{"id":"5c4d110c-7e89-4075-9e53-83851e0ebd9d","arxiv_id":"2505.08431","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A 3 µm YIG film with stripe domains exhibits multiple spin wave resonance modes localised in domain walls, flux closure caps, and bulk stripes, reproduced by micromagnetic simulations that include cubic anisotropy.","lead":"Researchers measured the magnetic resonances of a 3-micrometer yttrium iron garnet film whose magnetisation forms stripe domains, and used simulations to map where each resonance lives. The work maps out how domain walls, flux-closure caps, and bulk stripes each host distinct spin wave modes, with potential for reconfigurable magnonic devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Hard-axis mode assignments rely on a 2D simulation that cannot represent the zig-zag state present in the experiment at the same fields.","rationale":"The reader's weakest assumption already identifies both the hand-tuned simulation parameters and the 2D cross-section's inability to capture the zig-zag state. My stress-test sharpens this into a concrete, load-bearing concern: the hard-axis low-field mode assignments, including the highlighted splitting of Modes 1 and 2, are made in a field regime where the experimental ground state is a zig-zag pattern that the simulation cannot represent. This does not overturn the paper's overall conclusion, because the easy-axis data and high-field mode assignments may still be valid, but it does mean the claimed 'good correlation' is not established for a significant subset of the presented results. The condition already imposed by the reader remains appropriate; no verdict change is needed.","tokens_in":18341,"tokens_out":4804,"duration_ms":48835,"concrete_test":"Run a 3D mumax3 simulation of the same 3 µm film with lateral dimensions large enough to stabilise the zig-zag state (several stripe periods in both in-plane directions), using measured Ku = 1400 J/m^3, Kc = -600 J/m^3, and A = 6.5e-12 J/m, and compute the resonance spectra and mode profiles for Hext along [11-2] at 3 mT. Compare the low-frequency mode splitting and spatial localisation with Figs. 5(d) and 6(d). If the 3D zig-zag ground state produces a different splitting or different mode profiles, the 2D cross-section assignments in that regime are unreliable; if the splitting and profiles are reproduced, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that each observed resonance can be assigned a localisation with good correlation to simulation is least secure for the hard-axis low-field regime. The paper explicitly states in Section II.A that the 2D cross-section simulation 'prevents the possibility of simulating the zig-zag state' observed for Hext along [11-2]. Yet Fig. 5(d) assigns Modes 1-5 using power maps taken at 3 mT, and Fig. 1(d) shows that at 3 mT the experimental state is again approaching the zig-zag configuration. The claimed cubic-anisotropy-induced splitting of Modes 1 and 2 is derived from an asymmetry in E(1-10) that tilts stripes in the simulated 2D cross-section; but the real low-field state is modulated along the stripe direction, so the observed splitting could be a consequence of the zig-zag state rather than the proposed cubic-anisotropy mechanism. This is compounded by the parameter choices in Appendix A: Ku = 1000 J/m^3 is fitted to reproduce spectra rather than the measured 1400 ± 200 J/m^3, A = 6.5e-12 J/m is a compromise chosen to prevent domain division, and the simulated saturation field is about double the experimental value. If the simulated static ground state does not faithfully match the experimental one, the spatial mode profiles and hybridisations inferred from it are not independently secured.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports static and dynamic magnetic studies of a 3-µm-thick LPE-grown YIG film with small perpendicular anisotropy that hosts stripe domains. Wide-field MOKE imaging shows collinear and zig-zag stripe states depending on field direction and magnitude, and VNA-FMR measurements in four pumping/field geometries reveal multiple resonance modes in the stripe phase. Micromagnetic simulations (mumax3) of a 2D film cross-section are used to compute static domain evolution and resonance spectra, with power maps assigning each observed mode to a spatial localization: domain-wall modes, flux-closure cap modes, bulk stripe modes, and hybridized higher-order modes. The authors attribute the field-dependent domain evolution and a low-field mode splitting under hard-axis fields to cubic anisotropy. The central claim is that the simulated spectra and spatial power maps reproduce the