---
title: Zeeman-Kitaev Honeycomb Model
url: https://www.emergentmind.com/topics/zeeman-kitaev-honeycomb-model
type: topic
---

# Zeeman-Kitaev Honeycomb Model

The Zeeman–Kitaev honeycomb model is the Kitaev honeycomb spin model supplemented by an external magnetic field, typically treated through a Zeeman coupling to spin-\(\tfrac12\) moments on the honeycomb lattice. It is a canonical setting for studying how bond-directional exchange, fractionalization, and emergent gauge structure evolve under field tuning. In the pure Kitaev limit the model realizes a \(\mathbb Z_2\) quantum spin liquid with Majorana fermions and flux excitations, while field perturbations generate gapped non-Abelian descendants, polarized states, and—particularly for antiferromagnetic couplings—an intermediate-field regime that has been described in distinct ways, including as a gapless \(U(1)\) spin liquid, an abelian chiral phase with fermionic bulk excitations, and a symmetry-broken antiferromagnetic phase associated with anyon gap closings [1805.05953].

## 1. Microscopic definition

A standard form of the Zeeman–Kitaev honeycomb Hamiltonian is
\[
H \;=\;-K_x\!\sum_{\langle ij\rangle_x}\!\sigma_i^x\sigma_j^x\,-\,K_y\!\sum_{\langle ij\rangle_y}\!\sigma_i^y\sigma_j^y\,-\,K_z\!\sum_{\langle ij\rangle_z}\!\sigma_i^z\sigma_j^z \;-\;\sum_i\bigl(h_x\,\sigma_i^x+h_y\,\sigma_i^y+h_z\,\sigma_i^z\bigr)\,,
\]
where each bond \(\langle ij\rangle_\gamma\) carries an Ising-type exchange \(K_\gamma\) coupling the \(\gamma\)-components of two \(S=\tfrac12\) spins, and \({\bf h}=(h_x,h_y,h_z)\) is an arbitrary tilted magnetic field [1805.05953].

Several closely related formulations appear in the literature. A commonly studied isotropic version sets \(K_x=K_y=K_z\equiv K\), often with the field along \([111]\), so that \(\mathbf h\cdot \mathbf S_i = h(S_i^x+S_i^y+S_i^z)/\sqrt3\) [1905.11321]. In perturbative treatments the field also induces a three-spin “Haldane” term,
\[
H_{\kappa}\;=\;-\kappa\sum_{\hexagon}X_i\,Y_j\,Z_k\;\sim\;(h_xh_yh_z/K^2)\,
\sum_{\hexagon}X_iY_jZ_k,
\]
or equivalently a next-nearest-neighbor Majorana hopping that gaps the Majorana spectrum [2207.09492]. Extensions relevant to candidate materials further add isotropic Heisenberg exchange \(J\), off-diagonal symmetric \(\Gamma\) exchange, or second-neighbor Dzyaloshinskii–Moriya interaction \(H_{\rm DM}\) [1903.10026].

The model is therefore not a single isolated Hamiltonian but a core bond-directional spin model together with a hierarchy of field-induced and symmetry-allowed perturbations. This broader usage is especially important in the literature on honeycomb spin-orbit-entangled Mott insulators, where Zeeman, \(\Gamma\), \(J\), and DM terms are all treated as part of the effective low-energy description.

## 2. Fractionalization and emergent gauge structure

The exactly solvable Kitaev limit admits a Majorana-\(\mathbb Z_2\) gauge formulation. One representation is
\[
S_i^\alpha\;=\;\tfrac{i}{2}\,c_i\,b_i^\alpha,
\]
with the on-site constraint \(D_i\equiv c_i b_i^x b_i^y b_i^z=+1\), and bond variables
\[
u_{ij}\;=\;i\,b_i^\alpha\,b_j^\alpha\;=\;\pm 1
\]
on an \(\alpha\)-link \(\langle ij\rangle_\alpha\). The plaquette flux is
\[
W_p=\prod_{\langle ij\rangle\in p}u_{ij}=\pm1.
\]
In a fixed gauge-flux sector, the Kitaev Hamiltonian reduces to a quadratic Majorana problem, and different \(\{W_p\}\) sectors define distinct band structures [2207.09492].

