{"id":"88dc6fc5-94f1-455d-a44c-915c4634eac3","arxiv_id":"2505.07250","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"Magnetic proximity from FeAs/FeOx induces band inversion and a topological surface state in a 3-nm α-Sn film that is trivial by itself, as seen in Shubnikov-de Haas oscillations and DFT calculations.","lead":"A ferromagnetic layer placed next to a thin film of ordinary α-Sn turns the film into a topological material, creating a high-mobility two-dimensional surface band that is absent without the magnet. The result suggests a general recipe for making magnetic topological materials from trivial narrow-gap semiconductors.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No nonmagnetic control calculation: the DFT shows band inversion in ferromagnetic FeAs/α-Sn, but does not show that magnetism causes it.","rationale":"Reading the paper in good faith, the experiment is substantial: STEM and EDX confirm the intended heterostructure, SQUID shows ferromagnetism, SdH analysis yields two 2D bands with distinct mobilities and phase shifts, and the reference α-Sn sample shows only one lower-mobility band. The DFT calculation does show a surface-state-like green band in the inverted gap of the ferromagnetic heterostructure. These are real supporting results. The load-bearing weakness is the causal attribution to MPE. The paper never compares the ferromagnetic heterostructure against a nonmagnetic version of the same interface, so the band inversion is not shown to be a consequence of exchange splitting rather than of the chemical and structural proximity of the FeAs layer. This is distinct from, though adjacent to, the reader's flagged concern about U = -2.5 eV and the assumed ferromagnetic order: even if the magnetic order and U value are correct, the absence of a nonmagnetic control leaves the central 'by MPE' claim untested. A single constrained-moment DFT run would settle this. The reader's CONDITIONAL verdict remains appropriate, provided the conditions explicitly include this control and an independent, non-circular Fermi-level determination; therefore I do not change the verdict.","tokens_in":16047,"tokens_out":7686,"duration_ms":80784,"concrete_test":"Repeat the 1-ML FeAs/18-ML α-Sn slab calculation with the Fe local moments constrained to zero (nonmagnetic), using the same U = -2.5 eV, spin-orbit coupling, -0.76% strain, 8×8×1 k-mesh, and H termination. If the Γ-point band inversion and the green projected surface band persist in the nonmagnetic control, the topological transition is not caused by MPE and the abstract's causal claim is not supported. If a stable nonmagnetic state cannot be obtained, repeat the calculation with the Fe exchange splitting scaled continuously (e.g., 0%, 50%, 100% of the self-consistent value) and check whether the band-inversion gap closes and reopens as the exchange splitting increases.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is causal: magnetic proximity from FeAs/FeOx inverts the bands of a 3-nm α-Sn film and thereby creates a topological surface state. In the first-principles section (Methods, 'First-principles calculations'; Fig. 4f-g), the only heterostructure calculation shown has the FeAs monolayer in a fixed ferromagnetic state, with magnetization along [001] or [1-10]. There is no control calculation with the Fe moments constrained to zero, or with the exchange splitting artificially suppressed. Without such a control, the observed Γ-point band inversion cannot be attributed specifically to the magnetic proximity effect: it could instead arise from the nonmagnetic interface potential and orbital hybridization between the FeAs monolayer and α-Sn. The summary itself hedges by saying the transition occurs 'due to MPE and orbital hybridization,' but the paper's novelty is that a trivial semiconductor becomes topological specifically by MPE. This missing reference calculation leaves the causal claim unsupported. The robustness of GGA+U with U = -2.5 eV and the assumed 1-ML FeAs ferromagnetic order are secondary: even if those are correct, the band inversion may be nonmagnetic in origin.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a FeOx/FeAs/α-Sn(3 nm)/InSb heterostructure in which the α-Sn layer alone is a trivial narrow-gap semiconductor, and it claims that magnetic proximity from the FeAs/FeOx layers induces band inversion and a topological surface state (TSS). Shubnikov–de Haas measurements on the heterostructure reveal two two-dimensional bands: a low-frequency band (FLow = 12.3 T) with high mobility (28,900 cm²/Vs) and a nearly linear phase shift (γ = 0.28), assigned to a TSS, and a high-frequency band (FHigh = 33.4 T) with lower mobility and quadratic dispersion, assigned to a heavy-hole band. DFT calculations with GGA+U and ferromagnetic 1-ML FeAs show band inversion at Γ and a surface state localized on the vacuum side of the slab, which the authors equate with