{"id":"4413b418-e1f3-456e-91a2-0400ba865fa0","arxiv_id":"2505.16874","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Localized tip and laser writing can create single- and double-vortex states in artificial spin-vortex ice, and optically written double-vortices can seed chain-like magnetic reversals.","lead":"This paper shows two ways to write magnetic vortex textures into artificial spin-vortex ice: a scanning microscope tip writes single-vortex states, and a focused laser writes double-vortex states with polarization-dependent selectivity. The written vortices are then used to seed local chains of magnetic reversal at fields below the array's normal switching threshold, a step toward programmable nanomagnetic memory and neuromorphic devices.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'avalanche-like' chain claim rests on static before/after images; without time-resolved or field-step-resolved reversal order, the seeding interpretation is one of several that fit the data.","rationale":"The reader's verdict is CONDITIONAL with MODERATE confidence, and the identified weakest assumption matches the most load-bearing concern I find: the avalanche interpretation is not evidenced by static before/after images. I agree that the core writing demonstrations (tip-writing of single-vortices, laser-writing of double-vortices, polarization-dependent fidelity) are supported by raw per-row statistics in Supporting Tables S2-S3 and MFM images, and that the uncharacterized laser-writing mechanism is a limitation but not fatal for the writing claim. The most consequential weakness is the avalanche-seeding claim, because it appears in the abstract, the conclusions, and the motivation for future applications. The paper's own simulations (Figure 5h-i) show the demagnetizing-field reduction from a vortex extends over several neighboring islands, which is exactly the condition under which multiple macrospins could have their switching thresholds lowered simultaneously. Thus the data cannot distinguish between a propagating domino effect and a static landscape effect. A field-step-resolved or time-resolved reversal-order experiment would settle this. Because the concern is substantive but addressable and does not undermine the central local-write demonstrations, the verdict remains CONDITIONAL rather than REJECT or UNCHANGED. The authors should either provide reversal-order data supporting the avalanche interpretation or substantially soften the avalanche language in the abstract and conclusions.","tokens_in":23431,"tokens_out":4158,"duration_ms":29431,"concrete_test":"Repeat the Figure 4 field protocol in small field increments (e.g. 0.5-1 mT steps) from below Hc-start to above Hc-start, imaging after each step, so the order of individual island reversals is recorded. If the double-vortex islands and macrospin chain reverse in a clear sequential order starting at the written vortices and propagating outward, the avalanche-seeding claim is supported. If the first field step above the local threshold already shows several macrospins reversed simultaneously across the chain, the observation is better described as parallel switching in a locally modified switching-field landscape, not a seeded avalanche.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim of programmable local seeding of avalanche-like reversal events (abstract; Section 'Local seeding of avalanche-like reversal via double-vortex writing') depends on interpreting Figure 4b and Supporting Figures S14-S17 as a propagating chain. After applying a global field at 91% of Hc-start, adjacent unwritten macrospins switch, and the authors state this may be considered a one-dimensional avalanche reversal. However, only initial and final states are imaged; there is no time-resolved or field-step-resolved observation establishing that the double-vortex islands reversed first and that macrospin reversals propagated outward from them. The alternative is that the same stray-field landscape that lowers the switching field near a double-vortex also lowers it for several neighboring macrospins simultaneously, so those islands switch independently at the same applied field without domino-like propagation. The static data are equally consistent with this parallel-nucleation explanation. The paper's own simulations (Figure 5h-i) show a spatially extended reduction of the demagnetizing field near a vortex, extending over several neighboring islands, which actively supports the simultaneous-threshold-lowering picture. This avalanche claim is load-bearing because it motivates the abstract's programmable seeding, the conclusions' 'trigger long-range avalanche-like reversal events', and the proposed applications to avalanche studies and neuromorphic seeding. If the chain is merely simultaneous switching of islands with locally lowered thresholds, the novel functional claim of seeding a propagating event weakens, although local vortex writing and modification of nearby switching fields remain established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports two local, field-free methods for writing vortex textures in artificial spin-vortex ice (ASVI): scanning an MFM tip writes single-vortex states, and