Abstract
Elementary excitations in entangled states such as quantum spin liquids may exhibit exotic statistics different from those obeyed by fundamental bosons and fermions. NonAbelian anyons exist in a Kitaev spin liquid—the ground state of an exactly solvable model. A smokinggun signature of these excitations, namely a halfinteger quantized thermal Hall conductivity, was recently reported in αRuCl_{3}. While fascinating, a microscopic theory for this phenomenon remains elusive because the pure Kitaev model cannot display this effect in an intermediate magnetic field. Here we present a microscopic theory of the Kitaev spin liquid emerging between the low and highfield states. Essential to this result is an antiferromagnetic offdiagonal symmetric interaction which allows the Kitaev spin liquid to protrude from the ferromagnetic Kitaev limit under a magnetic field. This generic model displays a strong field anisotropy, and we predict a wide spin liquid regime when the field is perpendicular to the honeycomb plane.
Introduction
The Kitaev spin liquid (KSL) is a longrange entangled state on a honeycomb lattice^{1}, which hosts nonAbelian^{1,2,3} anyon excitations in a magnetic field. It has been proposed that topological quantum computation can be performed via braiding of nonAbelian anyons^{4}, meaning the KSL is of both practical, and fundamental interest. However, it has been challenging to find a solid state realization of Kitaev physics, which has been the focus of recent research. Several honeycomb materials have been suggested as KSL candidates, namely Mott insulators with strong spinorbit coupling (SOC) featuring 4d or 5d transition metal elements^{5,6,7,8,9}. Proposals so far include the iridates A_{2}IrO_{3}^{5,6,10,11,12,13} (A=Li, Na), and αRuCl_{3}^{14,15,16,17,18}. However, all these candidates exhibit magnetic ordering at low temperatures^{17,19,20,21,22,23,24,25,26} which masks potential Kitaev physics. Later theoretical^{27,28,29} and experimental^{30,31,32,33} results suggest that αRuCl_{3} may enter a fieldinduced spin liquid, but there has been no evidence that it is a chiral spin liquid until a halfinteger quantized thermal Hall conductivity was reported in αRuCl_{3}^{34}; a strong indication^{35,36,37} of chiral edge currents of Majorana fermions (MFs) predicted in a KSL.
While the observation of a halfinteger quantized thermal Hall conductivity is the first experimental evidence of chargeneutral nonAbelian anyons in spin systems, a microscopic theory describing their appearance under a field in αRuCl_{3} is missing. This is because, if the dominant interaction in αRuCl_{3} is the ferromagnetic (FM) Kitaev term (as shown through abinitio studies^{25,26} and spin wave analysis^{36}), the FM Kitaev phase is almost immediately destroyed, and the polarized state appears in an applied field^{38,39,40} with no intervening phase. This can be contrasted with the antiferromagnetic (AFM) Kitaev model which hosts a potentially gapless spin liquid under a field, supported by several numerical studies^{39,40,41,42,43,44,45,46,47}. However, this intermediate gapless spin liquid cannot explain the halfinteger thermal Hall effect observed in αRuCl_{3}. Thus, searching for a chiral spin liquid offering a halfinteger thermal Hall effect in an intermediate magnetic field remains a challenging task.
Here we present a microscopic theory in which the KSL is revealed under a magnetic field. The key to our result is an AFM symmetric offdiagonal Γ interaction, which is essential to stabilize the otherwise fragile KSL under intermediate fields. The intermediate phase emerges between the lowfield and highfield phases as Γ increases, and is adiabatically connected to the pure FM Kitaev phase at zero field, providing evidence that it is the KSL. We introduce a microscopic theory with a brief review of the generic nearest neighbor spin model for spinorbit coupled honeycomb materials, appropriate for αRuCl_{3}.
Results
Model
The nearest neighbor model has been derived in refs. ^{5,12,13} based on a strong coupling expansion of the Kanamori Hamiltonian. The combination of crystal field splitting and strong spinorbit coupling leads to a model based on pseudospin\({\textstyle{1 \over 2}}\) local moments with bonddependent interactions. On a bond of type γ ∈ {x, y, z} with sites j, k, the nearestneighbor spin Hamiltonian is taken to be of the JKΓΓ′ form^{13}
where α, β are the remaining spin components in {x, y, z}/{γ}. The spin components are directed along the cubic axes of the underlying ligand octahedra, so the honeycomb layer lies in a plane perpendicular to the [111] spin direction as shown in Fig. 1a. A small Γ′ is present due to trigonal distortion of ligand octahedra in the real material. Here we omit the Heisenberg J for simplicity, and its effects are discussed later. Earlier studies^{14,26,29,48,49} noted that the Γ interaction with AFM sign may play an important role near the FM Kitaev regime to stabilize the spin liquid^{48}. Since αRuCl_{3} has a dominant FM Kitaev interaction with AFM Γ, we focus on Γ/K ∈ [−1, 0] with Γ > 0 and K < 0. The remaining parameters of the Hamiltonian are expressed in units of \(\sqrt {K^2 + {\mathrm{\Gamma }}^2} \equiv 1\).
