{"id":"8a96ba03-d2ee-4743-a949-aec1b24cf5ac","arxiv_id":"2412.07993","paper_version":3,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"An asymmetric, flux-biased topological insulator nanowire SQUID is predicted to host Majorana zero modes over a wide, chemical-potential-independent range of magnetic flux.","lead":"A new device wraps a topological insulator nanowire in superconductors to form a tiny SQUID, and theory predicts that a magnetic field can drive it into a topological superconducting phase with Majorana particles at its ends. Experiments show the supercurrent flows only through the nanowire's surfaces, the key requirement for the predicted phase.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The topological phase is inferred from the two-junction model (Eq. 1) without computing a topological invariant or MZMs in the microscopic model; the top/bottom decoupling is the key unmeasured premise.","rationale":"I assessed the paper in good faith: the experiment is a clean demonstration of a surface-dominated columnar nano-SQUID, with five devices, period-area matching, gate tunability, and a Dirac-point signature; the theoretical idea is physically appealing and builds on established Fu-Kane physics. I find no internal inconsistency in the phase-winding argument itself; the asymmetry mechanism in Eq. (1) does put the weaker junction in the Fu-Kane window over the claimed flux range. My stress-test concern is that the step from this two-junction picture to the topological phase of the real device is not demonstrated at the level of a topological invariant. Supplement M's min|ε(k)| maps are necessary but not sufficient: a finite excitation gap is compatible with a topologically trivial BdG band structure. Because the device has two parallel channels that may reconnect at the front/back surfaces, the combined Pfaffian invariant must be computed, and the decoupling of top and bottom surfaces—the reader's flagged assumption—is exactly what makes the two-channel reduction legitimate. The data provide indirect support (no h/e even-odd effect), and the supplement's estimate suggests ξ_s < 15 nm, but neither measures ξ_s nor proves the invariant. These are addressable with a public Kwant model and would settle the question. I therefore agree with the reader's CONDITIONAL verdict; my concern sharpens the condition: compute the invariant and, if possible, measure ξ_s or its fingerprints.","tokens_in":24837,"tokens_out":20518,"duration_ms":207873,"concrete_test":"Using the Kwant BdG model of Eq. (2), implement an explicit top/bottom asymmetry (e.g., different pairing magnitudes or different junction transparencies for the two contacts) and compute the class-D Z2 Pfaffian invariant as a function of Φ together with the lowest-energy spectrum of a finite-length nanowire. Run this for the decoupled-junction parameters (large Δ, N=3 side-only pairing) and for the coherent-subband parameters (small Δ), and repeat with a few side-surface disorder realizations. If the Pfaffian is nontrivial and two localized zero-energy end states appear for (n−1/2)Φ_s^0 < Φ < (n+1/2)Φ_s^0 and vanish outside for both parameter sets, the central claim is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central prediction—flux-window topological superconductivity for (n−1/2)Φ_s^0 < Φ < (n+1/2)Φ_s^0 with odd n—rests on reducing the device to two independent Fu-Kane line junctions whose phase biases are set by Eq. (1), with the weaker junction driven through the π phase transition while the stronger junction stays near zero. That reduction requires the top and bottom surface junctions to be decoupled around the nanowire circumference; the only support is the Supplement L estimate that the Ar-etched side-surface coherence length ξ_s is much shorter than the 15 nm nanowire height, and that length is not measured. The absence of the h/e even-odd component in the measured Ic(Bx) oscillations is indirect evidence for decoupling, but it does not determine ξ_s. Independently, Section III verifies the topological region only through Eq. (1) and the single-channel gap formula E_gap ∝ cos(φ/2); the microscopic tight-binding simulations (Supplement M) report min|ε(k)| but never compute the class-D Z2 Pfaffian invariant or show localized Majorana end states in a finite wire. A gapped spectrum alone does not establish a nontrivial topological phase, and the two-channel structure with the front/back surface connections, explicitly left for future work, could make the combined system trivial even if each single junction lies in the Fu-Kane window. The load-bearing unverified step is therefore the inference from the two-junction phase-winding picture to a nontrivial topological invariant of the full device.