabata , where |E||E| has a gap. The other is a line-gap insulatorGong ; Kawabata , where Re[E]\left[E\right] or Im[E]\left[E\right] has a gap. In our model, we adopt the definition of the point-gap insulator due to 3\mathbb{Z}_{3} symmetry. We are able to determine the energy spectrum analytically at the Γ=(0,0)\Gamma=(0,0) point as

E(0,0)=(ta+tb),ω(ta+tb),ω2(ta+tb),E\left(0,0\right)=(t_{a}+t_{b}),\quad\omega(t_{a}+t_{b}),\quad\omega^{2}(t_{a}+t_{b}), (19)

and at the K=(4π/3,0)K=(4\pi/3,0) and K=(4π/3,0)K^{\prime}=(-4\pi/3,0) points as

E3(±4π/3,0)=(ta+tb)(ta2tb)(2tatb).E^{3}(\pm 4\pi/3,0)=\left(t_{a}+t_{b}\right)\left(t_{a}-2t_{b}\right)\left(2t_{a}-t_{b}\right). (20)

The point gap closes at the KK and KK^{\prime} points for ta/tb=1/2t_{a}/t_{b}=1/2, ta/tb=2t_{a}/t_{b}=2, and at the KK, KK^{\prime} and Γ\Gamma points for ta/tb=1t_{a}/t_{b}=-1, as in Fig.2(b). This is also confirmed numerically by calculating the band spectrum as in Fig.2(b).

Refer to caption

Figure 3: Energy spectrum of a triangle, where (a) the vertical axis is Re[EE] and the horizontal axis is Im[EE], while (b) the horizontal axis is ta/tbt_{a}/t_{b}. Color indicates the value of ta/tbt_{a}/t_{b}, where the color pallet is the same as in (b). C3vC_{3v} symmetry in the complex energy plane is manifest in (a). The emergence of zero-energy corner states is clear in (b). We have used a triangle with size 66.

Refer to caption


Figure 4: Complex energy spectrum (a1)\sim (h1) for the bulk, (a2)\sim (h2) for a nanoribbon and (a3)\sim (h3) for a triangle, where the hopping parameters are (a1)\sim (a3) ta=1.5tbt_{a}=-1.5t_{b}, (b1)\sim (b3) ta=tbt_{a}=-t_{b}, (c1)\sim (c3) ta=0.25tbt_{a}=-0.25t_{b}, (d1)\sim (d3) ta=0.25tbt_{a}=0.25t_{b}, (e1)\sim (e3) ta=0.5tbt_{a}=0.5t_{b}, (f1)\sim (f3) ta=1.25tbt_{a}=1.25t_{b}, (g1)\sim (g3) ta=2tbt_{a}=2t_{b} and (h1)\sim (h3) ta=2.5tbt_{a}=2.5t_{b}. The horizontal axis is Im[EE] and the vertical axis is Re[EE]. The numerical value on the horizontal axis in (a1)\sim (h1) is the energy in unit of tbt_{b}. Color indicates the momentum of along the nanoribbon direction, where red color indicates k=πk=\pi and blue color indicates k=0k=0 in a nanoribbon. We have used a nanoribbon with width 128128 for (a2)\sim (h2), and a triangle with size 1616 for (a3)\sim (h3). The magenta dots in (c3) and (d3) represent the topological corner states.

Refer to caption

Figure 5: (a1)\sim(h1) Bulk band structure, where the horizontal axis is kxk_{x}. Color in the spectrum indicates kyk_{y}, where red color indicates ky=0k_{y}=0 while cyan color indicates ky=2π/3k_{y}=2\pi/\sqrt{3}. (a2)\sim(h2) Band structure of a nanoribbon, where the horizontal axis is kk. The vertical axis is |E||E|. The values of tat_{a} and tbt_{b} are the same as in Fig.4. It is notable that there are no gapless edge states in the topological phase.

III.5 Topological number

The topological number is given by the Berry phase defined by

Q12πi02πψ0(kx,0)|kx|ψ0(kx,0)𝑑kx,Q\equiv\frac{1}{2\pi i}\int_{0}^{2\pi}\left\langle\psi_{0}\left(k_{x},0\right)\right|\partial_{k_{x}}\left|\psi_{0}\left(k_{x},0\right)\right\rangle dk_{x}, (21)

where ψ0\psi_{0} is the eigen function of the Hamiltonian (9) for the bulk, whose eigen energy is real along the kxk_{x} axis (i.e., ky=0k_{y}=0). We calculate it numerically, whose results are shown in Fig.2(a). We find Q=1Q=1 for 1<ta/tb<1/2-1<t_{a}/t_{b}<1/2, and Q=0Q=0 for ta/tb<1t_{a}/t_{b}<-1 and ta/tb>2t_{a}/t_{b}>2, while it continuously changes from 11 to 0 for 1/2<ta/tb<21/2<t_{a}/t_{b}<2. In fact, QQ is quantized in the insulator phases.

