{"id":"045c3487-be2f-42cc-b0fa-650fc893c9a2","arxiv_id":"2412.09310","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A mixed-spin Kitaev honeycomb model with alternating spin-1/2 and spin-3/2 ions hosts four quadrupolar-distinguished quantum spin liquid phases, confirmed by parton mean-field theory and DMRG, with Zr0.5Ru0.5Cl3 proposed as a candidate material.","lead":"This paper analyzes a honeycomb magnet made of two different magnetic ions, spin-1/2 and spin-3/2, and finds four distinct quantum spin liquid phases in its phase diagram. It also develops a microscopic model suggesting the material Zr0.5Ru0.5Cl3 could realize this exotic mixed-spin magnet.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Zero-flux sector assumed in the mean-field theory is not confirmed by the paper's own DMRG in the B phase; if the true ground state lies outside that sector, the mean-field phase diagram is incomplete in a large parameter region.","rationale":"The reader's weakest assumption is precisely the zero-flux-sector premise, and my independent reading confirms this is the most load-bearing issue. The central claim is that four quantum spin liquid phases are realized and that DMRG quantitatively confirms the mean-field phase diagram. That claim requires the parton mean-field calculation, which fixes the Z2 gauge fields to the zero-flux sector, to describe the true ground state. The paper's own numerical data show that in the B phase the DMRG ground state is not zero-flux, so the mean-field B phase is built on an unverified and internally contradicted sector choice. This is not a fatal error for the existence of nontrivial spin liquid physics, because DMRG still finds a spin liquid with the expected quadrupolar parameters, but it prevents the paper's strongest statement of quantitative confirmation from covering the B phase. The other concerns noted by the reader, such as the isotropic-point discrepancy and the fitted material-realization parameters, are secondary: the isotropic-point region is explicitly acknowledged, and the material section is presented as a proof of principle. The zero-flux issue, by contrast, affects a large region of the claimed phase diagram and is not acknowledged as a failure of the Ansatz. The recommended verdict remains CONDITIONAL, since the central four-phase structure is plausible and supported elsewhere, but the B phase requires either a flux-sector-resolved calculation or an explicit demonstration that the flux-disordered state is a finite-size artifact. This is the same conclusion as the reader, so no verdict adjustment is needed.","tokens_in":22277,"tokens_out":2814,"duration_ms":32803,"concrete_test":"At the representative B-phase point (K_z, D_z) = (2.0, 0.0), run DMRG on a cylinder with circumference L_x = 12 and length L_y = 16, measuring every plaquette expectation value <W_p>. Compare the unrestricted ground-state energy and flux pattern with a DMRG calculation initialized in the zero-flux sector (e.g., by using a flux penalty term lambda * sum_p (1 - W_p) with increasing lambda and extrapolating to lambda = 0). If the unrestricted ground state has bulk plaquette values deviating from +1 and a lower energy than the zero-flux-restricted state, then the zero-flux assumption fails in the B phase.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The parton mean-field treatment in Section II B is performed entirely in the zero-flux sector (W_p = +1 on every plaquette), justified by analogy with the spin-1/2 Kitaev model and higher-spin Kitaev models. But the paper's own DMRG results do not support this choice in the B phase. Table I reports that for (K_z, D_z) = (1.1, 0.0) and (1.05, 0.0), which lie inside the mean-field B phase, the DMRG ground state is a disordered-flux state rather than zero-flux. Section III explicitly states: 'In the B phase, however, our DMRG simulations do not converge to a unique ground-state flux configuration. Instead, they yield a disordered-flux state.' The text attributes this to a small flux-flip gap, but a small or vanishing flux gap means the zero-flux sector is not the isolated ground-state sector, so the mean-field B-phase Ansatz is not a controlled approximation