{"id":"050f59d6-920d-4b1d-8975-9b6bf427b81e","arxiv_id":"1908.04889","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"Orbital transmutation, driven by spin-orbit coupling or surface hybridization, explains why the xz and yz states at the M point of FeSe do not merge above the nematic transition.","lead":"This paper offers two explanations for a puzzling experimental observation in the iron-based superconductor FeSe, where two electron states that should become identical above a 90 K transition instead remain separate in measurements. Both explanations involve a subtle swapping of the orbital character of the states, and the authors propose polarized light experiments to tell the two mechanisms apart.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central resolution depends on unmeasured lambda/eta and on a fitted xz/xy level crossing; until a parameter-free test (polarized ARPES above Tnem or a full refit) is done, the claim is conditional, not established.","rationale":"The reader's conditional verdict is appropriate. I checked the diagonalizations in Sec. III: Eqs. (14)-(19) are internally consistent, and the branch-connectivity argument through the avoided crossing is sound. The main vulnerability is not a mathematical error but empirical underdetermination: lambda and eta are free parameters, and the crossing condition depends on fitted orbital energies and nematic order parameters. The paper is honest about this in Sec. V and proposes the correct discriminating experiment, but the central claim is not yet independently confirmed. My concern refines the reader's weakest assumption by identifying D(T) crossing zero as the precise condition that must hold, and the proposed refit tests that condition directly against existing data. Since the analysis is plausible and testable but not uniquely determined by current measurements, the conditional verdict should stand unchanged.","tokens_in":16794,"tokens_out":19050,"duration_ms":197553,"concrete_test":"Refit the published temperature-dependent ARPES peak positions and intensities at M using the full Hamiltonian of Sec. II with lambda, eta, phi1,0, and epsilon1,0 - epsilon3,0 as free parameters, including experimental uncertainties. Determine whether the best fit requires lambda or eta significantly nonzero and whether the fitted D(T) = epsilon1 - epsilon3 + phi1(T) + phi3(T) crosses zero below Tnem. If the best fit returns lambda ~ eta ~ 0, or D(T) > 0 for all T < Tnem, the orbital-transmutation scenario is falsified; if the fit requires nonzero couplings and a crossing, the central claim gains direct quantitative support.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that orbital transmutation resolves the ARPES puzzle—rests on two unverified conditions. First, the bare xz and xy levels at M must cross as nematic order grows, i.e. |phi1 + phi3| must exceed |epsilon1 - epsilon3|. If D(T) = epsilon1 - epsilon3 + phi1(T) + phi3(T) never changes sign, Eq. (16) gives phi approx 0 for all T, the low-T xz branch remains connected to the upper xz/yz doublet, and the standard-model contradiction reappears. The paper fixes phi1,0 = -24 meV and epsilon1 - epsilon3 = 7.4 meV, but phi1 is not measured independently of the very splitting that SOC/SIH modify (Eq. 21), and Sec. V explicitly states that neither lambda nor eta has been measured directly. Second, even if the crossing occurs, the spectral-weight transfer must be large enough to keep the transmuted branch observable; this depends on the same unmeasured lambda and eta. A smaller actual phi1, a larger actual epsilon1 - epsilon3, or negligible lambda/eta would suppress the transmutation. This does not make the paper wrong, but it makes the proposed resolution conditional on a parameter regime that is currently asserted rather than independently established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript proposes an explanation for the persistent splitting of the xz/yz-derived excitations at the M point in FeSe above the nematic transition. Working in the 2-Fe Brillouin zone, the authors add to the standard nematic k·p model either the spin-orbit coupling allowed at M or a surface-induced xz/yz hybridization. They show that in both cases the dominant orbital character of one low-energy branch changes between the tetragonal and deeply nematic regimes ('orbital transmutation'), so the state identified as xz at low T does not return to the upper xz/yz doublet as T approaches T_nem, but instead connects to the lower xy-dominated doublet. The model is used to compute M-point energies, spectral functions, Fermi-surface shapes, and orbital weights, and is compared with ARPES and STM data. The two scenarios are distinguished by predictions for polarized ARPES above T_nem.","tokens_in":17180,"tokens_out":17202,"duration_ms":169710,"significance":"If established, the orbital-transmutation mechanism would resolve a long-standing ARPES puzzle in FeSe and would be a useful concept for other small-Fermi-energy iron-based materials. The paper's strengths are its analytical transparency (the 2x2 diagonalizations leading to Eqs. (14)-(19) are clean), its symmetry-based derivation of the allowed SOC and surface terms, and the fact that it makes concrete falsifiable predictions: two degenerate doublets above T_nem in the SOC scenario versus split singlets in the SIH scenario, with distinct orbital contents measurable by polarized ARPES. The main limitation is that the effect is demonstrated in a parameter regime whose key inputs are not independently measured; the paper therefore establishes an internally consistent scenario rather than a closed explanation.","major_comments":[{"comment":"The central claim that the low-temperature xz branch does not merge at T_nem depends on the sign change of D(T) = eps1 - eps3 + phi1(T) + phi3(T) (Eq. (16)): only if |phi1+phi3| exceeds |eps1-eps3| does the mixing angle phi reach the transmuted branch. In the manuscript this inequality is guaranteed by choices that are fitted to the very data under study: phi1,0 = -24 meV, eps1,0 - eps3,0 is set to 7.4 meV in Sec. II and to 10.2 meV in the SOC scenario (Fig. 4), and phi3,0 is varied as 0 or +/-10 meV although its sign is stated to be experimentally unverified (Sec. I). Moreover, Sec. V explicitly states that neither lambda nor eta has been measured directly; the value 10 meV is assumed. If the actual lambda were much smaller, or if eps1 - eps3 were larger, the sign change would not occur and the standard-model contradiction would remain. Please provide an independent constraint on this parameter regime, for example a full fit to the M-point dispersions and orbital weights with lambda and eta as free parameters, or a threshold analysis showing how large lambda and eta must be for the transmutation to survive.","section":"Sec. III, Eqs. (13)-(19); Sec. V"},{"comment":"The quantitative comparison with ARPES is partly constructed. Eq. (21) shows that the low-temperature splitting Delta E is a combination of phi1, phi3, lambda, and eps1 - eps3, and the text (Sec. II) states that eps1,0 and eps3,0 are adjusted in each scenario to maintain the peanut-shaped pocket; the values differ between the SOC scenario (-24.8, -35.0 meV) and the SIH scenario (-26.3, -32.0 meV). Thus the measured M-point splitting constrains phi1 only through an assumed lambda (or eta), while the onsite energies are themselves re-fit to reproduce the pocket shapes. The agreement therefore demonstrates consistency, not prediction. I recommend an explicit statement of this circularity and, if possible, a two-parameter scan of (lambda, eta) showing the range over which the data are reproduced.","section":"Sec. III, Eq. (21); Sec. II"},{"comment":"The observability of the non-merging in the SOC scenario relies on the ad hoc assumption that Gamma_xy = 10 meV is much larger than Gamma_xz = Gamma_yz = 3 meV. If the xy orbital were coherent, the transmuted lower branch would remain a sharp peak and the spectral function would show four peaks; if Gamma_xy is made very large, the apparent merging is produced by broadening rather than by the physics of transmutation. The paper motivates this input by orbital-selective physics (Refs. [38,39]) but does not calibrate or test it. This premise should be stated as a separate assumption and its consequences checked by varying Gamma_xy over a wide range.","section":"Sec. III, Eq. (20) and Fig. 6"},{"comment":"The SIH scenario has the same status as a consistency check rather than an explanation, because the splitting above T_nem is exactly the assumed surface hybridization 2 eta and eta is not measured. The scenario is only distinguished from SOC by the orbital content above T_nem (Sec. V), which is not currently available. The authors should either soften the wording that these scenarios 'describe the data' to 'are consistent with the data' or provide additional evidence, for example a surface-sensitive measurement or a comparison between ARPES and bulk probes, that an xz/yz hybridization of order 10 meV exists.","section":"Sec. IV, Eq. (12); Sec. V"}],"minor_comments":[{"comment":"The mixing angle phi in Eq. (16) is defined through tan 2phi, which is invariant under 2phi -> 2phi + pi; the text's statement that phi varies from approximately 0 to approximately -pi/2 is only correct for one branch of the inverse tangent. Please state the chosen branch or define phi directly from the eigenvector components to avoid ambiguity.","section":"Eq. (16) and surrounding text"},{"comment":"There are several typographical errors: 'ro' should be 'to' in Sec. IV, 'intentisty' should be 'intensity' and 'variation with' should probably be 'contrast with' in Sec. V, and 'fartherst' should be 'farthest' in the Fig. 9 caption.","section":"Throughout"},{"comment":"The temperature-dependent onsite energies eps1(T) = eps1,0 + 0.083T and eps3(T) = eps3,0 + 0.083T are given without stating the units of the slope; please specify meV/K and note that this slope is also an input fitted to the ARPES data.","section":"Sec. II, Table I"},{"comment":"The caption color code references the same orbital colors as Fig. 2, but