{"id":"d815708b-4642-4f71-9717-1fb31382b0dc","arxiv_id":"2412.09704","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"BLS measurements validate TETRAX simulations of spin-wave dispersions in CoFeB/NiFe bilayers, and the simulations show how layer thickness and magnetization tune nonreciprocity to a maximum of 5.2 GHz.","lead":"Spin waves in a bilayer of two magnetic metals can have different frequencies when traveling in opposite directions. Researchers measured this effect in cobalt-iron-boron/nickel-iron stacks with laser light and used the data to validate a computer model that maps how to maximize the difference up to 5.2 GHz.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 5.2 GHz optimum is an unmeasured extrapolation: parameter sensitivity of the MSSW/PSSW hybridization at k≈40 rad/µm is not quantified, so the quantitative headline is not yet load-bearing.","rationale":"The paper's core demonstration—that TETRAX reproduces measured BLS dispersions for three NiFe/CoFeB bilayers and that nonreciprocity is governed by MSSW/PSSW hybridization—is credible. The code is open source, the data and scripts are deposited, and the field-dependent check at k = 6 rad/µm is a useful independent probe. My concern is not with the simulation method or with the existence of nonreciprocity, but with the precise quantitative headline: the 5.2 GHz optimum sits at k ≈ 40.5 rad/µm, outside the 20 rad/µm experimental window, at a composition close to but not identical to the measured stacks. Because the mechanism is a hybridization crossing, the result is sensitive to A and Ms values that are taken from the literature rather than fitted or error-bounded. The measured data constrain the model at low k but cannot constrain the exact location of an avoided crossing at twice the measured wave-vector range. A parameter-sensitivity rerun of the public scripts would settle whether this concern actually moves the prediction: if Δfmax remains near 5 GHz across the literature range of A, the concern is minor; if it varies by several GHz, the stated number should be downgraded. Either way, the reader's CONDITIONAL verdict remains appropriate, so the verdict is UNCHANGED.","tokens_in":15628,"tokens_out":5465,"duration_ms":60554,"concrete_test":"Re-run the published TETRAX optimization scripts for the 50 nm total-thickness map with the exchange constants moved to the endpoints of the literature range (A_NiFe = 10.5 and 15 pJ/m; A_CoFeB = 12 and 20 pJ/m), keeping Ms and all other inputs fixed. Record Δfmax, kmax, and the optimum NiFe fraction for each corner. If any corner shifts Δfmax by more than 1 GHz or moves the optimum ratio by more than 5 nm relative to the nominal 23/27 nm optimum, the headline value should be re-reported as illustrative and the 23/27 nm stack should be measured by BLS within the accessible k range to re-anchor the parameters.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is Δfmax = 5.2 GHz for the u=1 mode at k = 40.5 rad/µm for tNiFe=23 nm and tCoFeB=27 nm. The experimental validation covers only 0–20 rad/µm and three thickness combinations, so the peak is an extrapolation of a model that is never checked near the predicted optimum. The valid part of the model—TETRAX's dynamic-matrix calculation and the 1D thickness line—is not the main risk: for an infinite laterally homogeneous film, a thickness-only mesh is the natural translational-invariance reduction. The real load-bearing assumption is that fixed literature values (Ni81Fe19: Ms=830 kA/m, A=13 pJ/m; Co40Fe40B20: Ms=1260 kA/m, A=16 pJ/m) remain exact for the sputtered films over the whole sweep. The large nonreciprocity is attributed to crossing and hybridization of MSSW and the first PSSW mode; the PSSW frequency depends on A and thickness, and A for sputtered CoFeB/NiFe varies by tens of percent with composition and growth conditions. At k≈40 rad/µm, exchange contributions are substantial, so a modest A error shifts the avoided crossing and can change both Δfmax and kmax. The paper reports qualitative 'excellent agreement' without error bars or a parameter fit, so it does not bound this sensitivity. The mechanism itself is plausible and supported in the measured range, but the precise 5.2 GHz number is not yet established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates nonreciprocal spin-wave propagation in NiFe/CoFeB bilayer ferromagnetic