in that region. The DMRG agreement in quadrupolar parameters Q_z and Q_x does not resolve this issue, because those order parameters could be insensitive to the flux-sector structure. Since the B phase occupies a substantial fraction of the phase diagram, the central claim that the mean-field phase diagram is 'quantitatively confirmed by DMRG' is undermined exactly in that phase. The correct conclusion is not necessarily that the B phase is wrong, but that its characterization as a zero-flux Kitaev spin liquid is an unverified assumption, and the phase diagram's validity in the B region is conditional on a flux-sector choice that the numerical data contradict.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a mixed-spin Kitaev honeycomb model with S=1/2 and J=3/2 ions on alternating sublattices, H = sum K_gamma S_i^gamma J_j^gamma + D_z sum (J_j^z)^2. It rewrites the model in terms of SU(4) pseudospin/pseudo-orbital operators, identifies conserved Z2 plaquette fluxes, and uses an SO(6) Majorana parton construction to perform mean-field theory in the zero-flux sector. This yields four quantum spin liquid phases, labeled A0, Az, B, and C, distinguished by the quadrupolar order parameters Qz and Qx, with a full phase diagram shown in Fig. 2. The paper compares these mean-field results with DMRG calculations on cylinders (bond dimension up to 4000, truncation errors around 1e-6) and with exact diagonalization at the isotropic point. Good agreement is found in the A0 and Az phases and for most of the parameter plane, but the DMRG/ED results differ substantially near the isotropic point and the DMRG flux configuration in the B phase is not zero-flux. The paper also derives a superexchange Hamiltonian for Zr0.5Ru0.5Cl3 and identifies parameter regimes in which the Kitaev coupling dominates.","tokens_in":22495,"tokens_out":6145,"duration_ms":68337,"significance":"If the central phase diagram is correct, this work establishes ferrimagnetic mixed-spin honeycomb magnets as a new family of Kitaev spin liquids with coexisting quadrupolar order, which is a genuinely new contribution beyond uniform spin-1/2 and spin-3/2 Kitaev models. The paper is technically strong in several respects: the self-consistent mean-field procedure is documented in detail (200 initial guesses, tolerance 1e-12), the DMRG calculations have small truncation errors, and the comparison between mean-field and numerical order parameters is presented explicitly in Table I and Fig. 4. The authors also disclose the two main limitations of their approach: the isotropic-point discrepancy and the disordered-flux states in the B phase. These disclosures are important, because the central claim of quantitative confirmation is weaker exactly in the regions where the load-bearing zero-flux and parton Ansatz assumptions are not independently supported.","major_comments":[{"comment":"The parton mean-field theory is performed entirely in the zero-flux sector (u = +1 on every bond, W_p = +1 on every plaquette), justified by analogy with the spin-1/2 Kitaev model and by the statement that DMRG supports this choice. However, Table I lists (Kz, Dz) = (1.1, 0.0), (1.05, 0.0), (4.6, 1.0), and (5.0, 1.0) as having zero-flux configuration 'No', and Section III states that in the B phase DMRG yields a disordered-flux state rather than a unique flux configuration. Since the B phase occupies a substantial part of the phase diagram, the mean-field description of B is not controlled unless the flux gap is shown to be small and the zero-flux sector is shown to dominate the variational energy over other flux sectors. The sentence in the abstract and conclusion that the analytical results are 'quantitatively confirmed by DMRG' is therefore too strong for the B phase. The authors should either compare mean-field solutions in different flux sectors, provide a quantitative estimate of the flux gap, or explicitly restrict the claimed confirmation to sectors and parameter regions where the zero-flux assumption is verified.","section":"Section II B, Eq. (19), and Table I"},{"comment":"Near the isotropic point, the mean-field and numerical results disagree in a way that is qualitatively important. At (Kz, Dz) = (1, 