{"id":"1382b0ad-34cb-4c30-a643-58b87baeb172","arxiv_id":"2607.21218","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Proximity to NbSe2 switches the magnetic ground state of monolayer NiI2 from a spin spiral to a ferromagnet, detectable through Yu–Shiba–Rusinov states.","lead":"Monolayer NiI2 on a superconducting NbSe2 substrate loses its original spiral magnetic order and becomes ferromagnetic, as shown by scanning tunnelling microscopy and spectroscopy. The result demonstrates that a substrate can re-engineer magnetic order in a two-dimensional multiferroic, with Yu–Shiba–Rusinov states acting as the local magnetic probe.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Absence of ferroelectric stripes on a metallic substrate is not sufficient to prove loss of spin-spiral order; screening could hide the stripe while magnetism persists.","rationale":"The reader's weakest-assumption analysis correctly identifies the inference from the absence of ferroelectric stripes as a load-bearing but underdetermined step. The paper's own statement makes this explicit, and the alternative explanation—screening of the ferroelectric polarization by the metallic substrate without loss of spin-spiral order—is physically plausible. The DFT calculations could in principle rescue the claim, but the main text only presents them schematically, so they do not independently settle the experimental question. The YSR and edge-LDOS comparisons are suggestive but depend on a model whose ability to discriminate FM from spin-spiral is not shown with quantitative detail or parameter sensitivity in the main text. A direct spin-resolved measurement would resolve the ambiguity. Since this concern is the same one the reader identified, and the reader already assigned a CONDITIONAL verdict, my stress-test does not move the verdict.","tokens_in":9335,"tokens_out":7628,"duration_ms":97859,"concrete_test":"Perform spin-polarized STM/STS on the same NiI2/NbSe2 islands at 350 mK, using a magnetic tip and comparing maps with opposite tip magnetization. If the ground state is ferromagnetic, the differential conductance should show uniform or domain-like magnetic contrast with no periodic modulation at the free-standing spiral wavevector q≈2/5 ΓK. If a spin spiral persists with screened ferroelectric polarization, a periodic magnetic (YSR) modulation at that wavevector should appear across the island interior. This directly tests whether the missing stripe reflects loss of spiral order or merely screening of the polarization.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on the experimental inference that the absence of stripe-like modulation in NiI2 on NbSe2 means the non-collinear magnetic ground state is lost. In the Results, first paragraph, the authors state: 'In contrast to earlier work on a graphite substrates, we do not observe stripe-like modulation in NiI2 at any bias voltage... The absence of ferroelectric stripes indicates that monolayer NiI2 on NbSe2 does not host a non-collinear magnetic ground state.' This is load-bearing because it is the most direct experimental evidence for the helix-to-ferromagnet transition. However, the stripe pattern is a signature of the ferroelectric polarization induced by the spin spiral, not of the magnetic order itself. On a metallic substrate such as NbSe2, free-carrier screening can suppress the electrostatic/piezoelectric stripe contrast while the spin-spiral order remains intact. The paper does not rule out this alternative. The other experimental evidence—reduced coherence peaks, in-gap YSR states, and edge LDOS patterns—is model-dependent; the FM versus spin-spiral discrimination relies on a low-energy BdG model (Eqs. 1–4) whose parameters and uniqueness are not fully established in the main text. The DFT E(q) calculations are the direct theoretical support, but Fig. 1d,e show only schematic energy diagrams, with no numerical E(q) curves or energy differences in the main text, so the strength of the theoretical prediction cannot be independently assessed. Thus the experimental confirmation of the central claim hinges on a negative observation that is confounded by substrate screening.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports STM/STS measurements and DFT calculations on monolayer NiI2 grown on superconducting NbSe2. The authors claim that the metallic substrate renormalizes the exchange interactions in NiI2, driving a transition from the helimagnetic ground state of free-standing NiI2 to a ferromagnetic ground state. Experimental evidence includes the absence of the ferroelectric stripe pattern observed on graphite, reduced coherence peaks and in-gap YSR-like states in the bulk, and edge LDOS patterns with a symmetry argued to be incompatible with a spin spiral. A low-energy BdG model (Eqs. 1–4) is used to simulate YSR spectra and LDOS maps for ferromagnetic and spin-spiral configurations, with the ferromagnetic case claimed to match