Thermal Hall conductivity with sign change in the Heisenberg&ndash;Kitaev kagome magnet

Kangkang Li

doi:10.1088/1572-9494/aca00e

PDF(846 KB)

Communications in Theoretical Physics ›› 2023, Vol. 75 ›› Issue (1) : 15702. DOI: 10.1088/1572-9494/aca00e

Condensed Matter Theory

Thermal Hall conductivity with sign change in the Heisenberg–Kitaev kagome magnet

Kangkang Li

Author information +

History +

Abstract

The Heisenberg–Kitaev (HK) model on various lattices has attracted a lot of attention because it may lead to exotic states such as quantum spin liquid and topological orders. The rare-earth-based kagome lattice (KL) compounds ${M g}_{2} {R E}_{3} {S b}_{3} O_{14}$ $(R E = G d, E r)$ and $(R E = N d)$ have q = 0, 120° order and canted ferromagnetic (CFM) order, respectively. Interestingly, the HK model on the KL has the same ground state long-range orders. In the theoretical phase diagram, the CFM phase resides in a continuous parameter region and there is no phase change across special parameter points, such as the Kitaev ferromagnetic (KFM) point, the ferromagnetic (FM) point and its dual FM point. However, a ground state property cannot distinguish a system with or without topological nontrivial excitations and related phase transitions. Here, we study the topological magnon excitations and related thermal Hall conductivity in the HK model on the KL with CFM order. The CFM phase can be divided into two regions related by the Klein duality, with the self dual KFM point as their boundary. We find that the scalar spin chirality, which is intrinsic in the CFM order, changes sign across the KFM point. This leads to the opposite Chern numbers of corresponding magnon bands in the two regions, and also the sign change of the magnon thermal Hall conductivity.

Key words

canted ferromagnetic order / topological magnon / thermal Hall conductivity / sign change / Heisenberg–Kitaev kagome magnet

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

Kangkang Li. Thermal Hall conductivity with sign change in the Heisenberg–Kitaev kagome magnet[J]. Communications in Theoretical Physics, 2023, 75(1): 15702 https://doi.org/10.1088/1572-9494/aca00e

1. Introduction

The Heisenberg–Kitaev (HK) model on various lattices such as triangular, kagome, pyrochlore, hyperkagome and fcc lattices has attracted a lot of attention because there are frustrations from the competing exchange couplings coexisting with the geometric frustration of the underlying lattices, which may lead to exotic states such as quantum spin liquid and topological orders [1–11]. The studies were also connected to real materials, such as the layered honeycomb materials α-RuCl₃, Na₂IrO₃ and α-Ir₂IrO₃, the three-dimensional (3D) honeycomb material (β-γ)-Li₂IrO₃, and the iridium oxides with triangular lattices [12–42].

The rare-earth-based kagome lattice (KL) compounds

{M g}_{2} {R E}_{3} {S b}_{3} O_{14}

(R E = G d, E r)

and

(R E = N d)

have q = 0, 120° order and canted ferromagnetic (CFM) order [43, 44], respectively. These compounds, except for

R E = G d

, have an effective spin with S = 1/2 on the KL [45]. Since the exchange couplings between the nearest-neighbor (NN) spins are anisotropic, Morita et al studied the ground state of the classical and quantum spin HK model with anisotropic bond-dependent Kitaev interactions on the KL [46]. Interestingly, the ground state has the same type q = 0, 120° order and CFM order as those observed in the compounds. In the theoretical phase diagram, the CFM order resides in a continuous parameter region, and there is no phase change across special parameter points, such as the Kitaev ferromagnetic (KFM) point, the ferromagnetic (FM) point and its dual FM point. However, a ground state property cannot distinguish a system with or without topological nontrivial excitations and related phase transitions. The non-coplanar CFM order naturally has nonzero scalar spin chirality, which will provide a vector potential for magnons [47]. Then one may wonder if there are topological magnon excitations and the related thermal Hall effect in the system. Furthermore, whether is it possible for a topological phase transition of magnon excitations to happen on the continuous CFM ground state.

Here, we study the topological magnon excitations and related thermal Hall conductivity in the HK model on the KL with CFM order. In the CFM order, the solid angle spanned by the three spins in a unit cell changes continuously with the model parameter [48]. However, the continuous CFM phase can be divided into two regions related by the Klein duality, with the self dual KFM point as their boundary [48], and we are surprised to find that the scalar spin chirality which is intrinsic in the CFM order changes sign across the KFM point. This leads to the opposite Chern numbers of corresponding magnon bands in the two regions, and also the sign change of the magnon thermal Hall conductivity.

This paper is organized as follows. In section 2, we revisit the CFM phase in the model and investigate the scalar spin chirality. In section 3, we present the magnon band structures and discuss their topological properties. In section 4, we show the transverse thermal Hall conductivity with sign change phenomena. Finally, a summary is given in section 5.

2. Model, the CFM phase and scalar spin chirality

We first revisit the model and the CFM phase [46, 48]. Consider interacting spins that reside on the KL as shown in figure 1(a). The HK model is described by the spin Hamiltonian

\begin{array}{rcl} H = \sum_{⟨ i j ⟩} (J S_{i} \cdot S_{j} + {K S}_{i}^{γ_{i j}} S_{j}^{γ_{i j}}), \end{array}

(1)

where S_i and S_j are spin S = 1/2 spins reside on the NN lattice sites, and J and K denote the Heisenberg and Kitaev exchange couplings, respectively. The Cartesian components γ_ij equals x, y or z, depending on the bond type as shown in figure 1(a). The model can be parameterized by

\begin{array}{rcl} J = \cos ψ, K = \sin ψ, ψ \in [0, 2 π), \end{array}

(2)

with the energy unit J² + K² = 1.

Figure 1. (a) The structure of the KL. There are three spins that reside in a primitive cell, which are denoted by red, green and blue sites. Red, green and blue bonds between NN sites (i, j) carry three distinct Kitaev couplings $S_{i}^{x} S_{j}^{x}$ , $S_{i}^{y} S_{j}^{y}$ and $S_{i}^{x} S_{j}^{x}$ , respectively. Meanwhile, there are isotropic Heisenberg couplings in all the NN bonds. The KL sits on the (111) plane, and we define a new 2D frame $x^{'} y^{'}$ on the KL with basis vectors a₁ = (1, 0) and $a_{2} = (1, \sqrt{3}) / 2$ . (b) The first Brillouin zone of the KL.

