{"id":"6b1c9385-83dc-4cab-be36-9aafc0e2b0aa","arxiv_id":"2505.06449","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The bulk band structure of SmB6 violates the nodal pinning condition required for a topological Kondo insulator, indicating that SmB6 is topologically trivial.","lead":"Measurements of the three-dimensional electronic structure of SmB6 show a temperature-dependent shift of a bulk band at a high-symmetry point that should stay fixed if SmB6 were a topological Kondo insulator. The authors conclude that SmB6 is not the long-sought correlated topological insulator, challenging the current consensus.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Temperature-dependent chemical potential shift is not ruled out; 'pinned at X' refers to hybridization nodes, not absolute ARPES binding energy.","rationale":"The reader's weakest_assumption identifies exactly the gap in the paper's proof: the TKI pinning condition constrains the hybridization at X, not the absolute binding energy of the X feature relative to EF. The paper's own observation of temperature-dependent Sm valence makes a chemical-potential shift plausible, and no independent measurement of EF(T) is provided. This is a genuine logical gap in the central 'directly rules out' claim, but it does not invalidate the valuable experimental data or the alternative trivial explanations. The CONDITIONAL verdict is appropriate: the paper should be accepted only if the authors address the chemical-potential and bare-level shift interpretation, provide error bars for the ~10 meV shift, and temper the 'proof' wording. My independent reading agrees with the reader's assessment, so no verdict change is warranted.","tokens_in":10900,"tokens_out":3730,"duration_ms":40261,"concrete_test":"Measure the chemical potential shift μ(T) between 1 and 50 K using temperature-dependent core-level photoemission (Sm 3d or B 1s) or work-function measurements on the same crystals, then subtract μ(T) from the X-feature binding energy. If the corrected X shift is within experimental error of zero, the observed shift is a chemical-potential effect and does not violate TKI pinning. As a cross-check, impose a temperature-dependent chemical potential (constrained by the measured Sm valence in Fig. 2d) on a published TKI band structure (e.g., Lu et al.) and simulate the ARPES peak at X; if it moves by ~10 meV while the Z2 invariant stays nontrivial, the central falsification collapses.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central falsification rests on the statement in 'Connection Photoemission — 3-D Band Structure': 'the conduction band dispersion must be pinned at X regardless of hybridization strength V.' The argument treats the observed ~10 meV shift of the ARPES feature at X as a direct violation. The load-bearing assumption is that the ARPES binding energy at X measures the bare conduction band dispersion relative to a fixed reference. In a TKI, hybridization vanishes at X by parity, so the hybrid state there is the bare d-band, but its binding energy relative to EF depends on the chemical potential and on the bare d and f level positions, neither of which is fixed by topology. The paper's own XAS data (Fig. 2d) show a temperature-dependent Sm valence that can shift EF and/or the f-level relative to the d-band without changing the parity ordering or the Z2 invariant. A temperature-dependent chemical potential or self-energy shift could move the X feature by ~10 meV while SmB6 remains a TKI. The paper does not independently measure EF(T) or the bare level positions, so the 'directly rules out' conclusion is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports temperature-dependent ARPES, spin-resolved ARPES, and X-ray absorption measurements on SmB6. The central experimental observation is that the three-dimensional conduction band feature at the X point shifts by roughly 10 meV to lower binding energy upon cooling from 50 K to 1 K, while the bulk-like f-derived states at Gamma shift in the opposite direction. The authors simulate the photoemission intensity expected from a temperature-dependent hybridization gap in a topological Kondo insulator and argue that k-perp integration cannot reproduce the observed shift or line shape. From the assumption that a topological Kondo insulator must have a parity-enforced hybridization node at X that pins the conduction band dispersion regardless of hybridization strength, they conclude that the observed shift at X directly rules out SmB6 as a topological Kondo insulator. They further reinterpret the