III.6 Edge states

We calculate the energy spectrum in a nanoribbon numerically. Edge states are observed in Fig.4(a2)\sim(h2), where the complex energy spectrum is shown for various momentum kk specified by color. Three-fold symmetry is slightly broken in nanoribbon geometry. It is due to the finite size effect of a nanoribbon.

We present the band structure of a nanoribbon in Fig.5(a2)\sim(h2), where we observe the absence of gapless edge states in the topological phase.

III.7 Corner states

We calculate the energy spectrum in triangle geometry numerically. C3vC_{3v} symmetry is manifest as shown in Fig.3(a) and in Fig.4(a3)\sim(h3). It is because the triangle respects 3\mathbb{Z}_{3} clock symmetry. We find zero-energy states in the region 1<ta/tb<1/2-1<t_{a}/t_{b}<1/2, as indicated by a magenta line in Fig.2(d). We also show the energy spectrum as a function of ta/tbt_{a}/t_{b} in Fig.3(b), where the emergence of the zero-energy states is manifest in the topological phase.

Consequently, the present model is a second-order topological insulator in the region 1<ta/tb<1/2-1<t_{a}/t_{b}<1/2, being characterized by the emergence of topological corner states.

It is possible to obtain explicitly the wave functions of the three corner states by numerical calculation. We have numerically confirmed that they satisfy the relations (7) and (8). Therefore, they are 3\mathbb{Z}_{3} parafermions.

III.8 Electric-circuit implementation

Electric circuits are governed by the Kirchhoff current law. By making the Fourier transformation with respect to time, the Kirchhoff current law is expressed as

Ia(ω)=bJab(ω)Vb(ω),I_{a}\left(\omega\right)=\sum_{b}J_{ab}\left(\omega\right)V_{b}\left(\omega\right), (22)

where IaI_{a} is the current between node aa and the ground, while VbV_{b} is the voltage at node bb. The matrix Jab(ω)J_{ab}\left(\omega\right) is called the circuit Laplacian. Once the circuit Laplacian is given, we can uniquely setup the corresponding electric circuit. By equating it with the Hamiltonian HH asTECNature ; ComPhys

Jab(ω)=iωHab(ω),J_{ab}\left(\omega\right)=i\omega H_{ab}\left(\omega\right), (23)

it is possible to simulate various topological phases of the Hamiltonian by electric circuitsTECNature ; ComPhys ; Hel ; Lu ; YLi ; EzawaTEC ; EzawaLCR ; EzawaSkin ; Garcia ; Hofmann ; EzawaMajo ; Tjunc . The relations between the parameters in the Hamiltonian and in the electric circuit are determined by this formula.

The circuit Laplacian is constructed as follows. To simulate the positive and negative hoppings in the Hamiltonian, we replace them with the capacitance iωCi\omega C and the inductance 1/iωL1/i\omega L, respectively. We note that sink=(eikeik)/2i\sin k=(e^{ik}-e^{-ik})/2i represents an imaginary hopping in the tight-bind model. The imaginary hopping is realized by an operational amplifierHofmann .

We thus make the following replacements with respect to hoppings in the Hamiltonian to derive the circuit Laplacian: (i) +XiωCX+X\rightarrow i\omega C_{X} for X=taX=t_{a} and tbt_{b}, where CXC_{X} represents the capacitance whose value is XX [pF]. (ii) X1/iωLX-X\rightarrow 1/i\omega L_{X} for X=ta/2X=t_{a}/2 and tb/2t_{b}/2, where LXL_{X} represents the inductance whose value is XX [μ\muH].

Refer to caption

Figure 6: (a) Illustration of the breathing Kagome circuit corresponding to Fig.1(a). (b) Each node is grounded by an inductor. (c) Complex hoppings e±2πi/3\propto e^{\pm 2\pi i/3} are realized by a parallel connection of an inductor and an operational amplifier. We show a triangle with size 44.