To describe the effect of a magnetic field we consider a Zeeman term with isotropic gfactor
where h is the magnetic field strength, and \({\widehat{\mathbf{h}}}\) is a unit vector specifying the field direction. The effect of an anisotropic gfactor is discussed later. In order to make a connection with the thermal Hall measurements^{34} we focus on field directions in the \({\hat{\mathbf{a}}}{\hat{\mathbf{c}}}^ \ast\) plane spanned by \([11\bar 2]\) and [111]. The direction of the field in this plane is parameterized by an angle θ from the [111] direction, as shown in Fig. 1a.
Exact diagonalization
Our main results are shown in Fig. 1b, c. Phase diagrams in the Γ/K − h plane are shown for tilting angles (b) θ = 5° and (c) 90° obtained through numerical exact diagonalization (ED) with fixed Γ′ = −0.03 and J = 0. Details of the 24site honeycomb cluster used are discussed in Supplementary Note 1. Peaks in the susceptibilities \(\chi _{{\mathrm{\Gamma }}/K} =  \partial _{{\mathrm{\Gamma }}/K}^2e_0\) and \(\chi _h =  \partial _h^2e_0\), where e_{0} = E_{0}/N is the ground state energy density, are depicted as circles and triangles, respectively. There are three phases in the phase space, namely, zigzag (ZZ) magnetic order at low fields, the KSL, and a polarized state (PS) at high fields. Remarkably, we find an intermediate KSL sitting between ZZ order and the PS which is adiabatically connected to the pure K limit at h = 0.
The intermediate KSL begins from the pure FM K regime, i.e., bottom right corner of the phase diagram, which is unstable to a small magnetic field. However, it is stabilized by the AFM Γ term and extends above the ZZ phase in a magnetic field. For moderate Γ/K appropriate for αRuCl_{3}—for example, Γ/K ≃ −0.37 indicated by the green dashed line in Fig. 1b—we observe two phase transitions from ZZ order to the KSL, and then from KSL to the PS as the field increases. Close to the Kitaev limit, the peak positions of the magnetic (χ_{h}) and Γ/K (χ_{Γ/K}) susceptibilities agree well along the phase boundaries between the KSL and PS for each direction. At larger Γ/K the singular points determining the phase boundaries start to deviate for θ = 5°, as seen through differing positions of circular and triangular markers in Fig. 1b. However, the anomalous peaks in χ_{Γ/K} in this region shrink significantly while the peaks in χ_{h} retain their sharpness. Since these peaks are not seen in χ_{h}, and there are strong variations within the PS in the quantities discussed below, we determine the KSLPS phase boundary based on peaks in χ_{h} and attribute peaks in χ_{Γ/K} to large fluctuations above the KSL.
Interestingly, with a constant Γ′ both the ZZ and KSL phases widen with increasing Γ as suggested by the curvature of the transition line in Fig. 1b. This behavior survives for further tilting of the magnetic field away from [111] with increased \([11\bar 2]\) inplane component. However, the window of the KSL rapidly diminishes with tilting angle until a direct transition between ZZ order and the PS appears at large Γ/K, as shown in Fig. 1c for a \([11\bar 2]\) field. The critical field required to destroy the ZZ ordering drops dramatically with increasing θ. With an estimate of the energy unit as \(\sqrt {K^2 + {\mathrm{\Gamma }}^2} \sim 7\,{\mathrm{meV}}\), h = 0.1 corresponds to a field of ~10 T. This is within the range of fields required to kill the ZZ order in αRuCl_{3}^{34}.
