{"id":"6964380d-40ac-451b-bac5-c6e66cb95ed3","arxiv_id":"2608.05556","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Local spin excitations generated by the spin-flip and Ising parts of Hund's coupling are identified as the dynamical link that destroys coherent quasiparticles in the orbital-selective Mott phase.","lead":"Using computer simulations of a two-band Hubbard model, the authors find that local spin excitations appear exactly at the Fermi energy inside the orbital-selective Mott phase and tie the localized and itinerant electrons together. This mechanism is proposed as the cause of the quasiparticle breakdown seen in materials such as ruthenates, iron selenides, and V2O3.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The causal mechanism relies on a sharp LSE resonance at the Fermi level that is computed with only n_b=3 bath sites and unspecified broadening; without a convergence check, a finite-bath artifact would invalidate the central claim.","rationale":"The reader's weakest assumption and my primary concern coincide: the central mechanism depends on the LSE resonance at omega=0 being a physical property of the OSMP, not a finite-bath or broadening artifact. I agree with the conditional verdict because the claim is plausible and the component comparison is a valuable step, but the absence of a convergence check leaves the key observable unverified. I considered alternative objections: the HF-derived coupling term vanishes in a paramagnet, weakening the analytical support, and the Fig. 3 comparison is not controlled. These reinforce the need for a numerical convergence check but are secondary. If the proposed test shows the LSE peak persists for n_b>=5 and is stable against broadening, the verdict should move toward accept; if it does not, the causal mechanism is unsupported. Therefore the verdict remains CONDITIONAL.","tokens_in":12676,"tokens_out":6252,"duration_ms":56620,"concrete_test":"Repeat the DMFT-Lanczos calculation at U=3.50, J=U/4, R=0.5 with n_b=4, 5, and 6 bath sites per orbital, keeping beta=512 and using a fixed small broadening eta (e.g., 0.01 t1) to plot A_LSE(omega) and the DOS. Track the lowest-energy peak position and its spectral weight as a function of n_b. Also recompute A_LSE(omega) for a given n_b with eta=0.005, 0.01, and 0.02 to verify the zero-frequency weight is not an artifact of the broadening. If the peak position moves away from omega=0 or its weight decreases monotonically with n_b, the sharp Fermi-level resonance is a finite-bath artifact and the central mechanism is not supported. As a further check, compare with a complementary impurity solver such as NRG or CT-HYB at the same parameters to confirm the persistence of the resonance in the continuum limit.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The causal claim is pinned to the sharp LSE resonance at omega=0 inside the OSMP (Fig. 1b,c; Eq. 5). The sole numerical evidence for this resonance is Lanczos ED with n_b=3 bath sites per orbital and beta=512 (Models and methods), with no convergence check in n_b and no stated broadening eta for the plotted spectra. ED produces a discrete pole spectrum; with three bath sites per orbital the low-energy region is severely under-resolved, and a peak at omega=0 can be a single finite-bath pole rather than a true continuum resonance. If that peak shifts away from omega=0 or its weight collapses as n_b grows, the proposed LSE-mediated quasiparticle breakdown loses its only dynamical support. The analytical equations of motion in the Supplemental Material (Eqs. 8-9) do not rescue this: at the Hartree-Fock level the interorbital coupling appears through <d^dag_{2↓}d_{2↑}>, which vanishes by spin-rotation symmetry in the paramagnetic state, so the dynamical link is not established analytically. The component switch in Fig. 3 also changes Uc by nearly a factor of two, so the disappearance of LSEs is not a controlled causal test. The robustness of the zero-frequency LSE weight is therefore the load-bearing assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies a half-filled two-band Hubbard model on the Bethe lattice using DMFT with a Lanczos exact-diagonalization impurity solver (n_b = 3 bath sites per orbital, beta = 512). It introduces a local spin excitation (LSE) spectral function A_LSE(ω) and reports a sharp LSE resonance at the Fermi level inside the orbital-selective Mott phase (OSMP), together with non-half-integer values of the authors' local two-qubit fidelity (LTQF). The authors argue that LSEs generated by the spin-flip (J_sf) and Ising (J_z) components of Hund's coupling dynamically couple the wide and narrow bands, causing quasiparticle breakdown; removing J_sf and J_z restores a decoupled OSMP with quantized LTQF. The central claim is that the OSMP is not a coexistence of decoupled bands but a state whose charge dynamics are mediated by local spin