Full size|PPT slide

In the classical ground state phase diagram of the HK model on KL, there is a long-range ordered CFM phase for

ψ \in [π - \arctan (2), 2 π]

, which has eightfold degeneracy. At the phase boundary are the classical spin liquid (CSL) state and dual CSL state corresponding to the Heisenberg point with K = 0 and its dual point, respectively. Inside the CFM phase, there are other three special points, the KFM point with J = 0, the FM point with K = 0 and its dual FM point (see figure 2(a)). There is no phase transition of the ground state across the whole CFM phase. However, we find that the collective magnon excitations on the CFM ground state do have topological phase transition at the KFM point as shown in figure 2(a). Interestingly, we can divide the CFM phase into two regions related by the Klein duality [46, 48, 49], with the self-dual KFM point as the boundary. The Klein duality preserves the form of the Hamiltonian but alter the value of the parameters J and K, and in the KL it reads

\begin{array}{rcl} \tilde{H} = \sum_{⟨ i j ⟩} (\tilde{J} {\tilde{S}}_{i} \cdot {\tilde{S}}_{j} + \tilde{K} {\tilde{S}}_{i}^{γ_{i j}} {\tilde{S}}_{j}^{γ_{i j}}), \end{array}

(3)

where

\begin{array}{rcl} \tilde{J} = - J, \tilde{K} = 2 J + K, \end{array}

(4)

and the spin operators transform to

\begin{array}{rcl} {\tilde{S}}_{1} = (\begin{array}{c} - S_{1}^{x} \\ S_{1}^{y} \\ - S_{1}^{z} \end{array}), {\tilde{S}}_{2} = (\begin{array}{c} S_{2}^{x} \\ - S_{2}^{y} \\ - S_{2}^{z} \end{array}), {\tilde{S}}_{3} = (\begin{array}{c} S_{3}^{x} \\ S_{3}^{y} \\ S_{3}^{z} \end{array}) . \end{array}

(5)

Figure 2. (a) Evolution of λ (equation (7)) with ψ and the phase diagram of topological magnons in the CFM phase. In the region $ψ \in (π - \arctan (2), 1.5 π)$ shaded cyan, the three magnon bands have Chern numbers 1, 0, −1, while in the region ψ ∈ (1.5π, 2π) shaded pink they are −1, 0, 1. This denotes a topological phase transition at the KFM point. Note that the Chern numbers vanish at the FM, dual FM, CSL, dual CSL and KFM points where the magnon bands are gapless. The points ψ = 1.3789π and ψ = 1.5696π are a normal pair of dual points. (b) Schematic CFM spin order with χ = η = ζ = 1 and the evolution of the directions of the three spins in a primitive cell with ψ.

Full size|PPT slide

The spin structure of the CFM phase can be described by the directions of the three spins in a primitive cell [48]

\begin{array}{rcl} S_{1} = ξ S (\begin{array}{c} - λ χ \\ η \\ ζ \end{array}), S_{2} = ξ S (\begin{array}{c} χ \\ - λ η \\ ζ \end{array}), S_{3} = ξ S (\begin{array}{c} χ \\ η \\ - λ ζ \end{array}), \end{array}

(6)

where

ξ = 1 / \sqrt{2 + λ^{2}}

and

\begin{array}{rcl} λ = (\cos ψ + \sin ψ + \sqrt{5 + 4 \cos 2 ψ + \sin 2 ψ}) / (2 \cos ψ) . \end{array}

(7)

The eight degenerate spin orders are given by χ = ± 1, η = ± 1 and ζ = ± 1. The magnon excitations of the eight different spin orders have the same band structures, however different wave functions and therefore different topological properties. Here, we study the order given by χ = η = ζ = 1, which is the same as the one observed in experiment [44], as shown in figure 2(b).

Since the three spins in a primitive cell are non-coplanar in the CFM order, there is intrinsically non-zero scalar spin chirality. In the spin order with χ = η = ζ = 1, the scalar spin chirality is

\begin{array}{rcl} S_{1} \cdot (S_{2} \times S_{3}) {/ S}^{3} = \frac{- λ^{3} + 3 λ + 2}{{(λ^{2} + 2)}^{3 / 2}}, \end{array}

(8)

and its dual one is

\begin{array}{rcl} {\tilde{S}}_{1} \cdot ({\tilde{S}}_{2} \times {\tilde{S}}_{3}) {/ S}^{3} = \frac{- λ^{3} + 3 λ - 2}{{(λ^{2} + 2)}^{3 / 2}} . \end{array}

(9)

It turns out that the chirality is always positive except at the FM and CSL points, where it vanishes. Oppositely, the dual chirality is always negative and vanishes at the dual FM and dual CSL points, as shown in figure 3(a). Here we are surprised to find that the scalar spin chirality for the magnons changes sign at the KFM point, where it jumps from the original chirality to its dual one. We attribute it to the relative size of the corresponding solid angle (see appendix A) and its dual one. On the left side of the KFM point, the solid angle Ω spanned by spins S_1,2,3 is smaller than its dual one

\tilde{Ω}

(see figure 3(b)), which is spanned by the dual spins

{\tilde{S}}_{1, 2, 3}

, and the chirality for magnons is S₁ · (S₂ × S₃). However, the situation is opposite in the right side of the KFM point, and now the chirality for magnons will be

{\tilde{S}}_{1} \cdot ({\tilde{S}}_{2} \times {\tilde{S}}_{3})

Figure 3. (a) Evolutions of the scalar spin chirality (equation (8)) (blue line) and its dual one (equation (9)) (red line) with ψ. On the left side of the KFM point, the scalar spin chirality for magnons evolves with the blue line, while on the right side, it jumps to the red line. (b) Evolutions of the solid angle Ω and its dual one $\tilde{Ω}$ with ψ. On the left side of the KFM point, the solid angle Ω spanned by spins S_1,2,3 is much smaller, while the situation is opposite on the right side.