scanning a focused linearly polarized CW laser writes double-vortex states with a polarization-dependent selectivity for ±45°-oriented nanoislands. The authors characterize writing fidelity with per-row statistics, support the texture assignments with MuMax3 micromagnetic simulations and simulated MFM images, and show that laser-written double-vortices have reduced switching fields compared with macrospins. In a field-application experiment, laser-written double-vortices and several neighboring unwritten macrospins reverse at a field below the macrospin switching threshold; the authors interpret this as locally seeded, avalanche-like propagation of reversal. The paper claims an expanded six-state microstate space per nanoisland and proposes applications in avalanche physics, memory, and neuromorphic computing.","tokens_in":23637,"tokens_out":5743,"duration_ms":48734,"significance":"If the writing demonstrations hold, the paper provides useful experimental tools for local, reconfigurable control of vortex textures in strongly interacting nanomagnet arrays, extending prior single-vortex work to double-vortex writing and expanding the ASVI platform. Strengths include the use of standard micromagnetic parameters (Msat, Aex) rather than parameters fitted to the target outcomes, the provision of raw per-row switching statistics in the Supporting Information, and falsifiable quantitative predictions such as the lower switching field of the parallel double-vortex configuration and the spatially extended stray-field reduction around vortex textures. The main caveat is that the headline 'avalanche-like seeding' claim rests on static before/after imaging rather than on time-resolved or field-step-resolved observations, and the reported laser-writing 'fidelity' is a conditional rate whose denominator excludes several outcome categories.","major_comments":[{"comment":"The claim that the optically written double-vortices seed a propagating, avalanche-like reversal chain is not established by the presented data. The evidence is a comparison of MFM images before and after a single global field application at 91% of Hc-start, with no time-resolved or field-step-resolved observation of the reversal order. The alternative interpretation, that the same stray-field modification that lowers the switching field of a double-vortex island also lowers the thresholds of several neighboring macrospins simultaneously so that those islands reverse independently at the applied field, is fully consistent with the static before/after images and is in fact supported by the authors' own micromagnetic field-landscape plots in Fig. 5h–i, which show a spatially extended demagnetizing-field reduction covering several islands. Because the abstract and conclusions present 'programmable local seeding of avalanche-like reversal events' as a central result, this distinction is load-bearing. I recommend either softening the language to describe spatially correlated, locally seeded reversals without claiming sequential propagation, or adding an incremental-field experiment (e.g., applying the global field in small steps with MFM imaging between steps) to demonstrate the propagation order.","section":"Section 'Local seeding of avalanche-like reversal via double-vortex writing' (Fig. 4b; Supporting Figs. S14–S17)"},{"comment":"The reported double-vortex writing 'fidelity' is a conditional success rate: the denominator excludes islands that were damaged, islands written into single-vortex states, and islands with unidentifiable textures. Since single-vortex writing is a possible outcome of the same illumination rather than a damage event, this exclusion inflates the reported success rate relative to the probability that an illuminated island becomes a double-vortex. The raw per-row counts in Supporting Tables S1 and S2 allow unconditional rates to be recovered, so this is not a fatal flaw, but the main-text definition of fidelity should be stated explicitly, and the unconditional rates should be reported or plotted alongside the conditional rates so that readers can judge the actual writing selectivity.","section":"Section 'All-optical control of vortex textures' and Methods (Laser-writing; SI Tables S1–S2)"}],"minor_comments":[{"comment":"The text states 'Figures 2 a) and b) show the fidelities of double-vortex writing', but the relevant panels are in Figure 3; this cross-reference should be corrected.","section":"Section 'All-optical control of vortex textures'"},{"comment":"The main text references Supporting Figures S1–S4 for array dimensions and S10–S12 for repeated avalanche-like seeding tests, but the supplied Supporting Information uses S6–S9 for the dimension histograms and S14–S17 for the repeated seeding experiments; the numbering should be harmonized.","section":"Supporting Information figure numbering"},{"comment":"The tip-writing results are presented through selected examples without quantitative success or chirality-selectivity statistics; a compact table of attempted writes, successes, and outcome types would strengthen the comparison with the laser-writing characterization.","section":"Section 'Surface-probe control of vortex textures'"},{"comment":"The P/AP double-vortex