Spinon-Kondo Effect in Fractionalized Systems

• Spinon-Kondo effect is the screening of magnetic impurities by fractionalized spin excitations, where spinons replace conventional electrons.
• Experiments on quantum spin liquids and one-dimensional systems reveal subgap singularities and distinct spectroscopic signatures from standard Kondo resonances.
• Lattice generalizations show that Kondo dynamics can generate emergent spinon bands and hybridization gaps, offering novel insights into charge-insulating hosts.

Spinon-Kondo effect denotes a class of Kondo phenomena in which the screening bath is not an ordinary electron metal, but a sector of fractionalized spin excitations—most commonly spinons in a quantum spin liquid or in a critical one-dimensional spin system. In the current literature, the term spans several closely related settings: magnetic-impurity screening by a spinon Fermi sea in a charge-insulating host; boundary screening by spinons in spin chains; subgap threshold singularities in Mott insulators where an impurity is screened by a spinon rather than by an electron; and lattice problems in which Kondo dynamics itself generates or hybridizes spinon degrees of freedom (Gomilšek et al., 2019, Chen et al., 2022, Kattel et al., 2023, Kattel et al., 2024, Ge et al., 2022, Zheng et al., 2024, Pereira et al., 26 Aug 2025).

1. Definition and conceptual scope

In the conventional single-impurity Kondo problem, a localized spin-$\tfrac12$ moment is exchange-coupled to itinerant conduction electrons, and below a dynamically generated scale $T_K$ the impurity flows to a strongly correlated screened regime. The spinon-Kondo effect preserves the many-body logic of this mechanism while replacing the electronic bath by a spinon bath. In the formulation used for a spinon metal, the reference Hamiltonian is written in the standard Kondo form,

$$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$

but the operators $c^\dagger_{k\sigma},c_{k\sigma}$ now create and annihilate itinerant spinons rather than charged electrons (Gomilšek et al., 2019).

This replacement is physically nontrivial. A spinon bath can be charge insulating while remaining spinful and itinerant, so the impurity can be Kondo screened without metallic charge transport. That distinction is central in the experimental interpretation of charge-insulating quantum spin liquids and Mott insulators: the existence of screening does not require an electron Fermi surface, only a low-energy bath of itinerant spin-carrying excitations (Gomilšek et al., 2019, Chen et al., 2022).

The expression is also used more broadly for boundary and lattice settings. In one-dimensional antiferromagnetic spin chains, the impurity is screened by a many-spinon cloud rather than by electrons, yielding a literal spinon-based Kondo phenomenon at the edge (Kattel et al., 2023, Kattel et al., 2024). In two-channel Kondo lattices, the Kondo effect can dynamically generate propagating fractionalized modes and an emergent spinon dispersion, so that “spinon-Kondo effect” refers not only to screening by pre-existing spinons but also to the self-consistent production of mobile spinons by Kondo correlations (Ge et al., 2022).

The term should not be conflated with all spin-dependent variants of the ordinary Kondo effect. Electron-based Kondo problems modified by pure spin current, spin-orbit coupling, or arbitrary spin-momentum locking remain conventional impurity problems in nontrivial baths; they do not introduce fractionalized spinons as the screening degrees of freedom (Hamaya et al., 2016, Puel et al., 5 Apr 2025, Goto et al., 5 Jun 2025).

2. Screening by spinons in charge-insulating hosts

The most direct experimental realization reported so far is “Kondo screening in a charge-insulating spinon metal” in Zn-brochantite, $\mathrm{ZnCu_3(OH)_6SO_4}$, a gapless kagome quantum spin liquid whose low-energy excitations are interpreted as a spinon Fermi surface. The central result is that a magnetic impurity behaves as though it were screened by a Fermi sea, even though the host remains a charge insulator. The muon Knight shift

$K = (\nu_1-\nu_2)/\nu_2$

saturates at low temperature instead of following Curie behavior, and the data are described by NRG with a Kondo scale $T_K=1.3(1)\,\mathrm{K}$. At $B=0.3\,\mathrm{T}$, the low-$T$ plateau is about $K\simeq 0.18\%$, corresponding to $T_K$0 per magnetic Cu ion and roughly $T_K$1 per impurity for the quoted impurity concentration, far below the free-spin expectation (Gomilšek et al., 2019).