the printed figures use overlapping blue and green shades; for accessibility, consider using distinct line styles or explicitly labeling the pockets in the lower panels.","section":"Fig. 7 and Fig. 11"}],"recommendation":"major_revision","confidential_remarks":"The paper is a plausible and clearly written proposal. My main concern is that the word 'solves' is stronger than the evidence; the authors can address this by reframing the conclusion and adding a parameter-sensitivity analysis. I see no ground for rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know about this paper: it proposes a specific mechanism — orbital transmutation — to explain why the xz/yz doublet at the M point of FeSe does not merge above T_nem, and it does so with a clean 2x2 model. That is actually new. The ingredients (SOC and surface hybridization) were known, but nobody had identified that the dominant orbital character of a band can flip between tetragonal and nematic phases, so that the low-T xz excitation actually connects to the xy-dominated doublet above T_nem. The paper works through both SOC and SIH scenarios, computes spectral functions and Fermi surfaces, and gives a concrete way to distinguish them with polarized ARPES. Credit is due for that.\n\nWhat it does well: the symmetry analysis is careful, the diagonalization is transparent, and the qualitative match to ARPES data — two sharp peaks deep in the nematic phase, two weaker xy features in between, and the persistence of splitting above T_nem — is plausible. It also explains the missing outer electron pocket via xy incoherence, which is consistent with earlier work. The paper is honest about what is not known; Sec. V explicitly says neither lambda nor eta has been measured directly.\n\nWhere it is soft: the whole transmutation mechanism only operates if the bare xz and xy levels cross as nematic order grows, i.e. |phi1+phi3| > epsilon1-epsilon3. That condition is satisfied only for the chosen parameters: phi1,0 = -24 meV and epsilon1-epsilon3 = 7.4 meV. These are not independently pinned down. phi1 is treated as known, but in the presence of SOC the splitting at low T is not simply -2 phi1 (Eq. 21), so there is no clean experimental handle on phi1 separately. Similarly, lambda and eta are set to 10 meV. If the actual values are smaller, or the orbital splitting larger, the transmutation is suppressed and the puzzle re-emerges. This makes the explanation conditional, not wrong. The paper does not pretend to have a parameter-free prediction; it offers a credible scenario and a test. The two scenarios (SOC vs SIH) are not distinguished by current data, but the paper gives a clear experiment to do.\n\nIn short: this is a solid theory paper that resolves a known puzzle in a very specific parameter regime. That regime may be right — FeSe has unusually small Fermi energies and the level crossing is visible in the standard model itself — but it is asserted, not yet established. The paper deserves a serious referee and would be a good reading-group piece for exactly that reason: it's a clear example of a mechanism that can look like a prediction but is actually fitted to the data. I'd cite it if I write about FeSe nematicity, and I'd accept it for peer review with the expectation that the authors sharpen the parameter-dependence discussion.","headline":"Orbital transmutation is a genuinely new mechanism for the FeSe M-point puzzle, but the resolution is conditional on unmeasured couplings and fitted level crossings; still worth a serious referee.","tokens_in":17693,"tokens_out":3599,"would_cite":true,"duration_ms":32693,"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":"Spin-orbit coupling or surface hybridization can change an excitation's dominant orbital character in FeSe between the tetragonal and nematic phases, explaining why the xz and yz modes at the M point do not merge at T_nem.","keywords":["FeSe","nematic order","orbital transmutation","spin-orbit coupling","surface-induced hybridization","angle-resolved photoemission spectroscopy","iron-based superconductors","electron pockets"],"falsifier":"A decisive test is polarization-resolved ARPES at the $M$ point just above $T_{\\rm nem}$ with resolution better than the expected splittings: the spin-orbit scenario predicts one doubly degenerate $xz/yz$ doublet and one doubly degenerate $xy$ doublet, whereas the surface-hybridization scenario predicts two split $xz/yz$ modes with equal $xz$ and $yz$ orbital weight. Observing neither the predicted doublet structure nor the two split singlets would falsify the orbital-transmutation mechanism; seeing the two modes actually merge above $T_{\\rm nem}$ would indicate the puzzle was a thermal-broadening artifact.","tokens_in":16604,"feed_emoji":"🧲","tokens_out":10997,"duration_ms":95217,"temperature":0.7,"pith_summary":"The paper sets out to explain a specific ARPES anomaly in FeSe: deep in the nematic phase two sharp excitations at $M=(\\pi,\\pi)$ are dominated by the $xz$ and $yz$ orbitals, yet as the temperature rises through $T_{\\rm nem}$ they approach each other and remain split, even though $xz$ and $yz$ are degenerate in the tetragonal phase. The paper argues that the standard nematic model fails here because it ignores hybridization already present above $T_{\\rm nem}$. Including spin-orbit coupling, or the surface-allowed $xz/yz$ hybridization probed by ARPES and STM, produces an orbital transmutation: the excitation that looks $xz$ deep in the nematic phase becomes mostly $xy$ near $T_{\\rm nem}$, so it merges with a different doublet and the two observed modes never meet. A sympathetic reader would care because the mechanism turns a puzzling experimental observation into a sharp, testable prediction about how orbital character changes with temperature, and it also explains why one of FeSe's electron pockets is hard to see in photoemission.","feed_headline":"Orbital swap explains why FeSe's electron bands stay split","feed_subtitle":"Spin-orbit coupling turns an xz mode into xy near T_nem, so the two ARPES peaks at the M point never meet.","key_machinery":"The load-bearing object is the effective $2\\times2$ Hamiltonian at the $M$ point in the $xz$--$xy$ and $yz$--$xy$ orbital sectors, built from a $k\\cdot p$ expansion that respects the glide-plane symmetry of FeSe. The central identity is the mixing angle $\\tan 2\\phi = -\\lambda/(\\epsilon_1-\\epsilon_3+\\phi_1+\\phi_3)$; when the nematic combination $\\phi_1+\\phi_3$ grows large enough to overcome the small orbital splitting $\\epsilon_1-\\epsilon_3$, the angle moves from near zero to near $-\\pi/2$ and the dominant orbital weight of the two eigenstates is exchanged. That sign change, named orbital transmutation, is what lets one excitation from each doublet swap its orbital identity between high and low temperature. In the surface scenario the same work is done by the hybridization $\\eta\\,\\hat d^\\dagger_{xz,\\sigma} d_{yz,\\sigma}+\\mathrm{H.c.}$, which is forbidden in the bulk by glide-plane symmetry but allowed at the surface probed by ARPES and STM.","core_discovery":"The central claim is that an excitation from each doublet at the $M$ point undergoes an orbital transmutation below $T_{\\rm nem}$: its dominant orbital contribution changes relative to the tetragonal phase. In the spin-orbit-coupling scenario, the doublet closer to the Fermi level splits into a state that remains predominantly $yz$ and a state that becomes predominantly $xy$ at $T\\ll T_{\\rm nem}$; the other doublet splits into a state that remains predominantly $xy$ and a state that becomes predominantly $xz$. Because the $xz$-dominated low-temperature mode has become $xy$ as $T$ approaches $T_{\\rm nem}$, it does not merge with the $yz$ mode; instead it joins the lower $xy$ doublet. The paper argues this reproduces the ARPES spectra, including the two sharp peaks at low temperature, the two weaker intermediate peaks identified with $xy$ orbitals, and the missing outer electron pocket if $xy$ fermions are incoherent. In the complementary surface-hybridization scenario, the same non-merging follows from a tetragonal-phase splitting of the $xz/yz$ doublet into equal $xz\\pm yz$ mixtures, with the low-temperature eigenstates becoming almost pure $yz$ and $xz$.","pith_inferences":["If the mechanism is generic, orbital transmutation should be strongest in materials where the relevant orbital splitting is comparable to the nematic order parameter, as in FeSe; in materials with larger orbital splittings the same hybridization would produce only small weight transfers rather than an identity swap.","A testable extension would be to look for the predicted temperature dependence of spectral weight in detwinned crystals: the low-temperature $xz$-dominated peak should continuously lose $xz$ weight and gain $xy$ weight as $T$ approaches $T_{\\rm nem}$, which could be seen by tracking peak intensities under two polarizations.","The same level-crossing logic may apply to other multiorbital systems with a nematic-like order parameter that can flip the sign of a mixing-angle denominator, suggesting orbital transmutation as a general spectroscopic signature of orbital order."],"forward_implications":["The measured splitting between the $xz$ and $yz$ modes deep in the nematic phase is not simply $2|\\phi_1|$; it is the modified expression in Eq. (21) that involves $\\lambda$, $\\phi_1+\\phi_3$, and the orbital splitting, so extracting the nematic order parameter from ARPES requires accounting for spin-orbit coupling.","The absence of the outer electron pocket in ARPES finds a natural explanation: when $xy$ fermions are incoherent, a pocket that is predominantly $xy$ contributes only weak, broad spectral weight.","The $xz$ spectral weight measured along the diagonal of the peanut-shaped inner electron pocket is a direct fingerprint of spin-orbit coupling, and its magnitude is sensitive to the values of $\\lambda$ and the nematic order parameters.","The two scenarios can be separated by measurements in the tetragonal phase: SOC gives two doubly degenerate bands, while surface-induced hybridization splits