This \(\mathbb Z_2\) formulation underlies the conventional low-field Kitaev spin liquid. It also furnishes the language for visons, which are flux excitations of the emergent gauge field. In the exactly solvable model they are static, whereas perturbations such as a Zeeman field induce vison hopping and, depending on the sign of \(K\), distinct symmetry implementations on vison motion [2207.09492].

A complementary low-energy description employs fermionic partons coupled to an emergent \(U(1)\) gauge field. One introduces Abrikosov fermions \(\psi_i\) and writes
\[
\mathcal{L}=\psi^\dagger\Bigl(iD_t-\tfrac{1}{2m}D_i^2-\mu\Bigr)\psi \;+\;g\bigl(\psi^\dagger\,\Delta(\partial)\,\psi^\dagger+\mathrm{h.c.}\bigr)\;+\;\cdots ,
\]
with \(D_\mu=i\partial_\mu+a_\mu\) and a chiral \(p\)-wave pairing channel \(\Delta\sim(\partial_x\pm i\partial_y)\). In this language the Kitaev spin liquid corresponds to \(\langle\psi\psi\rangle\neq0\), which Higgses \(U(1)\) to \(\mathbb Z_2\), while an intermediate field can restore full \(U(1)\) gauge invariance and produce a spinon Fermi surface coupled to a massless gauge photon [1805.05953].

The same emergent \(U(1)\) framework is used to describe field-induced internal gauge flux in the presence of second-neighbor DM interactions. The scalar spin chirality
\[
\chi_{ijk}\;=\;\mathbf{S}_i\!\cdot\!(\mathbf{S}_j\times\mathbf{S}_k)
\]
generates an internal flux \(\phi\) through second-neighbor triangles according to
\[
\sin\phi \;=\; \tfrac12\,\chi_{ijk} \;\simeq\; \tfrac12\,\lambda\,D_{111}\,\chi\,h,
\qquad \chi=\partial m/\partial h,
\]
with \(\phi\approx \tfrac12\lambda D_{111}\chi h\) for small flux [1905.11321].

## 3. Field-driven phases and their diagnostics

Large-scale exact diagonalization on clusters up to \(N=32\) sites has been used to map the phase diagram in tilted magnetic fields. The standard diagnostics include the low-lying spectrum \(E_n(h,\theta)\), the static structure factor
\[
S({\bf q}) \;=\;\frac1N\sum_{ij}\langle0|\mathbf S_i\!\cdot\!\mathbf S_j|0\rangle\,e^{i{\bf q}\cdot({\bf r}_i-{\bf r}_j)},
\]
the dynamical structure factor
\[
S^{\alpha\alpha}({\bf q},\omega)\;=\;\sum_n\bigl|\langle n|S^{\alpha}_{\bf q}|0\rangle\bigr|^2\, \delta\bigl(\omega-(E_n-E_0)\bigr),
\]
the plaquette flux operator \(W_P\), and the finite-temperature specific heat
\[
C(T)\;=\;\frac{\langle H^2\rangle_T-\langle H\rangle_T^2}{T^2}.
\]
Phase boundaries are identified from peaks in \(\partial^2E_0/\partial h^2\) and minima in the ground-state fidelity \(F(h)=\langle\Psi_0(h)|\Psi_0(h+\delta h)\rangle\) [1805.05953].

| Regime | Principal characterization | Representative hallmarks |
|---|---|---|
| Small field | Kitaev spin liquid (KSL) | three-fold quasi-degenerate torus manifold; finite vison gap; two peaks in \(C(T)\) |
| Intermediate field | gapless phase / intermediate phase | enhanced low-energy levels; flux-gap collapse; no sharp Bragg peaks |
| Large field | polarized (PL) phase | product state; magnon excitations; one Schottky-like peak in \(C(T)\) |

For antiferromagnetic \(K>0\), three regimes were reported as \(h\) increases. At small field, the Kitaev spin liquid survives and becomes gapped once the field has a component along \([111]\), acquiring non-Abelian Ising topological order. At large field, the system enters a trivial polarized phase. Between them, for a wide range of tilt angles, exact diagonalization identifies an intermediate gapless phase with a strong enhancement of low-energy states, a featureless \(S({\bf q})\), strong \(W_P\) fluctuations, collapse of the zero-field flux gap in \(S({\bf q},\omega)\), and a downward shift of the low-\(T\) specific-heat peak while the high-\(T\) peak remains almost unchanged [1805.05953].