films, combining Brillouin light scattering (BLS) measurements with eigenmode simulations from the finite-element package TETRAX. The authors measure dispersions for three bilayer thickness combinations and a field-dependent spectrum, report good visual agreement with TETRAX simulations, and then use the validated model to sweep saturation magnetization, layer thickness ratio, and total thickness. They identify the hybridization of the magnetostatic surface wave (MSSW) with the first perpendicular standing spin wave (PSSW) as the origin of large nonreciprocity, and report a maximum frequency splitting of Δfmax = 5.2 GHz for the u = 1 mode at k = 40.5 rad/µm for tNiFe = 23 nm and tCoFeB = 27 nm, with similar magnitudes for u = 0. The paper also discusses mode profiles, anti-Larmor precession, and a comparison with CoFe/YIG-based bilayers.","tokens_in":15922,"tokens_out":2307,"duration_ms":26125,"significance":"If the findings hold, the paper offers a useful design rule for engineering large nonreciprocal spin-wave responses in all-metallic bilayers, with the measured data providing an independent validation of the TETRAX eigenmode approach. The experimental validation over three thickness combinations and a field range, together with the open data and scripts, is a concrete strength. The mechanistic explanation in terms of MSSW/PSSW hybridization is plausible and consistent with the measured dispersions. However, the headline quantitative result, the 5.2 GHz maximum, is a simulation prediction at wave vectors outside the BLS-accessible range, and the manuscript does not quantify the sensitivity of this prediction to the assumed material parameters. The central qualitative mechanism is well supported, but the precise numerical headline requires additional uncertainty analysis before it becomes load-bearing.","major_comments":[{"comment":"The headline quantitative claim, Δfmax = 5.2 GHz for the u = 1 mode at kmax = 40.5 rad/µm, is obtained from simulations at wave vectors beyond the measured range (the BLS data in Fig. 2 cover only up to 20 rad/µm). The paper does not provide error bars, a parameter sensitivity study, or a quantitative comparison with experiment in that regime. Since this number is the central 'optimized conditions' result of the abstract, the authors should either measure at larger wave vectors (e.g., using a different technique or a higher-index BLS setup) or provide a sensitivity analysis showing how Δfmax and kmax vary under reasonable uncertainties in Ms and A. Without this, the quantitative headline is an unverified extrapolation.","section":"Results, 'Changing layer thickness' (Fig. 7)"},{"comment":"The simulations fix the material parameters to literature values (Ni81Fe19: Ms = 830 kA/m, A = 13 pJ/m; Co40Fe40B20: Ms = 1260 kA/m, A = 16 pJ/m) for all thicknesses and wave vectors, with the statement 'we keep the same parameters for all simulations.' The PSSW frequency, and hence the MSSW/PSSW hybridization that drives the nonreciprocity, depends strongly on A and Ms. Sputtered films can deviate from nominal composition and bulk values by tens of percent. The paper should justify these values for the actual films (for example by independent FMR or SQUID characterization of the same samples) or quantitatively demonstrate that the predicted Δfmax and kmax are robust to within the expected parameter spread.","section":"Methods, 'Theory and simulation'"},{"comment":"The agreement between BLS data and TETRAX is described as 'excellent' and 'agree well,' but no quantitative metric is provided. The dispersions in Fig. 2 and the field dependence in Fig. 3 are overlaid as lines on color maps without extracted peak positions, residuals, or confidence intervals. A quantitative comparison (e.g., extracted BLS mode frequencies versus simulated curves, with residuals and a measure of scatter) would strengthen the validation claim and would also provide a baseline for assessing the uncertainty of the extrapolated predictions.","section":"Results, 'Experimental verification of the simulation model' and Fig. 3"}],"minor_comments":[{"comment":"The software name is inconsistently rendered as 'TETRAX' and 'TETRA X'; please use a single consistent spelling.","section":"Abstract and throughout"},{"comment":"The corresponding author email contains a typo: 'h.schutheiss@hzdr.de' should be 'h.schultheiss@hzdr.de'.","section":"Author