labels are defined relative to an applied field at +46° in Fig. 5a, while the simulations are described as averaging fields at ±45°; the relation between the schematic angle and the averaged simulation geometry should be clarified in the caption or text.","section":"Section 'Multi-level switching fields of macrospin, single-vortex and double-vortex states'"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the manuscript's core experimental writing demonstrations are valuable and likely sound, but the avalanche-seeding interpretation is overclaimed relative to the static imaging evidence, and the fidelity metric needs redefinition. The paper would be strengthened by a more cautious treatment of the propagation claim, either by adding incremental-field data or by reframing the result as spatially correlated local seeding. I also note that the relationship to the authors' prior work (refs. 30 and 47) should be stated more explicitly so that the incremental advance is clear."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new things here are the two local write techniques: polarization-selective all-optical writing of double-vortex states, and MFM-tip writing of single-vortex states in the width-modified ASVI geometry. The expanded six-states-per-island microstate space is real, and the authors provide per-row fidelity statistics in the SI, which is better than the norm for this field. The micromagnetic simulations use standard Permalloy parameters and measured geometries, not parameters fitted to the target outcome, so the switching-field ordering for macrospin, single-vortex, and double-vortex states is credible.\n\nThe soft spot is the avalanche claim. The paper images only initial and final states after applying a global field at 91% of Hc-start, so the reversal order is never observed. The text does hedge once (the chain “may be considered” a one-dimensional avalanche), but the abstract and conclusions say “programmable local seeding of avalanche-like reversal events” and “trigger long-range avalanche-like reversal events.” That is stronger than the data support. The plausible alternative is that the double-vortices lower the switching field of several neighboring macrospins at once, so those islands reverse independently at the same applied field. The authors’ own simulations (Figure 5h-i) show a spatially extended demagnetizing-field reduction over several islands, which actually favors the simultaneous-threshold-lowering picture. This does not break the local-writing result, but it weakens the “propagating event” novelty. To support the avalanche interpretation, they would need field-step-resolved or time-resolved imaging, or at least a controlled experiment varying vortex spacing.\n\nOther caveats are minor. The laser-writing mechanism is explicitly uncharacterized, which is honest but means “all-optical control” is a demonstration, not an explanation. The fidelity statistics exclude damaged, single-vortex, and unidentified outcomes, which could inflate success rates. No code or raw data are deposited. These are addressable.\n\nOverall: the local writing methods are the contribution, and they look solid. This deserves a serious referee. The referee should push for either time-resolved/step-resolved supporting data or softened avalanche language. For anyone in artificial spin ice or nanomagnonics, the writing techniques are worth knowing and citing; the avalanche physics should be read with caution.","headline":"Solid local-write techniques for vortex states in ASVI, but the avalanche-like reversal claim is stronger than the static MFM evidence supports.","tokens_in":24267,"tokens_out":1886,"would_cite":true,"duration_ms":17518,"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":"Focused laser light and a magnetic tip can locally write double-vortex and single-vortex textures in artificial spin-vortex ice, and the written vortices seed programmable avalanche-like reversal chains.","keywords":["artificial spin ice","artificial spin-vortex ice","magnetic vortex textures","all-optical writing","magnetic force microscopy","avalanche dynamics","nanomagnetic metamaterial","neuromorphic computing"],"falsifier":"Apply the same driving field in small increments, for example stepping from 0 to 18.2 mT in sub-mT steps, and image with magnetic force microscopy after each step. If unwritten islands reverse one by one in spatial order away from the double-vortex seeds, the avalanche interpretation is supported; if the whole cluster changes in a single step, or reversals appear in no order relative to the seed positions, then the propagating-avalanche claim is falsified.","tokens_in":23193,"feed_emoji":"🧲","tokens_out":12461,"duration_ms":99744,"temperature":0.7,"pith_summary":"This paper reports two ways to write magnetic vortex textures into artificial spin-vortex ice without applying a global magnetic field. A focused continuous-wave laser writes double-vortex states, with the laser polarisation selecting which subset of nanoislands is written, while a scanning magnetic-force-microscopy tip writes single-vortex states, with vortex chirality apparently set by where the tip crosses each island. The authors argue that these written vortices