experiment.","tokens_in":9703,"tokens_out":4572,"duration_ms":55654,"significance":"If established, this would be a valuable demonstration of substrate-driven magnetic phase control in a monolayer multiferroic and would extend the use of YSR states as a spin-averaged in situ probe of 2D magnetism. The work combines MBE growth, low-temperature STM/STS, DFT with the generalized Bloch theorem, and open-source BdG simulations (pyqula, ref 36), a strong and partly reproducible methodology. The edge-symmetry argument and the q-dependent energy framework are thoughtful. However, the central experimental inference is indirect: the absence of ferroelectric stripes is not a direct measurement of spin order, and the FM-versus-spiral discrimination relies on a parametrized model without quantitative fits. The DFT E(q) result, which is the independent theoretical backbone, is presented only schematically in the main text.","major_comments":[{"comment":"The q-dependent DFT E(q) calculation is the central theoretical support for the exchange renormalization, but the main text shows only schematic energy diagrams (Fig. 1d,e) and states the FM minimum at q=(0,0) without giving numerical E(q) curves, energy differences, or exchange couplings (J1/J3) for freestanding vs heterostructure. This prevents the reader from assessing the strength of the predicted transition. Please include the numerical E(q) data (at least in SI) and computational parameters (U, vdW corrections, k-point sampling).","section":"Results, first paragraph and Fig. 1d,e"},{"comment":"The statement 'The absence of ferroelectric stripes indicates that monolayer NiI2 on NbSe2 does not host a non-collinear magnetic ground state' is load-bearing, but the stripe contrast is a ferroelectric/piezoelectric signature, not a direct spin probe. On a metallic substrate, free-carrier screening can suppress the stripe pattern while a spin spiral remains, and the imaging conditions differ from those on graphite. This alternative is not excluded. A control measurement (e.g., identical imaging of NiI2 on a weakly screened substrate in the same setup) or a direct spin-resolved probe is needed to support the claim.","section":"Results, first paragraph"},{"comment":"The FM-vs-spin-spiral assignment rests on qualitative agreement between measured and simulated LDOS. The model contains freely adjustable parameters: the exchange distribution J(r), spiral vector q, moiré modulation, and edge-enhanced coupling. The main text says the spin-spiral 'consistently result[s] in strongly shifted and split coherence peak features' while FM is 'more in-line', but no quantitative comparison (e.g., residuals or fit quality) or parameter-sensitivity analysis is shown. Please provide quantitative metrics and demonstrate that a spin spiral with different q or J distribution cannot reproduce the data.","section":"YSR states subsection, Eqs. (1)–(4)"},{"comment":"The edge-symmetry argument—that the spin spiral should produce nonequivalent edge contrast on crystallographically identical edges—relies on the model's symmetry breaking. However, the observed 60°/120° edge difference could also arise from iodine termination (up vs down) or moiré registry, as the text itself notes. To support the conclusion, the authors should disentangle these effects, e.g., by comparing multiple edge types on the same island and by simulating termination-dependent coupling explicitly.","section":"YSR states subsection, edge LDOS paragraph"}],"minor_comments":[{"comment":"Please label the axes explicitly as q (or propagation vector) and E(q), and specify the energy scale; 'Energy diagrams' alone is ambiguous.","section":"Fig. 1d,e"},{"comment":"The dashed magenta lines are said to mark the moiré period in both panels c and e, but the text says they correspond to the conduction-band modulation; clarify whether the same positions are used in both panels and whether drift corrections were applied.","section":"Fig. 2e"},{"comment":"At 5 K the authors state the YSR features are 'challenging to resolve in detail', while Fig. 3a is measured at 350 mK. Please clarify how the bulk YSR band is established in the 350 mK data and whether the moiré modulation of the gap (Fig. 2e) is reproducible at sub-Kelvin temperatures.","section":"Results, bulk YSR paragraph"},{"comment":"The SI is referenced for parameter values and additional comparisons, but the main text does not state the numerical values of J, Δ, or the moiré-period modulation used in Figs. 3b,c,f,g. Please summarize these parameters in a table or in the main text.","section":"Methods / SI"},{"comment":"The paper uses 'NiI2' and 'NiI$_{2}$' inconsistently; please standardize formatting. Also, reference 36 is a GitHub repository without version/commit; please cite a version or persistent identifier for reproducibility.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern that screening could explain the absence of ferroelectric stripes is valid and directly weakens the central experimental inference. The quantitative E(q) data and a more critical assessment of the YSR model's discriminative power are necessary before the FM ground-state claim can be considered established."