{"id":"e91beac8-f910-423a-a658-eb50b414f22a","arxiv_id":"2509.05050","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Spin waves in Ga:YIG nanowaveguides propagate at up to 600 m/s nearly independent of waveguide width, with decay lengths of 7 to 10 micrometers, several times faster than in YIG.","lead":"Researchers measured spin waves in gallium-substituted yttrium iron garnet nanowires as narrow as 145 nm and found they travel up to 600 meters per second, much faster than in plain YIG. The result suggests a practical path for shrinking magnon-based computing circuits to the nanoscale.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Wavevector calibration rests on unvalidated adjusted Ms (22.76→17.51 kA/m) used in the same model that is compared with experiment; if incorrect, the reported vg(k) scaling and width-independence at fixed k weaken.","rationale":"The reader's weakest assumption is the adjusted Ms, and I agree it is the most load-bearing concern. The raw time-of-flight measurements are credible and not disputed; the issue is the model-dependent k-axis. Because Ms appears in both the frequency and the exchange-length terms, its 23% reduction is not a small perturbation. The paper flags the adjustment but provides no independent support, and the same model is used both to generate the theoretical curve and to assign the independent variable of the experimental points. This does not make the paper fraudulent or the 600 m/s measurement wrong; it makes the specific quantitative agreement and the fixed-k comparison conditional. A single independent measurement of the patterned-film Ms would settle this, and if it confirms 17.51 kA/m, the central claim is much stronger. I therefore keep the reader's CONDITIONAL verdict (UNCHANGED) rather than escalating to rejection.","tokens_in":10139,"tokens_out":11572,"duration_ms":132389,"concrete_test":"Measure the saturation magnetization of the same Ga:YIG film after nanopatterning (e.g., SQUID/VSM on a large array of waveguides, or FMR on the patterned film) to determine whether the effective Ms is 17.51 kA/m or remains near 22.76 kA/m. Then recompute Eq. (3) and the TetraX dispersion with that measured Ms, reassign kx for each experimental excitation frequency (e.g., 6.95 GHz and the red-dashed frequencies in Fig. 2(c)), and re-plot Fig. 2(b). A decisive outcome: if the measured patterned Ms is 22.76 kA/m and the reassigned experimental vg points move off the predicted curves by more than the stated error bars, the adjusted Ms is the load-bearing element and the central quantitative claim is not supported as stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section II.C states that Ms was adjusted from the VSM film value of 22.76 kA/m to 17.51 kA/m 'to account for potential modifications during the nanofabrication process,' but no measurement on the patterned film or waveguides is reported. This single parameter enters Eq. (3) both through ωM=γμ0Ms and through the exchange length λex=sqrt(2Aex/(μ0Ms^2)); lowering Ms by 23% therefore raises the exchange contribution and changes the frequency-to-wavevector map used to place every experimental group-velocity point in Fig. 2(b). The adjusted Ms is also used in the TetraX dispersion whose output is compared with the same experimental vg values. Because the model is used both to predict vg(k) and to assign k to the measured frequencies, the agreement in Fig. 2(b) and the claim that the 240/530-nm waveguides were measured at the same kx=11.25 rad/µm are not independent validations. This is a validation gap rather than an internal inconsistency: the time-of-flight vg values themselves are direct measurements, but the central quantitative comparison—'up to 600 m/s,' linear vg vs k, and YIG comparison at equal k—is conditional on an unverified parameter. If the true patterned Ms is closer to 22.76 kA/m, the experimental points will sit at different k values and may no longer fall on the model curves.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports time-resolved micro-focused Brillouin light scattering (µBLS) measurements of spin-wave group velocities and decay lengths in Ga:YIG nanowaveguides with widths of 145, 240, and 530 nm and thickness 73 nm, in backward-volume geometry. The experimental data are compared with an analytical dispersion relation (Eq. (3)) and finite-element micromagnetic simulations (TetraX). The central claim is that, because of the reduced saturation magnetization of Ga:YIG, the waveguides support exchange-dominated spin waves with group velocities up to 600 m/s, scaling almost linearly with wavevector and showing little dependence on waveguide width, in contrast to non-substituted YIG. The paper additionally reports decay lengths up to about 10 µm and argues that Ga:YIG enables faster and longer-range spin-wave transport in nanoscale magnonic devices.","tokens_in":10475,"tokens_out":6187,"duration_ms":67304,"significance":"If the quantitative comparisons are valid, the result is significant for nanoscale magnonics: it identifies a material route to high group velocities and long propagation distances in deeply scaled waveguides, with direct relevance to magnonic logic and interconnect concepts. The paper's main strengths are the direct time-of-flight measurement of group velocity (which is independent of the dispersion model), the combination of experiment, analytical theory, and full micromagnetic simulation, and the demonstration of propagation in waveguides as narrow as 145 nm. The principal weakness is that the