spin-polarized signal from the X surface state as originating from Sm 4f multiplets and propose that the Gamma-centered Fermi contour is a surface d-f hybrid rather than an umklapp or topological surface state.","tokens_in":11102,"tokens_out":7513,"duration_ms":82366,"significance":"If the central claim is correct, the paper would overturn the widely accepted classification of SmB6 as the only known topological Kondo insulator, a result of considerable importance for the search for correlated topological materials. The paper has clear strengths: it provides careful photon-energy-dependent measurements that establish the three-dimensional nature of the X feature, quantitative simulations that rule out a simple k-perp integration artifact, and new spin-resolved data relevant to the surface-state interpretation. The temperature-dependent data and the reinterpretation of the surface states are valuable in their own right. However, the logical step from the observed X-point shift to the conclusion 'directly rules out SmB6 as a topological Kondo insulator' rests on an unstated and unmeasured assumption about the temperature dependence of the chemical potential and the bare band positions. Because that assumption is load-bearing and is not established, the paper's headline conclusion is not yet proven.","major_comments":[{"comment":"","section":"Connection Photoemission — 3-D Band Structure"},{"comment":"","section":"Connection Photoemission — 3-D Band Structure; Fig. S2"}],"minor_comments":[{"comment":"The sentence 'As of today it appears also experimentally established as the only representative of this material class' is in tension with the paper's own conclusion and should be reworded, for example to 'widely regarded as.'","section":"Introduction"},{"comment":"The estimated ~10 meV shift between 50 K and 1 K is a key quantity, but the figure does not report fit uncertainties or the number of independent measurements; please add this information to the figure or caption.","section":"Fig. 2"},{"comment":"The main text refers to Fig. S2 for the central k-perp integration argument without summarizing its content; a brief description of the simulation geometry and parameters should be given in the main text or directly in the supplementary material.","section":"Fig. S2"},{"comment":"The sentence 'The higher binding energy valleys around X yield a plano-convex shape with the flat edge at the highest binding energy' is difficult to follow; a labeled schematic of the expected versus observed line shapes would improve clarity.","section":"Connection Photoemission — 3-D Band Structure"},{"comment":"Reference [20] is cited as an arXiv preprint from 2013; if a peer-reviewed version exists, please update the citation.","section":"References"},{"comment":"The claim that the spin polarization of a topological surface state 'must disappear with the closing of the hybridization gap' is an assertion that is not justified or referenced; adding a justification or citation would strengthen the argument.","section":"Surface states"}],"recommendation":"major_revision","confidential_remarks":"The manuscript addresses a high-stakes question and contains valuable data, but the central inference has a gap that is fixable in revision: the temperature dependence of the chemical potential and bare band positions must be addressed, either by new measurements or by a clear statement of the additional assumption and its plausibility. The paper would benefit from separating the robust experimental facts from the interpretive step leading to 'directly rules out.' The Discussion's use of the Alexandrov valence phase diagram also depends on the validity of that model, which is not central but should be labeled as such."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper argues that the temperature-dependent shift of the bulk X-point conduction band in SmB6 falsifies the topological Kondo insulator picture. The new data and the analysis are valuable, but the central 'proof' has a gap the authors don't close.\n\nWhat's genuinely new: the shift at X was seen before by Min et al. and Denlinger et al., but this is the first time it's used as a systematic falsification test of the TKI nodal-pinning condition. The paper adds new spin-resolved ARPES data that make a credible case that the reported surface spin polarization is dominated by Sm 4f multiplet emission, not a topological surface state. It also reinterprets the Gamma-centered Fermi surface feature as a surface d-f hybrid rather than an umklapp copy, which is a nice move. The k⊥-integration simulation is well constructed and convincingly rules out an instrumental artifact.