---
title: 3D Monitored Kitaev Models
url: https://www.emergentmind.com/topics/3d-monitored-kitaev-models
type: topic
---

# 3D Monitored Kitaev Models

Searching arXiv for recent and foundational papers on 3D Kitaev models, spectroscopy/monitoring, and multilayer stacked Kitaev systems.
“3D monitored Kitaev models” does not denote a single standardized model class in the present literature. The phrase instead intersects three tightly related but conceptually distinct directions: exactly solvable and perturbed **three-dimensional Kitaev spin liquids** on tricoordinated lattices; **monitored** 3D Kitaev physics in the condensed-matter sense of being probed by dynamical observables such as inelastic neutron scattering, electron spin resonance, resonant inelastic X-ray scattering, NMR, susceptibility, and thermal probes; and, more loosely, **multilayer or stacked Kitaev-honeycomb systems** that route two-dimensional Kitaev physics toward realistic quasi-three-dimensional materials [1604.05199], [1705.05894], [1705.07760], [2410.14338]. By contrast, the modern quantum-information meaning of “monitored dynamics”—continuous or repeated measurements producing nonunitary evolution and measurement-induced entanglement transitions—has so far been developed in the cited corpus only for **two-dimensional** Kitaev-type monitored circuits, not for genuine 3D lattices [2509.16758], [2511.20545].

## 1. Three-dimensional Kitaev systems as the substrate of “monitoring”

Three-dimensional Kitaev models are defined on **tricoordinated lattices** whose nearest-neighbor bonds can be partitioned into \(x\), \(y\), and \(z\) types, permitting the bond-directional Hamiltonian
\[
\hat{H} = -\frac{J_x}{2}\sum_{\langle jk\rangle_x}\hat{\sigma}_j^x\hat{\sigma}_k^x -\frac{J_y}{2}\sum_{\langle jk\rangle_y}\hat{\sigma}_j^y\hat{\sigma}_k^y -\frac{J_z}{2}\sum_{\langle jk\rangle_z}\hat{\sigma}_j^z\hat{\sigma}_k^z .
\]
This structure underlies the hyperhoneycomb, hyperoctagon, hyperhexagon, stripyhoneycomb, hypernonagon, and related \((p,3)\) lattices studied in the 3D Kitaev literature [1604.05199], [1511.05569], [1907.10241], [2006.07386].

The exact solution proceeds through Majorana fractionalization. In one common notation,
\[
\hat{\sigma}_j^\alpha = i \hat b_j^\alpha \hat c_j,
\]
with static \(\mathbb Z_2\) bond operators
\[
\hat u_{jk} = i \hat b_j^{\alpha_{jk}} \hat b_k^{\alpha_{jk}},
\]
so that the spin Hamiltonian becomes a free-Majorana hopping problem in a static gauge background [1604.05199], [1508.05324], [1511.05569]. The physically conserved gauge-invariant quantities are loop or plaquette operators such as
\[
\hat{\mathcal W}_\gamma=\prod_{l\in\gamma}\tilde K_l,\qquad \tilde K_l=-i\hat u_{jk},
\]
or, equivalently, products of bond terms around closed loops [1604.05199].

A defining three-dimensional feature is that gauge-flux excitations are generally **loop-like** rather than point-like. This distinction is central for thermodynamics, spectroscopy, and topological excitations. In the gapped strong-coupling limit of a 3D Kitaev model introduced on a trivalent lattice, the low-energy theory reduces to a commuting plaquette Hamiltonian on an effective diamond lattice, with excitations constrained to form closed loops in an embedded lattice; the shortest such loops are fermionic and braid nontrivially with larger loops, giving a phase \(\pi\) when a fermion winds through a loop [1101.3718]. In more general 3D Kitaev models, the same loop-based gauge structure reappears thermodynamically as vison-loop proliferation and spectroscopically as localized flux-loop insertions created by spin flips [1705.07760], [2006.07386].