We explicitly study the breathing Kagome lattice described by (9), where the electric circuit is given by Fig.6(a). The Hamiltonian (9) is decomposed into

H=H1+H2,H=H_{1}+H_{2}, (24)

with

H1=(0h1212h13h120ω2h2312h13ω2h230)H_{1}=\left(\begin{array}[]{ccc}0&h_{12}&-\frac{1}{2}h_{13}\\ h_{12}^{\ast}&0&\omega^{2}h_{23}\\ -\frac{1}{2}h_{13}^{\ast}&\omega^{2}h_{23}^{\ast}&0\end{array}\right) (25)

and

H2=32(00ih1300ih23ih13ih230),H_{2}=\frac{\sqrt{3}}{2}\left(\begin{array}[]{ccc}0&0&ih_{13}\\ 0&0&-ih_{23}\\ ih_{13}^{\ast}&-ih_{23}^{\ast}&0\end{array}\right), (26)

where H1H_{1} is Hermitian (H1=H1H_{1}^{\dagger}=H_{1}), and H2H_{2} is anti-Hermitian (H2=H2H_{2}^{\dagger}=-H_{2}). It is necessary to construct imaginary hopping Hamiltonians

32ta(00i000i00),\frac{\sqrt{3}}{2}t_{a}\left(\begin{array}[]{ccc}0&0&i\\ 0&0&0\\ i&0&0\end{array}\right), (27)

and

32ta(00000i0i0)\frac{\sqrt{3}}{2}t_{a}\left(\begin{array}[]{ccc}0&0&0\\ 0&0&-i\\ 0&-i&0\end{array}\right) (28)

for the magenta lines in Fig.6(a). They are constructed by using operational amplifiers and resistors.

Refer to caption

Figure 7: Impedance as a function of ω\omega. The vertical axis is log10Z\log_{10}Z and the horizontal axis is ω/ω0\omega/{\omega_{0}}. (a) topological phase with ta/tb=0.25t_{a}/t_{b}=0.25, (b) metallic phase with ta/tb=1t_{a}/t_{b}=1 and (c) trivial phase with ta/tb=2.5t_{a}/t_{b}=2.5. A prominent peak is found in the topological phase.

We review a negative impedance converter with current inversion based on an operational amplifier with resistorsHofmann . The voltage-current relation for the operational amplifier circuit is given byHofmann

(I1I2)=1R(νν11)(V1V2),\left(\begin{array}[]{c}I_{1}\\ I_{2}\end{array}\right)=\frac{1}{R}\left(\begin{array}[]{cc}-\nu&\nu\\ -1&1\end{array}\right)\left(\begin{array}[]{c}V_{1}\\ V_{2}\end{array}\right), (29)

with ν=Rb/Ra\nu=R_{b}/R_{a}, where RR, RaR_{a} and RbR_{b} are the resistances in an operational amplifier: See Fig.6(d). We note that the resistors in the operational amplifier circuit are tuned to be ν=1\nu=1 in the literatureHofmann so that the system becomes Hermitian, where the corresponding Hamiltonian represents a spin-orbit interaction.

In this paper, we use two negative impedance converters parallelly connected with the opposite direction as in Fig.6(c). The circuit Laplacian due to these two converters is given by

1R[(νν11)+(11νν)]=1R(1νν1ν11ν).\frac{1}{R}\left[\left(\begin{array}[]{cc}-\nu&\nu\\ -1&1\end{array}\right)+\left(\begin{array}[]{cc}1&-1\\ \nu&-\nu\end{array}\right)\right]=\frac{1}{R}\left(\begin{array}[]{cc}1-\nu&\nu-1\\ \nu-1&1-\nu\end{array}\right). (30)

It corresponds to the Hamiltonian

H=1iωR(1νν1ν11ν).H=\frac{1}{i\omega R}\left(\begin{array}[]{cc}1-\nu&\nu-1\\ \nu-1&1-\nu\end{array}\right). (31)

It is embedded in the 3×\times3 matrix as

H=1iωR(1ν0ν1000ν101ν),H=\frac{1}{i\omega R}\left(\begin{array}[]{ccc}1-\nu&0&\nu-1\\ 0&0&0\\ \nu-1&0&1-\nu\end{array}\right), (32)

where we have set

32ta=1νωR\frac{\sqrt{3}}{2}t_{a}=\frac{1-\nu}{\omega R} (33)

with ν<1\nu<1, and

H=1iωR(00001νν10ν11ν),H=\frac{1}{i\omega R}\left(\begin{array}[]{ccc}0&0&0\\ 0&1-\nu&\nu-1\\ 0&\nu-1&1-\nu\end{array}\right), (34)

where we have set

32ta=ν1ωR\frac{\sqrt{3}}{2}t_{a}=\frac{\nu-1}{\omega R} (35)

with ν>1\nu>1. These matrices are different from Eqs.(27) and (28) by the diagonal terms. They are cancelled by adding a resistor (for ν>1\nu>1) or an operational amplifier (for ν<1\nu<1) with the amount of

1νiωR\frac{1-\nu}{i\omega R} (36)

between a lattice site and the ground.

