{"id":"4e601d64-f28a-411f-b6ea-2ef00639f1da","arxiv_id":"2509.06180","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":12,"one_line_summary":"Monolayer FeSe's orbital-selective renormalization is symmetry-required: x^2-y^2 electrons at the M point couple to Néel fluctuations while xz,yz electrons do not.","lead":"This paper shows that orbital-selective electronic behavior in monolayer FeSe is fixed by crystal symmetry and band structure, not by a Mott transition. The mechanism explains which iron orbitals feel magnetic fluctuations and matches key features of the measured ARPES spectra.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The ARPES comparison assumes zero renormalization of the xz,yz chemical potential, but the symmetry-permitted interband M3-M4 magnon vertex is never computed and could offset the x2-y2 shift that produces the modeled gap.","rationale":"The reader's weakest_assumption identifies the same load-bearing assumption, and I agree with that assessment. The central symmetry mechanism is well supported: the M1 versus M3/M4 irreducible-representation decomposition, the B2u selection rules, and the static DFT Neel calculations are parameter-free and would survive even if the quantitative ARPES fit is non-unique. The concern targets the quantitative claim that the mechanism accounts for the observed gapped dispersion. The authors explicitly acknowledge the missing mu_{xz,yz} renormalization and provide no calculation of the symmetry-permitted interband M3-M4 self-energy, so the gap in Fig. 4 rests on an unchecked assumption. This is testable with the same one-loop formalism already used, so a targeted computation would settle whether the concern actually lands. Absent that check, the paper's conclusion should remain conditional rather than fully accepted. I do not see grounds to reject the symmetry argument, which is independently motivated, so the reader's CONDITIONAL verdict is appropriate and unchanged.","tokens_in":11632,"tokens_out":9037,"duration_ms":87528,"concrete_test":"Evaluate Eq. (4) for the xz,yz interband M3-M4 vertex using the same electron-magnon coupling g = 0.93 eV, the same single-frequency magnon omega0 = 0.05 eV, and the DFT-derived M3-M4 separation (about 1 eV) and band parameters. Compute Re Sigma_{xz,yz}(0) and the implied shift of mu_{xz,yz}. If this shift is larger than about 0.1 eV, the assumed absence of xz,yz renormalization in Fig. 4 is invalid and the ARPES fit is not uniquely explained by x2-y2-only renormalization.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The symmetry selection rules and static DFT Neel calculations are independent, parameter-free support for an orbital-selective coupling at the M point. The load-bearing soft spot is the step from that vertex to the ARPES comparison. In Eq. (3) the authors couple Neel fluctuations only to the x2-y2 orbital, and in the k.p model the xz,yz bands are left with unrenormalized parameters. Their own limitation statement concedes this: 'the renormalization of mu_x2-y2 can be offset by a renormalization of mu_{xz,yz}, and we have assumed that there is no such renormalization.' This is not merely a fitting detail: symmetry forbids only the intra-band M4-M4 component; it explicitly permits a Neel vertex between the M3 and M4 IRs of the xz,yz sector. Since M3 sits only about 1 eV below M4, an interband one-loop self-energy of order g'^2/Delta E (with Delta E ~ 1 eV and g' comparable to the x2-y2 value g = 0.93 eV) is not obviously negligible. If that real self-energy shifts mu_{xz,yz}, the upward motion of x2-y2 required for the gapped dispersion can be cancelled, and the Fig. 4 ARPES agreement would not be evidence for the mechanism. The DFT-based self-energy check in the left panels of Fig. 4 does not settle this, because it imposes the same assumption by applying a real self-energy only to x2-y2.