\n\nThe soft spot is the central inference. The argument assumes that the ARPES binding energy at X directly measures the bare d-band energy relative to a fixed reference. In a TKI, hybridization at X vanishes by parity, so the state there is the bare d-band, but its binding energy relative to EF depends on the chemical potential and on the bare d and f level positions. Neither is fixed by topology. The paper doesn't independently measure EF(T) or the bare level positions. Its own XAS data show a temperature-dependent Sm valence; even if the valence is roughly constant below 50 K, the ±0.05 systematic uncertainty leaves room for a meV-level shift. A band-dependent self-energy shift, common in Kondo systems, could move the X feature without changing the parity ordering. The observed non-rigid shift between X and Gamma rules out a simple rigid EF shift, but it doesn't eliminate the self-energy alternative. So 'directly rules out' is not supported; at best it's 'strongly suggests.'\n\nMinor points: the ~10 meV total shift is partly extrapolated (they directly see about 5 meV); the title's 'Proof' overstates the case. These are fixable.\n\nOverall: this is a serious experimental contribution that deserves referee time. The data are good, the question is important, and the challenge to the SmB6 TKI consensus is legitimate. The authors need to address the chemical-potential/self-energy issue or soften the claim. I'd send it to peer review, not desk reject.","headline":"Strong experimental challenge to the SmB6 topological Kondo insulator consensus, but the central 'proof' leaves the chemical-potential/self-energy door open.","tokens_in":11727,"tokens_out":7298,"would_cite":true,"duration_ms":74414,"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":"This paper claims that the temperature-driven shift of the three-dimensional conduction band at the $X$ point of samarium hexaboride violates the parity-pinning condition required of a topological Kondo insulator, and therefore that…","keywords":["samarium hexaboride","topological Kondo insulator","strongly correlated electrons","ARPES","hybridization node","parity","surface states","Sm valence"],"falsifier":"Measure the temperature-dependent $X$-point conduction-band binding energy while following a non-dispersive bulk core level that tracks the chemical potential: if the observed $\\sim10$ meV downward shift is fully accounted for by a chemical-potential shift, so that the dispersion at $X$ is actually pinned, the paper's central claim collapses. The converse control is a band calculation that preserves odd-parity hybridization nodes at $X$ and still reproduces the measured non-rigid shift.","tokens_in":10717,"feed_emoji":"⚛️","tokens_out":12268,"duration_ms":109398,"temperature":0.7,"pith_summary":"For decades samarium hexaboride (SmB$_6$) has been the only widely accepted candidate for a topological insulator driven by strong electron correlation. This paper argues that its three-dimensional electronic structure disproves that status. In a topological Kondo insulator, hybridization between opposite-parity $d$ and $f$ states must vanish at high-symmetry points, so the conduction band at the $X$ point is pinned to its unmixed constituents regardless of hybridization strength. The authors measure a shift of roughly 10 meV in that $X$-point band as the sample cools from 50 K to 1 K, together with an opposite shift at $\\Gamma$--a non-rigid band shift that the pinning condition forbids. They conclude that SmB$_6$ is either a nodeless hybridization insulator or an insulator produced by a mechanism unrelated to hybridization; either way it is not a topological Kondo insulator.","feed_headline":"X-point band shift rules out SmB6 as a topological Kondo insulator","feed_subtitle":"The 3D conduction band at X moves with temperature; a true Kondo TI must pin it at all hybridization strengths.","key_machinery":"The load-bearing condition is the parity-odd hybridization node. In a topological Kondo insulator, the $d$-$f$ hybridization term is odd under parity because it couples even-parity itinerant $d$ states to odd-parity localized $f$ states; being periodic in the reciprocal lattice, it must vanish at high-symmetry points. This pins the $X$-point conduction band to its unmixed position for every hybridization strength $V$, converting a topological invariant into a directly measurable rule: the $X$-point binding energy must not move with temperature. The paper tests