III.9 Impedance resonance

The zero-energy parafermion corner states are well observed by impedance resonance, which is definedHel by

Zab=Va/Ib=GabZ_{ab}=V_{a}/I_{b}=G_{ab} (37)

where G=J1G=J^{-1} is the Green function. It diverges at the frequency where the admittance is zero (J=0J=0). Taking the nodes aa and bb at two corners, we show the impedance in topological, metallic and trivial phases in Figs.7(a)\sim(c), respectively. A strong impedance peak is observed at the critical frequency ω01/LC\omega_{0}\equiv 1/\sqrt{LC} only in the topological phase. It signals the emergence of zero-energy parafermion corner states.

Refer to caption

Figure 8: Energy spectrum of a square, where (a) the vertical axis is Re[EE] and the horizontal axis is Im[EE], while (b) the horizontal axis is ta/tbt_{a}/t_{b}. Color indicates the value of ta/tbt_{a}/t_{b}, where the color pallet is the same as in (b). C4vC_{4v} symmetry in the complex energy plane is manifest in (a). The emergence of zero-energy corner states is clear in (b).

IV 4\mathbb{Z}_{4} Parafermion

We proceed to a model with d=4d=4. 4\mathbb{Z}_{4} parafermions are represented by the shift operatorZohar ; Fend ; AliceaPara

γ1τ=(0001100001000010),\gamma_{1}\equiv\tau=\left(\begin{array}[]{cccc}0&0&0&1\\ 1&0&0&0\\ 0&1&0&0\\ 0&0&1&0\end{array}\right), (38)

and the clock operatorZohar ; Fend ; AliceaPara

γ2σ=diag.(1,i,1,i).\gamma_{2}\equiv\sigma=\text{diag.}\left(1,i,-1,-i\right). (39)

Here, τ\tau and σ\sigma satisfy the 4\mathbb{Z}_{4} parafermion relations,

τ4=σ4=1,τσ=ωστ.\tau^{4}=\sigma^{4}=1,\qquad\tau\sigma=\omega\sigma\tau. (40)

In the 4\mathbb{Z}_{4} clock-symmetric model, the energy spectrum is composed of quartets En(0,1,2,3)E_{n}^{\left(0,1,2,3\right)}, n=0,1,2,n=0,1,2,\cdots,

En(0,1,2,3)=εn,iεn,εn,iεn.E_{n}^{\left(0,1,2,3\right)}=\varepsilon_{n},\quad i\varepsilon_{n},\quad-\varepsilon_{n},\quad-i\varepsilon_{n}. (41)

The system is necessarily non-Hermitian because the eigen energies are complex except for zero-energy states.

IV.1 Zero-energy parafermion states

It follows from Eq.(41) that only the zero-energy states form a set of degenerate states respecting 4\mathbb{Z}_{4} clock symmetry. They are 4\mathbb{Z}_{4} parafermion states. We denote them as |ψ0\left|\psi_{0}\right\rangle, |ψ1\left|\psi_{1}\right\rangle, |ψ2\left|\psi_{2}\right\rangle and |ψ3\left|\psi_{3}\right\rangle. They are characterized by the properties

τ|ψ0\displaystyle\tau\left|\psi_{0}\right\rangle =|ψ1,τ|ψ1=|ψ2,\displaystyle=\left|\psi_{1}\right\rangle,\quad\tau\left|\psi_{1}\right\rangle=\left|\psi_{2}\right\rangle, (42)
τ|ψ2\displaystyle\tau\left|\psi_{2}\right\rangle =|ψ3,τ|ψ3=|ψ0,\displaystyle=\left|\psi_{3}\right\rangle,\quad\tau\left|\psi_{3}\right\rangle=\left|\psi_{0}\right\rangle, (43)
σ|ψ0\displaystyle\sigma\left|\psi_{0}\right\rangle =|ψ0,σ|ψ1=i|ψ1,\displaystyle=\left|\psi_{0}\right\rangle,\quad\sigma\left|\psi_{1}\right\rangle=i\left|\psi_{1}\right\rangle, (44)
σ|ψ2\displaystyle\sigma\left|\psi_{2}\right\rangle =|ψ2,σ|ψ3=i|ψ3,\displaystyle=-\left|\psi_{2}\right\rangle,\quad\sigma\left|\psi_{3}\right\rangle=-i\left|\psi_{3}\right\rangle, (45)

from which the matrix representations (3) and (4) follow. Then, the 4\mathbb{Z}_{4} parafermion relations (5) are verified. Namely, it is necessary and sufficient to examine Eqs.(7) and (8) for a triplet set of zero-energy states in order to show that they are 4\mathbb{Z}_{4} parafermions.