locally reshape the array's magnetic energy landscape, so that a field below the normal macrospin switching field destroys the double-vortices and triggers chains of reversals in neighbouring unwritten islands, yielding a one-dimensional ground-state domain. If correct, this turns ASVI into a reconfigurable platform where avalanche-like dynamics can be seeded at chosen locations and raises the per-island state count to six, supporting future avalanche, memory, and neuromorphic applications.","feed_headline":"Written vortices seed programmable avalanches in spin ice","feed_subtitle":"Laser and magnetic-tip writing give six states per island and seed controlled reversal chains.","key_machinery":"The paper's central object is the width-modified artificial spin-vortex ice (ASVI): alternating rows of wide and narrow stadia-shaped magnetic nanoislands, with the wide islands dimensioned so that macrospin, single-vortex, and double-vortex textures are all metastable. The argument runs through the local stray-field landscape, encoded in the effective switching field $\\mu_0H_{c\\text{-eff}} = \\mu_0H_{c\\text{-int}} + \\mu_0H_{dip}$: because a vortex closes its flux internally, it emits less stray field than a macrospin and lowers the switching threshold of neighbouring islands by about 1.6 mT, while a double-vortex additionally has a lower intrinsic coercivity than a macrospin. The two writing tools are local perturbations to this landscape: a magnetic tip injects a vortex core directly beneath it, and a focused laser writes double-vortices with polarisation-dependent absorption selecting which island orientation responds. Written double-vortices thereby act as sacrificial templates that locally re-route reversal pathways under a global field.","core_discovery":"The central claim is that ASVI nanoislands can be engineered so that macrospin, single-vortex, and double-vortex textures are all metastable, and that each can be locally written in a controlled way. Laser illumination preferentially creates double-vortices when the linear polarisation lies along the island's short axis, and the written vortex configuration is set by the array's prior saturation direction; scanning a high-moment magnetic tip writes single-vortices, with the crossing point on the island influencing chirality and whether a macrospin reversal occurs. Written double-vortices, which have lower coercivity than macrospins, then act as local seeds: at 91% of the macrospin start-switching field they themselves switch to macrospins and, according to the authors, trigger a one-dimensional avalanche-like reversal chain through adjacent unwritten islands, producing a row of type-1 ground-state vertices. Simulation supports the mechanism by showing that a vortex reduces the effective switching field of a nearest-neighbour island by about 1.6 mT through the local demagnetising field, and together these results expand the accessible microstate space to six states per nanoisland.","pith_inferences":["We infer that the avalanche interpretation could be tested directly by applying the driving field in small increments and imaging between steps; a spatial ordering of reversals propagating outward from the seeded double-vortices would confirm the domino picture, whereas a near-simultaneous cluster would point to static threshold lowering.","We infer that the unresolved 0° versus 90° polarisation asymmetry, which cannot be explained by linear absorption alone, points to a magneto-optical or magnetic-state-dependent mechanism; a clean test would be laser-writing well-separated islands with controlled initial magnetic states so that stray fields from neighbours cannot mask the effect.","We infer that the dependence of written double-vortex configuration on the array's initial saturation direction could be exploited as an additional writing degree of freedom, since the written vortex polarity would then encode the prior global-field history.","We infer that the technique could serve as a fabrication-free probe of quenched disorder: repeated local writes across many arrays would map the distribution of switching fields island by island, separating intrinsic disorder from seeded effects."],"forward_implications":["Pairing a global field below the macrospin switching threshold with locally written double-vortices prepares one-dimensional chains of type-1 ground-state vertices at chosen positions in the array.","Avalanche studies become repeatable and location-controlled: the writing step fixes where a reversal chain starts, instead of relying on stochastic nucleation and many repeated imaging runs.","The laser's polarisation acts as a spatial mask, writing double-vortices preferentially into islands whose short axis is aligned with the polarisation, so either the +45° or −45° island subset can be targeted.","Because double-vortices switch below the macrospin field and end as macrospins, they act as sacrificial templates that can be repeatedly rewritten, while single-vortices, with higher switching fields than macrospins, act as persistent templates for steering later reversals.","The enlarged six-state-per-island microstate space and local writing give a route to direct physical implementation of