Full size|PPT slide

The scalar spin chirality can provides a vector potential for magnons, which will lead to topological magnons and thermal Hall effect, as previous studies on ferromagnets with Dzyaloshinskii–Moriya (DM) interaction induced vector potential [50–55]. The more interesting thing here is the scalar spin chirality changes sign across the KFM point. This will lead to the opposite Chern numbers of corresponding magnon bands in the two dual regions, and also the sign change of the magnon thermal Hall conductivity, as we show in the next sections.

3. Band structure and topological magnons

Now we turn to study the collective magnon excitations on the CFM order with χ = η = ζ = 1. Starting from the Hamiltonian (1), after the Holstein–Primakoff (HP) [56] and Fourier transformations, we get the magnon Hamiltonian matrix

h (k^{'})

(see appendix B). Then after diagonalizing

I_{-} h (k^{'})

, we get three magnon bands. The magnon bands are gapped in all the CFM phase, except at the CSL, dual CSL, FM, dual FM and KFM points as shown in the top panel of figure 4. Due to the same form of the Hamiltonian, the band structures are the same for any pair of dual points, except for the energy scale. At the FM and dual FM points, the magnon bands are the same as the isotropic Heisenberg model on KL [53]. At the CSL and dual CSL points, there are zero energy flat bands, which means the magnon modes now can excite without energy cost. Moreover, the possibilities for the excited magnon modes with different

k^{'}

are the same, and this will lead to the breakdown of the long-range order. Here the breakdown of the CFM order gives the CSL phase. At the KFM point, there is also a flat band, however with finite energy. Hence, to excite the flat band modes now will cost energy, and the CFM order is protected by the energy gap.

Figure 4. Top panel: magnon bands of the CFM order at the (a) FM point, (c) KFM point, (e) CSL point, and (b) and (d) a pair of dual points. The magnon bands are gapped in all the CFM phase, except at the CSL, dual CSL, FM, dual FM and KFM points. The points (b) ψ = 1.3789π and (d) ψ = 1.5696π also denoted in figure 2(a) are a pair of dual points, and their band structures are the same except for the energy scale. The bands in (b) carry Chern numbers 1, 0, −1, while they are −1, 0, 1 in (d). Bottom panel: the corresponding magnon bands in a strip geometry. There are in-gap edge modes for the gapped band structures, which are promised by the nonzero Chern numbers of the corresponding bulk bands.

Full size|PPT slide

Though the band structures of the dual points are the same, the wave functions and topological properties are not necessarily the same. To show the topological property of the gapped magnon bands, we calculate their Chern numbers. The Chern number of the nth band is defined by the integration of the Berry curvature over the first Brillouin zone [57]

\begin{array}{rcl} C_{n} = \frac{1}{2 π} \int_{B Z} d k_{x}^{'} d k_{y}^{'} B_{k_{x}^{'} k_{y}^{'}}^{n}, \end{array}

(10)

with

\begin{array}{rcl} B_{k_{x}^{'} k_{y}^{'}}^{n} = i \sum_{n^{'} \neq n} \frac{⟨ ϕ_{n} | \frac{\partial h (k^{'})}{\partial k_{x}^{'}} | ϕ_{n^{'}} ⟩ ⟨ ϕ_{n^{'}} | \frac{\partial h (k^{'})}{\partial k_{y}^{'}} | ϕ_{n} ⟩ - (k_{x}^{'} \leftrightarrow k_{y}^{'})}{{(E_{n} - E_{n^{'}})}^{2}}, \end{array}

(11)

where E_n and φ_n are the eigenvalue and eigenvector of the nth band respectively. It turns out that the Chern numbers are (C₁, C₂, C₃) = (1, 0, − 1) in the region

ψ \in (π - \arctan (2), 1.5 π)

, while they are (−1, 0, 1) in the region ψ ∈ (1.5π, 2π). This denotes a topological phase transition at the KFM point ψ = 1.5π (see figure 2(a)). We show the gapped magnon bands of a pair of dual points in figures 4(b) and (d), and their Chern numbers are opposite. As mentioned in the above section, it is the sign change of the scalar spin chirality for magnons that leads to the opposite Berry curvatures and Chern numbers of corresponding magnon bands in the two regions related by Klein duality. Due to the bulk-edge correspondence [58], the nonzero Chern numbers will promise in-gap edge modes in a strip geometry as shown in figures 4(g) and (i).

4. Transverse thermal Hall conductivity with sign change

As previous studies on ferromagnets with DM interaction induced vector potential, the scalar spin chirality here can also provide a vector potential for magnons, which will lead to the magnon thermal Hall effect. Furthermore, the sign change of the chirality here will also induce the sign change of the thermal conductivity simultaneously. The transverse thermal Hall conductivity can be calculated as [59–61]

\begin{array}{rcl} κ_{x^{'} y^{'}} = \frac{k_{B}^{2} T}{{(2 π)}^{2} ℏ} \sum_{n} \int_{B Z} c_{2} (ρ_{n}) B_{k_{x}^{'} k_{y}^{'}}^{n} d k_{x}^{'} d k_{y}^{'}, \end{array}

(12)

with the sum running over the three bands and the integral is over the first Brillouin zone.

ρ_{n} = 1 / (\exp (E_{n} / k_{B} T) - 1)

is the Bose distribution with E_n as the nth eigenvalue. c₂ is given by

\begin{array}{rcl} c_{2} (ρ_{n}) = (1 + ρ_{n}) {(l n \frac{1 + ρ_{n}}{ρ_{n}})}^{2} - {(l n ρ_{n})}^{2} - 2 {L i}_{2} (- ρ_{n}), \end{array}

(13)

where Li₂(x) is the polylogarithm function of order 2.

As shown in figures 5(a) and (b), the transverse thermal conductivity for

ψ \in (π - \arctan (2), 1.5 π)

is always positive, while it is always negative for ψ ∈ (1.5π, 2π). At the CSL, dual CSL, FM, dual FM and KFM points, the magnon bands are gapless, leading to vanishing thermal Hall conductivity. There is sign change of the thermal conductivity at the KFM point, and we attribute it to the sign change of the scalar spin chirality. As the chirality changes sign, the vector potential for magnons becomes opposite, and the propagating magnons will be deflected in the opposite direction, which results in the opposite transverse conductivity.