IV.2 Breathing square lattice

The quadrupole insulator has been proposed on the breathing square latticeScience . We propose a model possessing 4\mathbb{Z}_{4} parafermions on the breathing square lattice. The bulk Hamiltonian is given by

H=(0fx0ifyfx0ify00ify0fxify0fx0),H=\left(\begin{array}[]{cccc}0&f_{x}^{\ast}&0&if_{y}^{\ast}\\ -f_{x}&0&if_{y}^{\ast}&0\\ 0&-if_{y}&0&-f_{x}\\ -if_{y}&0&f_{x}^{\ast}&0\end{array}\right), (46)

with

fx=ta+tbeikx,fy=ta+tbeiky,f_{x}=t_{a}+t_{b}e^{ik_{x}},\qquad f_{y}=t_{a}+t_{b}e^{ik_{y}}, (47)

where we have introduced two hopping parameters tat_{a} and tbt_{b}, which are shown in Fig.1(b4).

The Hamiltonian (46) has 4\mathbb{Z}_{4} clock symmetryFend ,

τH(𝐤)τ=iH(R𝐤),\tau H\left(\mathbf{k}\right)\tau^{\dagger}=-iH\left(R\mathbf{k}\right), (48)

where RR rotates the momentum by 9090 degrees as

R(kx,0)\displaystyle R\left(k_{x},0\right) =(0,ky),\displaystyle=\left(0,-k_{y}\right), (49)
R(0,ky)\displaystyle R\left(0,k_{y}\right) =(kx,0),\displaystyle=\left(k_{x},0\right), (50)

making the energy spectrum have 4\mathbb{Z}_{4} symmetry in the complex plane as in Eq.(41).

IV.3 Edge and corner states

The topological number is defined by (21), where ψ0\psi_{0} is the eigen function of the Hamiltonian (46) for the bulk. We find the topological insulator phase for |ta/tb|<1\left|t_{a}/t_{b}\right|<1, where Q=1Q=1 and the trivial insulator phase for |ta/tb|>1\left|t_{a}/t_{b}\right|>1, where Q=0Q=0.

We calculate the energy spectrum in square geometry numerically. C4vC_{4v} symmetry is manifest as shown in Fig.8(a). It is because the square lattice respects 4\mathbb{Z}_{4} clock symmetry. We also show the energy spectrum as a function of ta/tbt_{a}/t_{b} in Fig.8(b), where the emergence of the zero-energy states is manifest in the topological phase.

Consequently, the present model is a second-order topological insulator in the region |ta/tb|<1|t_{a}/t_{b}|<1, being characterized by the emergence of topological corner states.

It is possible to obtain numerically the wave functions of the four corner states. We have confirmed that they satisfy the relations (7) and (8). Therefore, they are 4\mathbb{Z}_{4} parafermions.

Refer to caption

Figure 9: (a) Electric circuit implementation of the breathing square model shown in Fig.1(b), where the hopping parameters correspond to those in Fig.1(b4). (b) Configuration of negative impedance converter with current inversion realizing imaginary hoppingHofmann , and (c) the one realizing nonreciprocal hoppingNonR , corresponding to those in (a).

IV.4 Electric-circuit implementation

In the Hamiltonian (46), there are nonreciprocal hopping terms along the xx axis. Nonreciprocal hopping is constructed by a combination of operational amplifier and capacitorsNonR ,

(IijIji)=iωC(1111)(ViVj),\left(\begin{array}[]{c}I_{ij}\\ I_{ji}\end{array}\right)=i\omega C\left(\begin{array}[]{cc}-1&1\\ -1&1\end{array}\right)\left(\begin{array}[]{c}V_{i}\\ V_{j}\end{array}\right), (51)

It corresponds to the Hamiltonian

H=C(1111).H=C\left(\begin{array}[]{cc}-1&1\\ -1&1\end{array}\right). (52)

We add an inductor or a capacitor in order to cancel the diagonal term.

We construct an electric circuit as shown in Fig.9(a), where the xx axis is constructed by Fig.9(c) and the yy axis is constructed by Fig.9(b).