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's what you need to know. The paper argues that monolayer NiI2, normally a spin-spiral multiferroic on graphite, becomes ferromagnetic when grown on NbSe2, and that this transition can be read out via Yu-Shiba-Rusinov states in the superconductor. That claim is new: I'm not aware of any prior demonstration of substrate-induced switching of a monolayer multiferroic ground state, and using YSR states as a local magnetism probe is a genuinely nice extension of earlier work on CrBr3/CrCl3. If it holds up, it's a useful result for the vdW spintronics crowd.\n\nThe measurement quality is good. The MBE growth, atomic-resolved images, and moire characterization are consistent with the group's prior NiI2 work. The edge LDOS comparison between simulated FM and spin-spiral states is a clever idea, and the observed double-stripe pattern being present on all equivalent edges is a sensible symmetry argument. The DFT direction is also plausible: NbSe2 is known to have a large ferromagnetic spin susceptibility, so an RKKY-type renormalization pushing NiI2 toward FM is physically reasonable.\n\nBut the experimental evidence for the transition has a real weak spot. The main direct indicator is the absence of the ferroelectric stripe pattern that appears on graphite. That's a negative observation, and on a metallic substrate free-carrier screening could suppress the stripe contrast even if the spin-spiral order is still there. The paper doesn't rule that out, and it's a load-bearing point in the conclusion. The YSR spectra show reduced coherence peaks and in-gap states, which are consistent with magnetic coupling, but the discriminating power between FM and spin-spiral in the BdG simulation depends on the chosen J(r) distribution and other parameters; the main text only gives the qualitative spectra, not the full parameter sensitivity. The DFT E(q) calculation is the strongest independent support, but the main text shows schematic energy diagrams rather than the actual E(q) curves or energy differences. Without those numbers, it's hard to see how robust the FM state is against correlation or strain effects.\n\nOverall, this is a solid paper with a plausible central claim. The weakness is that the experimental confirmation hinges on a missing stripe, which screening could explain. A careful review should ask for the actual E(q) data, a quantitative estimate of stripe suppression by screening, and a fuller account of the YSR model parameters. I'd send it to referees rather than desk reject it—the idea is good and the experiments aren't sloppy. But I'd expect a revision that tightens the link between the experimental observations and the claim of ferromagnetism.\n\nFor whom: experimentalists working on 2D magnetism, YSR physics, and vdW heterostructures; theorists doing DFT and multiferroic prediction. I'd bring it to a reading group and would cite it if I were writing about substrate control of magnetic order.","headline":"Plausible and potentially important claim of substrate-driven helix-to-ferromagnet transition in NiI2/NbSe2, but the experimental proof rests on a negative observation that screening could explain; worth refereeing, with room for revision.","tokens_in":10170,"tokens_out":3051,"would_cite":true,"duration_ms":30980,"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":"Placing monolayer NiI2 on a superconducting NbSe2 substrate drives its magnetic ground state from a spin spiral to a ferromagnet, through substrate-induced renormalization of the exchange couplings.","keywords":["two-dimensional multiferroics","NiI2 monolayer","NbSe2 substrate","Yu-Shiba-Rusinov states","proximity-induced magnetism","RKKY exchange renormalization","spin-spiral to ferromagnetic transition","scanning tunneling microscopy/spectroscopy"],"falsifier":"A spin-resolved STM image of the same NiI2 islands showing a magnetic spiral modulation, or a YSR spectrum that matches the calculated spin-spiral response in a sample where the stripe pattern is absent, would disprove the ferromagnetic ground state.","tokens_in":9275,"feed_emoji":"🧲","tokens_out":3466,"duration_ms":35609,"temperature":0.7,"pith_summary":"This paper argues that the magnetic ground state of monolayer NiI2—a candidate 2D multiferroic that on inert substrates hosts a spin-spiral order—is overturned by proximity to the metallic superconductor NbSe2. Using low-temperature STM and STS, the authors find no sign of the ferroelectric stripe pattern expected for the spin spiral, and instead observe Yu–Shiba–Rusinov bound states inside NbSe2's superconducting gap whose spectra and real-space maps match a ferromagnetic NiI2 layer. Density-functional calculations support the picture: NbSe2's strongly ferromagnetic RKKY susceptibility renormalizes the Ni