The low-energy Majorana sector of 3D Kitaev spin liquids is richer than in the two-dimensional honeycomb model. Symmetry analysis and exact solutions show that depending on lattice geometry and the projective implementation of time-reversal and inversion, the itinerant Majorana fermions may form **Majorana Fermi surfaces**, **nodal lines**, or **Weyl points** [1511.05569]. Representative cases are the hyperoctagon lattice with a Majorana Fermi surface, the hyperhoneycomb lattice with nodal lines, and the hyperhexagon lattice with Weyl points [1604.05199].

## 2. “Monitoring” as spectroscopy: INS, ESR, and RIXS

In the 3D Kitaev literature, the most developed meaning of “monitored” is **experimentally probed**. The key observables are the dynamical spin structure factor for inelastic neutron scattering and ESR, and channel-resolved resonant inelastic X-ray scattering [1604.05199], [1508.05324], [1705.05894].

For INS and ESR, the central object is the dynamical spin structure factor
\[
S(\mathbf q,\omega) = \sum_{a,b,j,k} e^{-i\mathbf q\cdot(\mathbf r_j-\mathbf r_k)} \int_{-\infty}^{\infty}dt\,e^{i\omega t} S^{ab}_{jk}(t),
\]
with \(S^{ab}_{jk}(t)=\langle 0|\hat\sigma_j^a(t)\hat\sigma_k^b(0)|0\rangle\) [1604.05199]. In an ideal Kitaev model, a local spin operator does not create a bare magnon. Instead, it flips a bond variable, inserts flux excitations, and launches Majorana matter in the altered gauge background. In 3D this leads to a local quantum quench involving a **static flux-loop insertion** plus **dynamical Majorana matter fermions** [1604.05199]. As a result, the spin response is thresholded by a **flux gap**,
\[
S(\mathbf q,\omega)=0,\qquad \omega<\Delta,
\]
even when the Majorana sector is gapless [1604.05199], [1508.05324].

The threshold line shape diagnoses the type of 3D Majorana metal. If the Majorana density of states vanishes as
\[
\rho(\varepsilon)\propto \varepsilon^x,\qquad x>0,
\]
then the structure factor above threshold behaves as
\[
S(\omega)\propto (\omega-\Delta)^x.
\]
For the hyperhoneycomb nodal-line case, \(S(0,\omega)\) rises linearly above threshold, while for the hyperhexagon Weyl case it rises quadratically [1604.05199]. By contrast, for a Majorana Fermi surface with finite \(\rho(0)\), the threshold is a genuine **Majorana X-ray-edge singularity**
\[
S(\omega)\propto (\omega-\Delta)^{-\alpha},\qquad \alpha=2g-g^2,
\]
with
\[
g=\frac{\delta_0}{\pi}, \qquad \delta_0 = -\arctan\!\left( \frac{2J\,\Im[G_0^R(0)]} {1+\tilde J\,\Re[G_0^R(0)]} \right)
\]
[1604.05199]. This separation between divergent, linear, and quadratic threshold behavior is one of the most concrete monitored signatures of distinct 3D Kitaev spin liquids.

The hyperhoneycomb INS problem was analyzed in detail in a thermodynamic-limit exact treatment. There the response remains broad and continuum-like but exhibits a **response gap even in the gapless QSL phase**, because any spin flip necessarily creates flux loops [1508.05324]. The same work found that the response is dominated by **single-Majorana final states**, accounting for \(87\%\) of the response at \(\mathbf q=0\) for isotropic couplings, which is why INS acts as relatively direct spectroscopy of the Majorana density of states [1508.05324].