Since the pure Kitaev limit at h = 0 involves the fractionalization of spins into itinerant MFs and \({\Bbb Z}_2\) fluxes, another quantity that characterizes the KSL is the plaquette operator W_{p}^{1}
defined on sites belonging to a hexagonal plaquette p. The pure KSL with h = Γ = Γ′ = 0 (bottom right corner of the phase diagram) is a fluxfree state with W_{p} = +1 on all plaquettes. A finite Γ, Γ′ or h spoils the exact solubility of the Kitaev model, as they generate interactions among the MFs and \({\Bbb Z}_2\) fluxes. Despite the fact that plaquette operators are no longer conserved quantities 〈W_{p}〉 remains positive in the KSL, denoted by red colors in Fig. 1. At the same time the plaquette expectation value is negative in the ZZ ordered phase, as denoted by blue colors in Fig. 1, distinguishing it from the neighboring KSL. Due to this sign difference, the phase boundary between KSL and ZZ is accompanied by a vanishing 〈W_{p}〉 and seen through the rapid color change in Fig. 1. This is a generic feature which also appears for an inplane field of θ = 90°. Further details of the negative plaquette expectation in the ZZ phase can be found in Supplementary Note 3.
To confirm the ZZ magnetically ordered phase at low field, and the sequence of phase transitions, we compute the spin structure factor S(q) for increasing values of h along the green dashed line of Fig. 1b where Γ/K ≃ −0.37. For low fields of h < h_{c1}, where h_{c1} is the position of the first transition, S(q) displays sharp features at the Mpoint of the Brillouin zone (BZ) consistent with ZZ magnetic order as shown in Fig. 2a. Within the intermediate phase, h_{c1} < h < h_{c2} where h_{c2} is the position of the second transition, S(q) is diffuse with a soft peak at the Γpoint. Interestingly, an increased intensity at the Γpoint was observed in αRuCl_{3} under fields in a recent neutron scattering measurement^{50}. As expected, the PS exhibits a sharp feature at the Γpoint for h > h_{c2}. Note that the magnetization m = −∂_{h}e_{0} eventually saturates at \({\textstyle{1 \over 2}}\) in the PS at a field larger than h_{c2}, as shown in Fig. 2b indicating large spin fluctuations inside the PS.
Competition between ZZ and KSL
ZZ ordering at low fields can be traced back to the presence of other small interactions, such as such as a FM Γ′^{5,12,13,27} and/or FM (J) and AFM thirdnearest neighbor (J_{3}) Heisenberg interactions^{51}. We have considered Γ′ as a minimal nearestneighbor perturbation away from the K − Γ model which induces ZZ order at low fields. The combination of J < 0 and J_{3} > 0 has a similar effect to Γ′ in that their combination can also induce ZZ ordering^{13,51}. Inclusion of these small terms would not alter our main results as they further stabilize the ZZ phase. However, the combined strength of terms stabilizing the ZZ order should be small enough to maintain the intermediate KSL.
The particular values of Γ′ used in our calculations were chosen based on the following criteria. For Γ′ = 0 the phase to the left of the KSL at larger Γ/K is nonmagnetic, as discussed in the Underlying Phase Diagram section, and develops ZZ magnetic order as Γ′ < 0 is introduced. The magnitude Γ′ was chosen to be the smallest value such that ZZ magnetic order develops within this phase. If Γ′ becomes too large then the KSL at h = 0 will be wiped out entirely. To quantify this, we have calculated χ_{Γ/K} with ED on a 24site honeycomb cluster near the Kitaev limit at h = 0 for different values of Γ′. As Γ′ increases, the ZZ ordered phase expands while the KSL shrinks at h = 0 as shown in Fig. 3a. The KSL at h = 0 is found to disappear entirely beyond \({\mathrm{\Gamma }}_{\mathrm{c}}^\prime \simeq  0.1\). Evolution of the intermediatefield KSL with Γ′ is seen in Fig. 3b through the magnetic susceptibility χ_{h} in a 5° field at Γ/K = −1, where the intermediate phase is largest with Γ′ = −0.03. Emergence of the intermediatefield KSL depends crucially on the survival of the pure Kitaev phase at h = 0, because χ_{h} shows a single transition from ZZ to PS beyond \({\mathrm{\Gamma }}_{\mathrm{c}}^\prime \simeq  0.1\). Quantifying the strengths of Γ′, J, J_{3} required for an intermediate KSL is left for a detailed future study.
DMRG
In order to check the dependence of this result on cluster geometry, we have also studied a twoleg honeycomb strip using densitymatrix renormalization group (DMRG). We denote the total number of sites in the strip by N. This geometry has recently been used to study the KitaevHeisenberg model^{52}, where it was found that its phase diagram displays a striking similarity with that of the 2D honeycomb lattice. For the K − Γ model we find quantitative differences in the positions of the phase boundaries due to the cluster connectivity, but the main result of an emerging intermediatefield KSL remains the same. Further rationale for this choice of geometry is discussed in Supplementary Note 1.