contributions / contact"},{"comment":"The penetration-depth formula in Eq. (1) uses ω and Im{εxx} without defining them in the text; please state that ω is the laser angular frequency and specify the source and value of the permittivity used.","section":"Eq. (1)"},{"comment":"In the text near Fig. 7, 'fmax = −4.8 GHz' for the u = 0 mode should read 'Δfmax = 4.8 GHz' (with the sign convention for the splitting) to avoid confusion with a frequency value.","section":"Results, 'Changing layer thickness'"},{"comment":"The caption phrase 'second to lowest' is awkward; consider 'second-lowest mode' or 'u = 1 mode'.","section":"Fig. 7 caption"},{"comment":"The term 'hetero-symmetric spin-wave' is used with references 40 and 41 but is not defined in the text; a brief definition or a more explicit connection to the mode profiles would improve readability.","section":"Discussion"}],"recommendation":"major_revision","confidential_remarks":"The use of TETRAX, which the authors co-develop, is not by itself problematic because the central validation is against independent BLS measurements. The main concern is scope-wise: the paper's quantitative claim (5.2 GHz at k≈40 rad/µm) is a prediction beyond the measured range, and the manuscript should either soften the headline or add a sensitivity analysis. If the authors provide such an analysis or reframe the result as a qualitative design prediction, the paper would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Honest take: this is a useful paper, and the core experimental validation is real. The new content is the systematic parameter map—layer ratio, total thickness, magnetization contrast—for CoFeB/NiFe bilayers, plus the mode-profile analysis that explains the nonreciprocity through MSSW/PSSW hybridization. The BLS data for three film stacks and the field-dependent measurements agree visually with TETRAX; the data and scripts are openly available, and the model itself is open-source. Credit where due: the authors don't hide that the strongest splitting is a simulation result at k≈40 rad/µm, beyond the BLS-accessible range, and they explicitly say so in the discussion.\n\nMy main reservation is the one you'd expect: the 5.2 GHz number is load-bearing but depends on fixed literature values of Ms and A that are never fitted or varied. At k≈40 rad/µm the exchange contribution is large, and the PSSW frequency—which sets the avoided crossing—is sensitive to A. A 10–20% change in A could shift both Δf and the peak wave vector noticeably. The paper reports 'excellent agreement' without error bars or a quantitative figure of merit, so the extrapolation isn't bounded. The 1D line-trace approximation is fine for laterally homogeneous films; that's not the issue.\n\nA minor point: the saturation of Δf with total thickness and the optimum layer ratio are robust qualitative findings that survive even if the exact peak value shifts. The qualitative mechanism—hybridization of MSSW and first PSSW, with anti-Larmor precession in thicker films—is well supported by the measured range and the mode profiles.\n\nWorth a serious referee. The experimental core deserves publication; the quantitative peak should be reframed as a testable prediction, not a measured result. I'd ask the authors for a sensitivity analysis on A and Ms, or at least a clear statement of expected uncertainty. For magnonics researchers, it's a useful reference, and I'd cite it if I worked on spin-wave nonreciprocity.","headline":"A solid simulation-backed study of bilayer spin-wave nonreciprocity whose headline 5.2 GHz optimum is a credible but unmeasured extrapolation—worth refereeing, with a request for sensitivity analysis.","tokens_in":16507,"tokens_out":1941,"would_cite":true,"duration_ms":21485,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["75.30.Ds","75.70.Cn"],"model":"deepseek-v4-flash","headline":"Magnetic bilayer design predicts spin-wave frequency splitting of 5.2 GHz, verified against laser-scattering measurements.","keywords":["spin waves","magnonics","nonreciprocity","magnetic bilayer","Brillouin light scattering","TETRAX","MSSW","PSSW hybridization"],"falsifier":"Measure the spin-wave dispersion of a 23 nm NiFe / 27 nm CoFeB bilayer at a wave vector of 40.5 rad/µm (for example with propagating spin-wave spectroscopy in a narrow waveguide or grating-assisted BLS) and check whether the u = 1 mode