In the Zeeman-plus-DM model, exact-diagonalization phase diagrams place an intermediate \(U(1)\) spin liquid in the approximate field window
\[
h\;\approx\;0.35\,K\;\dots\;0.60\,K
\]
for antiferromagnetic \(K>0\) when the field is close to \([111]\). In that setting the gapless \(U(1)\) phase survives a finite second-neighbor DM interaction up to \(|D_{111}|\sim0.025\,K\) [1905.11321].

## 4. Intermediate-field regime and competing descriptions

The intermediate-field regime in the antiferromagnetic model is the main conceptual focus of the Zeeman–Kitaev literature. One interpretation is a field-driven gapless \(U(1)\) spin liquid with fermionic spinons and a massless gauge field. In the parton description, the low-field Kitaev spin liquid is a paired state with \(\langle\psi\psi\rangle\neq0\), while increasing field suppresses the pairing amplitude at a critical \(g_c\), restores \(U(1)\) gauge invariance, and produces a spinon Fermi surface. The transition into the polarized state is then described as confinement via monopole proliferation once the spinons are gapped [1805.05953].

A distinct microscopic gauge-theory description has been formulated in terms of a lattice mutual Chern–Simons theory. In that construction, both ferro- and antiferromagnetic models exhibit low-field nonabelian Ising topological order with \(c=\tfrac12\). For the antiferromagnetic case, an intermediate-field phase appears with effective \(K\)-matrix
\[
K_{\rm int}=\begin{bmatrix}-1&0\\0&0\end{bmatrix},
\]
no ground-state degeneracy, \(\gamma=0\), and fermionic bulk excitations; despite the change in topological order, the chiral central charge remains \(c=\tfrac12\), so the half-quantized thermal Hall response is unchanged across the transition [2401.12750].

A third line of work analyzes field-induced anyon polarons. In that approach, visons, fermionic polarons, and bosonic bound states are tracked perturbatively in small Zeeman field. For the antiferromagnetic model, single visons become gapless at \(h_v^{\rm AFM}\approx0.45\,K\), the first fermion closing occurs at \(h_{\chi1}^{\rm AFM}\approx0.5\,K\), and boson condensation occurs at \(h_d^{\rm AFM}\approx0.5\,K\). The bosonic excitation carries the quantum numbers of an in-plane Néel order orthogonal to \([111]\), and its soft mode transforms as an \(l=1\) irreducible representation of \(C_3\); this was used to argue that the intermediate phase is a genuine antiferromagnetic phase rather than a spin liquid [2411.08105].

These descriptions are not identical. The literature therefore contains a genuine disagreement over whether the antiferromagnetic intermediate-field regime is best understood as a gapless \(U(1)\) spin liquid, an abelian chiral phase with trivial two-dimensional topological order, or a symmetry-broken ordered phase. What is common across these analyses is that the intermediate regime is tied to vison-gap collapse, strong field sensitivity of gauge-sector excitations, and a nontrivial restructuring of the low-energy spectrum.

## 5. Thermal Hall response, visons, and band topology

The Zeeman–Kitaev model is closely connected to thermal Hall transport because magnetic field breaks time-reversal symmetry and can induce topological band structure for emergent quasiparticles. In a \(U(1)\) spin liquid with a spinon Fermi surface, second-neighbor DM interactions and Zeeman coupling generate an internal \(U(1)\) gauge flux that twists spinon motion. At mean-field level the spinons occupy flux-twisted bands \(\epsilon_n(\mathbf k,\phi)\) with Berry curvature \(\Omega_{n\mathbf k}\), and the thermal Hall conductivity takes the form
\[
\kappa_{xy}(T) \;=\; -\frac{1}{T}\int_{-\infty}^{+\infty}d\epsilon\;(\epsilon-\mu)^2 \frac{\partial f(\epsilon)}{\partial\epsilon} \;\sigma_{xy}(\epsilon),
\]
with
\[
\sigma_{xy}(\epsilon) \;=\; -\frac{1}{\hbar}\sum_{n,\mathbf{k}}\Theta\bigl(\epsilon-\epsilon_{n\mathbf{k}}\bigr)\,\Omega_{n\mathbf{k}}.
\]
For \(\phi\neq0\), \(\kappa_{xy}/T\) approaches a finite, noninteger plateau as \(T\to0\), then decreases monotonically with temperature and vanishes at high \(T\) [1905.11321].