the α-Sn/InSb interface. The paper concludes that this is a demonstration of a trivial-to-topological transition driven by the magnetic proximity effect.","tokens_in":16270,"tokens_out":7192,"duration_ms":73851,"significance":"If the central claim holds, the work would be significant: it would show that a magnetic topological material can be obtained from a trivial narrow-gap semiconductor by proximity exchange, avoiding the disorder introduced by magnetic doping. The transport analysis is careful, with internal cross-checks via Dingle and fan diagrams (Supplementary Notes 1 and 3), angle-dependent SdH measurements, and a plausible assignment of the two observed frequencies. The DFT calculations include explicit orbital projections and magnetization-direction dependence, which give a microscopic picture. However, two load-bearing issues currently prevent the claim from being fully supported: the absence of a nonmagnetic control calculation for the heterostructure, and the circular placement of the DFT Fermi level from the same SdH frequency that is then used to claim experiment-theory agreement. These issues are fixable within the manuscript's scope, but they must be addressed before the topological conclusion can be accepted.","major_comments":[{"comment":"The central causal claim that the magnetic proximity effect (MPE) induces the band inversion is not established because no nonmagnetic control calculation is shown. The only heterostructure calculation has the FeAs monolayer in a fixed ferromagnetic state; to attribute the Γ-point band inversion to exchange splitting, the authors need to also compute the band structure with the Fe moments constrained to zero (or with the exchange splitting artificially suppressed) and show that the inversion disappears. The manuscript's own summary ('due to MPE and orbital hybridization') concedes that hybridization is also involved, so without this control the band inversion could be a nonmagnetic interface effect. The sensitivity of the result to the choice U = −2.5 eV and to the assumed FeAs magnetic order should also be reported, since the topological conclusion depends on them.","section":"Methods, 'First-principles calculations'; Fig. 4f-g"},{"comment":"The claimed experiment-theory agreement on the distance between EF and the Dirac point is circular. The DFT Fermi level is placed at the pink dashed line using Eq. (S3) with the SdH frequency FLow; the same quantity is then quoted as agreement between experiment and theory. This cannot serve as independent validation of the FLow = TSS assignment. The assignment may still be plausible from the phase shift, mobility, and 2D character, but the 'good agreement' sentence in the main text should be removed or replaced by a genuinely independent comparison, such as a prediction of a quantity not used to set EF.","section":"Supplementary Note 5; Fig. 4f-g"},{"comment":"The DFT slab represents the α-Sn/InSb interface as a vacuum surface (right panels of Fig. 4f-g), yet the TSS is claimed to form at that interface in the experiment. InSb is a narrow-gap semiconductor with its own electronic states, not a vacuum barrier, and hybridization with the substrate could modify or destroy the surface state. The calculation therefore does not directly demonstrate that a TSS survives at the actual α-Sn/InSb interface; an explicit treatment of the InSb substrate, or at least a quantitative discussion of why the vacuum approximation is justified, is needed.","section":"Fig. 4f-g and Methods"}],"minor_comments":[{"comment":"The text states that appending zeros 'increase[s] the number of data points' to 'recover the resolution'; zero padding does not increase the intrinsic frequency resolution of the FFT, it only interpolates the spectrum. Please rephrase this sentence.","section":"Methods, 'Analysis of SdH oscillations'"},{"comment":"The statement that 'γ is 0 or 1 for a linear dispersion and 0.5 for a quadratic one' could be clarified, since γ and γ+1 are equivalent modulo the cosine period; specify the convention, such as whether a Berry phase of π corresponds to γ = 0.","section":"Eq. (1) and surrounding text"},{"comment":"The text states that FIn-plane is independent of the magnetic field angle, indicating 3D characteristics, but it appears only beyond 45 degrees and is attributed to the InSb layer; it would help to state explicitly in the main text that this component is from the InSb buffer and not from α-Sn.","section":"Fig. 4e and main text"},{"comment":"The use of 'U = −2.5 eV' as the on-site Coulomb potential is unusual; please clarify the sign convention and justify the negative value, or correct a possible typo, since this parameter directly affects the calculated band inversion.","section":"Methods, 'First-principles calculations'"},{"comment":"The SQUID magnetization curve is shown without a zero-field guide or quantitative values; the coercivity and saturation magnetization