Figure 5. (a) The transverse thermal conductivity as a function of temperature for different ψ. For $ψ \in (π - \arctan (2), 1.5 π)$ , the conductivity is always positive, while for ψ ∈ (1.5π, 2π) it is always negative. At the CSL, dual CSL, FM, dual FM and KFM points, the conductivity is always zero. (b) The transverse thermal conductivity as a function of ψ with T = 3. There is sign change at the KFM point. Here we set ℏ = k_B = 1.

Full size|PPT slide

The sign change of the thermal Hall conductivity is also consistent with the opposite Chern numbers of corresponding magnon bands. Consider the high temperature limit of the transverse thermal conductivity [55, 62]

\begin{array}{rcl} κ_{x^{'} y^{'}}^{l i m} = \underset{T \to \infty}{l i m} κ_{x^{'} y^{'}} = - \frac{k_{B}}{{(2 π)}^{2} ℏ} \sum_{n} \int_{B Z} E_{n} B_{k_{x}^{'} k_{y}^{'}}^{n} d k_{x}^{'} d k_{y}^{'}, \end{array}

(14)

it can be further simplified as

\begin{array}{rcl} κ_{x^{'} y^{'}}^{l i m} \propto - \frac{k_{B}}{{(2 π)}^{2} ℏ} \sum_{n} C_{n} {\bar{E}}_{n}, \end{array}

(15)

where the

k^{'}

-dependent band energy is replaced by the average

{\bar{E}}_{n}

. Then the high temperature thermal conductivity of the phase with Chern numbers (1, 0, −1) is

{\bar{E}}_{3} - {\bar{E}}_{1}

which is positive, while the conductivity of the phase with Chern numbers (−1, 0, 1) is

{\bar{E}}_{1} - {\bar{E}}_{3}

which is negative. Then the sign change of the chirality, the opposite Chern numbers of corresponding bands and the sign change of the thermal conductivity are consistent with each other.

In the end of this section, we will discuss briefly the eight degenerate CFM spin orders. We find that the chirality of the orders with [χηζ] = [111],

[1 \bar{1} \bar{1}]

[\bar{1} 1 \bar{1}]

, and

[\bar{1} \bar{1} 1]

\begin{array}{rcl} S_{1} \cdot (S_{2} \times S_{3}) {/ S}^{3} = \frac{- λ^{3} + 3 λ + 2}{{(λ^{2} + 2)}^{3 / 2}}, \end{array}

(16)

and its dual one is

\begin{array}{rcl} {\tilde{S}}_{1} \cdot ({\tilde{S}}_{2} \times {\tilde{S}}_{3}) {/ S}^{3} = \frac{- λ^{3} + 3 λ - 2}{{(λ^{2} + 2)}^{3 / 2}} . \end{array}

(17)

However, the chirality of the orders

[\bar{1} \bar{1} \bar{1}]

[\bar{1} 11]

[1 \bar{1} 1]

, and

[11 \bar{1}]

are opposite

\begin{array}{rcl} S_{1} \cdot (S_{2} \times S_{3}) {/ S}^{3} = \frac{λ^{3} - 3 λ - 2}{{(λ^{2} + 2)}^{3 / 2}}, \end{array}

(18)

\begin{array}{rcl} {\tilde{S}}_{1} \cdot ({\tilde{S}}_{2} \times {\tilde{S}}_{3}) {/ S}^{3} = \frac{λ^{3} - 3 λ + 2}{{(λ^{2} + 2)}^{3 / 2}} . \end{array}

(19)

Besides, the solid angles Ω and

\tilde{Ω}

keep the same for all eight orders. This suggests that the Chern numbers of the magnon bands and the sign of the thermal Hall conductivity are the same for the orders [111],

[1 \bar{1} \bar{1}]

[\bar{1} 1 \bar{1}]

, and

[\bar{1} \bar{1} 1]

, while they are opposite to that of the orders

[\bar{1} \bar{1} \bar{1}]

[\bar{1} 11]

[1 \bar{1} 1]

, and

[11 \bar{1}]

5. Summary

We have studied the topological magnon excitations and related thermal Hall conductivity in the HK model on the KL with CFM order. The CFM phase can be divided into two regions related by the Klein duality [46, 48, 49], with the self dual KFM point as their boundary. We find that the scalar spin chirality, which is intrinsic in the CFM order, changes sign across the KFM point. This leads to the opposite Chern numbers of corresponding magnon bands in the two regions, and also the sign change of the magnon thermal Hall conductivity. Moreover, we have checked that for the coplanar q = 0, 120° order [46] the magnon bands are always gapless, and there is no thermal Hall conductivity. Interestingly, the rare-earth-based KL compounds

{M g}_{2} {R E}_{3} {S b}_{3} O_{14}

(R E = G d, E r)

and

(R E = N d)

have the same q = 0, 120° order and CFM order [43, 44], respectively. Though the chiral edge modes are difficult to measure in an experiment, the sign structure of the thermal conductivity is more accessible [63]. Therefore, the study of the topological magnons and related thermal Hall conductivity here will contribute to the understanding of related compounds.

Acknowledgments

We thank Changle Liu for helpful discussions. This work was supported by the National Natural Science Foundation of China (Grant NO. 12 104 407) and the Natural Science Foundation of Zhejiang Province (Grant NO. LQ20A040004).

Appendix A. Solid angle of the three spins

The solid angle Ω spanned by the three spins in a primitive cell can be calculated as

\begin{array}{rcl} \tan (\frac{Ω}{4}) = \sqrt{\tan (\frac{s}{2}) \tan (\frac{s - α}{2}) \tan (\frac{s - β}{2}) \tan (\frac{s - γ}{2})}, \end{array}

(A1)

where

s = \frac{1}{2} (α + β + γ)

, and α, β and γ are the angle between spins S_1,2, S_2,3 and S_3,1 respectively. The solid angle

\tilde{Ω}

spanned by the dual spins

{\tilde{S}}_{1, 2, 3}

has the same definition.