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports a combined experimental and theoretical study of a topological-insulator nanowire side-contacted by two superconducting electrodes, forming a columnar nano-SQUID in which the top and bottom surfaces act as two SNS line junctions. Experimentally, the authors observe critical-current oscillations in an axial magnetic field with period h/2e across five devices, with near-complete suppression at half-integer flux in the symmetric tuning, which they interpret as evidence for surface-only supercurrent with negligible bulk contribution. Theoretically, they propose that for an asymmetric nano-SQUID the magnetic flux drives only one of the two junctions through the Fu-Kane π-phase transition, leading to a topological superconducting phase in the flux windows (n−1/2)Φ_s^0 < Φ < (n+1/2)Φ_s^0 for odd integer n, with Majorana zero modes at the nanowire ends. The central experimental observation is well supported, whereas the topological prediction is derived from a two-junction model and gapped-spectrum simulations rather than from a direct topological-invariant calculation.","tokens_in":24974,"tokens_out":4088,"duration_ms":41303,"significance":"If the topological phase prediction is confirmed, the platform would be an important addition to the Majorana toolbox: it is gate-tunable, uses a bulk-insulating TI, and the predicted phase is robust to chemical-potential disorder. Credit should be given for the multi-device dataset, the careful field-alignment and flux-focusing analysis, and the Kwant-based tight-binding modeling with parameters taken from the BHZ model rather than fitted to the target result. The prediction is falsifiable and clearly stated. However, the paper does not yet provide a direct topological-invariant calculation or a finite-wire Majorana end-state demonstration for the full microscopic model, and the decoupling of top and bottom junctions is supported by an unmeasured coherence-length estimate. These gaps currently prevent full acceptance.","major_comments":[{"comment":"The central topological claim—flux-window topological superconductivity for (n−1/2)Φ_s^0 < Φ < (n+1/2)Φ_s^0 with odd n—is established only through the phenomenological two-junction model Eq. (1) and the single-channel gap formula E_gap ∝ cos(φ/2), not through a topological invariant computed in the microscopic tight-binding model. The simulations in Supplement M report min|ε(k)|, which is a necessary but not sufficient condition for a nontrivial class-D phase, and no finite-wire Majorana end-state calculation is presented. The statement in Sec. III that 'it is useful to perform more detailed analysis to understand the exact role of the front and back surfaces' concedes an unresolved piece that could render the combined system trivial even if each single junction lies in the Fu-Kane window; please add a direct Pfaffian/topological-invariant calculation or a transparent finite-wire MZM localization calculation for the full model.","section":"Sec. III and Supplement M"},{"comment":"The load-bearing premise that top and bottom junctions are decoupled rests on the estimate ξ_s << 15 nm for the Ar-etched side surfaces, and this coherence length is not measured. The absence of the h/e even-odd component in Ic(Bx) is indirect evidence but does not bound ξ_s. Because the phase-winding argument that one junction sits at π while the other sits at 0, and hence the entire topological window, depends on this decoupling, please provide a direct measurement of the side-surface coherence length, a more robust justification such as explicit modeling of side-surface disorder, or a clearly stated condition under which the prediction holds.","section":"Supplement L"},{"comment":"The equilibrium phase differences used to construct the topological phase diagram in Fig. 4c are obtained from Eq. (1), which includes only the first harmonic of the current-phase relation. The tight-binding simulations themselves show higher harmonics in the CPR (Supplement Figs. S11 and S13), and the degree to which the phase-winding picture survives with realistic higher harmonics is not demonstrated. Since the π-phase condition is the key to the Fu-Kane transition, please show explicitly that the phase distribution obtained from the full tight-binding energy-phase relation still places the weaker junction in the topological window over the claimed flux range.","section":"Sec. II, Eq. (1) and Fig. 4c"}],"minor_comments":[{"comment":"The main text states that the VG-dependence of Ic shown in Fig. 2d is for device B, but the Fig. 2 caption identifies panel (d) as device A; please correct this inconsistency.","section":"Fig. 2 caption and main text"},{"comment":"Please correct the typographical errors: 'supercurent' in the abstract, 'measurenents' in Methods, and the spacing in the 'T ransport regime' heading.","section":"Abstract and Methods"},{"comment":"In the Supplement, 'devicea A–C' should read 'devices A–C'.","section":"Supplement Fig. S1"},{"comment":"The author listing contains the corrupted text 'Micha/suppress l Papaj'; this should read 'Michael Papaj'.","section":"Author list"}],"recommendation":"major_revision","confidential_remarks":"The experimental part of the manuscript is strong and likely publishable as a demonstration of a surface-dominated columnar nano-SQUID. The theoretical topological prediction is the main advertised result, but it is currently