V 6\mathbb{Z}_{6} Parafermion

Finally, we construct a 6\mathbb{Z}_{6} parafermion model. The parafermion operators are represented by the shift operatorZohar ; Fend ; AliceaPara

γ1τ=(000001100000010000001000000100000010),\gamma_{1}\equiv\tau=\left(\begin{array}[]{cccccc}0&0&0&0&0&1\\ 1&0&0&0&0&0\\ 0&1&0&0&0&0\\ 0&0&1&0&0&0\\ 0&0&0&1&0&0\\ 0&0&0&0&1&0\end{array}\right), (53)

and the clock operatorZohar ; Fend ; AliceaPara

γ2σ=diag.(1,ω,ω2,ω3,ω4,ω5).\gamma_{2}\equiv\sigma=\text{diag.}\left(1,\omega,\omega^{2},\omega^{3},\omega^{4},\omega^{5}\right). (54)

Here, τ\tau and σ\sigma satisfy the 6\mathbb{Z}_{6} parafermion relations,

τ6=σ6=1,τσ=ωστ.\tau^{6}=\sigma^{6}=1,\qquad\tau\sigma=\omega\sigma\tau. (55)

In the 6\mathbb{Z}_{6} clock-symmetric model, the energy spectrum is composed of sextets En(0,1,2,3,4,5)E_{n}^{\left(0,1,2,3,4,5\right)}, n=0,1,2,n=0,1,2,\cdots,

En(0,1,2,3,4,5)=εn,ωεn,ω2εn,ω3εn,ω4εn,ω5εn.E_{n}^{\left(0,1,2,3,4,5\right)}=\varepsilon_{n},\quad\omega\varepsilon_{n},\quad\omega^{2}\varepsilon_{n},\quad\omega^{3}\varepsilon_{n},\quad\omega^{4}\varepsilon_{n},\quad\omega^{5}\varepsilon_{n}. (56)

The system is necessarily non-Hermitian because the eigen energies are complex except for zero-energy states.

Refer to caption

Figure 10: Energy spectrum of a hexagon shown in Fig.1(c2). (a) the vertical axis is Re[EE] and the horizontal axis is Im[EE], while (b) the vertical axis is |E||E| and the horizontal axis is ta/tbt_{a}/t_{b}. Color indicates the value of ta/tbt_{a}/t_{b} in (a). C6vC_{6v} symmetry in the complex energy plane is manifest in (a). The emergence of zero-energy corner states is clear in (b). The green curves are calculated from the bulk Hamiltonian, while red curves are calculated based on the hexagon. The cyan curve is the energy at the Γ\Gamma point, while the blue curve is the energy at the KK point calculated based on the bulk Hamiltonian.

A Hermitian second-order topological insulator has been proposed on the breathing honeycomb latticeMizoguchi . We generalize it to a non-Hermitian model with 6\mathbb{Z}_{6} clock symmetry by introducing complex hoppings. The bulk Hamiltonian is defined on the breathing honeycomb lattice and given by

H=(0ω2tb0taeiky0ω4tbω5tb0ω3tb0ωtaei3kx+ky200tb0ω4tb0ω2taei3kx+ky2ω3taeiky0ωtb0ω5tb00ω4taei3kxky20ω2tb0tbωtb0ω5taei3kx+ky20ω3tb0).H=\left(\begin{array}[]{cccccc}0&\omega^{2}t_{b}&0&t_{a}e^{ik_{y}}&0&\omega^{4}t_{b}\\ \omega^{5}t_{b}&0&\omega^{3}t_{b}&0&\omega t_{a}e^{i\frac{-\sqrt{3}k_{x}+k_{y}}{2}}&0\\ 0&t_{b}&0&\omega^{4}t_{b}&0&\omega^{2}t_{a}e^{-i\frac{\sqrt{3}k_{x}+k_{y}}{2}}\\ \omega^{3}t_{a}e^{-ik_{y}}&0&\omega t_{b}&0&\omega^{5}t_{b}&0\\ 0&\omega^{4}t_{a}e^{i\frac{\sqrt{3}k_{x}-k_{y}}{2}}&0&\omega^{2}t_{b}&0&t_{b}\\ \omega t_{b}&0&\omega^{5}t_{a}e^{i\frac{\sqrt{3}k_{x}+k_{y}}{2}}&0&\omega^{3}t_{b}&0\end{array}\right). (57)

We diagonalize this Hamiltonian for a hexagon shown in Fig.1(c2). C6vC_{6v} symmetry is manifest in the complex energy plane as in Fig.10(a). Furthermore, we find six zero-energy topological corner states for |ta/tb|<1\left|t_{a}/t_{b}\right|<1, as shown in Fig.10(b). We also find the trivial insulator phase for ta/tb<2t_{a}/t_{b}<-2 and ta/tb>1t_{a}/t_{b}>1. Additionally, there is metallic phase for 2<ta/tb<1-2<t_{a}/t_{b}<-1.

VI Conclusion

We have constructed a 3\mathbb{Z}_{3} parafermion model on the breathing Kagome lattice, a 4\mathbb{Z}_{4} parafermion model on the breathing square lattice and a 6\mathbb{Z}_{6} parafermion model on the breathing honeycomb lattice. These model exhaust all the possible realization of d\mathbb{Z}_{d} parafermions since there are only three-fold, four-fold and six-fold rotational symmetries that are compatible with the periodic lattices. We note that two-fold symmetry corresponds to Majorana fermions.