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a band-theory origin for orbital-selective correlations in monolayer FeSe. Using DFT of the Néel-ordered state, it observes that checkerboard Néel fluctuations couple strongly to x^2-y^2 orbitals and only weakly to {xz,yz} orbitals at the M point. A symmetry analysis for space group P4/nmm shows that x^2-y^2 forms the M1 irreducible representation and can be chosen site-localized on a single Fe, so a B2u Néel order parameter couples to it at first order, while the {xz,yz} orbitals form the M3 and M4 representations and can couple to B2u only through an inter-M3-M4 vertex. The authors then build a k.p model with an electron-magnon self-energy only on x^2-y^2, fit it to ARPES data for FeSe1-xSx, and argue that the resulting mass renormalization and upward shift of x^2-y^2 produce the observed gapped dispersion. They also state that the mechanism generalizes to a set of tetragonal and hexagonal space groups.","tokens_in":12084,"tokens_out":7955,"duration_ms":69170,"significance":"The symmetry-selection-rule part is clean, parameter-free, and independently supported by the static DFT Néel calculations; it gives a concrete, falsifiable distinction between site-localized x^2-y^2 states and non-site-localized {xz,yz} states. If the mechanism survives quantitative scrutiny, it is a significant alternative to site-local Mott-based orbital selectivity in FeSe. The paper's experimental support is the weakest element: the parameters entering the self-energy comparison are partly fitted, and the conclusion depends on an explicitly acknowledged assumption about the renormalization of the {xz,yz} chemical potential. The group-theoretic core is nevertheless novel and worth publishing after the numerical support is hardened.","major_comments":[{"comment":"The effective theory in Eq. (3) couples the Néel fluctuations only to the x2-y2 bands, but the symmetry analysis in the previous section permits a B2u vertex between the M3 and M4 representations of the {xz,yz} orbitals (the product M3⊗M4 contains B2u). The authors explicitly assume that the resulting renormalization of μ_{xz,yz} is absent, and they acknowledge that a shift of μ_{xz,yz} could offset the upward shift of μ_{x2-y2}. This assumption is load-bearing for the Fig. 4 ARPES comparison: the gapped dispersion is produced by the relative band shift, and an interband M3-M4 self-energy of order g'^2/ΔE with ΔE≈1 eV and g' comparable to g=0.93 eV is not obviously negligible. The manuscript should estimate this contribution, or explicitly downgrade the ARPES comparison to an illustrative consistency check.","section":"Band renormalization from Néel order, Eq. (3)"},{"comment":"The quantitative agreement with ARPES is partly a fit. The density of states N(0)=0.21 states/eV per Fe is chosen to reproduce the experimental bands, and the model calculation uses Z^{-1}=2, whereas the parameter estimate from g, N(0), and ω0 gives Z^{-1}≈3.6. Because the chosen Z^{-1} is the value needed to produce the experimental gap, the agreement in Fig. 4 does not independently confirm the mechanism. The authors should either fix all parameters from independent inputs or present the agreement as consistency rather than as validation.","section":"Band renormalization from Néel order, text following Eq. (4)"},{"comment":"The DFT-based self-energy check does not resolve the previous concern, because it applies a real orbital-dependent self-energy to x2-y2 only. This imposes by construction the same assumption that μ_{xz,yz} is not renormalized by Néel fluctuations. A self-consistent calculation that includes the symmetry-allowed M3-M4 Néel vertex is needed to test whether the relative band shift and the gap survive.","section":"Fig. 4, left panels"}],"minor_comments":[{"comment":"The caption should define the color scale and specify which extracted dispersion corresponds to which orbital character; the green and dashed black curves are difficult to associate with the panels.","section":"Fig. 2 caption"},{"comment":"The Pauli matrices τ and σ are not defined in the text; clarify that τ acts in the two-dimensional orbital IR space and σ in spin space.","section":"Eq. (3)"},{"comment":"The generalization to space groups 136 and 194 is only a sketch; if it is intended as a concrete claim, provide the IR decompositions, and if it is a forward-looking remark, label it explicitly.","section":"Orbital selective coupling in other space groups"},{"comment":"The use of the bulk FeSe magnon frequency ω0=0.05 eV for monolayer FeSe should be flagged as an assumption at the point where it is introduced, even though the authors later note that the monolayer magnon has not been measured.","section":"Band renormalization from Néel order"}],"recommendation":"major_revision","confidential_remarks":"The group-theory core is strong and the authors are candid about the μ_{xz,yz} assumption. The main revision should either compute the M3-M4 interband self-energy or narrow the paper's claims so that the ARPES comparison is presented as an illustration. In its present form the title's 'symmetry-required' is justified for the qualitative coupling structure, but the quantitative orbital selectivity in the ARPES comparison rests on an unverified assumption."