The same study identifies a field scale associated with Kondo-resonance splitting. The muon relaxation anomaly occurs at

$T_K$2

with $T_K$3 and $T_K$4, consistent with a strong-coupling Kondo picture in which the resonance is Zeeman split once the field becomes comparable to the many-body scale. This is important because it reproduces a hallmark of electronic Kondo physics in a host whose low-energy carriers are argued to be chargeless spinons (Gomilšek et al., 2019).

A second experimental line comes from cobalt adatoms on single-layer $T_K$5-TaSe$T_K$6, treated as a gapless quantum spin liquid candidate with a spinon Fermi surface. STM/STS detects impurity-induced resonances near both Hubbard band edges when Co adatoms occupy the star-of-David center with maximal overlap to the Hubbard-band charge distribution: $T_K$7 near $T_K$8 and $T_K$9 near $$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$0, with reported widths of about $$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$1 and $$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$2. These peaks disappear for reduced spatial overlap and for nonmagnetic Au adatoms. The accompanying theory uses a modified Anderson impurity model on a gapless QSL, slave-rotor fractionalization $$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$3, and gauge-fluctuation-induced spinon-chargon binding. Within that description the impurity spin sector develops a spinon Kondo resonance with $$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$4, while the experimentally visible Hubbard-edge side peaks arise because the impurity-centered spinon Kondo cloud binds chargons through the emergent gauge field (Chen et al., 2022).

These two cases establish complementary versions of the same principle. In Zn-brochantite, the evidence is bulk magnetic and dynamical, expressed through impurity moment reduction and field-dependent strong-coupling behavior. In $$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$5-TaSe$$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$6, the evidence is local and spectroscopic, but the measured electronic resonances are not interpreted as a metallic zero-bias Kondo peak; they are the electron-channel manifestation of a spinon Kondo cloud recombined with charge-sector excitations (Gomilšek et al., 2019, Chen et al., 2022).

3. Boundary and spectroscopic manifestations in one dimension

One-dimensional systems sharpen the distinction between electron Kondo screening and spinon Kondo screening because spin-charge separation is explicit. In a half-filled Hubbard chain with a boundary Anderson impurity, the bulk is a Mott insulator with a charge gap

$$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$7

Below this gap, a zero-energy metallic Kondo resonance is impossible, but the spin sector remains gapless. The impurity is therefore screened by a spinon drawn from the spin sector, not by an electron. The predicted tunneling density of states near the boundary has a subgap threshold singularity

$$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$8

and near the strong-coupling fixed point the exponent becomes universal, $$H=\sum_{k\sigma} \epsilon_k c^\dag_{k\sigma} c_{k\sigma} + J \mathbf{S} \cdot \mathbf{s}(\mathbf{r}=0) + g \mu_B S_z B ,$$9, yielding

$c^\dagger_{k\sigma},c_{k\sigma}$0

The paper identifies this one-sided power law as the spectroscopic signature of a boundary spinon-Kondo effect rather than an ordinary metallic Kondo peak (Pereira et al., 26 Aug 2025).

The strong-coupling picture in that problem is explicit. The impurity and a boundary spinon form a singlet; tunneling breaks the singlet and creates a composite threshold excitation involving a holon and a spinon. DMRG finds a subgap peak at the impurity and boundary sites, rapidly decaying with distance from the boundary, with fitted exponents approaching $c^\dagger_{k\sigma},c_{k\sigma}$1 as the impurity coupling increases. This directly ties the threshold singularity to spinon-mediated screening in a gapped charge background (Pereira et al., 26 Aug 2025).