supported only by a reduction to the known Fu-Kane phase-biased junction and by gapped-spectrum simulations. The refereed community would expect either a direct topological invariant or Majorana end-state calculation for the full model, and some quantitative handle on the top/bottom decoupling assumption. I would therefore recommend major revision rather than rejection: the issues are specific and addressable within the scope of the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis paper is worth your time, but read the theory section with a skeptical eye. The experiment reporting a columnar TI nano-SQUID is solid: five devices show h/2e-periodic critical-current oscillations in axial field, with near-total suppression at half flux quantum and recovery at one flux quantum, consistent with surface-dominated supercurrent and negligible bulk. The gate-tunability of the asymmetry and the vanishing Ic minima in device B are convincing. The tight-binding simulations reproduce the SQUID behavior, including the 0–π transition that kills Ic at h/4e. This part is a genuine advance: a new side-contacted TI nanowire junction that acts as a nanometer-scale SQUID. Citation practice looks fine; the distinction from prior TINW proposals is explicit.\n\nThe theoretical claim—flux-window topological superconductivity with Majorana modes at the ends—is more tentative than the abstract suggests. The argument uses the Fu-Kane phase-biased junction result and assumes the top and bottom junctions are decoupled, so the flux can drive one junction through π while the other stays near zero. The support for decoupling is an estimate that the Ar-etched side-surface coherence length is much shorter than the 15-nm nanowire height; that length is not directly measured. The absence of an h/e even-odd component in the oscillations is suggestive but not a measurement of ξ_s.\n\nThe microscopic tight-binding model checks the excitation gap, min|ε(k)|, but does not compute a topological invariant (like the class-D Pfaffian) or show localized Majorana end states in a finite wire. So the step from 'the weaker junction is phase-wound' to 'the whole device is topological' is an inference, not a direct demonstration. The paper acknowledges this by leaving the front/back surface connections and exact MZM location to future work. That honesty counts for something.\n\nThe disorder-robustness claim is supported by parameter sweeps over μ, φ, and Φ showing broad gapped regions. That is reasonable evidence but not explicit disorder simulations.\n\nNet: the experimental platform is publishable and important. The topological prediction is plausible but unproven. A careful referee should demand either a microscopic invariant calculation or a clear restatement that the topological phase is conjectured on the basis of the two-junction model. I would send it to review. It will generate useful discussion, and the device itself is a step forward for TI-based Majorana research.","headline":"Solid SQUID experiment and a promising platform, but the topological phase is an inference from a two-junction model with an unmeasured decoupling assumption.","tokens_in":25709,"tokens_out":3304,"would_cite":true,"duration_ms":33831,"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 predicts that an asymmetric topological-insulator nano-SQUID becomes a topological superconductor, with Majorana zero modes at its ends, for every odd flux quantum window, independent of the chemical potential.","keywords":["topological insulator","Majorana zero modes","nano-SQUID","Josephson junction","topological superconductivity","Fu-Kane model","flux-tunable","surface states"],"falsifier":"A tunnel-spectroscopy probe at one end of the nanowire should show a zero-bias conductance peak throughout each odd flux window $(n-1/2)\\Phi_s^0<\\Phi<(n+1/2)\\Phi_s^0$ and none between them; a measurement showing no such flux-periodic peaks, or peaks of the wrong periodicity, would falsify the central prediction.","tokens_in":2015,"feed_emoji":"🧲","tokens_out":2310,"duration_ms":106614,"temperature":0.7,"pith_summary":"This paper proposes a simple device for topological superconductivity: a rectangular nanowire of a bulk-insulating three-dimensional topological insulator, side-contacted along its length by two superconductors, so that the top and bottom surfaces each form an SNS Josephson line junction and the nanowire cross-section acts as a tiny SQUID loop. The central theoretical claim is that when the two junctions are asymmetric, threading magnetic flux through the cross-section periodically drives only the weaker junction through a Fu-Kane-type topological phase transition, placing the whole columnar SQUID in a topological state for flux windows $(n-1/2)\\Phi_s^0<\\Phi<(n+1/2)\\Phi_s^0$ with odd integer $n$ ($\\Phi_s^0=h/2e$). If true, the two ends of the nanowire host Majorana zero modes that are insensitive to chemical-potential disorder, and the required ingredients, surface-dominated supercurrent and gate-tunable top/bottom asymmetry, are already demonstrated in the measured devices. The