Vison motion contributes a separate topological channel. In a Zeeman field, the vison hopping problem maps to a tight-binding model on the triangular dual lattice. The key distinction is between ferromagnetic and antiferromagnetic Kitaev exchange. For \(K<0\), the vison experiences \(\Phi_{\rm UC}=0\pmod{2\pi}\), translations act non-projectively, the vison band has zero Berry curvature, and there is no intrinsic vison thermal Hall contribution. For \(K>0\), the vison experiences \(\Phi_{\rm UC}=\pi\pmod{2\pi}\), translations act projectively, the unit cell doubles, and two gapped vison bands appear with Chern numbers \(C_-=-1\) and \(C_+=+1\), yielding an intrinsic vison contribution to \(\kappa_{xy}\) [2207.09492].

The low-field non-Abelian phase generated by a field-induced three-spin term has its own quantized response. In proximate Kitaev spin liquids of the \(K\)-\(J\)-\(\Gamma\) model, a field normal to the plane gaps the state into chiral spin liquids with Chern numbers \(\nu=5\) and \(\nu=4\), corresponding to
\[
\frac{\kappa_{xy}}{T}=\frac{\nu}{2}\,\frac{\pi k_B^2}{6h},
\]
before a transition to a trivial polarized phase [1903.10026]. This shows that thermal Hall plateaus in honeycomb Kitaev systems need not be limited to the conventional \(\nu=1\) response.

## 6. Perturbations, extensions, and broader formulations

The field-driven phenomenology of the Zeeman–Kitaev model has been tested against perturbations expected in materials. Exact diagonalization with an additional Heisenberg term \(J\sum_{\langle ij\rangle}\mathbf S_i\!\cdot\!\mathbf S_j\) and off-diagonal \(\Gamma\) exchange found that for modest \(|J|,|\Gamma|\lesssim0.1|K|\), both the Kitaev spin liquid and the intermediate gapless phase survive, with the main effect being a shift of critical fields. Larger perturbations favor magnetically ordered states including zig-zag, Néel, and “vortex” patterns [1805.05953].

Variational Monte Carlo studies of the \(K\)-\(J\)-\(\Gamma\) model further identify a “generic” Kitaev spin liquid and a single proximate Kitaev spin liquid. The proximate phase is a gapless \(\mathbb Z_2\) state with 14 Majorana cones and a gapless spin response; in a field along \(\hat c=(1,1,1)/\sqrt3\), it realizes a non-Abelian chiral spin liquid with \(\nu=5\), then an Abelian chiral spin liquid with \(\nu=4\), and only at larger field a trivial polarized phase [1903.10026]. This places the Zeeman perturbation within a broader landscape of nearby honeycomb spin liquids.

Recent work has also recast the model in more general duality frameworks. A staggered-Majorana or “dumbbell fermion” construction expresses Kitaev’s exact solution as a special case of a higher-dimensional duality between Pauli spins and Majorana fermions, with the one-particle problem in each \(\mathbb Z_2\) flux sector governed by
\[
i\sum_j A_{ij}\,\psi_j=\epsilon\,\psi_i.
\]
In this formulation the Zeeman term spoils exact solvability, but third-order perturbation theory again generates the effective chiral term proportional to \(h_xh_yh_z/(K_xK_yK_z)\) that gaps the Majorana spectrum [2502.13975].

Taken together, these developments place the Zeeman–Kitaev honeycomb model at the intersection of exactly solvable \(\mathbb Z_2\) gauge theory, emergent \(U(1)\) gauge dynamics, topological band theory of fractionalized quasiparticles, and materials-motivated extensions. The recurring technical themes are the competition between Majorana and vison sectors, the role of field-induced chirality and gauge flux, and the sensitivity of the intermediate regime to both microscopic perturbations and the theoretical framework used to describe it.

Source: https://www.emergentmind.com/topics/zeeman-kitaev-honeycomb-model