{"id":"4fc3eb3e-50c6-4424-9f12-7d540c684fa4","arxiv_id":"2505.16746","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Intersite electron hopping creates short-range, orthogonal orbital correlations in strongly spin-orbit coupled t2g systems, even with no static orbital order.","lead":"This theory paper argues that in crystals with strong spin-orbit coupling, electron hopping can revive short-range orbital polarization around a local disturbance, even when the undisturbed state is orbitally symmetric. If right, orbital fluctuations may hide inside the low-energy spectra of iridates and similar materials, affecting how RIXS and other probes are interpreted.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim is not isolated from Hoex: the model includes an orbital-exchange term derived from Jahn-Teller couplings with unspecified strength, so the 'g→0, hybridization-only' attribution is untested.","rationale":"The reader's weakest-assumption analysis focused on the neglect of vertex corrections (Upsilon=0 in Appendix E), which is a real approximation concern. My review, however, identifies a more direct threat to the central claim: the model contains an explicit orbital-exchange interaction Hoex that is derived from Jahn-Teller physics, with no stated value for its coupling J'. The central claim requires the effect to persist in the g→0 limit and to be driven solely by hopping. If Hoex is present with a nonzero J', then the short-range orbital correlations could be produced by that interaction rather than by hybridization, and the g→0 limit is not actually reached because Hoex does not automatically vanish. The paper does not provide the crucial control calculation with Hoex removed. This is a missing-evidence issue rather than a numerical inaccuracy, and it conditions the physical interpretation. Therefore the verdict remains conditional, but the condition should be the control calculation with Hoex=0, not only a vertex-correction benchmark.","tokens_in":24554,"tokens_out":7610,"duration_ms":65792,"concrete_test":"Repeat the self-consistent Dyson calculation with the orbital-exchange term Hoex removed (set J'=0, i.e., drop Σ^oex from Eq. A1) while keeping all other parameters unchanged, for fillings N=1-5 and t=0.2 and 1.2 eV. If the real-space nearest-neighbor orbital correlator ⟨δT_iδT_j⟩ remains negative and finite as in Fig. 3c, the hybridization-only claim survives. If the correlation vanishes or changes sign, the central attribution fails and the g→0 statement must be revised.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's novelty is that intersite hopping alone restores short-range orbital polarization, even as the Jahn-Teller coupling g→0 (Section III, final paragraphs). However, the lattice Hamiltonian in Section II includes Hoex, described as 'orbital exchange arising from Jahn-Teller interactions mediated by the lattice.' This term is retained in the self-consistent calculation: Eq. (A1) contains Σ^oex, and Appendix D.6 derives it with a coupling J' that is never specified or set to zero in any figure or parameter list. If J' is nonzero, a nonlocal orbital interaction derived from Jahn-Teller physics is present, and it can generate short-range orbital correlations by itself, independently of the kinetic term. The paper never shows results with Hoex removed (J'=0) or with both g=0 and J'=0, so the claim that the effect is 'hybridization-driven' and survives 'in the limit g→0' is not actually established. The cited 'data not shown' for g=0.01 eV cannot exclude this confound, because Hoex is still active at that point. The absence of any reported value for J' also prevents the reader from assessing the relative weight of the two mechanisms.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript studies t2g models with spin-orbit coupling, Kanamori Coulomb interactions, Jahn-Teller electron-phonon coupling, and an orbital-exchange term, within a Matsubara Green's function framework. The authors solve the lattice Dyson equation self-consistently with Hartree-Fock plus second-order Born self-energies and a Migdal electron-phonon self-energy, then compute orbital, spin-orbital, and spin-spin correlation functions. The central claim is that for fillings N=1..5, a local orbital perturbation induces short-range orbital polarization with an orthogonal nearest-neighbor response, driven by intersite hybridization rather than by Jahn-Teller coupling, and that this