0), the mean-field solution is threefold degenerate with Qz values proportional to +1 and -1/2, whereas exact diagonalization in the zero-flux sector finds the opposite sign pattern with much smaller magnitudes, and DMRG shows a sizable divergence from the mean-field curves for Dz below about 0.1. The text acknowledges that the nature of this DMRG/ED spin liquid is 'not yet tractable within our parton Ansatz.' Because this region contains the B-C boundary and the isotropic point that connects the A0 and Az phases, the phase diagram is incomplete in a region that is physically central rather than peripheral. This is a disclosed limitation, but the abstract and conclusions need to state more prominently that the four-phase diagram is established only outside a finite neighborhood of the isotropic point, and the boundary structure inside that neighborhood remains unresolved.","section":"Section III, Fig. 4(a), and Eqs. (26)-(27)"},{"comment":"The superexchange derivation for Zr0.5Ru0.5Cl3 is presented as a proof of principle, but the link between the microscopic model and the pure Kitaev plus SIA Hamiltonian of Eq. (1) is not quantitative. The authors fix the onsite parameters U2, JH, and lambda to the same values for Zr and Ru, introduce a large onsite potential V on Zr, and then tune the hopping parameters to t1 = -t3 and t4 = 0 to maximize the Kitaev coupling. Even in the tuned regime, the couplings include a significant Gamma term and several comparable higher-order multipolar terms (Fig. 6). The paper should clarify that these conditions are a proof of existence for dominant Kitaev interactions in the mixed-spin setting, not a concrete prediction for Zr0.5Ru0.5Cl3, and should state what would be needed to elevate the material claim to a realistic model.","section":"Section IV, Eq. (30), and Fig. 6"}],"minor_comments":[{"comment":"The caption of Fig. 4(b) is incomplete; it ends with '... over values such' and the sentence is cut off. The caption should state the full range of Kz used and the phase labels for each segment.","section":"Fig. 4 caption"},{"comment":"The notation r_gamma for the spin-3/2 nearest neighbor of site r is used without a precise definition before Eq. (22). The text should define r_gamma as the B-sublattice site reached from r along the gamma-bond.","section":"Section II B, after Eq. (22)"},{"comment":"The column 'zero-flux' in Table I would benefit from a definition of how the flux configuration is diagnosed in the cylindrical DMRG calculation, including how plaquette flux values are averaged and how the open boundaries are handled. This is important given that the B-phase flux disorder is a central qualitative result.","section":"Section III and Table I"},{"comment":"The approximation of using identical U2, JH, and lambda for Zr and Ru is stated in Section IV but the text later notes that these parameters should differ for the two ions. This should be flagged as a simplifying assumption from the outset, rather than only in the concluding paragraph of the section.","section":"Section IV"},{"comment":"The phrase 'quantitatively confirmed by DMRG' appears in the abstract and in the conclusions without the qualification that the confirmation excludes a finite region around the isotropic point and that the B phase has disordered flux in DMRG. The abstract should carry the same caveat as Section III.","section":"Abstract and Conclusions"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a solid, well-executed paper that opens a genuinely new corner of Kitaev physics. The mixed-spin S=1/2/S=3/2 honeycomb model is new; the four QSL phases distinguished by quadrupolar order parameters are new; and the DMRG check over most of the (Kz, Dz) plane is convincing. I believe the four-phase structure is real, and the paper deserves a serious referee.\n\nWhat it does well: the parton mean-field treatment is clearly laid out, the self-consistency is thorough, and the DMRG data with truncation error around 1e-6 give independent support for the A0, Az, B, and C phases across most of the plane. The isotropic-point ED result showing threefold degeneracy matches the mean-field relative values, which is a nice check. The superexchange section is a genuine derivation, though long and not fully checkable without a notebook.