RIXS refines this monitored picture because it resolves distinct channels. The full intensity is written as
\[
I(\omega,\mathbf q)=\sum_m |A(m,\mathbf q)|^2\,\delta(\omega-E_m),
\]
with amplitudes decomposed into spin-conserving and non-spin-conserving channels [1705.05894]. In the **non-spin-conserving** channels, the leading term reduces to the spin-polarized INS amplitude and is dominated by **flux-creating processes**. In the **spin-conserving** channel, the first nontrivial contribution creates **no fluxes** and probes the **Majorana matter sector directly** through a two-fermion continuum
\[
I_0(\omega,\mathbf q)\propto \sum_{\mathbf k,\mu,\mu'} \left|(\mathcal A_{\mathbf q,\mathbf k})_{\mu\mu'}\right|^2 \delta\!\left( \omega-\varepsilon_{\mathbf k,\mu} -\varepsilon_{\mathbf q-\mathbf k,\mu'} \right) \Theta(\varepsilon_{\mathbf k,\mu}) \Theta(\varepsilon_{\mathbf q-\mathbf k,\mu'}).
\]
This permits direct identification of whether the 3D spin liquid hosts nodal lines, Weyl points, or Fermi surfaces [1705.05894]. A universal additional signature is the suppression of the spin-conserving RIXS response near \(\Gamma\), traced to destructive interference between sublattices and interpreted as a spectroscopic manifestation of symmetry fractionalization [1705.05894].

## 3. Thermodynamic and dynamical monitoring of the 3D gauge sector

A second major sense of monitoring is through **temperature-dependent response functions** that track the emergent \(\mathbb Z_2\) gauge sector. The most complete such analysis for a true 3D Kitaev lattice is the QMC+CTQMC study of the hyperhoneycomb model [1705.07760].

There the monitored observables are the magnetic susceptibility,
\[
\chi = \frac{1}{N} \sum_{j,j'} \int_0^\beta d\tau\langle S^z_j (\tau) S^z_{j'}\rangle,
\]
the NMR relaxation rate \(1/T_1\), the dynamical spin structure factor
\[
S(\mathbf q, \omega) = \frac1N \sum_{j,j'} e^{i\mathbf q\cdot(\mathbf r_j-\mathbf r_{j'})}S^z_{j,j'}(\omega),
\]
and a direct gauge-flux fluctuation measure
\[
\Delta W_p = \frac{1}{N_p T^2}\Big( \big\langle \big(\sum_p W_p \big)^2 \big\rangle - \big\langle \sum_p W_p \big\rangle^2 \Big)
\]
[1705.07760]. The isotropic hyperhoneycomb model shows a high-temperature crossover
\[
T_H \simeq 0.375
\]
associated with itinerant matter Majoranas, and a genuine 3D finite-temperature transition
\[
T_c \simeq 0.0039
\]
driven by the proliferation of looplike gauge fluxes [1705.07760]. Near \(T_c\), \(d\chi/dT\), \(d(1/T_1)/dT\), and \(\Delta W_p\) all develop sharp peaks, and \(S(\Gamma,\omega)\) shows strong quasi-elastic weight above the transition and a shift of the low-energy peak to higher \(\omega\) below it, signaling flux-gap opening [1705.07760]. The comparison with the two-dimensional honeycomb model is explicit: in 2D the corresponding lower scale is only a crossover at \(T_L \simeq 0.012\), whereas in 3D the dynamical signatures are singular [1705.07760].

A broader thermodynamic classification across elementary 3D tricoordinated lattices was then developed by sign-free Majorana quantum Monte Carlo [2006.07386]. Across a broad class of bipartite lattices, the ground-state flux sector is organized by the elementary plaquette length \(p\): \(p\bmod 4=2\) gives **0-flux**, while \(p\bmod 4=0\) gives **\(\pi\)-flux**, extending Lieb-type intuition beyond the cases where Lieb’s theorem is rigorously applicable [2006.07386]. The low-temperature gauge transition correlates strongly with the **smallest vison gap \(\Delta\)** rather than simply with loop length. For example, \((10,3)b)\) has \(\Delta=0.0426(4)\) and \(T_c=0.00519(9)\), while \((8,3)b)\) has \(\Delta=0.0532(3)\) and \(T_c=0.0079(3)\) [2006.07386].