effect persists in the limit g→0. The paper also discusses imaginary-time dynamics and estimates the energy scale relevant to RIXS.","tokens_in":24814,"tokens_out":7715,"duration_ms":63827,"significance":"If correct, the result identifies a latent orbital instability in strongly spin-orbit-coupled t2g systems: a local probe could reveal short-range orbital correlations even when static Jahn-Teller order is quenched. This would be an interesting conceptual addition to the orbital-fluctuation literature and a useful caveat for interpreting RIXS spectra. The paper is transparent about its approximations: the vertex correction is explicitly set to zero, the self-energy channels are enumerated, and all material parameters are taken from the literature rather than fitted. The detailed appendices (DIIS, IR basis, self-energy derivations) support reproducibility. The main limitations are that the central correlation function is not benchmarked against any numerically exact method and that the role of the orbital-exchange term Hoex is not controlled; these limit the strength of the causal claim.","major_comments":[{"comment":"The lattice Hamiltonian in Section II contains Hoex, described as 'orbital exchange arising from Jahn-Teller interactions mediated by the lattice,' and the self-consistent Dyson equation (Eq. (A1)) includes the corresponding self-energy Σ^oex. Appendix D.6 defines the coupling J' but no numerical value is given anywhere, and no figure or parameter list sets J' = 0. Since Hoex is itself a nonlocal orbital interaction derived from JT physics, it can generate short-range orbital correlations by itself. The central claim that the effect is 'hybridization-driven' and survives 'in the limit g → 0' is therefore not established; the authors need to report the value of J' used and show calculations with both g = 0 and J' = 0 (or with Hoex removed). The 'data not shown' sentence for g = 0.01 eV cannot exclude this confound because Hoex is still included in those runs.","section":"Section II and Appendix D.6"},{"comment":"The quantitative evidence for the short-range orbital correlations is the mean-field factorization of the two-particle Green's function with the vertex correction Υ set to zero (Eq. (E3)), which yields the bubble expression in Eq. (3)/Eq. (E13). In the strongly correlated regime considered (U = 2.5 eV, J = 0.4 eV, t up to 1.2 eV), vertex corrections are potentially large, and the paper provides no benchmark against a numerically exact method (e.g., exact diagonalization on a small cluster, DCA, or CDMFT). A concrete test would be to compute the same nearest-neighbor correlator on a small cluster with the same parameters and J' = 0; without such a test, the persistence and sign of the correlations are not fully established.","section":"Appendix E, Eqs. (E2)-(E3), and Eq. (3)"},{"comment":"The interpretation of the negative nearest-neighbor correlations as a cooperative 'orthogonal' response should be separated from a mathematical property of the operator definitions. For m ≠ n, the rotation matrices in Eq. (E15) give cos(2(m−n)π/3) = −1/2, so the sign of the m ≠ n correlators in Eq. (E17) is fixed by construction before the self-consistent calculation is performed. The physical content is the magnitude, momentum dependence, and decay length of the correlation; the paper should verify that the orthogonal sign is not simply inherited from the choice of rotated basis, and should state this distinction explicitly in the discussion of Fig. 3c.","section":"Section III, Fig. 3, and Appendix E, Eqs. (E15)-(E17)"}],"minor_comments":[{"comment":"In the displayed HJT term, the phonon factor (b_iη + b_iη†) appears twice, which appears to be a typo; only one factor should be present in the t⊗E coupling.","section":"Section II"},{"comment":"The caption states that panel (a) varies ξ = 0.3 and 0.5 eV but then lists ξ = 0.1 eV among the parameters used; this inconsistency should be resolved.","section":"Figure 6 caption"},{"comment":"The conclusion that t = 0.2 eV corresponds to 'gapless or weakly gapped dynamics' is inferred only from the shape of C(τ) without analytic continuation; the claim should be softened or supplemented with a quantitative gap estimate, as the authors themselves acknowledge.","section":"Section IV and Fig. 5"},{"comment":"Equation (D25) labels the off-diagonal Hartree-Fock component as sHF_d; given the context, this is presumably a typo for