Appendix B. Magnon Hamiltonian matrix

We denote the directions of spins S_i=1,2,3 by their polar angles θ_1,2,3 and azimuthal angles φ_1,2,3 in the global frame. The HP transformation for a spin in its local frame reads

\begin{array}{rcl} S_{x}^{0} = \frac{\sqrt{2 S}}{2} (a + a^{†}), \end{array}

(B1)

\begin{array}{rcl} S_{y}^{0} = \frac{\sqrt{2 S}}{2 i} (a - a^{†}), \end{array}

(B2)

\begin{array}{rcl} S_{z}^{0} = S - a^{†} a, \end{array}

(B3)

where a^† and a are the magnon creation and annihilation operators respectively, which obey the boson commutation rules. Then by multiplying a rotation matrix we get the HP transformation for spin S_i in the global frame

\begin{array}{rcl} (\begin{array}{c} S_{i x} \\ S_{i y} \\ S_{i z} \end{array}) = (\begin{array}{ccc} \cos θ_{i} \cos ϕ_{i} & - \sin ϕ_{i} & \sin θ_{i} \cos ϕ_{i} \\ \cos θ_{i} \sin ϕ_{i} & \cos ϕ_{i} & \sin θ_{i} \sin ϕ_{i} \\ - \sin θ_{i} & 0 & \cos θ_{i} \end{array}) (\begin{array}{c} S_{i x}^{0} \\ S_{i y}^{0} \\ S_{i z}^{0} \end{array}) . \end{array}

(B4)

Substituting S_ix,y,z into the Hamiltonian (1) and then do the Fourier transformation, we get the quadratic Hamiltonian in momentum space

\begin{array}{rcl} H = \frac{1}{2} \sum_{k^{'}} Ψ_{k^{'}}^{†} h (k^{'}) Ψ_{k^{'}}, \end{array}

(B5)

where

Ψ_{k^{'}}^{†} = (a_{1 k^{'}}^{†}, a_{2 k^{'}}^{†}, a_{3 k^{'}}^{†}, a_{1 - k^{'}}, a_{2 - k^{'}}, a_{3 - k^{'}})

. The magnon Hamiltonian matrix is

\begin{array}{rcl} h (k^{'}) = (\begin{array}{cc} A_{k^{'}} & B_{k^{'}}^{†} \\ B_{k^{'}} & A_{k^{'}}^{*} \end{array}) S, \end{array}

(B6)

with

A_{k^{'}}

and

B_{k^{'}}

3 × 3 matrices. Their elements are as follows

\begin{array}{rcl} A_{11} = - 2 J (c_{z} + e_{z}) - 2 K (c_{8} + e_{7}), \end{array}

(B7)

\begin{array}{rcl} A_{22} = - 2 J (c_{z} + d_{z}) - 2 K (c_{8} + d_{6}), \end{array}

(B8)

\begin{array}{rcl} A_{33} = - 2 J (d_{z} + e_{z}) - 2 K (d_{6} + e_{7}), \end{array}

(B9)

\begin{array}{rcl} A_{12} = [J (c_{x} + c_{y} - i c_{x y} + i c_{y x}) + {K c}_{3}] \cos k^{'} \cdot δ_{1}, \end{array}

(B10)

\begin{array}{rcl} A_{21} = A_{12}^{*}, \end{array}

(B11)

\begin{array}{rcl} \begin{array}{rcl} A_{13} & = & [J (e_{x} + e_{y} + i e_{x y} - i e_{y x}) \\ + K (e_{2} + e_{5} + i e_{10} - i e_{12})] \cos k^{'} \cdot δ_{3}, \end{array} \end{array}

(B12)

\begin{array}{rcl} A_{31} = A_{13}^{*}, \end{array}

(B13)

\begin{array}{rcl} \begin{array}{rcl} A_{23} & = & [J (d_{x} + d_{y} - i d_{x y} + i d_{y x}) \\ + K (d_{1} + d_{4} - i d_{9} + i d_{11})] \cos k^{'} \cdot δ_{2}, \end{array} \end{array}

(B14)

\begin{array}{rcl} A_{32} = A_{23}^{*}, \end{array}

(B15)

\begin{array}{rcl} B_{12} = B_{21} = [J (c_{x} - c_{y} - i c_{x y} - i c_{y x}) + {K c}_{3}] \cos k^{'} \cdot δ_{1}, \end{array}

(B16)

\begin{array}{rcl} \begin{array}{rcl} B_{13} & = & B_{31} = [J (e_{x} - e_{y} - i e_{x y} - i e_{y x}) \\ + K (e_{2} - e_{5} - i e_{10} - i e_{12})] \cos k^{'} \cdot δ_{3}, \end{array} \end{array}

(B17)

\begin{array}{rcl} \begin{array}{rcl} B_{23} & = & B_{32} = [J (d_{x} - d_{y} - i d_{x y} - i d_{y x}) \\ + K (d_{1} - d_{4} - i d_{9} - i d_{11})] \cos k^{'} \cdot δ_{2}, \end{array} \end{array}

(B18)

\begin{array}{rcl} B_{11} = B_{22} = B_{33} = 0, \end{array}

(B19)

where

\begin{array}{rcl} c_{1} = \cos θ_{1} \cos ϕ_{1} \cos θ_{2} \cos ϕ_{2}, \end{array}

(B20)

\begin{array}{rcl} c_{2} = \cos θ_{1} \sin ϕ_{1} \cos θ_{2} \sin ϕ_{2}, \end{array}

(B21)

\begin{array}{rcl} c_{3} = \sin θ_{1} \sin θ_{2}, \end{array}

(B22)

\begin{array}{rcl} c_{4} = \sin ϕ_{1} \sin ϕ_{2}, \end{array}

(B23)

\begin{array}{rcl} c_{5} = \cos ϕ_{1} \cos ϕ_{2}, \end{array}

(B24)

\begin{array}{rcl} c_{6} = \sin θ_{1} \cos ϕ_{1} \sin θ_{2} \cos ϕ_{2}, \end{array}

(B25)

\begin{array}{rcl} c_{7} = \sin θ_{1} \sin ϕ_{1} \sin θ_{2} \sin ϕ_{2}, \end{array}