exchange couplings, moving the minimum of the spin-spiral energy to zero wavevector. The work matters because it demonstrates substrate engineering as a handle on magnetic order in atomically thin multiferroics, and introduces YSR states as a spin-averaged in situ probe of 2D magnetism.","feed_headline":"Superconductor turns a 2D multiferroic ferromagnetic","feed_subtitle":"NiI2's spin spiral vanishes on NbSe2; Yu–Shiba–Rusinov states reveal the new magnetic order.","key_machinery":"The argument hinges on the exchange-coupling renormalization of NiI2 by the substrate, captured in two complementary calculations: DFT spin-spiral energy curves E(q) using the generalized Bloch theorem, whose minimum moves from finite q to q=0 in the heterostructure; and a Bogoliubov–de Gennes model of NbSe2 with a spatially modulated exchange field J(r), which generates Yu–Shiba–Rusinov (YSR) bands—bound quasiparticle states that a magnetic moment induces inside a superconductor's gap—whose energy spectra and real-space maps distinguish ferromagnetic from spin-spiral order.","core_discovery":"The central discovery is that the spin-spiral (helimagnetic) ground state of monolayer NiI2 is replaced by a ferromagnetic ground state when the layer is grown on NbSe2, and that this transition is visible in the YSR states of the underlying superconductor. The authors show that the NbSe2 substrate, being close to a ferromagnetic instability, transfers a ferromagnetic RKKY interaction that renormalizes the first- and third-neighbor exchange couplings in NiI2, shifting the minimum of the spin-spiral energy from a finite wavevector to zero. Experimentally, the ferromagnetic state is inferred from the absence of ferroelectric stripes, from YSR bands inside the superconducting gap, and from edge","pith_inferences":["If the mechanism is generic, other nickel dihalide monolayers (e.g., NiBr2, NiCl2) on strongly susceptible metals should show similar helimagnet-to-ferromagnet transitions, which could be tested with the same YSR readout.","The YSR band modulation over the moiré period could be used as a local probe of the exchange field's spatial variation, effectively mapping magnetic coupling at the nanoscale.","The absence of ferroelectric stripes might also result from substrate screening of the electric polarization even if a spiral persisted, so a direct spin-resolved measurement would cleanly separate these effects.","The edge-dependent double-stripe LDOS pattern suggests island shape and edge termination control the magnetic coupling, implying that growth geometry could pattern magnetic textures at the nanoscale."],"forward_implications":["YSR states can serve as an in situ, spin-averaged probe of the magnetic ground state of a 2D magnet placed on a superconductor.","Substrate engineering—proximity to a metal with large ferromagnetic susceptibility—can switch magnetic order in atomically thin layers.","NiI2 becomes a substrate-tunable frustrated magnet whose competing exchange interactions can be biased toward ferromagnetism.","The edges of NiI2 islands show enhanced magnetic coupling to NbSe2, giving edge-localized YSR states modulated by the moiré pattern.","This opens a route to combining multiferroic order and superconductivity in van der Waals heterostructures."],"fun_headline_variants":["Superconductor flips NiI2 spin spiral to ferromagnet","NbSe2 substrate turns monolayer NiI2 ferromagnetic","YSR states reveal ferromagnetism in NiI2 on NbSe2","Proximity to superconductor switches NiI2 to ferromagnet","Spin spiral gone: NiI2 on NbSe2 is now ferromagnetic"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The conclusion rests on reading the absence of ferroelectric stripes and the shape of the YSR spectra as proof that the spin spiral is gone; if those stripes are suppressed by other substrate effects while the spiral survives, the central claim would not hold.","fun_headline_variants_meta":{"raw":{"variants":["Superconductor flips NiI2 spin spiral to ferromagnet","NbSe2 substrate turns monolayer NiI2 ferromagnetic","YSR states reveal ferromagnetism in NiI2 on NbSe2","Proximity to superconductor switches NiI2 to ferromagnet","Spin spiral gone: NiI2 on NbSe2 is now ferromagnetic"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000661,"raw_usage":{"total_tokens":2853,"prompt_tokens":734,"completion_tokens":2119,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":478,"completion_tokens_details":{"reasoning_tokens":2037}},"tokens_in":478,"tokens_out":2119,"duration_ms":14378,"temperature":1.0,"reasoning_tokens":2037,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T08:05:13.940593+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A spin-resolved STM image of the same NiI2 islands showing a magnetic spiral modulation, or a YSR spectrum that matches the calculated spin-spiral response in a sample where the stripe pattern is absent, would disprove the ferromagnetic ground state.","supporting_citations":[],"review_version":1}