The author is very much grateful to Y. Tanaka and N. Nagaosa for helpful discussions on the subject. This work is supported by the Grants-in-Aid for Scientific Research from MEXT KAKENHI (Grants No. JP17K05490 and No. JP18H03676). This work is also supported by CREST, JST (JPMJCR16F1 and JPMJCR20T2).

References

  • (1) G. Moore and N. Read, Nucl. Phys. B 360, 362 (1991).
  • (2) S. Das Sarma, M. Freedman, and C. Nayak, Phys. Rev. Lett. 94, 166802 (2005).
  • (3) A. Kitaev, Annals of Physics 321, 2 (2006).
  • (4) C. Nayak, S. H. Simon, A. Stern, M. Freedman, and S. Das Sarma, Rev. Mod. Phys. 80, 1083 (2008).
  • (5) A. Stern, Ann. Physics 323, 204 (2008).
  • (6) A. Stern, Nature 464, 187 (2010).
  • (7) S. Das Sarma, M. Freedman, C. Nayak, npj Quantum Information 1, 15001 (2015).
  • (8) C. W.J. Beenakker, Annu. Rev. Condens. Matter Phys. 4, 113 (2013).
  • (9) T. D. Stanescu and S. Tewari, J. Phys. Condens. Matter 25, 233201 (2013).
  • (10) S.R. Elliott and M. Franz, Rev. Mod. Phys. 87, 137 (2015).
  • (11) D. Aasen, M. Hell, R. V. Mishmash, A. Higginbotham, J. Danon, M. Leijnse, T. S. Jespersen, J. A. Folk, C. M. Marcus, K. Flensberg, and J. Alicea, Phys. Rev. X 6, 031016 (2016).
  • (12) D. A. Ivanov Phys. Rev. Lett. 86, 268, (2001).
  • (13) B. I. Halperin, Y. Oreg, A. Stern, G. Refael, J. Alicea and F. von Oppen, Phys. Rev. B 85, 144501 (2012).
  • (14) J. Alicea, Y. Oreg, G. Refael, F. von Oppen and M.P.A. Fisher, Nat. Phys. 7, 412 (2011).
  • (15) X.-L. Qi, S.-C. Zhang, Rev. Mod. Phys. 83, 1057 (2011).
  • (16) M. Leijnse and K. Flensberg, Semicond. Sci. Technol. 27, 124003 (2012).
  • (17) J. Alicea, Rep. Prog. Phys. 75, 076501 (2012).
  • (18) M. Sato and Y. Ando, Rep. Prog. Phys. 80, 076501 (2017).
  • (19) Y. Tanaka, M. Sato and N. Nagaosa, J. Phys. Soc. Jpn. 81, 011013 (2012).
  • (20) Y. Kasahara, T. Ohnishi, Y. Mizukami, O. Tanaka, Sixiao Ma, K. Sugii, N. Kurita, H. Tanaka, J. Nasu, Y. Motome, T. Shibauchi, Y. Matsuda, Nature 559, 227 (2018).
  • (21) A. Ahlbrecht, L. S. Georgiev, R. F. Werner, Phys. Rev. A 79, 032311 (2009).
  • (22) P. Fendley, J. Stat. Mech. 2012, 11020 (2012).
  • (23) P. Fendley, J. Phys. A 47, 075001 (2014).
  • (24) J. Alicea, P. Fendley, Annual Review of Condensed Matter Physics 7, 119 (2016).
  • (25) Adam S. Jermyn, Roger S. K. Mong, Jason Alicea, and Paul Fendley, Phys. Rev. B 90, 165106 (2014).
  • (26) H. Ebisu, E. Sagi, Y. Tanaka, and Y. Oreg, Phys. Rev. B 95, 075111 (2017)
  • (27) A. Hutter and D. Loss, Phys. Rev. B 93, 125105 (2016).
  • (28) R. J. Baxter, Phys. Lett. A 140, 155 (1989); J. Statist. Phys. 57, 1 (1989).
  • (29) N. Read and E. Rezayi, Phys. Rev. B 59, 8084 (1999).
  • (30) D. J. Clarke, J. Alicea, K. Shtengel, Nat. Com. 4, 1348 (2013).
  • (31) K. Laubscher, D. Loss, J. Klinovaja, Phys. Rev. Research 1, 032017 (2019).
  • (32) K. Laubscher, D. Loss, J. Klinovaja, Phys. Rev. Research 2, 013330 (2020).
  • (33) F. Zhang, C.L. Kane and E.J. Mele, Phys. Rev. Lett. 110 , 046404 (2013).