Phase diagrams in the Γ/K − h plane with N = 200 and open boundary conditions (OBC) for tilting angles θ = 0°, 5°, 10°, and 90° with fixed Γ′ = −0.1 and J = 0 are shown in Fig. 4. The phase boundaries in Fig. 4, determined by peaks in χ_{h} or χ_{Γ/K}, are represented by red lines. We find a qualitative similarity with the ED phase diagram of Fig. 1, showing a region of KSL which extends above the ZZ ordered phase and below the PS under a magnetic field. The KSL phase space shrinks rapidly as the inplane component of the field becomes larger. As found with ED on the 24site honeycomb cluster, the intermediate KSL at large Γ/K disappears as the field tilts towards θ = 90°, leaving a single direct transition from ZZ to the PS. Crucially, a small region of KSL remains intact at smaller Γ/K. Thus, with the inplane \([11\bar 2]\) field the KSL is confined to a narrow range of low fields near the pure FM Kitaev limit. The same qualitative behavior is seen for another inplane field direction of \([\bar 110]\) only for the 24site honeycomb cluster used with ED. On the twoleg honeycomb strip the KSL (in the region Γ/K < Γ/K_{c} where Γ/K_{c} is the transition point between ZZ and KSL at h = 0) is immediately destroyed by any nonzero field in this direction, as shown in Supplementary Fig. 3. While this is consistent with the observation that there is no \({\Bbb Z}_2\) topological order when a certain mirror symmetry is preserved^{46}, the phase boundary is an artifact of the strip geometry.
The first row in Fig. 4 shows \(\left\langle {S_j^xS_k^x} \right\rangle\) at separation k − j = 50 along the leg of the strip with maximum correlations as a function of field and Γ/K for different tilting angles of the field in the \({\hat{\mathbf{a}}}{\hat{\mathbf{c}}}^ \ast\) plane. As expected, correlations are appreciable within the magnetically ordered and polarized states. The KSL phase is clearly distinguished from the surrounding ordered states by nearly vanishing \(\left\langle {S^xS^x} \right\rangle\) (\(= \left\langle {S^yS^y} \right\rangle\)) spin correlations. However, spinspin correlations need not be identically zero except at h = 0 due to a component of the spin aligning with the field, which is more pronounced when h is large. The PS away from the pure Kitaev region shows large spin fluctuations, which is similar to the 24site ED result where saturation of the magnetization occurs at higher fields above h_{c2}, and small peaks in χ_{Γ/K} only appear above the KSLPS phase boundary.
In the bottom row of Fig. 4 we show plaquetteplaquette correlations \(\left\langle {W_pW_{p\prime }} \right\rangle\) at separation p′ − p = 30. Close to the Kitaev limit these correlations are nearly unity, consistent with 〈W_{p}〉 = +1 in the K limit, and decrease with increasing field and Γ/K within the KSL. This is expected because the magnetic field, as well as Γ, Γ′, introduces interactions among the MF and flux degrees of freedom. Interestingly, the plaquette–plaquette correlations, which approach \(\left\langle {W_p} \right\rangle ^2\) in the KSL at large separations, show large fluctuations in the PS above the KSL phase for 5° and 10° tilting angles. We note that these fluctuations in the PS for θ = 5° and 10° disappear above the transition line between the KSL and PS determined by θ = 0°.
A representative cut of the phase diagram is presented in Fig. 5c, d as a function of a 5° tilted field with Γ/K = −0.325, which corresponds to the green line in Fig. 4b. With increasing field, a sequence of transitions from ZZ order to the KSL and finally the PS are evidenced by strong singular behavior in χ_{h} in Fig. 5a. The transition between ZZ order and the KSL is accompanied by a sharp increase in plaquette–plaquette correlations shown in Fig. 5b, and a larger value of 〈W_{p}〉 accordingly. Note that the maximum value of \(\left\langle {W_pW_{p\prime }} \right\rangle \simeq 0.1\) corresponds to 〈W_{p}〉 ≃ 0.32, as \(\left\langle {W_pW_{p\prime }} \right\rangle\) approaches \(\left\langle {W_p} \right\rangle ^2\) at large distances. Components \(\left\langle {S^xS^x} \right\rangle\) and \(\left\langle {S^zS^z} \right\rangle\) of the spin–spin correlators are plotted in Fig. 5c, d. While the \(\left\langle {S^xS^x} \right\rangle\) correlations are small in the KSL, the \(\left\langle {S^zS^z} \right\rangle\) correlations are slightly larger. This is similar to the honeycomb cluster with ED, where a finite magnetization m is present in the KSL phase, as shown in Fig. 2b. The source of asymmetry between the S^{x} and S^{z} components of the spin is twofold. One is due to the twoleg honeycomb strip connectivity, and the second is the finite tilting of the magnetic field which further enhances their difference. The preceding properties are shown for N = 100, 200, 300, 400, and iDMRG in Fig. 5 with different colors, and are seen to be relatively insensitive to the system size.