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know: the symmetry selection rule at the M point is the real result, and it is probably correct. The authors show that Néel order couples within the M1 IR (x^2-y^2), giving a site-local basis, while {xz,yz} (M3/M4) only couple inter-band. That is parameter-free group theory plus a clean DFT demonstration with static Néel order. That part holds up.\n\nWhat is new: previous literature attributes orbital-selective renormalization in FeSe to site-local Mott physics. Here the dichotomy is enforced by the non-symmorphic space group and the intertwining of orbital and sublattice degrees of freedom. I have not seen that argument in the orbital-selective Mott literature, and the generalization claim to other space groups, though sketched, is plausible.\n\nThe soft spot is the ARPES comparison. The model uses Z^-1=2 to reproduce the gapped dispersion, while the parameter estimate gives ~3.6, and the authors explicitly assume no renormalization of mu_{xz,yz}. The stress-test objection is real: symmetry permits a Néel vertex between M3 and M4, and if that self-energy is comparable to g^2/omega_0, it can shift the xz,yz chemical potential and cancel the upward motion of x^2-y^2. The paper never computes that inter-band vertex. The DFT check in Fig. 4 (left) does not settle it, because it applies a self-energy only to x^2-y^2. So the agreement with ARPES is a fit with an acknowledged assumption, not a prediction. This does not break the central symmetry argument, but it means the paper is strongest as a band-structure mechanism and weaker as a quantitative account of the ARPES gap.\n\nThe citation pattern is fine, building on the authors' prior k.p work without hiding it. Raw data and code are not provided, which is normal for this kind of paper but limits independent checking.\n\nWho should read it: anyone working on Fe-based superconductors, orbital selectivity, or non-symmorphic band structures. It deserves a serious referee; the referee should press for a computation of the M3-M4 vertex and a sensitivity analysis of the mu shift. I would accept it with revisions rather than desk reject.","headline":"The symmetry selection rule at the M point is clean, new, and likely right; the ARPES comparison is a fitted illustration rather than a test, because the interband xz,yz vertex is never computed.","tokens_in":12635,"tokens_out":2230,"would_cite":true,"duration_ms":20359,"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":"In monolayer FeSe, crystal symmetry forces orbital-selective renormalization without any site-local Mott transition.","keywords":["orbital-selective correlations","monolayer FeSe","checkerboard antiferromagnetic fluctuations","symmetry-required orbital selectivity","iron-based superconductors","ARPES","irreducible representations","effective k.p model"],"falsifier":"A self-consistent DFT calculation with checkerboard antiferromagnetic order at a small local moment (say $0.05\\,\\mu_B$) should already show a splitting of the $x^2-y^2$ bands at $M$ while the $\\{xz,yz\\}$ bands remain degenerate; if both orbital sets split comparably, or neither does, the symmetry-required mechanism would be falsified.","tokens_in":11441,"feed_emoji":"🧲","tokens_out":13129,"duration_ms":102215,"temperature":0.7,"pith_summary":"This paper tries to establish that the orbital-selective physics in monolayer FeSe—where electrons in $x^2-y^2$ orbitals are much more strongly renormalized than those in $\\{xz,yz\\}$ orbitals—is required by the crystal's symmetry and band structure, not by a site-local Mott transition. At the $M$ point of the Brillouin zone, the $x^2-y^2$ Bloch states can be localized on individual Fe sites, so they couple strongly to checkerboard antiferromagnetic fluctuations, while the $\\{xz,yz\\}$ states cannot be site-localized and couple only weakly through an inter-band channel. A one-loop self-energy calculation shows these fluctuations renormalize the $x^2-y^2$ band, push it up through the $\\{xz,yz\\}$ bands, and produce the gapped dispersion observed in angle-resolved photoemission, with Fermi momenta held fixed by Luttinger's theorem. If correct, this replaces a strong-correlation explanation with a symmetry-required band-theory mechanism that extends to other crystal space groups.","feed_headline":"Symmetry, not Mott physics, drives orbital selectivity in FeSe","feed_subtitle":"In-plane d-orbital bands