nanomagnetic weights and memory states in neuromorphic devices."],"supporting_citations":[{"why":"It establishes artificial spin-vortex ice as a platform where macrospin and vortex textures are both metastable, the system this paper extends with double-vortex states and local writing.","marker":"30"},{"why":"It motivates the work by showing ASVI's use in neuromorphic and reservoir computing, which the authors aim to make locally programmable.","marker":"31"},{"why":"It supplies the tip-writing method for artificial spin ice via topological defect injection that the surface-probe writing builds on.","marker":"16"},{"why":"It provides the technique for injecting and manipulating magnetic states with a scanning tip, the basis of the single-vortex writing.","marker":"47"},{"why":"It shows that moving magnetic tips can create vortices, supporting the mechanism claimed for MFM-tip single-vortex writing.","marker":"49"},{"why":"It supplies the simulation result that tip crossing position controls vortex chirality, used to explain the chirality transition in the tip-written line.","marker":"45"},{"why":"It establishes low-power continuous-wave all-optical magnetic switching in ferromagnetic nanoarrays, the foundation of the laser-writing approach.","marker":"59"},{"why":"It supplies the one-dimensional random-field Ising model interpretation used to describe the seeded reversal chain as an avalanche-like event.","marker":"9"},{"why":"It provides the micromagnetic simulation method used to calculate switching fields of each texture and the local demagnetising-field changes.","marker":"61"},{"why":"It provides the optical absorption simulation used to check whether polarisation-dependent absorption explains the laser-writing selectivity.","marker":"62"}],"fun_headline_variants":["Laser and tip write vortices that trigger ice avalanches","Spin-ice vortices written on demand, then seed reversals","Six spin states per island via laser and magnetic-tip writing","Programmable vortex seeds control avalanche chains in spin ice","All-optical and MFM tip writing seeds avalanche reversals"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The avalanche claim rests on the assumption that adjacent unwritten islands reverse in a domino-like sequence starting from the written double-vortices, rather than reversing independently because the static magnetic landscape and fabrication defects have already lowered their switching thresholds; the paper images only before and after the field step, not the order of events.","fun_headline_variants_meta":{"raw":{"variants":["Laser and tip write vortices that trigger ice avalanches","Spin-ice vortices written on demand, then seed reversals","Six spin states per island via laser and magnetic-tip writing","Programmable vortex seeds control avalanche chains in spin ice","All-optical and MFM tip writing seeds avalanche reversals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000552,"raw_usage":{"total_tokens":2657,"prompt_tokens":993,"completion_tokens":1664,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":609,"completion_tokens_details":{"reasoning_tokens":1579}},"tokens_in":609,"tokens_out":1664,"duration_ms":10007,"temperature":1.0,"reasoning_tokens":1579,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:53:01.259696+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Apply the same driving field in small increments, for example stepping from 0 to 18.2 mT in sub-mT steps, and image with magnetic force microscopy after each step. If unwritten islands reverse one by one in spatial order away from the double-vortex seeds, the avalanche interpretation is supported; if the whole cluster changes in a single step, or reversals appear in no order relative to the seed positions, then the propagating-avalanche claim is falsified.","supporting_citations":[{"cited_title":"D.et al.Neuromorphic overparameterisation and few-shot learning in multilayer physical neural networks","cited_arxiv_id":null,"evidence_quote":"It motivates the work by showing ASVI's use in neuromorphic and reservoir computing, which the authors aim to make locally programmable."},{"cited_title":"C., Burn, D","cited_arxiv_id":null,"evidence_quote":"It provides the technique for injecting and manipulating magnetic states with a scanning tip, the basis of the single-vortex writing."},{"cited_title":"P., Hucht, A., Hinrichsen, H., Dahmen, S","cited_arxiv_id":null,"evidence_quote":"It shows that moving magnetic tips can create vortices, supporting the mechanism claimed for MFM-tip single-vortex writing."},{"cited_title":"D.et al.Magnonic bending, phase shifting and interferometry in a 2d reconfigurable nanodisk crystal.ACS Nano15, 674–685, DOI: 10.1021/acsnano.0c06894 (2021)","cited_arxiv_id":null,"evidence_quote":"It supplies the simulation result that tip crossing position controls vortex chirality, used to explain the chirality transition in the tip-written line."},{"cited_title":"S.et al.Experimental realization of the 1d random field ising model.Phys","cited_arxiv_id":null,"evidence_quote":"It supplies the one-dimensional random-field Ising model interpretation used to describe the seeded reversal chain as an avalanche-like event."}],"review_version":1}