A related but distinct realization occurs in integrable spin chains, where the bath consists purely of spinons. In the isotropic antiferromagnetic Heisenberg chain with a boundary spin-$c^\dagger_{k\sigma},c_{k\sigma}$2 impurity, the impurity is screened by a many-spinon cloud for antiferromagnetic coupling $c^\dagger_{k\sigma},c_{k\sigma}$3. The Kondo scale is defined from the impurity density of states and is given by

$c^\dagger_{k\sigma},c_{k\sigma}$4

with $c^\dagger_{k\sigma},c_{k\sigma}$5 the maximum energy carried by a single spinon. When $c^\dagger_{k\sigma},c_{k\sigma}$6, the screening mechanism changes: the impurity is no longer screened by a Kondo cloud but by an exponentially localized boundary mode with energy

$c^\dagger_{k\sigma},c_{k\sigma}$7

The transition is therefore not the disappearance of screening but a reorganization from many-body spinon screening to single-particle boundary-mode screening (Kattel et al., 2023).

The $c^\dagger_{k\sigma},c_{k\sigma}$8 chain displays an analogous boundary transition in a quadratic setting. For $c^\dagger_{k\sigma},c_{k\sigma}$9, the impurity is screened by a multiparticle Kondo cloud of spinons. For $\mathrm{ZnCu_3(OH)_6SO_4}$0, a boundary string appears and the impurity is screened by a localized bound mode with energy

$\mathrm{ZnCu_3(OH)_6SO_4}$1

The impurity spinon density of states has a Lorentzian-like low-energy peak in the Kondo phase and a delta-function contribution at $\mathrm{ZnCu_3(OH)_6SO_4}$2 in the bound-mode phase. The chain remains critical in the bulk, but boundary conformal invariance is lost once the massive boundary mode appears (Kattel et al., 2024).

Taken together, these one-dimensional works show that “spinon-Kondo effect” need not imply a metallic-like resonance at the Fermi level. Depending on whether the host is a Mott insulator or a pure spin chain, the natural signature may be a subgap threshold singularity, a Lorentzian impurity spinon density of states, or a boundary eigenstate phase transition between collective and localized screening (Pereira et al., 26 Aug 2025, Kattel et al., 2023, Kattel et al., 2024).

4. Lattice generalizations and emergent spinon bands

Beyond isolated impurities, the literature has developed lattice-scale versions of the same idea. In the two-channel Kondo lattice, the Kondo effect can induce dynamic magnetic correlations that generate an emergent spinon dispersion. Using a dynamical large-$\mathrm{ZnCu_3(OH)_6SO_4}$3 formulation with Schwinger bosons $\mathrm{ZnCu_3(OH)_6SO_4}$4 for local moments and Grassmann holons $\mathrm{ZnCu_3(OH)_6SO_4}$5 for the Kondo sector, the coupled self-energies are

$\mathrm{ZnCu_3(OH)_6SO_4}$6

with

$\mathrm{ZnCu_3(OH)_6SO_4}$7

Because each self-energy depends on the propagator of the other fractionalized field, the Kondo process self-consistently generates spin dynamics even when the bare spinon hopping is absent or very small (Ge et al., 2022).

This mechanism is dimension dependent. In the $\mathrm{ZnCu_3(OH)_6SO_4}$8-dimensional limit, channel asymmetry is relevant because $\mathrm{ZnCu_3(OH)_6SO_4}$9, so the system flows toward channel symmetry breaking. In one dimension, the analysis yields $K = (\nu_1-\nu_2)/\nu_2$0, channel anisotropy becomes irrelevant, and the system supports a gapless fractionalized conformal fixed point with restored channel symmetry. The critical mode is explicitly fractionalized: the low-energy theory is built from bosonic spinons $K = (\nu_1-\nu_2)/\nu_2$1 and fermionic holons $K = (\nu_1-\nu_2)/\nu_2$2, not from electrons or magnons (Ge et al., 2022).