That classification also isolates two important exceptions. The \((8,3)c\) lattice exhibits **gauge frustration**, because local flux preferences cannot be simultaneously satisfied; the low-temperature average flux becomes
\[
\overline{W_p}=-\frac13,
\]
and the transition is strongly suppressed to \(T_c=0.0020(2)\) [2006.07386]. The odd-loop hypernonagon \((9,3)a\) instead has \(W_p=\pm i\), intertwining gauge ordering with spontaneous time-reversal breaking and producing a **single first-order transition** at
\[
T_c=0.00244(4)
\]
into a crystalline \(\mathbb Z_2\) gauge-ordered phase [2006.07386], [1907.10241].

Thermodynamic tensor-network work on the hyperhoneycomb Kitaev model reinforces the same 3D picture. In the thermodynamic limit, the specific heat shows a high-temperature crossover at
\[
T' \approx 0.256
\]
for the isotropic case and a lower gauge-ordering transition estimated at
\[
T_c \approx 0.006,
\]
to be compared with a quoted QMC benchmark
\[
T_c^{\rm QMC}=0.0024
\]
[2011.11577]. The entropy drops from \(\ln 2\) to \(\ln 2/2\) at the high-temperature crossover, then to zero below the gauge-ordering scale, expressing 3D fractionalization as a two-stage process [2011.11577].

## 4. Gauge order, topological excitations, and Majorana topology in 3D

The monitored signatures of 3D Kitaev matter are inseparable from the structure of the gauge sector and the topology of the Majorana bands. In the gapped strong-coupling phase of a 3D Kitaev model on a trivalent lattice, degenerate perturbation theory produces an effective Hamiltonian on the diamond lattice,
\[
H_{\mathrm{eff}} = -\frac{7}{256J_z^5} \left( J_x^4J_y^2\sum_p B_p + J_x^2J_y^4\sum_p' B_p \right),
\]
with commuting loop operators \(B_p\) and loop excitations constrained by local, surface, and volume constraints [1101.3718]. The elementary shortest-loop excitations are fermions, and noncontractible loop and surface operators produce an \(8\)-fold ground-state degeneracy on the three-torus [1101.3718]. This provides a canonical example of 3D Kitaev topological order in the low-energy gapped regime.

In gapless phases, the classification is instead in terms of Majorana semimetals. A symmetry-based analysis of elementary 3D tricoordinated Kitaev lattices shows that projective time-reversal and inversion symmetries determine whether the itinerant Majoranas realize Fermi surfaces, nodal lines, or Weyl points [1511.05569]. Inversion may act projectively as
\[
\hat h(\mathbf k) = U_{\rm I}\,\hat h(-\mathbf k+\tilde{\mathbf k}_0)\,U_{\rm I}^{-1},
\]
and time reversal as
\[
\hat h(\mathbf k) = U_{\rm T}\,\hat h^\ast(-\mathbf k+\mathbf k_0)\,U_{\rm T}^{-1},
\]
so the zero-energy manifold is set by the pair \((\mathbf k_0,\tilde{\mathbf k}_0)\) rather than by nonprojective Bloch-band symmetry alone [1511.05569]. This explains, for example, why the hyperhoneycomb \((10,3)b)\) carries nodal lines, while \((8,3)b)\) can host Weyl nodes even with physical time-reversal and inversion symmetry preserved at the spin level [1511.05569].

Odd-loop 3D Kitaev lattices introduce a further layer of structure. On the hypernonagon lattice, the odd plaquette length forces
\[
W_p^2=-1,\qquad W_p=\pm i,
\]
so spontaneous time-reversal breaking occurs already at the level of the flux sector [1907.10241]. Extensive QMC and variational analysis identify at least five distinct chiral spin-liquid phases—A0F, AF, AFII, SI, and SII—all with **crystalline ordering of the \(\mathbb Z_2\) fluxes** [1907.10241]. At the isotropic point the system enters the AFII phase at
\[
T_c = 0.00244(4),
\]
with a flux-loop gap
\[
\Delta = 0.0339(1),
\qquad T_c/\Delta \approx 0.07
\]
[1907.10241]. Depending on flux-crystal pattern and anisotropy, the itinerant Majorana sector may be fully gapped, contain Weyl nodes with Fermi arcs, or host nodal lines with drumhead states [1907.10241].