Underlying phase diagram
To understand the microscopic mechanism of the emerging KSL, we study the K − Γ model without Γ′—i.e., in the absence of small interactions that induce ZZ order—at θ = 0. At zero field, there is a finite region of the KSL when the AFM offdiagonal symmetric Γ interaction is introduced. Then in zero field at Γ/K_{c} there is a transition from the KSL to another possible spin liquid dubbed KΓ spin liquid (KΓSL)^{28}. Components of the spinspin correlators, \(\left\langle {S^xS^x} \right\rangle\) and \(\left\langle {S^zS^z} \right\rangle\) shown in Fig. 6a, b, demonstrate a lack of magnetic order in the KSL and KΓSL at h = 0, and finite correlations building with increasing field.
The KΓSL is characterized by a finite \(\langle {W_pW_{p\prime }} \rangle\) like the KSL, but with negative 〈W_{p}〉 as shown in Fig. 6c, d. While 〈W_{p}〉 is positive in the KSL, a negative 〈W_{p}〉 in the KΓSL indicates a phase with a finite flux density. Strikingly, when the field is applied along the [111] direction the KSL sits above the KΓSL for a fixed Γ/K, leading to two phase transitions with increasing field: KΓSL → KSL → PS. The KΓSL is extremely fragile to additional interactions that stabilize ZZ order. For example, when a small Γ′ interaction is introduced the KΓSL is replaced by the ZZ ordered phase as shown in Fig. 1. Importantly, the ZZ order does not extend all the way to the Kitaev limit and leaves a finite region of the KSL at zero field.
Discussion
Here we presented a microscopic theory, based on dominant Kitaev and offdiagonal symmetric Γ interactions, which offers an intermediate KSL emerging between the lowfield and highfield states. The lowfield ZZ state is induced via small perturbations beyond K and Γ, such as Γ′ due to trigonal distortion of ligand octahedra. Our numerical data indicates that this additional interaction can wipe out the KSL phase (including the pure Kitaev limit at h = 0) when its strength is too large, leading to a single transition between ZZ and PS under a magnetic field. Experimental reports of a halfquantized thermal Hall conductivity in αRuCl_{3} imply that the strengths of interactions beyond K and Γ are small enough to leave the intermediate KSL intact, yet finite to induce the ZZ order. In the absence of these interactions the KΓ model exhibits another possible spin liquid called the KΓSL, which is then replaced by the ZZ due to these small interactions. It is possible that a region of the KΓSL survives with smaller Γ′ while developing ZZ order, resulting in two spin liquids between ZZ and PS under a field. We leave these questions for a future study.
As the magnetic field is tilted away from the outofplane [111] direction towards the inplane \([11\bar 2]\) direction the intermediate KSL region shrinks rapidly—independent of the strength of Γ′ and for both cluster shapes studied here. What remains is a small intermediate phase at fields an order of magnitude smaller for moderate Γ/K, showing a dramatic magnetic anisotropy. Considering an anisotropic gfactor, due to a combination of the layered structure and SOC, the magnetic anisotropy is further enhanced by the ratio between the inplane g_{ab} and the outofplane \(g_{{\mathrm{c}}^ \ast }\) components. While smaller tilting angles are less effective at destroying the ZZ magnetic order, they offer a much larger region of the KSL. To enlarge the intermediate KSL phase, we therefore propose that a field should be applied at smaller tilting angles. Further thermal Hall transport measurements for different inplane components would be desirable in order to test our microscopic theory.