experimental spectra well enough to identify the origin of each observed resonance.","tokens_in":18623,"tokens_out":6695,"duration_ms":64073,"significance":"If the mode assignments hold, the paper provides a useful demonstration of reconfigurable magnon modes in a low-damping garnet, with systematic variation of field and pumping geometry and direct real-space imaging of static states. The spatial power maps that distinguish domain-wall, flux-closure, bulk, and hybridized modes are a strength, and the anisotropy-energy analysis gives a concrete mechanism for the observed domain evolution. The mode spatial profiles are emergent from the simulations rather than fitted mode-by-mode, which provides some independence. However, the comparison is weakened by partly fitted simulation parameters and by the explicit exclusion of the experimentally observed zig-zag state from the hard-axis simulations.","major_comments":[{"comment":"The hard-axis low-field mode assignments in Fig. 5(d) are not securely connected to the experimental state. Section II.A states that the 2D cross-section simulation 'prevents the possibility of simulating the zig-zag state,' and Fig. 1(d) shows that at 3 mT (the field at which the Fig. 5(d) power maps are taken) the experimental domain pattern is again approaching the zig-zag configuration. Consequently, the simulated 2D ground state and the real 3D state can differ qualitatively, and the claimed cubic-anisotropy-induced splitting of Modes 1-2 could equally arise from the unmodelled zig-zag modulation along the stripe direction rather than from the proposed E(1¯10) asymmetry. The authors should either provide 3D simulations that capture the zig-zag state, or explicitly restrict the hard-axis mode identification to collinear-stripe fields and present the low-field hard-axis assignments as tentative.","section":"Section II.A, Fig. 1(d), Fig. 5(d)"},{"comment":"The simulation parameters are partly fitted to the target observables: Appendix A reports Ku = 1000 J/m^3 chosen to best reproduce the experimental results, whereas the measured value is 1400 ± 200 J/m^3; A = 6.5 × 10^-12 J/m is a compromise chosen to prevent domain division rather than a measured value; and the simulated saturation field is roughly twice the experimental value. Since the ground state on which the mode profiles sit is determined by these parameters, the simulated spectra are not an independent prediction. The claim in Section II.B that the simulations 'still allow us to elucidate the origin of the resonant mode observed' should be supported by a sensitivity analysis demonstrating that the mode ordering, spatial profiles, and the hard-axis splitting are robust to variations of Ku and A within their stated uncertainties, and by an explicit discussion of how the saturation-field mismatch affects the comparison of field-dependent mode trajectories.","section":"Appendix A, Section II.B"},{"comment":"The simulations impose a fixed lateral period via periodic boundary conditions and a fixed simulation width, while the experimental stripe period varies substantially with field (e.g., from 2.8 to 5.2 µm for the [11¯2] direction in Fig. 1(f)). The authors mention this as a possible cause of the overestimated saturation field, but they do not quantify its impact on the resonance frequencies or the field positions of the mode anticrossings. A comparison of simulated spectra at several periods, or an explicit statement of which conclusions are insensitive to the lateral period, is needed to assess the quantitative reliability of the simulated field-frequency maps.","section":"Section II.A, Fig. 1(e-f)"}],"minor_comments":[{"comment":"The Fig. 1 caption appears to swap the field directions for panels (c) and (d): it reads 'for the field being applied along either [11¯2] direction (c) or [1¯10] direction (d)', whereas the text states that Fig. 1(c) shows the [1¯10] (easy-axis) evolution and Fig. 1(d) shows the [11¯2] (hard-axis) evolution.","section":"Fig. 1 caption"},{"comment":"The software name is inconsistently rendered: the text says 'micromagnetic simulations (mu-max3)' while the reference list gives 'MuMax3'; please use a single consistent spelling.","section":"Section II.A"},{"comment":"The Fig. 2 caption does not describe panels (g) and (h), although the text refers to Fig. 2(g) and Fig. 2(h) for the cubic anisotropy energy density plots.","section":"Fig. 2 caption"},{"comment":"There are minor spacing/grammar issues, including 'anisotropy,with good correlation' in the abstract, 'energy density,E,(Fig. 2(g))' in Section II.A, and 'The qualitative agreement lead us' which should be 'leads