this rule with temperature-dependent ARPES, and to close the loophole of imperfect momentum resolution it simulates the photoemission intensity for a $k_\\perp$-integrated, weakly hybridized W-shaped band. That simulated plano-convex shape is not seen experimentally, so the observed shift must be a genuine motion of the dispersion at $X$.","core_discovery":"The paper's central claim is that the three-dimensional band structure of SmB$_6$ fails the defining test of a topological Kondo insulator. Because the hybridization that opens the gap couples opposite-parity states, the hybridization matrix element must be odd under parity and therefore vanish at high-symmetry points; at $X$ the hybrid bands remain pinned to their bare $d$- and $f$-level positions no matter how strong or weak the hybridization $V$ is. Temperature-dependent ARPES shows instead that the bulk conduction-band feature at $X$ shifts by roughly 10 meV to lower binding energy between 50 K and 1 K, while bulk-like $f$ states near $\\Gamma$ move the opposite way. The authors rule out a W-shaped to U-shaped hybridization change (which would move spectral weight even with the node) by simulating $k_\\perp$-integrated photoemission: the expected plano-convex contours are not observed. They conclude that SmB$_6$ is topologically trivial, either with nodeless hybridization or with a non-rigid shift unrelated to hybridization, and the same data are used to reinterpret the spin-polarized photoemission and the $\\bar{\\Gamma}$ surface feature as non-topological $d$-$f$ surface hybrids.","pith_inferences":["Inference beyond the paper: the parity-node pinning test transfers to any other inversion-symmetric Kondo insulator candidate; if its $X$-point (or equivalent time-reversal-invariant momentum) band shifts with temperature or doping, the topological Kondo insulator assignment should be rejected no matter how convincing its surface states look.","Inference beyond the paper: if the non-rigid shift reflects a temperature-dependent chemical potential or Sm valence rather than a hybridization change, then binding-energy shifts measured by ARPES in other mixed-valent $f$-electron compounds should be referenced to a core level before being read as band-structure changes.","Inference beyond the paper: applying uniaxial strain or chemical pressure to vary hybridization at fixed temperature would give a cleaner test--if the $X$-point band follows the hybridization strength, the gap is nodeless and trivial, whereas if it stays pinned while the gap changes, the topological Kondo insulator scenario survives.","Inference beyond the paper: the strategy inverts the usual search order for correlated topological insulators--screen candidate materials with a quick three-dimensional band-structure test before investing in surface-state spin or tunneling measurements."],"forward_implications":["If the claim holds, SmB$_6$ cannot serve as the sole experimental confirmation of a correlated topological insulator; the search must restart from candidates whose three-dimensional bands pass the pinning test.","The surface features at $\\bar{X}$ and $\\bar{\\Gamma}$, previously cited as topological surface states, are reinterpreted as two-dimensional surface $d$-$f$ hybrids, removing the even-number-of-Dirac-cones problem at $\\bar{\\Gamma}$.","The spin polarization reported for the $\\bar{X}$ state is reassigned to the atomic Sm $4f$ multiplet structure, so spin-resolved ARPES on this material does not demonstrate spin-momentum locking.","The measured temperature-dependent Sm valence places SmB$_6$ on the trivial side of the predicted topological phase diagram, in line with the trivial conclusion.","Theoretical calculations that unanimously predict a topological Kondo insulator for SmB$_6$ are left facing an unexplained discrepancy with the measured three-dimensional band structure."],"supporting_citations":[{"why":"Introduces the topological Kondo insulator concept: band inversion between even-parity itinerant $d$ states and odd-parity localized $f$ states.","marker":"[10]"},{"why":"Supplies the parity argument that hybridization must vanish at high-symmetry points, which is the pinning condition the experiment tests.","marker":"[31]"},{"why":"One of the core TKI calculations: first published surface band structure, with a small inverted gap, used as a comparison target.","marker":"[12]"},{"why":"Provides the Fermi surface and band structure prediction with W-shaped to U-shaped hybridization evolution that the paper's simulation rules out.","marker":"[13]"},{"why":"Earlier