sHF_od.","section":"Appendix D.4"},{"comment":"The notation η, ζ, m, n and θ/ϕ would benefit from a summarizing table that connects the mode labels, the spatial directions, and the Gell-Mann matrices, since the current text is difficult to follow.","section":"Appendix E"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the paper is within the journal's scope and the central idea is interesting, but the missing J' parameter, the unbenchmarked mean-field correlation function, and the partially definitional origin of the orthogonal sign make the central claim insufficiently supported. A major revision with small-cluster benchmarks and an explicit J' = 0 calculation seems the right route."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version: the core result is likely true — in spin-orbit-entangled t2g lattices, finite hopping restores short-range nearest-neighbor orbital correlations even when static JT order is quenched — and it deserves a serious referee. But the headline attribution ('hybridization-driven, independent of orbital-lattice coupling, in the limit g→0') is not actually isolated. The lattice Hamiltonian includes an orbital-exchange term Hoex, described as arising from Jahn-Teller interactions mediated by the lattice, and its coupling J' is never given a value and never set to zero in any figure. There is no control with J' = 0.\n\nCredit where it's earned. The approximations are stated plainly: Hartree-Fock plus second Born, Migdal electron-phonon, and the Upsilon = 0 vertex truncation in the two-particle channel. The appendices give enough explicit algebra — the self-energies in D27 and D38, the V² = 2I + V decomposition — that a patient reader could reproduce the calculation. The t → 0 limit recovering the symmetric ground state is the right minimal control and shows hopping is necessary for the effect. The single-site SOC-versus-JT maps for N = 1–5 are a useful recap, and the paper is properly cautious about how hard these meV-scale fluctuations would be to resolve in RIXS.\n\nThe soft spots, in proportion. Largest: the missing Hoex control. Because J' is unspecified, a JT-derived nonlocal orbital interaction of unknown strength sits in the self-consistent loop of every figure, so the 'even in the limit g → 0' claim is untested, and the key 5d case rests on 'data not shown.' This is fixable: rerun with Hoex removed, state J', and show the weak-g case in an actual figure. Second: the correlations are bare bubbles of dressed propagators (Eq. E3), and the orbital channel at U = 2.5 eV is exactly where vertex corrections can matter; with no benchmark against ED or cluster DMFT, the 'latent instability' language is uncalibrated. I'd rank this second to the Hoex issue, contrary to the reader's ordering. Third: part of the 'orthogonal response' is a projection artifact — the cos factor in Eq. E17 fixes the opposite sign between orientations once the bubble is positive — so the interpretation overstates what is dynamical rather than structural. Minor: Section I reports J = 0.2 while the lattice figures use J = 0.4, never reconciled.\n\nBottom line: referee it, conditionally. The requested controls are cheap and concrete, and if the authors supply them the qualitative claim likely survives in weaker, better-attributed form. I would not cite the headline attribution as it stands, but the paper would make a decent reading-group discussion about how attribution can outrun a controlled calculation.","headline":"A careful mean-field susceptibility calculation whose likely-true core result — hopping restores short-range orthogonal orbital correlations in SOC t2g systems — is undercut by an unspecified JT-derived exchange term the paper never controls for.","tokens_in":25338,"tokens_out":9865,"would_cite":false,"duration_ms":82548,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Intersite electron hopping can restore short-range orbital polarization in strongly spin-orbit-coupled t2g systems, even without Jahn-Teller coupling.","keywords":["spin-orbit coupling","orbital polarization","t2g electrons","Jahn-Teller effect","Matsubara Green's functions","intersite hopping","orbital correlations","4d and 5d transition-metal oxides"],"falsifier":"Run an unbiased many-body calculation on a small t2g cluster (four to eight sites) with the paper's parameters, keeping vertex corrections via exact diagonalization or determinant quantum Monte Carlo, and inspect the nearest-neighbor