excitations.","tokens_in":13005,"tokens_out":4582,"duration_ms":40557,"significance":"If the central numerical observation is robust, the paper offers a concrete microscopic mechanism for quasiparticle breakdown in the OSMP and a quantitative diagnostic (the LTQF anomaly) tied to that mechanism. The phase-selective evolution of the LSE weight across the metal-OSMP-insulator transitions is a clean falsifiable prediction, and the component-wise comparison of Hund's coupling terms is a useful organizing principle. However, the current evidence is not fully convincing: the sharp Fermi-level LSE resonance rests on a very small impurity bath with no convergence check, and the analytical support in the supplement relies on a Hartree-Fock truncation that, as written, yields a vanishing interorbital coupling in the paramagnetic state. The paper is therefore potentially significant but needs substantial technical strengthening before the causal claim can be accepted.","major_comments":[{"comment":"The central LSE resonance at ω = 0 is computed with the Lanczos ED impurity solver using only n_b = 3 bath sites per orbital and beta = 512, with no convergence check in n_b and no reported broadening parameter for the plotted spectra. ED produces a discrete pole spectrum, and with three bath sites per orbital the low-energy region is severely under-resolved; a peak at the Fermi level can be a single finite-bath pole rather than a true continuum resonance. Because the entire causal claim is pinned to this resonance, the authors must show that its position and weight are stable as n_b is increased (e.g., n_b = 4, 5, 6) and must specify the broadening η used in ρ_α(ω) and δ in Eq. (5).","section":"Models and methods; Fig. 1b,c; Eq. (5)"},{"comment":"The analytical support for interorbital coupling states that G_1(ω) depends on J_sf ⟨d†_{2↓}d_{2↑}⟩, but in the paramagnetic, spin-rotationally invariant state treated throughout the paper this expectation value vanishes by symmetry. At the Hartree-Fock level, the advertised dynamical link is therefore exactly zero, so the equation-of-motion derivation does not establish a nonzero LSE-mediated coupling. The supplement acknowledges the Hartree-Fock truncation but does not resolve this contradiction; the authors should either provide a symmetry-preserving higher-order closure or explicitly state that the HF expression is only a schematic and cannot be used as quantitative support.","section":"Supplemental Material, Eqs. (8)-(9)"},{"comment":"The causal test in Fig. 3 compares J_sf = J_z = U/4, J_ph = 0 with J_sf = J_z = 0, J_ph = U/4, but these two cases have very different critical couplings (U_c1 = 3.74 vs. 7.21) and the second case violates spin-rotational invariance because U = U′ + 2J with only J_ph retained does not satisfy the Kanamori symmetry conditions. The disappearance of LSEs and restoration of half-integer LTQF values may therefore reflect a different interaction point or a broken-symmetry artifact rather than the causal role of the LSEs. A controlled test should vary J_sf and J_z at fixed U and fixed total J, or at least compare both cases at the same ratio U/U_c.","section":"Fig. 3 and Models and methods"},{"comment":"The LSE operator b†_α = d†_{α↓}d_{α↑} is precisely the spin-flip operator that appears in the J_sf term, and the LTQF is the authors' own diagnostic introduced in their previous work. Consequently, showing that J_sf generates spectral weight in the J_sf spectral function is partly definitional, and the correlation between LSE weight at ω = 0 and the LTQF anomaly does not by itself establish a causal mechanism. The claim that LSEs mediate quasiparticle breakdown needs an independent test, for example by showing that the LSE propagator enters the charge self-energy through a two-particle vertex in a controlled approximation, or by tuning the LSE energy away from the Fermi level at fixed J_sf and observing the predicted restoration of coherence.","section":"Eq. (5); Figs. 1-3; Refs. [41,42]"}],"minor_comments":[{"comment":"The text states 'We perform calculations at zero temperature, where β is treated as a large but finite parameter'; this is contradictory, and the authors should state that they work at low but finite temperature and comment on the convergence of β = 512.","section":"Models and methods"},{"comment":"The broadening δ in Eq. (5) and η in the DOS definition are never assigned numerical values, so the plotted spectra cannot be reproduced or compared quantitatively; please specify both and describe any smoothing procedure.","section":"Eq. (5) and DOS definition"},{"comment":"The citation to the supplemental material is vague; please cite the specific equations (e.g., Eqs. (S8)-(S9)) where the Green's function equations of motion are derived.","section":"Reference [57]"},{"comment":"The caption states 'The LSE peaks at the Fermi level in the NB Mott gap' but this applies to panel (c) only; the WB LSE resonance in panel (b) lies at the Fermi level but not inside a Mott gap, so the wording should be clarified.","section":"Fig. 1 caption"},{"comment":"The pair-hopping term J_ph is written with spin indices that may not match the standard Kanamori form; please verify the operator ordering and the summation convention.","section":"Eq. (2)"},{"comment":"There are several grammatical slips, e.g., 'the spin-flip and Ising-type components constitute' (subject-verb agreement) and 'signaling the emergence of the interorbital coupling'; a careful proofread is needed.