the $xz/yz$ doublet into distinct singlet excitations even above $T_{\\rm nem}$."],"supporting_citations":[{"why":"It supplies the symmetry-allowed k·p Hamiltonian and the momentum-independent spin-orbit coupling at the M point that the paper extends.","marker":"[10]"},{"why":"It provides the measured spin-orbit coupling of about 20 meV on the hole pockets, the scale that motivates lambda = 10 meV.","marker":"[12]"},{"why":"It gives the STM data on the peanut-shaped inner electron pocket and its orbital composition used to set the model parameters.","marker":"[2]"},{"why":"It reports the ARPES behavior of the excitations near M, including the strongly temperature-dependent chemical potential.","marker":"[16]"},{"why":"It documents the emergence of the nematic electronic state and the persistence of the splitting above T_nem.","marker":"[18]"},{"why":"It introduces the surface-induced xz/yz hybridization term allowed only at the surface, which is the basis of the second scenario.","marker":"[20]"},{"why":"It reports two additional weaker excitations deep in the nematic phase, which the paper assigns to predominantly xy modes.","marker":"[23]"},{"why":"It is the polarized ARPES study that identifies the low-temperature excitations as xz- and yz-dominated and sharpens the non-merging puzzle.","marker":"[26]"},{"why":"It follows the two excitations from deep in the nematic phase to above T_nem and argues that the residual splitting is physical.","marker":"[34]"}],"fun_headline_variants":["Orbital shift keeps FeSe bands split above T_nem","FeSe's split bands persist via orbital transmutation","Spin-orbit coupling explains FeSe's non-merging bands","Orbital character change stops FeSe band merging","Why FeSe electron bands never meet: orbital swap"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument needs the spin-orbit coupling $\\lambda$ (or the surface hybridization $\\eta$) to be large enough—the paper uses 10 meV for each—and the orbital splitting $\\epsilon_1-\\epsilon_3$ to be small enough that the nematic combination $\\phi_1+\\phi_3$ overturns the sign of the mixing-angle denominator; neither $\\lambda$ nor $\\eta$ has been measured directly on the electron pockets.","fun_headline_variants_meta":{"raw":{"variants":["Orbital shift keeps FeSe bands split above T_nem","FeSe's split bands persist via orbital transmutation","Spin-orbit coupling explains FeSe's non-merging bands","Orbital character change stops FeSe band merging","Why FeSe electron bands never meet: orbital swap"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000246,"raw_usage":{"total_tokens":1691,"prompt_tokens":1247,"completion_tokens":444,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":863,"completion_tokens_details":{"reasoning_tokens":363}},"tokens_in":863,"tokens_out":444,"duration_ms":4625,"temperature":1.0,"reasoning_tokens":363,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:29:36.458981+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is polarization-resolved ARPES at the $M$ point just above $T_{\\rm nem}$ with resolution better than the expected splittings: the spin-orbit scenario predicts one doubly degenerate $xz/yz$ doublet and one doubly degenerate $xy$ doublet, whereas the surface-hybridization scenario predicts two split $xz/yz$ modes with equal $xz$ and $yz$ orbital weight. Observing neither the predicted doublet structure nor the two split singlets would falsify the orbital-transmutation mechanism; seeing the two modes actually merge above $T_{\\rm nem}$ would indicate the puzzle was a thermal-broadening artifact.","supporting_citations":[{"cited_title":"Cvetkovic and O","cited_arxiv_id":null,"evidence_quote":"It supplies the symmetry-allowed k·p Hamiltonian and the momentum-independent spin-orbit coupling at the M point that the paper extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the measured spin-orbit coupling of about 20 meV on the hole pockets, the scale that motivates lambda = 10 meV."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It gives the STM data on the peanut-shaped inner electron pocket and its orbital composition used to set the model parameters."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It reports the ARPES behavior of the excitations near M, including the strongly temperature-dependent chemical potential."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It documents the emergence of the nematic electronic state and the persistence of the splitting above T_nem."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It introduces the surface-induced xz/yz hybridization term allowed only at the surface, which is the basis of the second scenario."},{"cited_title":"Fedorov, A","cited_arxiv_id":null,"evidence_quote":"It reports two additional weaker excitations deep in the nematic phase, which the paper assigns to predominantly xy modes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It follows the two excitations from deep in the nematic phase to above T_nem and argues that the residual splitting is physical."}],"review_version":1}