There are several aspects of this work that require further study. The first is the presence of large fluctuations in \(\left\langle {W_pW_{p\prime }} \right\rangle\) and \(\left\langle {S_j^xS_k^x} \right\rangle\) just above the KSL phase into the PS, which is also seen by χ_{Γ/K} and an unsaturated magnetization in the 24site honeycomb cluster. This is suggestive of a nontrivial crossover region into the fully polarized phase. We also note the presence of a nonmagnetic phase dubbed KΓSL in the underlying phase diagram of the KΓ model on the twoleg geometry next to the KSL phase. The KΓSL at zero field is differentiated from the KSL by a sharp drop from 〈W_{p}〉 = 1 in the KSL to \(\left\langle {W_p} \right\rangle \simeq  \frac{1}{3}\) in the KΓSL, accompanied by a singular χ_{Γ/K}. Nature of the KΓSL, numerical studies of this phase in the honeycomb geometry, and the transition to the KSL are excellent subjects for future study. For instance, studies of possible vortex patterns due to strong interactions among MFs and \({\Bbb Z}_2\) vortices would be highly interesting to pursue.
Methods
Details of simulations
Numerical exact diagonalization (ED), and density matrix renormalization group (DMRG) are used to study the parameter space appropriate for αRuCl_{3}. In the ED and DMRG calculations, we consider the twoleg honeycomb strip geometry, and the 24site honeycomb shape in ED only, both shown in Supplementary Fig. 1. The choice of these clusters is discussed in Supplementary Note 1, and is related to hidden points of SU(2) symmetry present in the 2D limit.
ED was performed on the 24site honeycomb cluster with periodic boundary conditions, where the Lanczos method was used to obtain the lowestlying eigenvalues and eigenvectors of the Hamiltonian in Eq. (1).
Part of the numerical calculations were performed using the ITensor library (http://itensor.orghttp://itensor.org) typically with a target precision of 10^{−11} using up to 1000 states. All DMRG calculations were performed using open boundary conditions (OBC). The iDMRG calculations were performed using a target precision of 5 × 10^{−11} and up to 1000 states.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Code availability
The code used to generate the data used in this study is available from the corresponding author upon reasonable request.
References
Kitaev, A. Anyons in an exactly solved model and beyond. Ann. Phys. 321, 2–111 (2006).
Balents, L. Spin liquids in frustrated magnets. Nature 464, 199 (2010).
Zhou, Y., Kanoda, K. & Ng, T.K. Quantum spin liquid states. Rev. Mod. Phys. 89, 025003 (2017).
Nayak, C., Simon, S. H., Stern, A., Freedman, M. & Sarma, S. D. NonAbelian anyons and topological quantum computation. Rev. Mod. Phys. 80, 1083–1159 (2008).
Jackeli, G. & Khaliullin, G. Mott insulators in the strong spinorbit coupling limit: from Heisenberg to a quantum compass and Kitaev models. Phys. Rev. Lett. 102, 017205 (2009).
Chaloupka, J., Jackeli, G. & Khaliullin, G. KitaevHeisenberg model on a honeycomb lattice: possible exotic phases in iridium oxides A_{2}IrO_{3}. Phys. Rev. Lett. 105, 027204 (2010).
WitczakKrempa, W., Chen, G., Kim, Y. B. & Balents, L. Correlated quantum phenomena in the strong spinorbit regime. Annu. Rev. Condens. Matter Phys. 5, 57–82 (2014).
Rau, J. G., Lee, E.H. & Kee, H.Y. Spinorbit physics giving rise to novel phases in correlated systems: Iridates and related materials. Annu. Rev. Condens. Matter Phys. 7, 195–221 (2016).
Winter, S. M. et al. Models and materials for generalized Kitaev magnetism. J. Phys. Condens. Matter 29, 493002 (2017).
Singh, Y. et al. Relevance of the HeisenbergKitaev model for the honeycomb lattice iridates A_{2}IrO_{3}. Phys. Rev. Lett. 108, 127203 (2012).
Modic, K. A. et al. Realization of a threedimensional spinanisotropic harmonic honeycomb iridate. Nat. Commun. 5, 4203 (2014).
Rau, J. G., Lee, E.H. & Kee, H.Y. Generic spin model for the honeycomb iridates beyond the Kitaev limit. Phys. Rev. Lett. 112, 077204 (2014).
Rau, J. G. & Kee, H.Y. Trigonal distortion in the honeycomb iridates: Proximity of zigzag and spiral phases in Na_{2}IrO_{3.} Preprint at http://arXiv.org/abs/1408.4811 (2014).
Plumb, K. W. et al. αRuCl_{3}: A spinorbit assisted Mott insulator on a honeycomb lattice. Phys. Rev. B 90, 041112 (2014).
Sandilands, L. J., Tian, Y., Plumb, K. W., Kim, Y.J. & Burch, K. S. Scattering continuum and possible fractionalized excitations in αRuCl_{3}. Phys. Rev. Lett. 114, 147201 (2015).