A more direct “spinon Kondo lattice” is obtained by coupling a U(1) spin liquid with a spinon Fermi surface to a lattice of Anderson impurities. In slave-rotor variables,

$K = (\nu_1-\nu_2)/\nu_2$3

and the interlayer coupling generates self-consistent hybridization fields $K = (\nu_1-\nu_2)/\nu_2$4 between QSL spinons $K = (\nu_1-\nu_2)/\nu_2$5 and impurity spinons $K = (\nu_1-\nu_2)/\nu_2$6, and $K = (\nu_1-\nu_2)/\nu_2$7 between chargons $K = (\nu_1-\nu_2)/\nu_2$8 and $K = (\nu_1-\nu_2)/\nu_2$9. The central result is a finite threshold: for $T_K=1.3(1)\,\mathrm{K}$0, $T_K=1.3(1)\,\mathrm{K}$1; for $T_K=1.3(1)\,\mathrm{K}$2, $T_K=1.3(1)\,\mathrm{K}$3, heavy quasiparticles form near the Fermi level, and the system enters a spinon Kondo lattice phase. At half filling this phase is described as a spinon Kondo insulator with two hybridized spinon bands separated by a small gap (Zheng et al., 2024).

Gauge fluctuations remain essential in the lattice problem. The same work argues that the emergent U(1) gauge field mediates an effective Coulomb attraction between spinons and chargons, producing spinon-chargon bound states that reshape the spectral density near the Hubbard-band edges. Thermodynamically, the uncoupled QSL has $T_K=1.3(1)\,\mathrm{K}$4 and $T_K=1.3(1)\,\mathrm{K}$5, whereas the spinon Kondo lattice shows strong low-$T_K=1.3(1)\,\mathrm{K}$6 suppression of specific heat consistent with a hybridization gap. Neutron scattering is likewise modified through changes in the spinon susceptibility and the RPA response (Zheng et al., 2024).

These lattice constructions broaden the subject in two directions. First, spinons can serve as the bath that screens or hybridizes local degrees of freedom, as in the impurity language. Second, Kondo dynamics can itself create coherent spinon motion and heavy spinon bands, so the spinon-Kondo effect becomes a mechanism for reorganizing the low-energy Hilbert space of an entire lattice (Ge et al., 2022, Zheng et al., 2024).

5. Gauge-, chirality-, and current-mediated variants

Some works extend the spinon-Kondo idea beyond direct screening of a magnetic impurity by a pre-existing spinon Fermi sea. In a frustrated spin liquid with Chern–Simons gauge structure, a non-magnetic bond defect can locally enhance gauge fluctuations and generate an effective pseudospin exchange,

$T_K=1.3(1)\,\mathrm{K}$7

where the impurity is a pseudospin-$T_K=1.3(1)\,\mathrm{K}$8 formed by local occupancy of two sublattice states rather than a real spin. The exchange is highly anisotropic, involving only transverse pseudospin components, and NRG identifies a stable asymmetric Kondo fixed point at $T_K=1.3(1)\,\mathrm{K}$9. Because the bath consists of neutral spinons and the coupling is gauge-mediated, the predicted observables are not the standard Kondo resistivity minimum but a low-temperature thermal conductivity with $B=0.3\,\mathrm{T}$0 and anisotropic pseudospin correlations (Wang et al., 2021).

Frustrated impurity clusters introduce another variant, the spin current Kondo mechanism. In the three-impurity Kondo model on an equilateral triangle, magnetic frustration produces a nonzero Schwinger-boson hopping amplitude $B=0.3\,\mathrm{T}$1, so that spin currents of conduction electrons couple to the vector chirality of the impurity cluster. The emergent nonlocal term has the form

$B=0.3\,\mathrm{T}$2

with $B=0.3\,\mathrm{T}$3. The intermediate phases are characterized by holon phase-shift plateaus

$B=0.3\,\mathrm{T}$4

where $B=0.3\,\mathrm{T}$5 and $B=0.3\,\mathrm{T}$6 describe partially screened non-Fermi-liquid states. In the authors’ terminology, the movable Kondo singlets are “holons,” and their motion couples directly to local vector chirality, providing a microscopic route from frustration to a broadened notion of spinon-Kondo physics (Wang et al., 2021).