us'.","section":"Abstract and Section II.A"},{"comment":"The sentence 'Figure 3 presents spinwave spectra and simulation results for the perpendicular pumping with Hext applied along the easy axis Fig. 3(a).' is awkwardly worded; consider revising to '... along the easy axis (Fig. 3(a)).'","section":"Section II.B"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the journal's scope and presents a useful experimental data set. The main risk is that the hard-axis low-field mode assignments and the inferred role of cubic anisotropy rest on a 2D simulation that explicitly cannot represent the zig-zag state observed at those fields. This is reparable by restricting the claims or adding 3D simulations, but as it stands the load-bearing comparison for one of the four geometries is not fully supported. A parameter-sensitivity study would also strengthen the circularity concerns raised by the fitted Ku and A."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is a careful experimental study of spin-wave resonances in 3 µm YIG stripe domains, backed by micromagnetic simulations that assign each observed mode to a spatial region. It is the first detailed mode-resolved map in pure YIG, and the combination of MOKE imaging, two pumping geometries, and two in-plane field directions gives a thorough picture. The easy-axis spectra and mode profiles (domain-wall modes, flux-closure caps, bulk stripes, hybridized modes) look convincing, and the paper is honest about its simulation limitations.\n\nThe main soft spot is the hard-axis low-field regime. The paper admits the 2D cross-section simulation cannot capture the zig-zag state, but Fig. 5(d) assigns Modes 1–5 using power maps at 3 mT, where the experimental state is close to zig-zag. The claimed cubic-anisotropy-induced splitting of the two stripe modes rests on an asymmetry in E(1-10), which is a real mechanism in the 2D geometry, but the actual 3D state is modulated along the stripe direction. That splitting could come from the zig-zag rather than the proposed mechanism. This does not sink the paper, but the hard-axis low-field assignments are not independently secured.\n\nThe parameter choices in Appendix A are also a concern, though not a hidden one. Ku is fitted to reproduce the spectra, A is a compromise to avoid domain division, and Kc is from literature. The saturation field in simulation is roughly double the experiment. The mode spatial profiles and hybridizations are emergent, so they carry some weight, but the agreement is qualitative rather than predictive. The paper would be stronger with a parameter sensitivity study and a discussion of whether the hard-axis mode splitting survives in a 3D simulation.\n\nWho this is for: magnonics people working on reconfigurable spin-texture devices, and simulation groups looking for benchmark data. It is not a breakthrough, but it is useful data and a plausible mode catalog. The lack of raw data and code is a minor omission.\n\nMy recommendation: send it to peer review. The experimental core is solid, the limitations are stated, and the hard-axis ambiguity is addressable with additional simulations or a softened claim. I would expect heavy revision, not a desk reject.","headline":"Solid experimental map of stripe-domain magnon modes in pure YIG, with credible but partly hand-tuned simulations; the hard-axis low-field mode assignments are the weakest link because the 2D model cannot represent the zig-zag state.","tokens_in":19204,"tokens_out":1254,"would_cite":true,"duration_ms":15069,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["75.30.Ds","75.70.Kw","75.50.Gg"],"model":"deepseek-v4-flash","headline":"This paper argues that a 3-µm yttrium iron garnet film with weak perpendicular anisotropy hosts a rich set of spin-wave resonances in its stripe-domain state, and that micromagnetic simulations reproduce the spectra well enough to assign…","keywords":["yttrium iron garnet","stripe domains","spin wave resonance","micromagnetic simulation","cubic anisotropy","magneto-optic Kerr effect","mode hybridisation","magnonics"],"falsifier":"Phase-resolved imaging of the spin-wave amplitude across the film thickness during resonance (for example time-resolved scanning transmission X-ray microscopy or depth-sensitive Brillouin light scattering) would directly test the predicted localisation of the lowest mode on the 90° domain walls and the second mode at the flux-closure caps; if the measured amplitude maxima appear at different positions or with different field dependence, the assignments are wrong.","tokens_in":18202,"feed_emoji":"🧲","tokens_out":6659,"duration_ms":61823,"temperature":0.7,"pith_summary":"This