Kim, H.S., Shankar, V. V., Catuneanu, A. & Kee, H.Y. Kitaev magnetism in honeycomb αRuCl_{3} with intermediate spinorbit coupling. Phys. Rev. B 91, 241110 (2015).
Banerjee, A. et al. Proximate Kitaev quantum spin liquid behaviour in a honeycomb magnet. Nat. Mater. 15, 733 (2016).
Sandilands, L. J. et al. Spinorbit excitations and electronic structure of the putative Kitaev magnet αRuCl_{3}. Phys. Rev. B 93, 075144 (2016).
Choi, S. K. et al. Spin waves and revised crystal structure of honeycomb iridate Na_{2}IrO_{3}. Phys. Rev. Lett. 108, 127204 (2012).
Chaloupka, J., Jackeli, G. & Khaliullin, G. Zigzag magnetic order in the iridium oxide Na_{2}IrO_{3}. Phys. Rev. Lett. 110, 097204 (2013).
Fletcher, J. M., Gardner, W. E., Fox, A. C. & Topping, G. Xray, infrared, and magnetic studies of α and βruthenium trichloride. J. Chem. Soc. A, 1038–1045 (1967).
Sears, J. A. et al. Magnetic order in αRuCl_{3}: A honeycomblattice quantum magnet with strong spinorbit coupling. Phys. Rev. B 91, 144420 (2015).
Johnson, R. D. et al. Monoclinic crystal structure of αRuCl_{3} and the zigzag antiferromagnetic ground state. Phys. Rev. B 92, 235119 (2015).
Cao, H. B. et al. Lowtemperature crystal and magnetic structure of αRuCl_{3}. Phys. Rev. B 93, 134423 (2016).
Kim, H. S. & Kee, H. Y. Crystal structure and magnetism in αRuCl_{3}: an ab initio study. Phys. Rev. B 93, 155143 (2016).
Janssen, L., Andrade, E. & Vojta, M. Magnetization processes of zigzag states on the honeycomb lattice: identifying spin models for αRuCl_{3} and Na_{2}IrO_{3}. Phys. Rev. B 96, 064430 (2017).
Yadav, R. et al. Kitaev exchange and fieldinduced quantum spinliquid states in honeycomb αRuCl_{3}. Sci. Rep. 6, 37925 (2016).
LampenKelley, P. et al. Fieldinduced intermediate phase in αRuCl_{3}: Noncoplanar order, phase diagram, and proximate spin liquid. Preprint at http://arXiv.org/abs/1807.06192 (2018).
Liu, Z.X. & Normand, B. Dirac and chiral quantum spin liquids on the honeycomb lattice in a magnetic field. Phys. Rev. Lett. 120, 187201 (2018).
Baek, S.H. et al. Evidence for a fieldinduced quantum spin liquid in αRuCl_{3}. Phys. Rev. Lett. 119, 037201 (2017).
Wolter, A. U. B. et al. Fieldinduced quantum criticality in the Kitaev system αRuCl_{3}. Phys. Rev. B 96, 041405 (2017).
Zheng, J. et al. Gapless spin excitations in the fieldinduced quantum spin liquid phase of αRuCl_{3}. Phys. Rev. Lett. 119, 227208 (2017).
Janša, N. et al. Observation of two types of fractional excitation in the Kitaev honeycomb magnet. Nat. Phys. 14, 786–790 (2018).
Kasahara, Y. et al. Majorana quantization and halfinteger thermal quantum Hall effect in a Kitaev spin liquid. Nature 559, 227–231 (2018).
VinklerAviv, Y. & Rosch, A. Approximately quantized thermal Hall effect of chiral liquids coupled to phonons. Phys. Rev. X 8, 031032 (2018).
Cookmeyer, J. & Moore, J. E. Spinwave analysis of the lowtemperature thermal Hall effect in the candidate Kitaev spin liquid αRuCl_{3}. Phys. Rev. B 98, 060412 (2018).
Ye, M., Halász, G. B., Savary, L. & Balents, L. Quantization of the thermal Hall conductivity at small Hall angles. Phys. Rev. Lett. 121, 147201 (2018).
Jiang, H.C., Gu, Z.C., Qi, X.L. & Trebst, S. Possible proximity of the Mott insulating iridate Na_{2}IrO_{3} to a topological phase: phase diagram of the HeisenbergKitaev model in a magnetic field. Phys. Rev. B 83, 245104 (2011).