should be stated in the text or caption to substantiate the claim of ferromagnetic hysteresis behavior.","section":"Fig. 2c"}],"recommendation":"major_revision","confidential_remarks":"The central idea is attractive and the transport data are careful, but the missing nonmagnetic control calculation and the circular Fermi-level placement are exactly the kinds of issues that need to be resolved before publication. I lean toward major revision rather than rejection because both issues are fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — quick take on arXiv:2505.07250. The paper claims that magnetic proximity from a FeAs/FeOx bilayer drives a 3-nm α-Sn film (trivial alone) into a topological phase, with a TSS seen in transport as a high-mobility linear 2D band. That would be a genuinely new result: earlier work proposed the mechanism, but this is the first transport demonstration in a trivial α-Sn film, and the specific heterostructure is new.\n\nWhat the paper does well: the SdH characterization is careful. The reference 3-nm α-Sn is measured, shows one frequency with γ ≈ 0.97; the heterostructure shows two frequencies, with angular dependence 1/cosθ confirming 2D character, Dingle and fan plots cross-checking μ and γ, and a plausible assignment of FLow to the TSS and FHigh to the HH band. The OMR observation (~100% at 1 T) is a nice extra and is consistent with time-reversal breaking in the α-Sn layer. The DFT does show band inversion and surface states in the FeAs/α-Sn slab, with a clear change with magnetization direction.\n\nThe soft spots are real. The biggest one is the one in the stress-test note: there is no control DFT calculation with the Fe moments forced to zero or exchange suppressed. The only heterostructure calculation has FeAs ferromagnetic, so the band inversion could in principle come from the nonmagnetic interface potential and orbital hybridization rather than from the magnetic proximity effect. The paper's own summary says \"due to MPE and orbital hybridization,\" which is honest but does not isolate the magnetic contribution. This is fixable: run the same slab with nonmagnetic FeAs or with U applied to suppress moments.\n\nSecond is the circular EF placement that the reader flags. Supplementary Note 5 says the Fermi level is set from the SdH frequency of FLow via the linear-dispersion relation; then Fig. 4f/g show EF at -0.25 eV and the distance from EF to the Dirac point is quoted as agreement between experiment and theory. That particular comparison is circular for those quantities. It does not invalidate the band inversion or the TSS assignment, but it should be reframed.\n\nMinor: no error bars on μ and γ from the LK fits, no deposited data (only \"on request\"), and the DFT relies on a hand-picked U = -2.5 eV with no sensitivity check. None of these sink the paper, but they add uncertainty.\n\nOverall: the core experimental observation — a new high-mobility 2D linear band in the heterostructure that is absent in the reference — is solid and valuable. The topological interpretation is plausible but not yet proven. I'd send this to a serious referee, with the expectation of major revision: add the nonmagnetic control, fix the circularity in the EF comparison, and release the data. If the control calculation still shows inversion without moments, the claim should be scaled back to \"interface-induced\" rather than \"magnetic-proximity-induced.\" Would I cite it? I'd cite the transport data, but I'd hold off on citing the topological phase transition until the control appears.","headline":"A careful transport study of a plausible MPE-induced topological phase in α-Sn, but the causal role of magnetism needs a control calculation and the EF comparison is partly circular.","tokens_in":16885,"tokens_out":2945,"would_cite":true,"duration_ms":27817,"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":"Magnetic proximity from FeAs/FeOx turns a trivial 3-nm α-Sn film into a topological material with a high-mobility surface state.","keywords":["magnetic proximity effect","topological surface state","α-Sn","band inversion","Shubnikov–de Haas oscillations","narrow-gap semiconductor","ferromagnetic heterostructure","odd-parity magnetoresistance"],"falsifier":"Recompute the FeAs/α-Sn slab with the correlation correction set to zero and with the FeAs magnetization rotated or disordered; if the band inversion at the Γ point disappears, the claimed phase transition is an artifact of the calculation parameters. On the transport side, a gate-voltage sweep that moves the Fermi level through the inverted gap should change the $F_\\mathrm{low}$ SdH component's frequency and Berry phase in a way consistent with a Dirac surface state.","tokens_in":15804,"feed_emoji":"🧲","tokens_out":14399,"duration_ms":124162,"temperature":0.7,"pith_summary":"This paper attempts to establish that a topological material is not a necessary starting point