paper studies a 3-micrometre yttrium iron garnet film whose small perpendicular anisotropy makes its magnetisation break into stripe domains. The authors measure the domain patterns with magneto-optic Kerr microscopy and the spin-wave resonances with broadband microwave absorption, then interpret both with micromagnetic simulations. Their central claim is that the simulations reproduce both the static domain evolution and the resonance spectra closely enough to assign every observed mode to a specific part of the texture: 90-degree domain walls, flux-closure caps at the film surfaces, the interior of the stripes, and hybridised higher-order modes. The point of doing this is that non-collinear textures could serve as reconfigurable magnonic elements, controlled by small magnetic fields and by the cubic anisotropy of the garnet.","feed_headline":"Stripe-domain YIG hides a spectrum of spin-wave resonances","feed_subtitle":"Simulations assign each resonance to a domain wall, flux-closure cap, or stripe bulk, pointing to reconfigurable magnonics.","key_machinery":"The argument is carried by a micromagnetic model of a single stripe-period cross-section of the film, with periodic boundary conditions, in which the static equilibrium is computed for each field and the resonance modes are extracted from the Fourier transform of the time-dependent magnetisation. The model supplies the spatial power maps that let the authors identify each experimental resonance with a localised excitation, and it also reproduces the static domain evolution, including the slanting and asymmetric widening of stripes. The load-bearing energy term is the first-order cubic anisotropy expression for a (111)-oriented film, whose asymmetry under polar rotation explains the different static evolution along the easy and hard in-plane directions and hence the different resonance spectra observed in the two field geometries.","core_discovery":"The authors find that the stripe-domain state of a 3 µm YIG film supports a much richer set of standing spin-wave resonances than the saturated state, and that the individual resonances are localised: the lowest mode sits on the 90° domain walls between the stripe moments and the flux-closure caps, another mode concentrates at the top and bottom surfaces of the stripes, a third is a bulk mode of the stripe interior, and the higher modes are hybridised combinations of bulk, cap, and edge modes. Pumping geometry selects which family is excited: perpendicular rf pumping excites acoustic modes, while parallel pumping excites optical modes. The field evolution of the static pattern—stripe slanting along the easy axis and asymmetric stripe widening along the hard axis—is traced to the odd-in-polar-angle terms of the cubic anisotropy energy in the (11̄0) plane, and the same asymmetry produces a frequency splitting between modes localised on stripes of opposite out-of-plane magnetisation. With the chosen simulation parameters, computed spectra and power maps correlate well with experiment, which is the evidence for the mode assignments.","pith_inferences":["If the spatial profiles are confirmed by direct imaging, the 90° domain-wall mode could act as a reconfigurable spin-wave conduit whose path is set by the stripe orientation, which the cubic anisotropy already rotates with the saturation direction.","The predicted frequency splitting between modes on opposite-polarity stripes could be read out as a field-controlled frequency difference, suggesting a compact magnetic-field sensor or a tunable two-tone magnonic source.","A testable extension would be to vary film thickness across the critical thickness for flux-closure caps: the theory here would predict that cap-localised and hybridised modes change strength and frequency systematically, which could be checked in a single wedged film.","Because the simulations are two-dimensional and cannot capture the zig-zag state, extending to three-dimensional simulations could reveal whether the zig-zag itself hosts additional resonances not seen in the present cross-section; the paper leaves this open."],"forward_implications":["If the mode assignments are right, the spin-wave response of a garnet film can be reshaped without any etching or irradiation, simply by moving the stripe domains with a few millitesla of in-plane field.","The selection of acoustic versus optical modes by the pumping geometry gives a way to excite different standing-wave patterns in the same magnetic texture.","The cubic-anisotropy-driven splitting of modes on stripes of opposite magnetisation means that the anisotropy provides a field-tunable knob for mode frequencies, not