Zhu, Z., Kimchi, I., Sheng, D. N. & Fu, L. Robust nonAbelian spin liquid and a possible intermediate phase in the antiferromagnetic Kitaev model with magnetic field. Phys. Rev. B 97, 241110 (2018).
Liang, S., Jiang, M.H., Chen, W., Li, J.X. & Wang, Q.H. Intermediate gapless phase and topological phase transition of the Kitaev model in a uniform magnetic field. Phys. Rev. B 98, 054433 (2018).
Gohlke, M., Moessner, R. & Pollmann, F. Dynamical and topological properties of the Kitaev model in a [111] magnetic field. Phys. Rev. B 98, 014418 (2018).
Nasu, J., Kato, Y., Kamiya, Y. & Motome, Y. Successive Majorana topological transitions driven by a magnetic field in the Kitaev model. Phys. Rev. B 98, 060416 (2018).
Hickey, C. & Trebst, S. Emergence of a fielddriven U(1) spin liquid in the Kitaev honeycomb model. Nat. Commun. 10, 530 (2019).
Ronquillo, D. C., Vengal, A. & Trivedi, N. Fieldorientationdependent spin dynamics of the Kitaev honeycomb model. Preprint at http://arXiv.org/abs/1805.03722 (2018).
Jiang, H.C., Wang, C.Y., Huang, B. & Lu, Y.M. Field induced quantum spin liquid with spinon Fermi surfaces in the Kitaev model. Preprint at http://arXiv.org/abs/1809.08247 (2018).
Zou, L. & He, Y.C. Fieldinduced neutral Fermi surface and QCD_{3}ChernSimons quantum criticalities in Kitaev materials. Preprint at http://arXiv.org/abs/1809.09091 (2018).
Patel, N.D. & Trivedi, N. Magnetic field induced intermediate quantum spinliquid with a spinon Fermi surface. Preprint at http://arXiv.org/abs/1812.06105 (2018).
Catuneanu, A., Yamaji, Y., Wachtel, G., Kim, Y. B. & Kee, H.Y. Path to stable quantum spin liquids in spinorbit coupled correlated materials. npj Quantum Mater. 3, 23 (2018).
Gohlke, M., Wachtel, G., Yamaji, Y., Pollmann, F. & Kim, Y. B. Quantum spin liquid signatures in Kitaevlike frustrated magnets. Phys. Rev. B 97, 075126 (2018).
Banerjee, A. et al. Excitations in the fieldinduced quantum spin liquid state of αRuCl_{3}. npj Quantum Mater. 3, 8 (2018).
Winter, S. M., Li, Y., Jeschke, H. O. & Valentí, R. Challenges in design of Kitaev materials: Magnetic interactions from competing energy scales. Phys. Rev. B 93, 214431 (2016).
Catuneanu, A., Sørensen, E.S. & Kee, H.Y. Nonlocal string order parameter in the s = 1/2 KitaevHeisenberg ladder. Phys. Rev. B 99, 195112 (2019).
Acknowledgements
This work was supported by the Natural Sciences and Engineering Research Council of Canada and the Center for Quantum Materials at the University of Toronto. This research was enabled in part by support provided by Sharcnet (http://www.sharcnet.ca) and Compute Canada (http://www.computecanada.ca). Computations were performed on the GPC and Niagara supercomputers at the SciNet HPC Consortium. SciNet is funded by: the Canada Foundation for Innovation under the auspices of Compute Canada; the Government of Ontario; Ontario Research Fund—Research Excellence; and the University of Toronto.
Author information
Authors and Affiliations
Contributions
Exact diagonalization calculations were performed by J.S.G. and A.C. Densitymatrix renormalization group calculations were performed by E.S.S. H.Y.K. planned and supervised the project. All authors wrote the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Journal peer review information: Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Gordon, J.S., Catuneanu, A., Sørensen, E.S. et al. Theory of the fieldrevealed Kitaev spin liquid. Nat Commun 10, 2470 (2019). https://doi.org/10.1038/s41467019104058
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467019104058
This article is cited by

Thermal Hall conductivity of αRuCl3
Nature Materials (2023)

Weakcoupling to strongcoupling quantum criticality crossover in a Kitaev quantum spin liquid αRuCl3
npj Quantum Materials (2023)

A magnetic continuum in the cobaltbased honeycomb magnet BaCo2(AsO4)2
Nature Materials (2023)

Planar thermal Hall effect of topological bosons in the Kitaev magnet αRuCl3
Nature Materials (2023)

Possible intermediate quantum spin liquid phase in αRuCl3 under high magnetic fields up to 100 T
Nature Communications (2023)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.