(B26)

\begin{array}{rcl} c_{8} = \cos θ_{1} \cos θ_{2}, \end{array}

(B27)

\begin{array}{rcl} c_{9} = - \cos θ_{1} \cos ϕ_{1} \sin ϕ_{2}, \end{array}

(B28)

\begin{array}{rcl} c_{10} = \cos θ_{1} \sin ϕ_{1} \cos ϕ_{2}, \end{array}

(B29)

\begin{array}{rcl} c_{11} = - \sin ϕ_{1} \cos θ_{2} \cos ϕ_{2}, \end{array}

(B30)

\begin{array}{rcl} c_{12} = \cos ϕ_{1} \cos θ_{2} \sin ϕ_{2}, \end{array}

(B31)

\begin{array}{rcl} c_{x} = c_{1} + c_{2} + c_{3}, \end{array}

(B32)

\begin{array}{rcl} c_{y} = c_{4} + c_{5}, \end{array}

(B33)

\begin{array}{rcl} c_{z} = c_{6} + c_{7} + c_{8}, \end{array}

(B34)

\begin{array}{rcl} c_{x y} = c_{9} + c_{10}, \end{array}

(B35)

\begin{array}{rcl} c_{y x} = c_{11} + c_{12}, \end{array}

(B36)

and change the corresponding subscripts in θ_1,2 and φ_1,2 to θ_2,3 and φ_2,3, we get the corresponding expressions for d_i with i = 1 − 12, x, y, z, xy, yx. Similarly, change θ_1,2 and φ_1,2 to θ_3,1 and φ_3,1, we get the corresponding expressions for e_i. The vectors δ_1,2,3 are the NN vectors of the KL with δ₁ = (1/2, 0),

δ_{2} = (- 1, \sqrt{3}) / 4

δ_{3} = (- 1, - \sqrt{3}) / 4

, which are defined in the new 2D frame

x^{'} y^{'}

. Note that to get the eigenvalues and eigenvectors of bosonic quadratic Hamiltonian, we need to diagonalize the matrix

I_{-} h (k^{'})

instead of

h (k^{'})

, where

\begin{array}{rcl} I_{-} = (\begin{array}{cc} I & 0 \\ 0 & - I \end{array}), \end{array}

(B37)

with I the 3 × 3 identity matrix.

References

List( Publishing order | Descend order by publishing year | Descend order by cited within ) Chart analysis

1	Kitaev A 2006 Anyons in an exactly solved model and beyond Ann. Phys. 321 2 111 https://doi.org/10.1016/j.aop.2005.10.005 Cited in this article [1]

2	Rousochatzakis I Reuther J Thomale R Rachel S Perkins N B 2015 Phase diagram and quantum order by disorder in the kitaev K₁ − K₂ honeycomb magnet Phys. Rev. X 5 041035 https://doi.org/10.1103/PhysRevX.5.041035

3	Lee E K-H Schaffer R Bhattacharjee S Kim Y B 2014 Heisenberg–Kitaev model on the hyperhoneycomb lattice Phys. Rev. B 89 045117 https://doi.org/10.1103/PhysRevB.89.045117

4	Nasu J Udagawa M Motome Y 2014 Vaporization of kitaev spin liquids Phys. Rev. Lett. 113 197205 https://doi.org/10.1103/PhysRevLett.113.197205

5	Takayama T Kato A Dinnebier R Nuss J Kono H Veiga L S I Fabbris G Haskel D Takagi H 2015 Hyperhoneycomb iridate β − Li₂IrO₃ as a platform for kitaev magnetism Phys. Rev. Lett. 114 077202 https://doi.org/10.1103/PhysRevLett.114.077202

6	Yao H Kivelson S A 2007 Exact chiral spin liquid with non-abelian anyons Phys. Rev. Lett. 99 247203 https://doi.org/10.1103/PhysRevLett.99.247203

7	Kargarian M Bombin H Martin-Delgado M A 2010 Topological color codes and two-body quantum lattice hamiltonians New J. Phys. 12 025018 https://doi.org/10.1088/1367-2630/12/2/025018

8	Jahromi S S Kargarian M Masoudi S F Langari A 2016 Topological spin liquids in the ruby lattice with anisotropic kitaev interactions Phys. Rev. B 94 125145 https://doi.org/10.1103/PhysRevB.94.125145

9	Jahromi S S Orús R Kargarian M Langari A 2018 Infinite projected entangled-pair state algorithm for ruby and triangle-honeycomb lattices Phys. Rev. B 97 115161 https://doi.org/10.1103/PhysRevB.97.115161

10	Kishimoto M Morita K Matsubayashi Y Sota S Yunoki S Tohyama T 2018 Ground state phase diagram of the kitaev-heisenberg model on a honeycomb-triangular lattice Phys. Rev. B 98 054411 https://doi.org/10.1103/PhysRevB.98.054411

11	Kimchi I Vishwanath A 2014a Kitaev-heisenberg models for iridates on the triangular, hyperkagome, kagome, fcc, and pyrochlore lattices Phys. Rev. B 89 014414 https://doi.org/10.1103/PhysRevB.89.014414 Cited in this article [1]

12	Chaloupka J Jackeli G Khaliullin G 2010 Kitaev-heisenberg model on a honeycomb lattice: Possible exotic phases in iridium oxides A₂IrO₃ Phys. Rev. Lett. 105 027204 https://doi.org/10.1103/PhysRevLett.105.027204 Cited in this article [1]

13	Jiang H-C Gu Z-C Qi X-L Trebst S 2011 Possible proximity of the mott insulating iridate Na₂IrO₃ to a topological phase: Phase diagram of the Heisenberg–Kitaev model in a magnetic field Phys. Rev. B 83 245104 https://doi.org/10.1103/PhysRevB.83.245104

14	Reuther J Thomale R Trebst S 2011 Finite-temperature phase diagram of the Heisenberg–Kitaev model Phys. Rev. B 84 100406 https://doi.org/10.1103/PhysRevB.84.100406

15	Kimchi I You Y-Z 2011 Kitaev-heisenberg-J₂ − J₃ model for the iridates A₂IrO₃ Phys. Rev. B 84 180407 https://doi.org/10.1103/PhysRevB.84.180407

16	Schaffer R Bhattacharjee S Kim Y B 2012 Quantum phase transition in Heisenberg–Kitaev model Phys. Rev. B 86 224417 https://doi.org/10.1103/PhysRevB.86.224417