  • (34) W. A. Benalcazar, B. A. Bernevig, and T. L. Hughes, Science 357, 61 (2017).
  • (35) F. Schindler, A. Cook, M. G. Vergniory, and T. Neupert, in APS March Meeting (2017).
  • (36) Y. Peng, Y. Bao, and F. von Oppen, Phys. Rev. B 95, 235143 (2017).
  • (37) J. Langbehn, Y. Peng, L. Trifunovic, F. von Oppen, and P. W. Brouwer, Phys. Rev. Lett. 119, 246401 (2017).
  • (38) Z. Song, Z. Fang, and C. Fang, Phys. Rev. Lett. 119, 246402 (2017).
  • (39) W. A. Benalcazar, B. A. Bernevig, and T. L. Hughes, Phys. Rev. B 96, 245115 (2017).
  • (40) F. Schindler, A. M. Cook, M. G. Vergniory, Z. Wang, S. S. P. Parkin, B. A. Bernevig, and T. Neupert, Science Advances 1, eaat0346 (2018).
  • (41) C. Fang, L. Fu, Science Advances 5, eaat2374 (2019).
  • (42) M. Ezawa, Phys. Rev. Lett. 120, 026801 (2018).
  • (43) M. Ezawa, Phys. Rev. B 98, 045125 (2018).
  • (44) M. Geier, L. Trifunovic, M. Hoskam, and P. W. Brouwer, Phys. Rev. B 97, 205135 (2018).
  • (45) E. Khalaf, Phys. Rev. B 97, 205136 (2018).
  • (46) M. Ezawa, Phys. Rev. B 100, 045407 (2019).
  • (47) T. Liu, Y.-R. Zhang, Q. Ai, Z. Gong, K. Kawabata, M. Ueda, F. Nori, Phys. Rev. Lett. 122, 076801 (2019).
  • (48) M. Ezawa, Phys. Rev. B 99, 201411(R) (2019).
  • (49) M. Ezawa, Phys. Rev. B 99, 121411(R) (2019).
  • (50) E. Edvardsson, F. K. Kunst, E. J. Bergholtz, Phys. Rev. B 99, 081302 (2019).
  • (51) G. Ortiz, E. Cobanera, Z. Nussinov, Nuc. Phys. B 854, 780 (2012).
  • (52) X. Ni, M. Weiner, A. Alu and A. B. Khanikaev, Nature Materials 18, 113 (2019).
  • (53) Z. Gong, Y. Ashida, K. Kawabata, K. Takasan, S. Higashikawa and M. Ueda, Phys. Rev. X 8, 031079 (2018).
  • (54) K. Kawabata, K. Shiozaki, M. Ueda, and M. Sato, Phys. Rev. X 9, 041015 (2019).
  • (55) S. Imhof, C. Berger, F. Bayer, J. Brehm, L. Molenkamp, T. Kiessling, F. Schindler, C. H. Lee, M. Greiter, T. Neupert, R. Thomale, Nat. Phys. 14, 925 (2018).
  • (56) C. H. Lee , S. Imhof, C. Berger, F. Bayer, J. Brehm, L. W. Molenkamp, T. Kiessling and R. Thomale, Communications Physics, 1, 39 (2018).
  • (57) T. Helbig, T. Hofmann, C. H. Lee, R. Thomale, S. Imhof, L. W. Molenkamp and T. Kiessling, Phys. Rev. B 99, 161114 (2019).
  • (58) Y. Lu, N. Jia, L. Su, C. Owens, G. Juzeliunas, D. I. Schuster and J. Simon, Phys. Rev. B 99, 020302 (2019).
  • (59) Y. Li, Y. Sun, W. Zhu, Z. Guo, J. Jiang, T. Kariyado, H. Chen and X. Hu, Nat. Com. 9, 4598 (2018).
  • (60) M. Ezawa, Phys. Rev. B 98, 201402(R) (2018).
  • (61) M. Serra-Garcia, R. Susstrunk and S. D. Huber, Phys. Rev. B 99, 020304 (2019).
  • (62) T. Hofmann, T. Helbig, C. H. Lee, M. Greiter, R. Thomale, Phys. Rev. Lett. 122, 247702 (2019).
  • (63) M. Ezawa, Phys. Rev. B 102, 075424 (2020).
  • (64) T. Helbig, T. Hofmann, S. Imhof, M. Abdelghany, T. Kiessling, L. W. Molenkamp, C. H. Lee, A. Szameit, M. Greiter, R. Thomale, Nature Physics 16, 747 (2020)
  • (65) T. Mizoguchi, H. Araki and Y. Hatsugai, J. Phys. Soc. Jpn. 88, 104703 (2019)