frequencies for opposite propagation directions differ by 5.2 GHz at 30 mT, as the simulation predicts.","tokens_in":15433,"feed_emoji":"🧲","tokens_out":6670,"duration_ms":59930,"temperature":0.7,"pith_summary":"This paper establishes that the nonreciprocity of spin-wave dispersion in a bilayer of two different ferromagnets can be predicted and deliberately engineered by choosing layer thicknesses and saturation magnetizations. The authors verify the open-source finite-element package TETRAX against Brillouin light scattering measurements on three NiFe/CoFeB bilayers, finding excellent agreement between simulated and measured dispersions. Using the validated model, they sweep the parameter space and show that counter-propagating spin waves can differ in frequency by several gigahertz, with a maximum predicted splitting of 5.2 GHz for a 50 nm bilayer with 23 nm NiFe and 27 nm CoFeB. The origin of the effect is the hybridization of the Damon–Eshbach surface mode with the first perpendicular standing spin-wave mode. The result provides a concrete design rule for unidirectional spin-wave devices.","feed_headline":"Bilayer design hits 5.2 GHz spin-wave frequency split","feed_subtitle":"Thickness and magnetization choices make spin waves travel one way, with a predicted 5.2 GHz split.","key_machinery":"The engine of the paper is the TETRAX finite-element dynamic-matrix solver, which computes spin-wave eigenmodes directly from a one-dimensional line trace across the film thickness, so a full dispersion takes seconds on a laptop. The physical mechanism it exposes is the hybridization of the asymmetric Damon–Eshbach surface wave with the first perpendicular standing spin wave; this hybridization creates regions of anti-Larmor precession and makes the mode localization shift between the two layers for opposite propagation directions, producing the frequency splitting.","core_discovery":"The central claim is that in a NiFe/CoFeB bilayer the nonreciprocal spin-wave dispersion is primarily produced by the magnetostatic surface wave (MSSW) and its hybridization with the first perpendicular standing spin wave (PSSW), and that this nonreciprocity is tunable through the layer thicknesses and the saturation magnetization difference. The paper shows that the simulated dispersions match the measured ones for three bilayer thicknesses, and that the same simulation, extended to wave vectors beyond the experimental range, predicts a maximum counter-propagation frequency difference of 5.2 GHz for the u = 1 mode at k = 40.5 rad/µm in a 23 nm NiFe / 27 nm CoFeB stack. It further finds that the nonreciprocity saturates near 5 GHz once the total thickness exceeds about 40 nm, while the wave vector of the maximum splitting decreases with thickness, which is why the thickest measured film shows the strongest effect in the accessible wave-vector range.","pith_inferences":["The same MSSW–PSSW hybridization mechanism suggests that tuning the crossing point of these modes via exchange constant, anisotropy, or an applied field gradient could shift the optimum to lower wave vectors or higher frequencies, a testable extension the paper does not explore.","The predicted 5.2 GHz optimum at k ≈ 40 rad/µm sits beyond the momentum transfer of conventional BLS, so a direct check would require propagating spin-wave spectroscopy or grating-assisted BLS; a successful measurement would close the loop between prediction and experiment.","Since fixed literature parameters are used for all thicknesses, the real optimum in sputtered films could shift if interface intermixing or strain changes Ms or A; measuring the dispersion of a 23/27 nm bilayer is the cleanest way to test this.","The saturation of the splitting with thickness hints that dipolar mechanisms alone cap the nonreciprocity for this material pair, suggesting that reaching larger splittings will require combining thickness asymmetry with interfacial Dzyaloshinskii–Moriya interaction."],"forward_implications":["A designer can select layer thicknesses and materials for a target nonreciprocity before fabrication, rather than tuning by trial and error.","The optimum wave vector for maximum splitting moves from about 40 rad/µm down into the measurable range as total thickness grows beyond 40 nm, explaining why thick films are the best candidates for BLS-verifiable devices.","Higher magnetization contrast, such as