for magnetic topological electronics. It argues that placing a ferromagnetic FeAs/FeOx stack against a 3-nm film of α-Sn, which by itself is a trivial narrow-gap semiconductor, induces band inversion and a topological surface state through the magnetic proximity effect. The experimental evidence is a high-mobility ($\\mu = 28{,}900\\ \\mathrm{cm^2\\,V^{-1}s^{-1}}$), quasi-two-dimensional, nearly linear band observed in Shubnikov–de Haas oscillations that is absent in the α-Sn reference sample, together with first-principles calculations showing band inversion at the Γ point. If correct, the result would let many ordinary narrow-gap semiconductors be converted into magnetic topological materials without magnetic doping.","feed_headline":"Magnetic neighbor turns a trivial film into a topological material","feed_subtitle":"A 3-nm α-Sn film, trivial on its own, hosts a high-mobility linear surface state when capped with FeAs/FeOx.","key_machinery":"The load-bearing mechanism is the magnetic proximity effect: the exchange field of the adjacent ferromagnetic FeAs layer spin-splits the bands of α-Sn, and the material's strong spin-orbit coupling then inverts the s/p band order around the Γ point, creating a topological gap and a surface state. The transport analysis runs on Lifshitz–Kosevich theory, whose phase shift $\\gamma$ distinguishes linear ($\\gamma \\approx 0$) from quadratic ($\\gamma \\approx 0.5$) bands and whose angular dependence $F \\propto 1/\\cos\\theta$ establishes two-dimensionality. The DFT slab calculation with a GGA+U correction identifies which side of the slab hosts the surface state and how its Dirac cone responds to the magnetization direction.","core_discovery":"The central claim is that exchange splitting from a neighboring ferromagnet, combined with the strong spin-orbit coupling of α-Sn, inverts the s- and p-like bands of a 3-nm α-Sn film that is topologically trivial on its own. The inverted band structure supports a topological surface state on the vacuum/InSb side of the α-Sn slab, while the FeAs/Sn interface states hybridize with Fe d orbitals and are pushed away from Γ. Transport shows two two-dimensional SdH components: $F_\\mathrm{low} = 12.3\\ \\mathrm{T}$ with $\\mu = 28{,}900\\ \\mathrm{cm^2\\,V^{-1}s^{-1}}$ and $\\gamma = 0.28$, assigned to the surface state, and $F_\\mathrm{high} = 33.4\\ \\mathrm{T}$ with lower mobility and $\\gamma = 0.42$, assigned to the heavy-hole band. DFT further shows that the surface state's Dirac point opens a small exchange gap when the FeAs magnetization points along [001] and shifts in $k_x$ when the magnetization lies in-plane, consistent with a massive Dirac surface state under broken time-reversal symmetry. The paper concludes that the transition from a trivial to a nontrivial phase in α-Sn is driven by magnetic proximity rather than by film thickness or doping.","pith_inferences":["A systematic thickness study of α-Sn under the same FeAs/FeOx cap should map where the proximity-induced band inversion turns on and off, sharpening the predicted phase boundary beyond the single 3-nm point.","Electrostatic gating of the heterostructure is a test the paper does not report: sweeping the Fermi level through the inverted gap should change the $F_\\mathrm{low}$ SdH frequency and Berry phase in a way unique to a Dirac surface state.","The paper leaves the origin of the complex angular dependence of the odd-parity magnetoresistance explicitly unexplained; if that signal is intrinsic to the α-Sn surface, it could become a complementary probe of the same proximity-induced gap, but that link is not established.","The predicted sensitivity of the surface state to the FeAs magnetization direction implies that samples with multidomain FeAs should show broadened or absent $F_\\mathrm{low}$ oscillations; single-domain samples would be required for the cleanest test."],"forward_implications":["A 3-nm α-Sn film, which is topologically trivial on its own, becomes a host of a topological surface state when interfaced with ferromagnetic FeAs/FeOx, so a thick topological host is not required.","Magnetic proximity preserves the high carrier mobility of the surface band ($28{,}900\\ \\mathrm{cm^2\\,V^{-1}s^{-1}}$), avoiding the disorder that magnetic doping typically introduces.","Rotating the FeAs magnetization should controllably open or shift the surface-state Dirac cone, giving a handle for switching topological transport properties.","The same recipe — exchange field plus strong spin-orbit coupling in a narrow-gap semiconductor — should generalize to other trivial narrow-gap semiconductors, expanding the pool of candidate magnetic topological materials.","The observed ~100% odd-parity magnetoresistance under in-plane field is presented as independent evidence of time-reversal breaking by the magnetic proximity effect in this heterostructure."],"supporting_citations":[{"why":"Establishes the α-Sn/InSb