17	Singh Y Manni S Reuther J Berlijn T Thomale R Ku W Trebst S Gegenwart P 2012 Relevance of the Heisenberg–Kitaev model for the honeycomb lattice iridates A₂IrO₃ Phys. Rev. Lett. 108 127203 https://doi.org/10.1103/PhysRevLett.108.127203

18	Choi S K 2012 Spin waves and revised crystal structure of honeycomb iridate Na₂IrO₃ Phys. Rev. Lett. 108 127204 https://doi.org/10.1103/PhysRevLett.108.127204

19	Reuther J Thomale R Rachel S 2014 Spiral order in the honeycomb iridate Li₂IrO₃ Phys. Rev. B 90 100405 https://doi.org/10.1103/PhysRevB.90.100405

20	Chaloupka J Jackeli G Khaliullin G 2013 Zigzag magnetic order in the iridium oxide Na₂IrO₃ Phys. Rev. Lett. 110 097204 https://doi.org/10.1103/PhysRevLett.110.097204

21	Okamoto S 2013 Global phase diagram of a doped kitaev-heisenberg model Phys. Rev. B 87 064508 https://doi.org/10.1103/PhysRevB.87.064508

22	Price C Perkins N B 2013 Finite-temperature phase diagram of the classical kitaev-heisenberg model Phys. Rev. B 88 024410 https://doi.org/10.1103/PhysRevB.88.024410

23	Sela E Jiang H-C Gerlach M H Trebst S 2014 Order-by-disorder and spin-orbital liquids in a distorted Heisenberg–Kitaev model Phys. Rev. B 90 035113 https://doi.org/10.1103/PhysRevB.90.035113

24	Rau J G Lee E K-H Kee H-Y 2014 Generic spin model for the honeycomb iridates beyond the kitaev limit Phys. Rev. Lett. 112 077204 https://doi.org/10.1103/PhysRevLett.112.077204

25	Yamaji Y Nomura Y Kurita M Arita R Imada M 2014 First-principles study of the honeycomb-lattice iridates Na₂IrO₃ in the presence of strong spin-orbit interaction and electron correlations Phys. Rev. Lett. 113 107201 https://doi.org/10.1103/PhysRevLett.113.107201

26	Sizyuk Y Price C Wölfle P Perkins N B 2014 Importance of anisotropic exchange interactions in honeycomb iridates: Minimal model for zigzag antiferromagnetic order in Na₂IrO₃ Phys. Rev. B 90 155126 https://doi.org/10.1103/PhysRevB.90.155126

27	Chaloupka J Khaliullin G 2015 Hidden symmetries of the extended kitaev-heisenberg model: Implications for the honeycomb-lattice iridates A₂IrO₃ Phys. Rev. B 92 024413 https://doi.org/10.1103/PhysRevB.92.024413

28	Okubo T Shinjo K Yamaji Y Kawashima N Sota S Tohyama T Imada M 2017 Ground-state properties of Na₂IrO₃ determined from an ab initio hamiltonian and its extensions containing kitaev and extended heisenberg interactions Phys. Rev. B 96 054434 https://doi.org/10.1103/PhysRevB.96.054434

29	Gotfryd D Rusnačko J Wohlfeld K Jackeli G Chaloupka J Oleś A M 2017 Phase diagram and spin correlations of the kitaev-heisenberg model: Importance of quantum effects Phys. Rev. B 95 024426 https://doi.org/10.1103/PhysRevB.95.024426

30	Dey T Mahajan A V Khuntia P Baenitz M Koteswararao B Chou F C 2012 Spin-liquid behavior in $J_{e f f} = \frac{1}{2}$ triangular lattice compound Ba₃IrTi₂O₉ Phys. Rev. B 86 140405 https://doi.org/10.1103/PhysRevB.86.140405

31	Becker M Hermanns M Bauer B Garst M Trebst S 2015 Spin-orbit physics of $j = \frac{1}{2}$ mott insulators on the triangular lattice Phys. Rev. B 91 155135 https://doi.org/10.1103/PhysRevB.91.155135

32	Catuneanu A Rau J G Kim H-S Kee H-Y 2015 Magnetic orders proximal to the kitaev limit in frustrated triangular systems: Application to Ba₃IrTi₂O₉ Phys. Rev. B 92 165108 https://doi.org/10.1103/PhysRevB.92.165108

33	Jackeli G Avella A 2015 Quantum order by disorder in the kitaev model on a triangular lattice Phys. Rev. B 92 184416 https://doi.org/10.1103/PhysRevB.92.184416

34	Rousochatzakis I Rössler U K van den Brink J Daghofer M 2016 Kitaev anisotropy induces mesoscopic $z_{2}$ vortex crystals in frustrated hexagonal antiferromagnets Phys. Rev. B 93 104417 https://doi.org/10.1103/PhysRevB.93.104417

35	Lee W-J 2017 Putative spin liquid in the triangle-based iridate Ba₃IrTi₂O₉ Phys. Rev. B 96 014432 https://doi.org/10.1103/PhysRevB.96.014432

36	Plumb K W Clancy J P Sandilands L J Vijay Shankar V Hu Y F Burch K S Kee H-Y Kim Y-J 2014 α − RuCl₃: A spin-orbit assisted mott insulator on a honeycomb lattice Phys. Rev. B 90 041112 https://doi.org/10.1103/PhysRevB.90.041112

37	Kubota Y Tanaka H Ono T Narumi Y Kindo K 2015 Successive magnetic phase transitions in α − RuCl₃: Xy-like frustrated magnet on the honeycomb lattice Phys. Rev. B 91 094422 https://doi.org/10.1103/PhysRevB.91.094422

38	Kim H-S Vijay Shankar V Catuneanu A Kee H-Y 2015 Kitaev magnetism in honeycomb RuCl₃ with intermediate spin-orbit coupling Phys. Rev. B 91 241110 https://doi.org/10.1103/PhysRevB.91.241110

39	Johnson R D 2015 Monoclinic crystal structure of α − RuCl₃ and the zigzag antiferromagnetic ground state Phys. Rev. B 92 235119 https://doi.org/10.1103/PhysRevB.92.235119