These generalized mechanisms share a structural theme with direct spinon screening: the screening channel is carried by emergent or fractionalized spin-sector variables, and the effective Kondo interaction is shaped by gauge structure, chirality, or nonlocal current operators rather than by a simple onsite $B=0.3\,\mathrm{T}$7 exchange. At the same time, the low-energy outcome can differ substantially from the textbook one-channel impurity problem, yielding anisotropic fixed points, partial screening, or stable non-Fermi-liquid phases (Wang et al., 2021, Wang et al., 2021).

6. Diagnostics, misconceptions, and relation to adjacent Kondo phenomena

A persistent interpretive issue is that not every Kondo signature observed in a spin-liquid candidate constitutes evidence for spinon-Kondo screening. This is illustrated by single-layer $B=0.3\,\mathrm{T}$8-TaSe$B=0.3\,\mathrm{T}$9 contacted to metallic $T$0-TaSe$T$1. In the $T$2 heterostructure, STM/STS shows a pronounced zero-bias peak fitted by a thermally convolved Fano line shape, with $T$3 and an estimated $T$4. The purpose of that experiment is to establish that each star-of-David cluster in the $T$5 layer hosts a localized spin that can couple to an ordinary metallic bath; the paper is explicit that the Kondo resonance is not itself a direct image of spinons. The spinon evidence instead comes from long-wavelength Hubbard-band modulations interpreted as signatures of a spinon Fermi surface instability (Ruan et al., 2020).

The cobalt-on-$T$6-TaSe$T$7 study therefore makes a stronger claim than the $T$8 heterostructure, but also a more conditional one. Its interpretation requires three linked ingredients: a gapless QSL substrate, impurity spin screening by spinons, and gauge-field-induced spinon-chargon binding to make the effect visible in STM. The side peaks near the Hubbard edges are thus not a direct spinon spectrum, and not a conventional zero-bias Kondo resonance either; they are a recombined electronic signature of a spinon-sector many-body state (Chen et al., 2022).

A second misconception is to equate any spin-selective or spin-textured Kondo phenomenon with a spinon-Kondo effect. The “spin-selective Kondo insulator” is a ferromagnetic phase of the Kondo lattice in which majority-spin conduction electrons remain metallic while minority-spin electrons develop a Kondo gap satisfying the dynamically generated commensurability condition

$T$9

This is conceptually relevant because it shows that Kondo screening can be partial, emergent, and spin selective, but it does not introduce a fractionalized spinon bath in the sense used by the spinon-Kondo literature (Peters et al., 2012).

Likewise, a pure spin current can suppress an ordinary Kondo singlet by producing a spin accumulation $K\simeq 0.18\%$0 that makes spin flips energetically costly; the low-temperature plateau of the nonlocal spin signal in Cu(Fe) is interpreted as the Fermi-liquid Kondo regime below $K\simeq 0.18\%$1, and increasing spin accumulation increases the spin diffusion length by reducing Kondo spin-flip scattering (Hamaya et al., 2016). Electron Kondo physics in a persistent spin helix 2DEG and in Weyl-type systems with arbitrary spin-momentum locking similarly shows that bath spin structure can strongly tune $K\simeq 0.18\%$2, but the screening particles remain electrons, not spinons (Puel et al., 5 Apr 2025, Goto et al., 5 Jun 2025).

The most defensible operational diagnostics for a literal spinon-Kondo effect are therefore the following. First, the host must be charge insulating or otherwise fractionalized while retaining low-energy spinful excitations. Second, the impurity response must show screening or strong-coupling behavior inconsistent with free moments and not explainable by impurity clustering alone. Third, the measured spectral signature need not be a metallic zero-bias resonance; in spinon systems it may instead be a Hubbard-edge resonance, a subgap threshold singularity, a Lorentzian impurity spinon density of states, or a boundary-mode transition. Fourth, when gauge fields are active, the electron-channel signature can differ qualitatively from the underlying spinon-sector fixed point because STM or related probes observe spin-charge recombination rather than bare spinons (Gomilšek et al., 2019, Chen et al., 2022, Pereira et al., 26 Aug 2025, Kattel et al., 2023, Kattel et al., 2024).