platform, the -0.76% compressive strain, the thickness-driven transition from trivial 3-nm film to Dirac semimetal, and the high quantum mobility of TSS in thicker α-Sn; the reference sample properties and strain parameters rest on it.","marker":"[13]"},{"why":"Shows that a zinc-blende FeAs monolayer embedded in a semiconductor is ferromagnetic, which is the premise for expecting magnetic proximity from the 1-ML FeAs layer.","marker":"[25]"},{"why":"Provides the theoretical proposal that exchange splitting larger than a narrow gap can make trivial narrow-gap semiconductors topologically nontrivial, the mechanism this paper applies.","marker":"[11]"},{"why":"Supplies the Lifshitz-Kosevich theory used to fit the SdH oscillations and extract cyclotron masses, quantum mobilities, and phase shifts.","marker":"[26]"},{"why":"Used to identify the exchange gap that opens at the Dirac point of a surface state when time-reversal symmetry is broken.","marker":"[30]"},{"why":"Provides the GGA+U parameter value and validates its description of α-Sn electronic structure, grounding the DFT method behind the band-inversion calculation.","marker":"[36]"},{"why":"Confirms that strained α-Sn on InSb(001) is an elemental topological insulator in DFT, validating the method's ability to reproduce topological phase transitions.","marker":"[40]"}],"fun_headline_variants":["Magnetic proximity flips a trivial film into a topological one","Trivial α-Sn turns topological via magnetic neighbor","Exchange splitting creates topological surface state in α-Sn","Band inversion without a topological starting material","Magnetic cap turns α-Sn into a topological surface state"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The band inversion and surface state predicted by the calculation depend on a specific electronic-correlation correction and on the assumption that the FeAs layer is ferromagnetic with a fixed magnetization direction; if either is wrong, the predicted topological phase may not exist in the real sample.","fun_headline_variants_meta":{"raw":{"variants":["Magnetic proximity flips a trivial film into a topological one","Trivial α-Sn turns topological via magnetic neighbor","Exchange splitting creates topological surface state in α-Sn","Band inversion without a topological starting material","Magnetic cap turns α-Sn into a topological surface state"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00052,"raw_usage":{"total_tokens":2565,"prompt_tokens":1036,"completion_tokens":1529,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":652,"completion_tokens_details":{"reasoning_tokens":1452}},"tokens_in":652,"tokens_out":1529,"duration_ms":11404,"temperature":1.0,"reasoning_tokens":1452,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:21:20.909265+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the FeAs/α-Sn slab with the correlation correction set to zero and with the FeAs magnetization rotated or disordered; if the band inversion at the Γ point disappears, the claimed phase transition is an artifact of the calculation parameters. On the transport side, a gate-voltage sweep that moves the Fermi level through the inverted gap should change the $F_\\mathrm{low}$ SdH component's frequency and Berry phase in a way consistent with a Dirac surface state.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the α-Sn/InSb platform, the -0.76% compressive strain, the thickness-driven transition from trivial 3-nm film to Dirac semimetal, and the high quantum mobility of TSS in thicker α-Sn; the reference sample properties and strain parameters rest on it."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows that a zinc-blende FeAs monolayer embedded in a semiconductor is ferromagnetic, which is the premise for expecting magnetic proximity from the 1-ML FeAs layer."},{"cited_title":"& Nomura, K","cited_arxiv_id":null,"evidence_quote":"Provides the theoretical proposal that exchange splitting larger than a narrow gap can make trivial narrow-gap semiconductors topologically nontrivial, the mechanism this paper applies."},{"cited_title":"Magnetic Oscillations in Metals","cited_arxiv_id":null,"evidence_quote":"Supplies the Lifshitz-Kosevich theory used to fit the SdH oscillations and extract cyclotron masses, quantum mobilities, and phase shifts."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Used to identify the exchange gap that opens at the Dirac point of a surface state when time-reversal symmetry is broken."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the GGA+U parameter value and validates its description of α-Sn electronic structure, grounding the DFT method behind the band-inversion calculation."},{"cited_title":"↑” and “↓","cited_arxiv_id":null,"evidence_quote":"Confirms that strained α-Sn on InSb(001) is an elemental topological insulator in DFT, validating the method's ability to reproduce topological phase transitions."}],"review_version":1}