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper leans heavily on the authors' own LTQF construction from Refs. [41,42] without an independent benchmark, and the supplement itself acknowledges that the Hartree-Fock truncation is uncontrolled. The most urgent fix is numerical: the sharp Fermi-level LSE resonance must be shown to survive with larger n_b and specified broadening. If the authors cannot provide that convergence evidence, I would not be comfortable recommending acceptance even after revision. The component-switch comparison also needs to be redesigned to preserve spin-rotational invariance and match interaction strengths."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nRead the paper. My take: the proposed mechanism is plausible and the paper is a reasonable step, but the central numerical evidence is thinner than the conclusions claim. The sharp LSE resonance at the Fermi level inside the OSMP comes from Lanczos ED with n_b = 3 bath sites per orbital and beta = 512, with no convergence check in n_b and no stated broadening. That is a single-pole finite-bath spectrum; it could be an artifact. This is the load-bearing result, so it needs to be checked with larger baths or an independent solver before I'd trust the causal claim.\n\nWhat's actually new: the LSE spectral function (the composite spin-flip propagator) and the observation that its weight condenses at omega = 0 selectively in the OSMP, correlated with the authors' LTQF anomaly. That's a concrete diagnostic and it does offer a unified way to talk about ruthenate, iron-selenide, and V2O3 ARPES. The paper also cites the earlier spin-fluctuation work (Refs 27, 34, 60) that already proposed interorbital spin fluctuations as a source of non-Fermi-liquid behavior; the new piece is the quantitative LSE spectrum and the link to their entanglement metric. The self-citations are appropriate here because the LTQF is their own construction.\n\nSoft spots, in order of importance. First, the n_b=3 issue. Second, the analytical equations of motion in the supplement: the Hartree-Fock decoupling makes the interorbital term proportional to <d^dag_{2down}d_{2up}>, which vanishes by spin-rotation symmetry in the paramagnetic state. So the supplement doesn't actually demonstrate the coupling analytically; it only restates the interaction structure. Third, the component comparison in Fig. 3 changes Uc by nearly a factor of two, so the disappearance of LSEs is not a controlled causal test. The qualitative contrast is suggestive, but the shift in critical couplings muddies the interpretation. Fourth, the introduction calls the mechanism \"unknown\" while citing papers that already contain it; that overstatement should be fixed.\n\nNone of this kills the paper. The mechanism is physically reasonable and the numerical trend across the phase diagram is consistent. But the central claim needs the n_b convergence check. I'd send it to peer review with a request for that check, not desk-reject it. The audience is the DMFT/correlated-electron community, and they'll get value from the LSE diagnostic even if the causal story is refined.\n\nRecommendation: serious referee, conditional on numerical convergence.","headline":"Plausible mechanism, but the central resonance rests on a three-bath-site ED calculation with no convergence check; worth refereeing if that is fixed.","tokens_in":13495,"tokens_out":3267,"would_cite":true,"duration_ms":28254,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.30.+h","71.27.+a","71.10.-t"],"model":"deepseek-v4-flash","headline":"This paper claims that local spin excitations generated by Hund's coupling destroy coherent quasiparticles in the orbital-selective Mott phase.","keywords":["orbital-selective Mott phase","Hund's coupling","local spin excitations","quasiparticle breakdown","dynamical mean-field theory","local two-qubit fidelity","two-band Hubbard model","interorbital coupling"],"falsifier":"Repeat the DMFT calculation at $U=3.5$, $J=U/4$, $R=0.5$ (deep inside the OSMP) with bath sizes $n_b=4$ and $n_b=5$ at the same $\\beta$, keeping the same broadening convention, and check whether the LSE spectral function still has a peak exactly at $\\omega=0$; if the zero-energy resonance moves or disappears, the central claim is not