ARPES study reporting a rigid parallel band shift on cooling; the paper's non-rigid shift measurement directly contradicts that interpretation.","marker":"[20]"},{"why":"Earlier ARPES report of the bulk conduction band near $X$ at elevated temperature, which the paper extends to a temperature-dependent shift.","marker":"[21]"},{"why":"Prior work establishing the trivial surface-hybrid explanation for the $\\bar{X}$ and $\\bar{\\Gamma}$ features that this paper builds on.","marker":"[23]"},{"why":"Supplies the topological phase diagram as a function of Sm valence; the measured low-temperature valence places SmB$_6$ on the trivial side.","marker":"[11]"},{"why":"Provides the $L_3$-edge Sm valence temperature dependence that the authors compare with their $M_5$-edge measurements.","marker":"[34]"},{"why":"Reported spin texture of the $\\bar{X}$ surface state; the paper's spin-resolved data reinterpret that signal as Sm $4f$ multiplet polarization.","marker":"[22]"}],"fun_headline_variants":["SmB6 band shift at X: topological Kondo insulator ruled out","Nodeless hybridization shows SmB6 is trivial","X-point movement kills SmB6 topological insulator claim","SmB6 fails nodeless test for topological Kondo insulator","Hybridization node missing: SmB6 topologically trivial"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that the pinning of the $X$-point conduction band to its unmixed constituents holds for the quantity ARPES actually measures--binding energy relative to the Fermi level--so that a chemical-potential shift or a temperature-induced shift of the bare $d$- and $f$-levels cannot move the $X$ feature while preserving the parity ordering and the $\\mathbb{Z}_2$ invariant.","fun_headline_variants_meta":{"raw":{"variants":["SmB6 band shift at X: topological Kondo insulator ruled out","Nodeless hybridization shows SmB6 is trivial","X-point movement kills SmB6 topological insulator claim","SmB6 fails nodeless test for topological Kondo insulator","Hybridization node missing: SmB6 topologically trivial"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000204,"raw_usage":{"total_tokens":1365,"prompt_tokens":897,"completion_tokens":468,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":383}},"tokens_in":513,"tokens_out":468,"duration_ms":4788,"temperature":1.0,"reasoning_tokens":383,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:42:54.027710+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the temperature-dependent $X$-point conduction-band binding energy while following a non-dispersive bulk core level that tracks the chemical potential: if the observed $\\sim10$ meV downward shift is fully accounted for by a chemical-potential shift, so that the dispersion at $X$ is actually pinned, the paper's central claim collapses. The converse control is a band calculation that preserves odd-parity hybridization nodes at $X$ and still reproduces the measured non-rigid shift.","supporting_citations":[{"cited_title":"Dzero, K","cited_arxiv_id":null,"evidence_quote":"Introduces the topological Kondo insulator concept: band inversion between even-parity itinerant $d$ states and odd-parity localized $f$ states."},{"cited_title":"Takimoto, J","cited_arxiv_id":null,"evidence_quote":"One of the core TKI calculations: first published surface band structure, with a small inverted gap, used as a comparison target."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Fermi surface and band structure prediction with W-shaped to U-shaped hybridization evolution that the paper's simulation rules out."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier ARPES report of the bulk conduction band near $X$ at elevated temperature, which the paper extends to a temperature-dependent shift."},{"cited_title":"Hlawenka, K","cited_arxiv_id":null,"evidence_quote":"Prior work establishing the trivial surface-hybrid explanation for the $\\bar{X}$ and $\\bar{\\Gamma}$ features that this paper builds on."},{"cited_title":"Alexandrov, M","cited_arxiv_id":null,"evidence_quote":"Supplies the topological phase diagram as a function of Sm valence; the measured low-temperature valence places SmB$_6$ on the trivial side."},{"cited_title":"Mizumaki, S","cited_arxiv_id":null,"evidence_quote":"Provides the $L_3$-edge Sm valence temperature dependence that the authors compare with their $M_5$-edge measurements."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reported spin texture of the $\\bar{X}$ surface state; the paper's spin-resolved data reinterpret that signal as Sm $4f$ multiplet polarization."}],"review_version":1}