40	Bastien G 2018 Pressure-induced dimerization and valence bond crystal formation in the kitaev-heisenberg magnet α − RuCl₃ Phys. Rev. B 97 241108 https://doi.org/10.1103/PhysRevB.97.241108

41	Zhu Z Kimchi I Sheng D N Fu L 2018 Robust non-abelian spin liquid and a possible intermediate phase in the antiferromagnetic kitaev model with magnetic field Phys. Rev. B 97 241110 https://doi.org/10.1103/PhysRevB.97.241110

42	Gohlke M Wachtel G Yamaji Y Pollmann F Kim Y B 2018 Quantum spin liquid signatures in kitaev-like frustrated magnets Phys. Rev. B 97 075126 https://doi.org/10.1103/PhysRevB.97.075126 Cited in this article [1]

43	Dun Z L 2016 Magnetic ground states of the rare-earth tripod kagome lattice ${M g}_{2} {R E}_{3} {S b}_{3} O_{14}$ ( $R E = G d, D y, E r$ Phys. Rev. Lett. 116 157201 https://doi.org/10.1103/PhysRevLett.116.157201 Cited in this article [2]

44	Scheie A Sanders M Krizan J Qiu Y Cava R J Broholm C 2016 Effective spin- $\frac{1}{2}$ scalar chiral order on kagome lattices in Nd₃Sb₃Mg₂O₁₄ Phys. Rev. B 93 180407 https://doi.org/10.1103/PhysRevB.93.180407 Cited in this article [3]

Dun

Z L

Trinh

Lee

Choi

E S

Y F

Wang

Y X

Blanc

Ramirez

A P

Zhou

H D

2017 Structural and magnetic properties of two branches of the tripod-kagome-lattice family A₂R₃Sb₃O₁₄ (A = Mg, Zn; R = Pr, Nd, Gd, Tb, Dy, Ho, Er, Yb Phys. Rev. B 95 104439

https://doi.org/10.1103/PhysRevB.95.104439

Cited in this article [1]

46	Morita K Kishimoto M Tohyama T 2018 Ground-state phase diagram of the kitaev-heisenberg model on a kagome lattice Phys. Rev. B 98 134437 https://doi.org/10.1103/PhysRevB.98.134437 Cited in this article [5]

47	Owerre S A 2017 Topological thermal hall effect in frustrated kagome antiferromagnets Phys. Rev. B 95 014422 https://doi.org/10.1103/PhysRevB.95.014422 Cited in this article [1]

48	Yang Y Perkins N B Koç F Lin C-H Rousochatzakis I 2020 Quantum-classical crossover in the spin- $\frac{1}{2}$ Heisenberg–Kitaev kagome magnet Phys. Rev. Res. 2 033217 https://doi.org/10.1103/PhysRevResearch.2.033217 Cited in this article [6]

49	Kimchi I Vishwanath A 2014b Kitaev-heisenberg models for iridates on the triangular, hyperkagome, kagome, fcc, and pyrochlore lattices Phys. Rev. B 89 014414 https://doi.org/10.1103/PhysRevB.89.014414 Cited in this article [2]

50	Katsura H Nagaosa N Lee P A 2010 Theory of the thermal hall effect in quantum magnets Phys. Rev. Lett. 104 066403 https://doi.org/10.1103/PhysRevLett.104.066403 Cited in this article [1]

51	Onose Y Ideue T Katsura H Shiomi Y Nagaosa N Tokura Y 2010 Observation of the magnon hall effect Science 329 297 299 https://doi.org/10.1126/science.1188260

52	Hirschberger M Chisnell R Lee Y S Ong N P 2015 Thermal hall effect of spin excitations in a kagome magnet Phys. Rev. Lett. 115 106603 https://doi.org/10.1103/PhysRevLett.115.106603

53	Chisnell R Helton J S Freedman D E Singh D K Bewley R I Nocera D G Lee Y S 2015 Topological magnon bands in a kagome lattice ferromagnet Phys. Rev. Lett. 115 147201 https://doi.org/10.1103/PhysRevLett.115.147201 Cited in this article [1]

54	Owerre S A 2016 A first theoretical realization of honeycomb topological magnon insulator J. Phys. Condens. Matter 28 386001 https://doi.org/10.1088/0953-8984/28/38/386001

55	Cao X Chen K He D 2015 Magnon hall effect on the lieb lattice J. Phys. Condens. Matter 27 166003 https://doi.org/10.1088/0953-8984/27/16/166003 Cited in this article [2]

56	Holstein T Primakoff H 1940 Field dependence of the intrinsic domain magnetization of a ferromagnet Phys. Rev. 58 1098 1113 https://doi.org/10.1103/PhysRev.58.1098 Cited in this article [1]

57	Xiao D Chang M-C Niu Q 2010 Berry phase effects on electronic properties Rev. Mod. Phys. 82 1959 2007 https://doi.org/10.1103/RevModPhys.82.1959 Cited in this article [1]

58	Hatsugai Y 1993 Chern number and edge states in the integer quantum hall effect Phys. Rev. Lett. 71 3697 3700 https://doi.org/10.1103/PhysRevLett.71.3697 Cited in this article [1]

59	Matsumoto R Murakami S 2011a Theoretical prediction of a rotating magnon wave packet in ferromagnets Phys. Rev. Lett. 106 197202 https://doi.org/10.1103/PhysRevLett.106.197202 Cited in this article [1]

60	Matsumoto R Murakami S 2011b Rotational motion of magnons and the thermal hall effect Phys. Rev. B 84 184406 https://doi.org/10.1103/PhysRevB.84.184406

61	Matsumoto R Shindou R Murakami S 2014 Thermal hall effect of magnons in magnets with dipolar interaction Phys. Rev. B 89 054420 https://doi.org/10.1103/PhysRevB.89.054420 Cited in this article [1]

62	Mook A Henk J Mertig I 2014 Magnon hall effect and topology in kagome lattices: A theoretical investigation Phys. Rev. B 89 134409 https://doi.org/10.1103/PhysRevB.89.134409 Cited in this article [1]

63	Yokoi T 2021 Half-integer quantized anomalous thermal hall effect in the kitaev material candidate α − RuCl₃ Science 373 568 572 https://doi.org/10.1126/science.aay5551 Cited in this article [1]