supported. A complementary check is to compute the same spectra with a different impurity solver (for example, continuous-time quantum Monte Carlo) and see whether the Fermi-level LSE peak survives.","tokens_in":12508,"feed_emoji":"🧲","tokens_out":8818,"duration_ms":63369,"temperature":0.7,"pith_summary":"Orbital-selective Mott phases are usually read as a simple coexistence: one band of electrons localized, the other itinerant, with the two effectively decoupled. The paper argues this picture is incomplete, and that the spin-flip and Ising-type pieces of Hund's coupling create local spin excitations—on-site electron-hole pairs across the two spin channels—that sit exactly at the Fermi energy inside the OSMP. These excitations dynamically couple the two orbitals, transfer spectral weight, and destroy well-defined quasiparticles. The quantitative fingerprint is a local two-qubit fidelity that takes non-half-integer values instead of the ideal 0 and 0.5. If the mechanism is right, the OSMP is an emergent coupled state rather than a decoupled band mixture, with direct consequences for understanding Hund metals and photoemission anomalies.","feed_headline":"Spin flips break quasiparticles in orbital-selective Mott phase","feed_subtitle":"Hund's spin-flip and Ising terms create zero-energy excitations that couple the two bands and erase electron coherence.","key_machinery":"The load-bearing object is the local spin-excitation spectral function $A^{\\mathrm{LSE}}_\\alpha(\\omega) = -\\frac{1}{\\pi}\\, \\mathrm{Im}\\langle b_\\alpha^\\dagger (\\omega - H + i\\delta)^{-1} b_\\alpha\\rangle$, built from the composite spin-flip operator $b_\\alpha^\\dagger = d^\\dagger_{\\alpha\\downarrow} d_{\\alpha\\uparrow}$, which creates a local electron-hole pair in orbital $\\alpha$ across the spin channel. Equation-of-motion analysis of the impurity Green's functions shows that $G_1(\\omega)$ and $G_2(\\omega)$ depend on the other orbital's spin-flip correlation $\\langle d^\\dagger_{\\alpha\\downarrow}d_{\\alpha\\uparrow}\\rangle$ through $J_{sf}$ and on its occupations through $J_z$; this is the formal route by which LSEs couple the charge dynamics of the two bands. The numerical vehicle is DMFT with a Lanczos exact-diagonalization impurity solver on a Bethe lattice, with the local two-qubit fidelity (LTQF) as the quantitative order parameter distinguishing decoupled ($0$ or $0.5$) from coupled (non-half-integer) regimes.","core_discovery":"The central claim is that local spin excitations generated by the spin-flip ($J_{sf}$) and Ising-type ($J_z$) terms of Hund's coupling are the microscopic cause of quasiparticle breakdown in the orbital-selective Mott phase. Within DMFT for the half-filled two-band Hubbard model on a Bethe lattice, the LSE spectral function $A^{\\mathrm{LSE}}_\\alpha(\\omega)$ shows a sharp resonance precisely at $\\omega=0$ only inside the OSMP: the wide band's resonance sits inside the narrow band's Mott gap, and the narrow band's resonance coexists with the wide band's coherent quasiparticle peak. The LSEs act as dynamical intermediaries that renormalize quasiparticle lifetimes and binding energies, giving the nominally insulating narrow band partial itinerant weight and depleting the wide band's coherence. Turning off $J_{sf}$ and $J_z$ (keeping only pair hopping) eliminates the LSE spectra entirely, restores exact $L_o=0$ and $L_o=0.5$ local two-qubit fidelity values, and fully decouples the two bands. The paper concludes that the OSMP is therefore not a coexistence of decoupled subbands but an emergent many-body state, with non-half-integer LTQF values as the quantitative signature of LSE-mediated breakdown.","pith_inferences":["If correct, the same mechanism predicts that any probe sensitive to local spin-flip excitations (resonant inelastic x-ray scattering, neutron scattering) should see a low-energy spin-excitation mode inside the Mott gap of the narrow band in OSMP materials, not just in the model.","A testable extension: the LSE-induced quasiparticle scattering rate in the wide band should track the weight of the narrow band's LSE peak at $\\omega=0$ across the whole OSMP window, giving a quantitative relation between the two bands' spectra.","The distinction the paper draws between Ising-only and full Hund's coupling suggests that materials with more isotropic Hund's exchange should show stronger interorbital coupling than those with Ising anisotropy, a comparison that could be checked by alloying or pressure studies."],"forward_implications":["The OSMP should be viewed as a dynamically coupled state, not two effectively independent subband systems.","Switching off the spin-flip and Ising-type components of Hund's coupling removes the LSEs and restores the decoupled OSMP, shifting the Mott critical points to larger $U$.","Non-half-integer LTQF values are a direct quantitative signature of quasiparticle