correlation $\\langle \\delta T_i \\delta T_j\\rangle$. The central claim stands if a negative nearest-neighbor correlation appears; it fails if the correlation is positive or zero.","tokens_in":1681,"feed_emoji":"⚛️","tokens_out":3279,"duration_ms":86161,"temperature":0.7,"pith_summary":"This paper argues that in the three low-lying d orbitals (the t2g manifold) of transition-metal oxides, ordinary electron hopping between sites can locally bring back orbital polarization that strong spin-orbit coupling otherwise wipes out. The effect needs no lattice distortion: the polarization is still there when the Jahn-Teller coupling is reduced to zero, and it appears only in response to a local perturbation. The response is short-ranged and cooperative: nearest-neighbor orbitals prefer to point orthogonally to the perturbed orbital, while correlations beyond the first neighbors die away. If the claim holds, a spin-orbit-entangled ground state that is globally unpolarized can still hide a latent orbital instability, and spectroscopic data on 4d and 5d compounds should account for these hybridization-driven fluctuations.","feed_headline":"Electron hopping alone revives orbital polarization","feed_subtitle":"A local nudge makes nearest-neighbor orbitals align orthogonally, even with zero Jahn-Teller coupling.","key_machinery":"The central object is the orbital charge moment $T_i^{\\eta}(n)$, a projection of the t2g occupation onto Eg-symmetric combinations of Gell-Mann matrices $\\lambda_3$ and $\\lambda_8$, rotated to the three Cartesian bond directions. Its correlations are computed in momentum space as a mean-field bubble, $\\langle \\delta T_q^{\\eta} \\delta T_{-q}^{\\zeta}\\rangle \\approx \\frac{1}{N}\\sum_k \\mathrm{Tr}[\\hat{\\lambda}^{\\eta}(m)\\hat{G}_k(0)\\hat{\\lambda}^{\\zeta}(n)\\hat{G}_{k-q}(\\beta)]$, i.e., two dressed single-particle Green's functions with all vertex corrections dropped. The dressed Green's functions come from a self-consistent Dyson equation that includes Hartree-Fock, second-order Born, electron-phonon, and orbital-exchange self-energies, exploiting the two-block algebraic structure of the propagators to keep the solver tractable. The sign of the computed real-space nearest-neighbor correlator is what carries the argument: negative values mean the neighboring orbital moment sits orthogonal to the perturbed one, and its distance dependence shows the response is confined to nearest neighbors.","core_discovery":"The paper's central claim is that intersite hybridization, not orbital-lattice coupling, is sufficient to create short-range orbital polarization in a strongly spin-orbit-coupled t2g system. Solving the lattice model with a self-consistent Matsubara Green's function approach, the authors find that a local orbital perturbation induces nonzero orbital-orbital and spin-orbital correlations whose real-space signature is a negative peak at nearest-neighbor distance, indicating orthogonal alignment of the neighboring orbital moments. The correlations remain when spin-orbit coupling is strong enough to quench static Jahn-Teller distortions in the atomic limit, and they survive even as the Jahn-Teller coupling $g$ goes to zero, which the authors take as evidence for a latent orbital instability of purely electronic origin. Their energy-scale estimate, $\\Delta_{\\mathrm{orbital}} \\sim (Zt)^2/(U_{\\mathrm{eff}}+\\Delta_{\\mathrm{SOC}})$, places the effect in the few-to-tens of meV range, below current RIXS resolution but potentially visible through indirect low-energy spectral signatures.","pith_inferences":["Editorial inference: the same bubble-correlator machinery could serve as a diagnostic for orbital instabilities in related multiorbital settings, such as systems with trigonal or orthorhombic crystal fields or t2g-eg mixing, where the authors themselves anticipate novel collective instabilities.","Editorial inference: because the predicted energy scale is only a few to tens of meV, very-low-temperature thermodynamic or transport measurements may reveal signs of short-range orbital fluctuations even where RIXS cannot resolve them directly.","Editorial inference: the claim implies a sharp crossover in hopping amplitude, with nearest-neighbor orthogonal correlations vanishing below roughly $t=0.05$ eV, a prediction that could be mapped in optical-lattice emulations of t2g bands with synthetic spin-orbit coupling."],"forward_implications":["A globally symmetric