\n\nThe main soft spots are the ones the reader flagged. First, the abstract says 'quantitatively confirmed' but the body reports a qualitatively different spin liquid near the isotropic point, with the sign of Qz inverted, which the parton Ansatz cannot represent. That is an overstatement and should be fixed. Second, the zero-flux sector assumption: in the B phase DMRG sees a disordered-flux ground state, not zero flux. The paper attributes this to a small flux gap. That means the mean-field Ansatz in the B phase is not controlled in that region, even if the quadrupolar parameters agree. The stress-test is right to call this out, but I do not think it kills the phase diagram: the agreement in Qz/Qx suggests the dominant correlations are still captured. Still, the authors should either show the flux gap is small but nonzero in the thermodynamic limit, or soften the claim that the B phase is a zero-flux KSL. Third, the material-realization section tunes hopping parameters and assumes a Zr on-site V of order 7-10 eV; that is a proof-of-principle, not a prediction, and the text says so. That is a minor concern.\n\nWho is it for: researchers working on Kitaev spin liquids, higher-spin generalizations, and numerical studies of frustrated magnets. It is a meaningful advance within the subfield.\n\nMy recommendation: send to peer review. The central claims are supported by two independent methods over most of the parameter space, the paper is transparent about its limitations, and the mixed-spin direction will attract interest. I would ask the referee to insist on fixing the abstract and to dig into the flux-sector issue in the B phase, but neither is a reason to desk-reject.","headline":"Solid mixed-spin Kitaev paper with a new four-phase diagram; the isotropic-point discrepancy and B-phase flux-sector assumption need scrutiny, but the core result is real and deserves peer review.","tokens_in":23221,"tokens_out":3029,"would_cite":true,"duration_ms":26616,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A honeycomb magnet with alternating spin-1/2 and spin-3/2 ions and Kitaev exchange interactions can host four distinct quantum spin liquid phases, distinguished by quadrupolar order parameters, the paper claims.","keywords":["Kitaev honeycomb model","quantum spin liquid","mixed-spin system","quadrupolar order","Majorana partons","single-ion anisotropy","ferrimagnetism","Zr0.5Ru0.5Cl3"],"falsifier":"A direct all-flux-sector calculation of the model for $K_z$ near 1 and small $D_z$--for instance, exact diagonalization on a 24-site torus including sectors with $W_p = -1$ and flux-disordered superpositions--would settle the zero-flux assumption. If the ground state there is not the $W_p = +1$ sector, the mean-field phases A0, Az, B, and C do not exhaust the phase diagram; a second check is to study the inverted $Q_z$ value on larger DMRG cylinders in the disputed region and show that it survives with increasing bond dimension.","tokens_in":21938,"feed_emoji":"🧲","tokens_out":12219,"duration_ms":97617,"temperature":0.7,"pith_summary":"This paper argues that a honeycomb lattice in which spin-1/2 ions occupy one sublattice and spin-3/2 ions the other, coupled by bond-direction-dependent Kitaev exchange plus a single-ion anisotropy $D_z$, realizes four distinct quantum spin liquid phases. These phases are distinguished by the quadrupolar order parameters $Q_z$ and $Q_x$, and the paper maps their stability in the $K_z$--$D_z$ plane using a parton mean-field theory in which conserved plaquette fluxes become static $Z_2$ gauge fields. DMRG simulations confirm the phase diagram quantitatively away from the isotropic point, while at $K_z = 1$, $D_z = 0$ the numerics find a different spin liquid with an inverted $Q_z$ that the analytic Ansatz cannot represent. The paper also derives a microscopic superexchange Hamiltonian for the candidate material Zr$_{0.5}$Ru$_{0.5}$Cl$_3$ and identifies conditions under which Kitaev-like exchange dominates, suggesting that ferrimagnetic mixed-spin compounds are a new family of Kitaev spin-liquid platforms.","feed_headline":"Mixed-spin