split into a gap at M while the other d orbitals stay fixed, matching photoemission data.","key_machinery":"The central object is the little-group representation content of the electron pockets at the $M$ point of space group 129, in particular which orbitals can form site-local Bloch states. The $x^2-y^2$ orbitals transform as the $M_1$ representation with a basis localized on individual Fe sites, while the $\\{xz,yz\\}$ orbitals form the $M_3$ and $M_4$ representations that cannot be site-localized without mixing. Checkerboard antiferromagnetic order transforms as $B_{2u}$, and the direct-product rules put this representation inside $M_1\\otimes M_1$ (intra-band coupling for $x^2-y^2$) but only inside $M_3\\otimes M_4$ (inter-band coupling for $\\{xz,yz\\}$). That selection rule, combined with a one-loop magnon self-energy in the effective $k\\cdot p$ model, produces the orbital-selective renormalization and the observed gapped dispersion.","core_discovery":"The central claim is that the orbital-selective renormalization in monolayer FeSe originates from the symmetry of the Bloch wavefunctions at the $M$ point, not from a Mott transition. In the non-magnetic state, the $x^2-y^2$ orbitals on the two Fe sites form the $M_1$ irreducible representation, whose basis states can be chosen as site-localized $|x^2-y^2,A\\rangle$ and $|x^2-y^2,B\\rangle$; checkerboard antiferromagnetic order (symmetry $B_{2u}$) appears in $M_1\\otimes M_1$, so it couples within the $M_1$ pair and splits the $x^2-y^2$ bands. The $\\{xz,yz\\}$ orbitals instead form the $M_3$ and $M_4$ representations, and the same $B_{2u}$ order appears only in $M_3\\otimes M_4$, so their coupling is purely inter-band and ineffective at low momenta. Including only the $x^2-y^2$ coupling to antiferromagnetic magnons at one-loop order renormalizes the effective mass and the $\\alpha$ term of the $k\\cdot p$ Hamiltonian by a common factor $Z^{-1}=N(0)g^2/\\omega_0$ (estimated near 2 to 3.6), and applying Luttinger's theorem to reset $\\mu_{x^2-y^2}$ pushes the band upward into a gapped crossing with $\\{xz,yz\\}$, matching the ARPES spectra. The authors conclude the mechanism is symmetry-required and generalizes to other space groups with four-fold and six-fold screw axes.","pith_inferences":["If correct, the mechanism predicts that any perturbation changing the little-group classification at $M$, such as strain that lowers the crystal symmetry, should weaken or destroy the orbital-selective renormalization; this is testable in uniaxial-strain experiments on monolayer FeSe.","The predicted renormalization factor $Z^{-1}=N(0)g^2/\\omega_0$ makes a quantitative prediction for how the effective-mass enhancement should track the magnon frequency and coupling constant as doping or strain tunes the system.","The same symmetry lens could be used as a design rule: materials whose Fermi-surface hot spots have site-localizable orbitals in the relevant irreducible representation should exhibit orbital-selective correlations, while those without should not.","The one-loop, single-frequency-magnon model is the minimal carrier of the effect; a momentum-dependent magnon spectrum and vertex corrections would be needed to see whether the estimated $Z^{-1}\\approx2$–$3.6$ window survives and whether it fully accounts for the experimental mass enhancement."],"forward_implications":["The apparent gap at the $M$ point in monolayer FeSe ARPES is a direct consequence of $x^2-y^2$ bands being pushed up through $\\{xz,yz\\}$ bands by checkerboard antiferromagnetic-fluctuation renormalization, with $k_F$ values locked by Luttinger's theorem.","Sulfur substitution weakens the checkerboard antiferromagnetic fluctuations, which explains the observed decrease in effective mass and the downward shift of band positions at $M$ with increasing S concentration.","Orbital-selective physics in FeSe does not require an orbital-selective Mott transition; the band structure alone enforces the hierarchy that $x^2-y^2$ electrons are more strongly correlated than $\\{xz,yz\\}$ electrons.","The extra $x^2-y^2$ spectral weight brought to the chemical potential amplifies the pairing interaction from the magnetic fluctuations, explaining why this pairing becomes more energetic than the usual $s_\\pm$ gap.","The same symmetry argument predicts analogous orbital-selective coupling to magnetic or nematic order in other tetragonal and hexagonal space groups, such as the rutile structure at its $M$ or $A$ point and hexagonal close-packed at $H$."],"supporting_citations":[{"why":"The earlier ARPES observation of the M-point gap and the orbital-selective-Mott interpretation this paper argues against; the band-interchange picture is reused and given a band-theory origin.","marker":"[5]"},{"why":"Provides the low-energy k.p Hamiltonian for the M-point electron pockets and the magnetic-fluctuation pairing model on which the self-energy calculation is built.","marker":"[19]"},{"why":"Spin-spiral calculations of the Heisenberg parameters for single-layer FeSe, used to argue that checkerboard antiferromagnetic fluctuations are strong there.","marker":"[24]"},{"why":"Gives J2/J1 = 0.59 for FeSe versus 0.69 for FeS, the basis for the statement that checkerboard antiferromagnetic fluctuations are more important in FeSe than in S-doped samples.","marker":"[33]"},{"why":"Standard crystallographic group-theory source for the irreducible-representation notation used to label the M-point states of space group 129.","marker":"[34]"},{"why":"Companion source providing representation product tables that yield the M1-M1 and M3-M4 selection rules.","marker":"[35]"},{"why":"The one-loop self-energy approach to Fermi-surface shrinking from interband coupling, used here to justify renormalizing only one orbital channel and applying Luttinger's theorem.","marker":"[41]"},{"why":"Measured magnon frequency in FeSe, used to estimate omega0 = 0.05 eV and hence the renormalization factor Z^{-1} = N(0)g^2/omega0.","marker":"[43]"}],"fun_headline_variants":["FeSe orbital selectivity: symmetry, not Mott, is the key","Symmetry dictates which orbital renormalizes in FeSe","In FeSe, symmetry picks the orbital that feels magnetism","No Mott needed: FeSe's orbital split comes from symmetry"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The comparison to ARPES assumes that the renormalization of $\\mu_{x^2-y^2}$ is not offset by a renormalization of $\\mu_{\\{xz,yz\\}}$; if both chemical potentials shift together, the $x^2-y^2$ band would not rise relative to $\\{xz,yz\\}$ and the predicted gap at $M$ would not appear.","fun_headline_variants_meta":{"raw":{"variants":["FeSe orbital selectivity: symmetry, not Mott, is the key","Symmetry dictates which orbital renormalizes in FeSe","In FeSe, symmetry picks the orbital that feels magnetism","No Mott needed: FeSe's orbital split comes from symmetry"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000935,"raw_usage":{"total_tokens":4090,"prompt_tokens":1125,"completion_tokens":2965,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":741,"completion_tokens_details":{"reasoning_tokens":2895}},"tokens_in":741,"tokens_out":2965,"duration_ms":18066,"temperature":1.0,"reasoning_tokens":2895,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:19:00.151244+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A self-consistent DFT calculation with checkerboard antiferromagnetic order at a small local moment (say $0.05\\,\\mu_B$) should already show a splitting of the $x^2-y^2$ bands at $M$ while the $\\{xz,yz\\}$ bands remain degenerate; if both orbital sets split comparably, or neither does, the symmetry-required mechanism would be falsified.","supporting_citations":[{"cited_title":"Yi, Z.-K","cited_arxiv_id":null,"evidence_quote":"The earlier ARPES observation of the M-point gap and the orbital-selective-Mott interpretation this paper argues against; the band-interchange picture is reused and given a band-theory origin."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the low-energy k.p Hamiltonian for the M-point electron pockets and the magnetic-fluctuation pairing model on which the self-energy calculation is built."},{"cited_title":"Phys.1, 8 (2018)","cited_arxiv_id":null,"evidence_quote":"Spin-spiral calculations of the Heisenberg parameters for single-layer FeSe, used to argue that checkerboard antiferromagnetic fluctuations are strong there."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives J2/J1 = 0.59 for FeSe versus 0.69 for FeS, the basis for the statement that checkerboard antiferromagnetic fluctuations are more important in FeSe than in S-doped samples."},{"cited_title":"Ivantchev, G","cited_arxiv_id":null,"evidence_quote":"Standard crystallographic group-theory source for the irreducible-representation notation used to label the M-point states of space group 129."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Measured magnon frequency in FeSe, used to estimate omega0 = 0.05 eV and hence the renormalization factor Z^{-1} = N(0)g^2/omega0."}],"review_version":2}