spin-orbit-entangled ground state with no static orbital order can still respond to local perturbations with orbital polarization on nearest-neighbor bonds.","The orbital response is cooperative: nearest-neighbor orbitals align orthogonally to the perturbed orbital, producing staggered short-range patterns rather than uniform ferroelectric-like order.","At realistic 4d and 5d hopping amplitudes ($t=0.2$-$1.2$ eV) the effect grows with hopping, while at very weak hopping ($t=0.05$ eV) it vanishes, confirming hybridization as the driver.","Imaginary-time correlators suggest that larger hopping opens a gap in orbital and spin-orbital fluctuations, while smaller hopping keeps them gapless or weakly gapped; analytic continuation would settle the distinction.","Hybridization-driven orbital dynamics may contribute to low-energy spectral features and polarization-dependent RIXS channels, even though direct detection is challenging."],"supporting_citations":[{"why":"Supplies the Matsubara Green's function and self-energy formalism used throughout the lattice calculation.","marker":"[18–21]"},{"why":"Provides the single-site spin-orbit versus Jahn-Teller phase picture that the lattice model extends to intersite hopping.","marker":"[26]"},{"why":"Give the Coulomb and spin-orbit parameter values used for 4d and 5d t2g systems.","marker":"[29, 30]"},{"why":"Defines the orbital charge moment operator used to construct the orbital-orbital and spin-orbital correlators.","marker":"[41]"},{"why":"Supply the realistic hopping amplitude range $t=0.2$-$1.2$ eV used in the calculations and energy estimate.","marker":"[42, 43]"},{"why":"Provide the Galitskii-Migdal formula used to compute the Jahn-Teller stabilization energy $E_{jt}$.","marker":"[50, 51]"},{"why":"Provide the intermediate-representation basis that makes the imaginary-time and imaginary-frequency transforms in the self-consistent solver accurate.","marker":"[64–66]"},{"why":"Provides the quasiparticle weight and effective Coulomb values used in the estimate of the hybridization-driven orbital energy scale.","marker":"[13]"}],"fun_headline_variants":["Electron motion alone rekindles orbital polarization","Hopping alone revives orbital order without Jahn-Teller","Pure hopping drives local orbital polarization sans Jahn-Teller","Spin-orbit systems: hopping polarizes orbitals alone"],"cache_read_input_tokens":27520,"weakest_assumption_plain":"The calculation assumes that two-particle orbital correlations are just products of two dressed single-particle Green's functions, discarding all vertex corrections; if those corrections are large in the strongly correlated regime, the predicted short-range orthogonal polarization could change, weaken, or disappear.","fun_headline_variants_meta":{"raw":{"variants":["Electron motion alone rekindles orbital polarization","Hopping alone revives orbital order without Jahn-Teller","Pure hopping drives local orbital polarization sans Jahn-Teller","Spin-orbit systems: hopping polarizes orbitals alone"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000692,"raw_usage":{"total_tokens":3144,"prompt_tokens":969,"completion_tokens":2175,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":585,"completion_tokens_details":{"reasoning_tokens":2110}},"tokens_in":585,"tokens_out":2175,"duration_ms":14201,"temperature":1.0,"reasoning_tokens":2110,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:55:47.060772+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run an unbiased many-body calculation on a small t2g cluster (four to eight sites) with the paper's parameters, keeping vertex corrections via exact diagonalization or determinant quantum Monte Carlo, and inspect the nearest-neighbor correlation $\\langle \\delta T_i \\delta T_j\\rangle$. The central claim stands if a negative nearest-neighbor correlation appears; it fails if the correlation is positive or zero.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the single-site spin-orbit versus Jahn-Teller phase picture that the lattice model extends to intersite hopping."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the orbital charge moment operator used to construct the orbital-orbital and spin-orbital correlators."},{"cited_title":"Gotfryd, E","cited_arxiv_id":null,"evidence_quote":"Provides the quasiparticle weight and effective Coulomb values used in the estimate of the hybridization-driven orbital energy scale."}],"review_version":1}