honeycomb lattice hosts four quantum spin liquids","feed_subtitle":"Alternating spin-1/2 and spin-3/2 ions with Kitaev exchange may realize spin liquids with distinct quadrupolar order.","key_machinery":"The load-bearing construction is an exact mapping of the model's conserved plaquette operators onto static $Z_2$ gauge fields. Spin-1/2 sites are written in the Kitaev Majorana representation $S^\\gamma = -\\tfrac{i}{2}\\eta^\\gamma c$, while spin-3/2 sites use an SO(6) Majorana representation in which the pseudospin operators $\\sigma^\\gamma$, the pseudo-orbital quadrupole/octupole operators $T^\\alpha$, and their products become Majorana bilinears; in this representation the constraint $D_i = 1$ restricts to the physical Hilbert space. Working in the zero-flux sector ($W_p = +1$ everywhere), the paper decouples the quartic term $ic_i\\theta^0_j$ into pairing parameters $\\Delta^t$ and onsite quadrupolar order parameters $Q^t = \\langle T^t\\rangle$, then solves the resulting quadratic Majorana Hamiltonian self-consistently from about two hundred initial conditions to produce the phase diagram. The single-ion anisotropy $D_z\\sum (J^z_j)^2$ acts as the control knob: as $D_z\\to\\infty$, pseudo-orbital fluctuations freeze out and the model maps onto the spin-1/2 Kitaev model with modified couplings, giving an exactly solvable limit against which the mean-field bands are checked.","core_discovery":"The central claim is that the mixed-spin Kitaev honeycomb model $H = \\sum_{\\langle ij\\rangle\\gamma} K_\\gamma S^\\gamma_i J^\\gamma_j + D_z \\sum_j (J^z_j)^2$ hosts four quantum spin liquid phases--labeled A0, Az, B, and C--distinguished by the quadrupolar order parameters $Q_z = \\langle T_z\\rangle$ and $Q_x = \\langle T_x\\rangle$. In the large-$D_z$ limit the A0 and Az phases are adiabatically connected to the gapless and gapped spin-1/2 Kitaev spin liquids; the B phase is a gapped toric-code-like liquid with $Q_z > 0$, $Q_x = 0$, and the C phase is a twofold-degenerate gapped liquid with $Q_z < 0$ and two possible signs of $Q_x$. The paper shows that the plaquette fluxes $W_p$ remain exact conserved quantities in the mixed-spin model, that the single-ion anisotropy commutes with them, and that the mean-field ground states in the zero-flux sector reproduce this structure. Exact diagonalization at the isotropic point gives a threefold-degenerate ground state whose quadrupolar values, $(Q_z,Q_x) = (-0.1214, 0)$ and $(\\pm 0.0607, \\pm 0.1051)$, match the mean-field relative pattern but not its magnitude, which the paper interprets as a qualitatively different spin liquid that the parton Ansatz does not describe.","pith_inferences":["A systematic scan of all flux sectors would likely locate the parameter boundary at which the zero-flux choice fails, and may reveal that the isotropic spin liquid is a flux-fluctuating state rather than a fixed-gauge state.","Extending the same parton construction to the Heisenberg, $\\Gamma$, and Dzyaloshinskii-Moriya terms of the derived superexchange Hamiltonian could stabilize chiral spin liquids or magnetically ordered phases that the pure-Kitaev model excludes.","Since the superexchange calculation uses identical $U_2$, $J_H$, and $\\lambda$ for both ions, a first-principles determination of the onsite parameters and the Zr potential $V$ is the crucial next step in testing whether the Kitaev-dominant regime survives in the real material."],"forward_implications":["If the phase diagram is right, the quadrupolar parameters $Q_z$ and $Q_x$ serve as measurable bulk signatures that tell the four spin liquids apart in candidate crystals.","In the large-$D_z$ limit the model reduces to the spin-1/2 Kitaev model with modified couplings, so the A0 and Az phases inherit the exact Majorana spectrum of the solvable limit.","The superexchange derivation indicates that hopping parameters with $t_1 = -t_3$ and small $t_4$ can make the Kitaev coupling the dominant exchange in Zr$_{0.5}$Ru$_{0.5}$Cl$_3$, identifying a concrete ferrimagnetic compound to search for these states.","The threefold-degenerate ground state at the isotropic point, with its quadrupolar values matching the mean-field relative