CoFe instead of CoFeB, increases the splitting, while too large a contrast (as with YIG) suppresses it, so moderate contrast is the design sweet spot.","Because the splitting saturates with thickness, further gains beyond roughly 5 GHz would require changing the material system or adding other asymmetries.","The hybridized modes carry negative group velocity for one propagation direction, which could be used to create frequency-selective directional filtering."],"supporting_citations":[{"why":"Supplies the one-dimensional line-trace dynamic-matrix approach that makes the systematic parameter sweeps computationally feasible.","marker":"[31]"},{"why":"The open-source TETRAX package that the paper uses for all eigenmode and dispersion calculations.","marker":"[32]"},{"why":"Defines the magnetostatic surface wave whose asymmetric thickness profile is the seed of the nonreciprocity.","marker":"[1]"},{"why":"Introduces the propagating-wave dynamic matrix method that TETRAX implements for direct dispersion calculation.","marker":"[25]"},{"why":"Demonstrates a spin-wave diode in a bilayer, the application class this design rule targets.","marker":"[26]"},{"why":"Reports the hybridization of the MSSW with perpendicular standing spin waves that the paper identifies as the origin of the large splitting.","marker":"[33]"},{"why":"Describes the anti-Larmor precession or heterosymmetric mode character that appears in the hybridized modes.","marker":"[40]"}],"fun_headline_variants":["5.2 GHz spin-wave split in NiFe/CoFeB bilayers","Predicted 5.2 GHz spin-wave nonreciprocity","Thickness tunes unidirectional spin waves in bilayers","Bilayer design yields one-way spin waves","Spin-wave split predicted at 5.2 GHz in bilayers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The simulations keep the same literature values of saturation magnetization and exchange constant for NiFe and CoFeB at every thickness and assume a one-dimensional line trace across the thickness captures the physics at all wave vectors up to 100 rad/µm; if the sputtered films differ from these parameters or the one-dimensional approximation fails at large wave vectors, the predicted optimum and the 5.2 GHz value would change.","fun_headline_variants_meta":{"raw":{"variants":["5.2 GHz spin-wave split in NiFe/CoFeB bilayers","Predicted 5.2 GHz spin-wave nonreciprocity","Thickness tunes unidirectional spin waves in bilayers","Bilayer design yields one-way spin waves","Spin-wave split predicted at 5.2 GHz in bilayers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001006,"raw_usage":{"total_tokens":4235,"prompt_tokens":905,"completion_tokens":3330,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":521,"completion_tokens_details":{"reasoning_tokens":3243}},"tokens_in":521,"tokens_out":3330,"duration_ms":23484,"temperature":1.0,"reasoning_tokens":3243,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T16:49:35.099791+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the spin-wave dispersion of a 23 nm NiFe / 27 nm CoFeB bilayer at a wave vector of 40.5 rad/µm (for example with propagating spin-wave spectroscopy in a narrow waveguide or grating-assisted BLS) and check whether the u = 1 mode frequencies for opposite propagation directions differ by 5.2 GHz at 30 mT, as the simulation predicts.","supporting_citations":[{"cited_title":"K\\\" o rber , author A","cited_arxiv_id":null,"evidence_quote":"Supplies the one-dimensional line-trace dynamic-matrix approach that makes the systematic parameter sweeps computationally feasible."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the magnetostatic surface wave whose asymmetric thickness profile is the seed of the nonreciprocity."},{"cited_title":"Propagating spin-wave normal modes: A dynamic matrix approach using plane-wave demagnetizating tensors","cited_arxiv_id":"1611.06153","evidence_quote":"Introduces the propagating-wave dynamic matrix method that TETRAX implements for direct dispersion calculation."},{"cited_title":"Grassi , author M","cited_arxiv_id":null,"evidence_quote":"Demonstrates a spin-wave diode in a bilayer, the application class this design rule targets."},{"cited_title":"Tacchi , author R","cited_arxiv_id":null,"evidence_quote":"Reports the hybridization of the MSSW with perpendicular standing spin waves that the paper identifies as the origin of the large splitting."}],"review_version":1}