This separation between **gauge-crystal order** and **Majorana topology** is especially important. The hypernonagon study shows explicitly that flux-order boundaries and Majorana gapped/gapless boundaries do not generally coincide; a fixed flux crystal can support different Majorana topologies as couplings are varied [1907.10241]. A plausible implication is that any future genuine 3D measurement-driven “monitored” construction will need to track gauge-sector ordering and matter-sector entanglement or band topology as distinct, though coupled, structures.

## 5. Stacked and multilayer Kitaev systems as a quasi-3D route

A different but physically important route into “3D monitored Kitaev models” is through **stacks of 2D Kitaev-honeycomb layers**. The multilayer Kitaev model of stacked honeycomb planes coupled by interlayer Heisenberg exchange is not a fully three-dimensional bond-dependent Kitaev lattice, but it is explicitly motivated by the three-dimensional structure of layered candidate materials such as \(\alpha\)-RuCl\(_3\) and H\(_3\)LiIr\(_2\)O\(_6\) [2410.14338].

The model is
\[
H = H_K + H_J ,
\]
with
\[
H_K = -K \sum_{l=1}^{n}\sum_{\langle i,j\rangle_a} \hat S^a_{l,i}\hat S^a_{l,j},
\]
and
\[
H_J = -J \sum_{i,l} \hat{\mathbf S}_{l,i}\cdot \hat{\mathbf S}_{l+1,i}.
\]
Here \(n\) is the number of honeycomb layers and \(J\) is an ordinary interlayer Heisenberg exchange [2410.14338]. An Abrikosov-fermion mean-field treatment plus projective symmetry analysis finds self-consistent quantum spin-liquid states built from layer-by-layer Kitaev ansätze whose bond amplitudes **alternate in sign from one layer to the next** [2410.14338].

The principal result is an **even–odd effect** in the number of layers. For small \(|J|\), the system is just a direct product of \(n\) single-layer Kitaev spin liquids. Beyond a first critical coupling \(J_{c1}\), a new multilayer QSL appears. For **even \(n\)** it is **gapped**, because the Dirac cones of the single-layer KSL gap out over an intermediate regime \(J_{c1}<|J|<|J_{c2}|\). For **odd \(n\)** it remains **gapless** even at arbitrarily large \(|J|\), because one effective half-odd-integer degree of freedom remains unpaired [2410.14338]. The paper calls these states **hybrid QSLs**, emphasizing that they are not just independent copies of the single-layer KSL [2410.14338].

For concrete cases, with ferromagnetic interlayer coupling the bilayer has \(J_{c1}\approx 2\) and \(J_{c2}\approx 4\), while the trilayer has a single transition near \(J_{c1}\approx 2\). With antiferromagnetic coupling the bilayer has \(J_{c1}\approx -0.75\) and \(J_{c2}\approx -1.45\), and the trilayer again has only one transition near \(J_{c1}\approx -0.75\) [2410.14338]. Under a \([111]\) magnetic field,
\[
\mathbf B = \frac{(1,1,1)}{\sqrt 3}\, B,
\]
the multilayer systems inherit a field-induced gapped chiral descendant of the Kitaev spin liquid, and **antiferromagnetic interlayer exchange stabilizes this gapped chiral KSL over a broader \(B\)-\(J\) window** [2410.14338].

This stacked-lattice route is relevant to 3D Kitaev phenomenology, but the paper is explicit that it does **not** study monitoring in the sense of quantum measurement dynamics, monitored circuits, or measurement-induced transitions [2410.14338]. Its relevance is instead that realistic 3D Kitaev materials are often structurally layered, so finite stacks already display parity-dependent gapped and gapless hybrid QSL behavior controlled by projective layer symmetries [2410.14338].