pattern, implies a distinct spin liquid that the parton Ansatz does not capture and that is not adiabatically connected to A0 or Az."],"supporting_citations":[{"why":"Provides the exactly solvable spin-1/2 Kitaev honeycomb model whose Majorana representation and conserved plaquette fluxes the mixed-spin construction extends.","marker":"[2]"},{"why":"Introduces the conserved-flux pseudospin representation for arbitrary spin, the step that carries the flux conservation to the spin-3/2 sublattice.","marker":"[10]"},{"why":"Shows that large-spin Kitaev models host $Z_2$ spin liquids, giving the numerical evidence the paper relies on for the zero-flux sector.","marker":"[11]"},{"why":"Establishes the SO(6) Majorana parton mean-field theory and its DMRG validation for the spin-3/2 Kitaev honeycomb model, the method adapted here.","marker":"[12]"},{"why":"Maps the spin-3/2 Kitaev model with single-ion anisotropy into the quadrupolar-ordered spin liquids A0, Az, and B against which the mixed-spin B and C phases are identified.","marker":"[13]"},{"why":"States the theorem that the zero-flux sector is the ground state of the spin-1/2 Kitaev model, which the paper invokes for the analogous mixed-spin assumption.","marker":"[44]"},{"why":"Supplies the superexchange mechanism for bond-anisotropic Kitaev interactions in edge-sharing octahedra that the microscopic derivation for Zr0.5Ru0.5Cl3 is built on.","marker":"[19]"},{"why":"Proposes $\\alpha$-ZrCl$_3$ as a $j=3/2$ spin-orbital liquid candidate, the material context that motivates the mixed-spin compound and its effective Hamiltonian.","marker":"[30, 31]"}],"fun_headline_variants":["Four Kitaev spin liquids from ferrimagnetic honeycomb magnets","A spin liquid quartet in a mixed-spin honeycomb","Mixed spin-1/2 and 3/2 honeycomb yields four spin liquids","Ferrimagnetic honeycomb: a quartet of spin liquids"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assumption that the zero-flux sector, with $W_p = +1$ on every plaquette, contains the true ground state of the mixed-spin model, a fact imported by analogy from the spin-1/2 Kitaev model and from numerical evidence in uniform higher-spin models rather than proved for this particular system.","fun_headline_variants_meta":{"raw":{"variants":["Four Kitaev spin liquids from ferrimagnetic honeycomb magnets","A spin liquid quartet in a mixed-spin honeycomb","Mixed spin-1/2 and 3/2 honeycomb yields four spin liquids","Ferrimagnetic honeycomb: a quartet of spin liquids"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001627,"raw_usage":{"total_tokens":6516,"prompt_tokens":1033,"completion_tokens":5483,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":649,"completion_tokens_details":{"reasoning_tokens":5409}},"tokens_in":649,"tokens_out":5483,"duration_ms":36643,"temperature":1.0,"reasoning_tokens":5409,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T17:06:26.652437+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct all-flux-sector calculation of the model for $K_z$ near 1 and small $D_z$--for instance, exact diagonalization on a 24-site torus including sectors with $W_p = -1$ and flux-disordered superpositions--would settle the zero-flux assumption. If the ground state there is not the $W_p = +1$ sector, the mean-field phases A0, Az, B, and C do not exhaust the phase diagram; a second check is to study the inverted $Q_z$ value on larger DMRG cylinders in the disputed region and show that it survives with increasing bond dimension.","supporting_citations":[{"cited_title":"BY1czi8zWl2WX28+FymVk5U1tgQ=","cited_arxiv_id":null,"evidence_quote":"Provides the exactly solvable spin-1/2 Kitaev honeycomb model whose Majorana representation and conserved plaquette fluxes the mixed-spin construction extends."},{"cited_title":"Broholm, R","cited_arxiv_id":null,"evidence_quote":"Establishes the SO(6) Majorana parton mean-field theory and its DMRG validation for the spin-3/2 Kitaev honeycomb model, the method adapted here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the superexchange mechanism for bond-anisotropic Kitaev interactions in edge-sharing octahedra that the microscopic derivation for Zr0.5Ru0.5Cl3 is built on."}],"review_version":1}