just a static background.","The hybridisation between bulk, edge, and cap modes in the intermediate-field regime implies that avoided crossings can be engineered in the resonance spectrum by controlling the stripe period.","The simulation strategy of using a two-dimensional periodic cross-section can be carried over to other low-damping films with stripe domains to predict their resonance spectra."],"supporting_citations":[{"why":"Supplies the micromagnetic simulation engine used to compute static domain states and resonance spectra.","marker":"[57]"},{"why":"Provides the domain-theory framework, including flux-closure caps, stripe widths, and the critical-thickness condition.","marker":"[34]"},{"why":"Earlier calculation of ferromagnetic resonance spectra in weak stripe domains that this work's mode assignments build on and extend.","marker":"[43]"},{"why":"Identifies resonances of two-dimensional wall structures and flux-closure caps in stripe domains, the basis for the low-order mode assignments.","marker":"[42]"},{"why":"Earlier study of magnetostatic wave propagation in YIG stripe domains, including the zig-zag state observed here.","marker":"[48]"},{"why":"Source of the cubic anisotropy energy expression used to explain the field evolution of the domain pattern.","marker":"[59]"},{"why":"Supports the growth-induced origin of the perpendicular anisotropy in (111)-oriented YIG films.","marker":"[53]"},{"why":"Provides reference values for YIG exchange stiffness against which the paper's chosen value is set.","marker":"[68]"}],"fun_headline_variants":["Stripe-domain YIG hosts localized spin-wave resonance modes","Pumping geometry selects spin-wave families in stripe-domain YIG","Cubic anisotropy splits spin-wave modes in YIG stripe domains","Stripe-domain YIG reveals domain-wall-localized spin waves"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire mode-identification scheme rests on computer simulations whose magnetic parameters were hand-tuned (anisotropy set below the measured value, exchange stiffness raised above the literature value, and cubic anisotropy taken as an approximation), so inaccurate parameters would shift or invalidate the spatial profiles that label each resonance.","fun_headline_variants_meta":{"raw":{"variants":["Stripe-domain YIG hosts localized spin-wave resonance modes","Pumping geometry selects spin-wave families in stripe-domain YIG","Cubic anisotropy splits spin-wave modes in YIG stripe domains","Stripe-domain YIG reveals domain-wall-localized spin waves"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000702,"raw_usage":{"total_tokens":3141,"prompt_tokens":893,"completion_tokens":2248,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":509,"completion_tokens_details":{"reasoning_tokens":2178}},"tokens_in":509,"tokens_out":2248,"duration_ms":16006,"temperature":1.0,"reasoning_tokens":2178,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:54:26.370502+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Phase-resolved imaging of the spin-wave amplitude across the film thickness during resonance (for example time-resolved scanning transmission X-ray microscopy or depth-sensitive Brillouin light scattering) would directly test the predicted localisation of the lowest mode on the 90° domain walls and the second mode at the flux-closure caps; if the measured amplitude maxima appear at different positions or with different field dependence, the assignments are wrong.","supporting_citations":[{"cited_title":"Vukadinovic, M","cited_arxiv_id":null,"evidence_quote":"Earlier calculation of ferromagnetic resonance spectra in weak stripe domains that this work's mode assignments build on and extend."},{"cited_title":"Ebels, L","cited_arxiv_id":null,"evidence_quote":"Identifies resonances of two-dimensional wall structures and flux-closure caps in stripe domains, the basis for the low-order mode assignments."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier study of magnetostatic wave propagation in YIG stripe domains, including the zig-zag state observed here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the cubic anisotropy energy expression used to explain the field evolution of the domain pattern."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supports the growth-induced origin of the perpendicular anisotropy in (111)-oriented YIG films."},{"cited_title":"Measurements of the exchange stiffness of YIG films using broadband ferromagnetic resonance tech- niques","cited_arxiv_id":null,"evidence_quote":"Provides reference values for YIG exchange stiffness against which the paper's chosen value is set."}],"review_version":1}