## 6. Relation to measured materials, perturbed Hamiltonians, and the current boundary of genuine “monitored dynamics”

Several works place 3D Kitaev monitoring into direct material context. The review of Kitaev materials identifies \(\beta\)-Li\(_2\)IrO\(_3\) and \(\gamma\)-Li\(_2\)IrO\(_3\) as the first truly three-dimensional Kitaev materials, built from tricoordinated edge-sharing IrO\(_6\) octahedra and expected to realize dominant bond-directional exchange [1701.07056]. It emphasizes that the 3D gauge sector differs from 2D because vison excitations are **closed flux loops**, leading to a genuine finite-temperature gauge transition around
\[
T \approx 0.004\,K
\]
for the hyperhoneycomb Kitaev model, whereas the higher fractionalization crossover occurs around
\[
T \approx 0.6\,K
\]
[1701.07056]. This establishes why thermodynamic and dynamical probes in 3D can monitor separate matter-fermion and gauge-loop scales.

Beyond the solvable limit, the nearest-neighbor 3D hyperhoneycomb Kitaev–Heisenberg model has been mapped out both semiclassically and by pseudofermion functional renormalization group. The PFFRG phase diagram contains QSL regions near both AFM and FM Kitaev limits and four ordered phases—Néel, zigzag, FM, and stripy—with isotropic QSL windows
\[
0.4625\lesssim \varphi/\pi \lesssim 0.5375,
\qquad
1.2625\lesssim \varphi/\pi \lesssim 1.6375
\]
[2303.09156]. That work also stresses a specifically 3D finite-temperature topological transition estimate
\[
T_c\sim 0.0078
\]
for the pure hyperhoneycomb Kitaev model, corresponding heuristically to \(\Lambda_c\sim 0.005\) in the PFFRG cutoff language [2303.09156]. A broader tensor-network study of the 3D hyperoctagon Kitaev–Heisenberg model similarly finds narrow QSL windows around the pure Kitaev points and conventional Landau thermal transitions in ordered phases, contrasting them with the non-Landau gauge-ordering physics of the Kitaev QSL [2011.11577].

What remains largely absent in the cited 3D literature is the **modern measurement-theoretic meaning** of monitored dynamics. The two explicitly monitored Kitaev works in the corpus—on qudit generalizations of the monitored Kitaev model and on the measurement-only honeycomb Kitaev model with added single-, three-, and four-qubit checks—are entirely **two-dimensional** [2509.16758], [2511.20545]. They show that measurement-only Kitaev circuits can support topological area-law, critical-law, and distinct volume-law phases, and that commuting versus noncommuting added checks organize whether plaquette flux memory is preserved [2509.16758], [2511.20545]. A plausible implication is that a genuine future “3D monitored Kitaev model” in the quantum-information sense would likely have to decide whether measurements preserve 3D loop- or membrane-like gauge invariants, whether they remain quadratic in Majoranas, and whether monitored dynamics disorders the 3D gauge sector by a process analogous to vison-loop proliferation. But those questions are not answered in the present 3D corpus.

The current state of the field therefore supports a precise summary. “3D monitored Kitaev models” is well established if “monitored” means **spectroscopically probed** or **thermodynamically diagnosed**: 3D Kitaev spin liquids are monitored by INS, ESR, RIXS, NMR, susceptibility, specific heat, and thermal transport, and these observables reveal flux thresholds, vison-loop physics, gauge-ordering transitions, and the topology of Majorana Fermi surfaces, nodal lines, and Weyl nodes [1604.05199], [1508.05324], [1705.05894], [1705.07760], [2006.07386]. It is also meaningful in a broader materials sense through stacked multilayer Kitaev models that interpolate between 2D and realistic quasi-3D layered compounds [2410.14338]. But as a class of **measurement-induced, nonunitary 3D quantum dynamical models**, it remains essentially an open direction within the cited literature.

Source: https://www.emergentmind.com/topics/3d-monitored-kitaev-models