paper thus offers a concrete, accessible platform for realizing and eventually braiding Majorana modes.","feed_headline":"Asymmetric nano-SQUID enters a topological phase at odd flux quanta","feed_subtitle":"Surface-only supercurrent and gate-controlled asymmetry keep the Majorana phase stable against disorder.","key_machinery":"The central object is the columnar nano-SQUID: a rectangular TI nanowire sandwiched laterally by superconductors over its full length, so the nanowire cross-section is the SQUID loop and the top and bottom surfaces are two independent SNS line junctions sharing a common phase bias $\\phi$. The load-bearing mechanism is the flux-induced phase winding: axial flux $\\Phi$ sets the two junction phase differences to $\\phi \\mp \\pi\\Phi/\\Phi_s^0$, as in the phenomenological current model $I(\\phi)=I_t\\sin(\\phi-\\pi\\Phi/\\Phi_s^0)+I_b\\sin(\\phi+\\pi\\Phi/\\Phi_s^0)$. When the junctions are asymmetric, the weaker junction is the one that acquires a $\\pi$ phase difference at odd half flux quanta, undergoing a Fu-Kane-type transition driven by the perfectly transmitted, $4\\pi$-periodic Andreev bound states that cross zero energy at odd multiples of $\\pi$. The paper supports this picture with a tight-binding model of the nanowire in a magnetic field, which reproduces the SQUID critical-current oscillations and the $0$–$\\pi$ transition in the equilibrium phase difference, and with spectral-gap calculations showing $E_{\\mathrm{gap}}\\propto \\cos(\\phi/2)$ for the topological surface.","core_discovery":"The discovery claim is that a columnar TI nano-SQUID becomes topological by flux alone. The paper argues that at half-integer flux quanta, $\\Phi=(n+\\tfrac12)\\Phi_s^0$, time-reversal symmetry (neglecting magnetic-field effects beyond the flux) forces one Josephson junction to carry a phase difference of $\\pi$ and the other a phase of 0; because a $\\pi$ phase difference switches the sign of the junction's Josephson energy, the weaker junction takes the $\\pi$, its spectral gap $E_{\\mathrm{gap}}\\propto \\cos(\\phi/2)$ closes, and the line junction passes into its topological regime while the stronger junction stays trivial. The topological window then extends over the full flux-quantum range $(n-1/2)\\Phi_s^0<\\Phi<(n+1/2)\\Phi_s^0$ for odd $n$, independent of the chemical potential and of the exact asymmetry magnitude. Experimentally, the paper reports that five nanowire devices show critical-current oscillations with period $h/2e$ in axial fields, with $I_c$ nearly vanishing at $h/4e$ and recovering at $h/2e$, which it reads as proof that supercurrent flows only through the top and bottom surfaces; a back gate tunes the asymmetry, and at the symmetric point $I_c$ goes to zero at the minima, confirming the surface-only picture.","pith_inferences":["The decoupling premise could be tested by making nanowires with varied heights: if the height were reduced below the side-surface coherence length, top and bottom junctions would couple coherently and the flux window for the topological phase should shift or close.","The same flux-biased asymmetric two-junction mechanism could transfer to other spin-momentum-locked surface systems, such as higher-order topological insulators, wherever two line junctions are separated by a proximitized side wall, though the explicit calculation would need to be redone.","The observed contrast of the critical-current oscillations as a function of gate voltage offers a quantitative, gate-based diagnostic of the top/bottom asymmetry parameter, which the paper does not explicitly develop.","A concrete braiding protocol for the Majorana modes at the nanowire ends is left open; one possible route is a network of coupled columnar nano-SQUIDs with flux-controlled phase biases, but the paper only flags the need for braiding."],"forward_implications":["A tunnel probe at the nanowire ends should find zero-bias conductance peaks in each odd flux window, providing a direct experimental test of the predicted Majorana zero modes.","Because the topological condition is set by phase bias rather than chemical potential, the platform is expected to be resilient to the Coulomb disorder that has hampered semiconductor nanowire platforms.","The same device can be tuned between symmetric and asymmetric regimes with a back gate, making asymmetry a control knob for switching the topological phase.","The simulations predict the gapped topological phase survives for chemical potentials across the full bulk band gap, in both the well-defined-subband and decoupled-junction limits, so fabrication variations that alter subband structure should not destroy it.","The absence of an even-odd effect in the measured critical-current oscillations indicates well-defined transverse subbands are not formed, which the paper argues is consistent with the decoupled-junction description."],"supporting_citations":[{"why":"Supplies the Fu-Kane line-junction result: perfectly transmitted, $4\\pi$-periodic Andreev bound states in a topological-insulator surface junction are topological for phase differences $\\pi<\\phi<3\\pi$, which is the basis of the predicted transition.","marker":"[4]"},{"why":"Establishes that a one-dimensional topological superconductor hosts Majorana zero modes at its ends; the paper uses this to conclude the topological line junction gives Majorana modes at the nanowire ends.","marker":"[46]"},{"why":"Provides the fermion-parity anomaly and zero-energy crossing at odd multiples of $\\pi$ in topological-insulator Josephson junctions, used to argue that only one junction turns topological at half flux quanta.","marker":"[44]"},{"why":"Describes helical Andreev bound states and superconducting Klein tunneling in topological-insulator Josephson junctions, supporting the Andreev spectrum used in the spectral-gap calculation.","marker":"[45]"},{"why":"Argues that topological superconductivity from topological-insulator surface states is resilient to disorder, backing the paper's robustness claim against chemical-potential fluctuations.","marker":"[9]"},{"why":"Reports a compact SQUID realized in a double-layer graphene heterostructure, serving as the earlier nano-SQUID realization to which this device is compared.","marker":"[15]"},{"why":"Reports a lateral sandwich junction on a topological insulator made by a diffusion process, the prior device that was not gate-tunable; the new side-contacting technique is positioned against it.","marker":"[16]"},{"why":"Gives the maximum-entropy analysis of irregular Fraunhofer patterns in topological-insulator Josephson junctions, which the paper uses to interpret the measured out-of-plane magnetic-field patterns.","marker":"[28]"}],"fun_headline_variants":["Flux-tuned topological phase emerges in asymmetric TI nano-SQUID at odd half-flux","Surface-only supercurrent: TI nano-SQUID turns topological at odd flux steps","Odd half-flux quanta: flux-driven topological phase in asymmetric TI nano-SQUID","Topological phase at odd flux steps in TI nano-SQUID with gate-tuned asymmetry"],"cache_read_input_tokens":27648,"weakest_assumption_plain":"The load-bearing premise is that the top and bottom surface Josephson junctions are decoupled, so the columnar SQUID can be treated as two independent line junctions whose phase differences are set individually by the flux; if electrons can coherently tunnel between the surfaces before being reflected, the flux-driven $\\pi$-phase arrangement that creates the topological phase can shift or disappear.","fun_headline_variants_meta":{"raw":{"variants":["Flux-tuned topological phase emerges in asymmetric TI nano-SQUID at odd half-flux","Surface-only supercurrent: TI nano-SQUID turns topological at odd flux steps","Odd half-flux quanta: flux-driven topological phase in asymmetric TI nano-SQUID","Topological phase at odd flux steps in TI nano-SQUID with gate-tuned asymmetry"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000966,"raw_usage":{"total_tokens":4158,"prompt_tokens":1040,"completion_tokens":3118,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":656,"completion_tokens_details":{"reasoning_tokens":3024}},"tokens_in":656,"tokens_out":3118,"duration_ms":22470,"temperature":1.0,"reasoning_tokens":3024,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T18:20:38.947635+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A tunnel-spectroscopy probe at one end of the nanowire should show a zero-bias conductance peak throughout each odd flux window $(n-1/2)\\Phi_s^0<\\Phi<(n+1/2)\\Phi_s^0$ and none between them; a measurement showing no such flux-periodic peaks, or peaks of the wrong periodicity, would falsify the central prediction.","supporting_citations":[{"cited_title":"The mean gauge-invariant phase difference jumps (or quickly crosses over) from 0 to π at Φ = Φ 2 0/ 2 flux bias","cited_arxiv_id":null,"evidence_quote":"Provides the fermion-parity anomaly and zero-energy crossing at odd multiples of $\\pi$ in topological-insulator Josephson junctions, used to argue that only one junction turns topological at half flux quanta."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes helical Andreev bound states and superconducting Klein tunneling in topological-insulator Josephson junctions, supporting the Andreev spectrum used in the spectral-gap calculation."},{"cited_title":"Long-range crossed Andreev reflection in topological insulator nanowires proximitized by a superconductor","cited_arxiv_id":"2407.02383","evidence_quote":"Reports a compact SQUID realized in a double-layer graphene heterostructure, serving as the earlier nano-SQUID realization to which this device is compared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports a lateral sandwich junction on a topological insulator made by a diffusion process, the prior device that was not gate-tunable; the new side-contacting technique is positioned against it."},{"cited_title":"Flensberg, J","cited_arxiv_id":null,"evidence_quote":"Gives the maximum-entropy analysis of irregular Fraunhofer patterns in topological-insulator Josephson junctions, which the paper uses to interpret the measured out-of-plane magnetic-field patterns."}],"review_version":1}