breakdown, not a generic property of the OSMP.","The same LSE channel can account for suppressed quasiparticle coherence and Fermi-surface reconstruction reported in Hund metals and in ARPES on $V_2O_3$ and $FeTe_{1-x}Se_x$.","LSE spectral weight concentrates at the Fermi level only inside the OSMP, so the breakdown is phase-selective and tied to the Mott transition."],"supporting_citations":[{"why":"Supplies the dynamical mean-field framework in which the two-band lattice model is mapped to an impurity problem.","marker":"[31]"},{"why":"Shows the impurity mapping becomes exact in infinite dimensions, justifying the DMFT self-consistency used here.","marker":"[54]"},{"why":"Provides the exact-diagonalization impurity-solver approach and self-consistency relation used for the numerics.","marker":"[55]"},{"why":"Prior demonstration that interorbital spin fluctuations can induce non-Fermi-liquid behavior in the paramagnetic OSMP, the baseline this paper makes concrete.","marker":"[27]"},{"why":"Introduced the local two-qubit fidelity and its quantized values 0 and 0.5 for decoupled metallic and Mott-insulating bands.","marker":"[41]"},{"why":"Found the non-half-integer LTQF values and narrow-band quantum entanglement in the OSMP that the present paper sets out to explain.","marker":"[42]"},{"why":"Documents an in-gap band from interorbital Coulomb interactions, supporting the idea of low-energy interorbital excitations.","marker":"[47]"},{"why":"Distinguishes Ising-type from full Hund's coupling in OSMPs, the classification this work refines.","marker":"[60]"}],"fun_headline_variants":["Hund's spin terms drive quasiparticle breakdown in orbital-selective Mott phase","Hund's spin coupling breaks quasiparticles in orbital-selective Mott phase","Local spin excitations erase quasiparticle coherence in orbital-selective Mott phase","Orbital-selective Mott phase: spin excitations cause quasiparticle breakdown","Quasiparticles die in orbital-selective Mott phase due to spin excitations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that a Lanczos impurity solver with only three bath sites per orbital and an inverse temperature of 512 resolves a real, sharp local-spin-excitation resonance exactly at the Fermi level inside the OSMP, with no convergence check in bath size and no stated broadening; if that resonance is a finite-bath artifact, the claimed coincidence between LSE weight and the LTQF anomaly carries no causal weight.","fun_headline_variants_meta":{"raw":{"variants":["Hund's spin terms drive quasiparticle breakdown in orbital-selective Mott phase","Hund's spin coupling breaks quasiparticles in orbital-selective Mott phase","Local spin excitations erase quasiparticle coherence in orbital-selective Mott phase","Orbital-selective Mott phase: spin excitations cause quasiparticle breakdown","Quasiparticles die in orbital-selective Mott phase due to spin excitations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001295,"raw_usage":{"total_tokens":5300,"prompt_tokens":972,"completion_tokens":4328,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":588,"completion_tokens_details":{"reasoning_tokens":4224}},"tokens_in":588,"tokens_out":4328,"duration_ms":46559,"temperature":1.0,"reasoning_tokens":4224,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T10:55:38.132470+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the DMFT calculation at $U=3.5$, $J=U/4$, $R=0.5$ (deep inside the OSMP) with bath sizes $n_b=4$ and $n_b=5$ at the same $\\beta$, keeping the same broadening convention, and check whether the LSE spectral function still has a peak exactly at $\\omega=0$; if the zero-energy resonance moves or disappears, the central claim is not supported. A complementary check is to compute the same spectra with a different impurity solver (for example, continuous-time quantum Monte Carlo) and see whether the Fermi-level LSE peak survives.","supporting_citations":[{"cited_title":"Biermann, L","cited_arxiv_id":null,"evidence_quote":"Prior demonstration that interorbital spin fluctuations can induce non-Fermi-liquid behavior in the paramagnetic OSMP, the baseline this paper makes concrete."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduced the local two-qubit fidelity and its quantized values 0 and 0.5 for decoupled metallic and Mott-insulating bands."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Found the non-half-integer LTQF values and narrow-band quantum entanglement in the OSMP that the present paper sets out to explain."},{"cited_title":"Aucar Boidi, H","cited_arxiv_id":null,"evidence_quote":"Documents an in-gap band from interorbital Coulomb interactions, supporting the idea of low-energy interorbital excitations."},{"cited_title":"Song, X.-C","cited_arxiv_id":null,"evidence_quote":"Distinguishes Ising-type from full Hund's coupling in OSMPs, the classification this work refines."}],"review_version":1}