wavevector axis used to compare experiment with theory relies on a single free parameter, the adjusted saturation magnetization, which is not independently validated for the patterned structures.","major_comments":[{"comment":"The paper adjusts the saturation magnetization from the VSM film value of 22.76 kA/m to Ms = 17.51 kA/m \"to account for potential modifications during the nanofabrication process,\" but no measurement on the patterned film or waveguides is provided. Ms enters Eq. (3) both through ωM = γ μ0 Ms and through the exchange length λex = sqrt(2Aex/(μ0 Ms^2)); the 23% reduction increases the exchange contribution by about 30%. The same dispersion is then used to assign the experimental frequencies to wavevectors (\"the specific wavevectors corresponding to the excited frequencies were calculated from the numerically obtained dispersion curve\"). Consequently, the agreement between the measured group velocities and the model curves in Fig. 2(b) is partly endogenous, and the quantitative claims—the linear vg(k) behavior, the 600 m/s value at a specific k, and the comparison with YIG at kx = 11.25 rad/","section":"Section II.C, Eq. (3), Fig. 2(b)"},{"comment":"The claim of width-independent group velocity is based on two measurements at 240 nm and 530 nm, reported only as numbers (529 ± 56 m/s and 524 ± 41 m/s) with no plotted data or table. The text states that these waves were \"excited at the same wavevector of kx = 11.25 rad/µm which roughly corresponds to the frequency of 6.95 GHz.\" However, Eq. (3) depends on waveguide width through the demagnetizing factors Fy and Fz, so the same frequency does not automatically correspond to the same wavevector in different widths. The manuscript does not show the dispersion curves for the 240 nm and 530 nm waveguides or the kx values actually used. Since width-independence is a central conclusion and the basis for the comparison with YIG, the data and the kx-assignment procedure must be presented explicitly for each width.","section":"Section III, Fig. 3 and text after Fig. 2"}],"minor_comments":[{"comment":"Typo: \"Ga:YIG as an suitable platform\" should be \"Ga:YIG as a suitable platform.\"","section":"Conclusion"},{"comment":"In the caption, \"1 st width mode\" should be \"1st width mode\" for consistency with the text.","section":"Fig. 2 caption"},{"comment":"Equation (3) is said to be \"implemented\" to analyze the fundamental mode; it may be clearer to state that the dispersion relation was solved or evaluated. Also, please ensure all symbols in Eqs. (4) and (5) are defined in the text (e.g., the integration variable ky is used without being explicitly introduced).","section":"Section II.C"},{"comment":"The individual experimental group-velocity values for the 145 nm waveguide are not listed with uncertainties in the text or figure caption; only the 240/530 nm values are given with error bars. Please provide a table of the measured vg, excitation frequency, and assigned kx for all waveguides so the reader can assess the scatter and the agreement with theory.","section":"Fig. 2(b) and Section III"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and addresses a timely topic in nanoscale magnonics. The direct time-of-flight group-velocity measurements are a strength and are not invalidated by the model-dependent kx assignment. However, the central quantitative comparisons all pass through the unvalidated adjusted Ms, and the width-independence claim lacks supporting data. I recommend major revision rather than rejection: the authors can address the concern by adding an independent measurement of Ms in the patterned structure or a sensitivity analysis, and by showing the 240/530 nm data and dispersion curves. If the authors can supply this, the paper would be an acceptable contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things before reading: the main experimental claim is direct time-of-flight BLS showing group velocities up to ~600 m/s in Ga:YIG nanowaveguides as narrow as 145 nm, and that is a genuinely new result that is independent of the model. The second thing is that the wavevector axis used to present those velocities is calibrated using an adjusted Ms (22.76→17.51 kA/m) with no independent measurement of the patterned film, and that same adjusted Ms enters the model curves the data are compared against. That is the paper's real soft spot, and the stress-test note is right about it.\n\nWhat is genuinely good: the time-resolved BLS methodology is careful—sigmoid fits to 1000 ns pulses over 8 µm, clear exponential decay fits, and the measured vg values are robust regardless of how Ms is set. The width-independence result (529±56 and 524±41 m/s for 240 and 530 nm wide guides) is striking and, if the kx assignment holds, a real advantage over YIG. The paper is well-written, the analytical and TetraX results agree, and the comparison to prior Ga:YIG film work (Böttcher et al., Carmiggelt et al.) is fair. This is a legitimate step beyond those studies: direct nanoscale propagation, not just film dispersion.\n\nThe soft spots are proportionate to the concern but not fatal. The adjusted Ms is a free parameter used twice: it sets the exchange length and the demag tensor, and it sets the k-axis for the experimental points in Fig. 2(b). If the true Ms is closer to the film value, the measured points shift in kx, and the nice linear vg vs k agreement could weaken. The authors flag this in Section II.C but don't provide any independent check—no patterned-film VSM, no FMR on the milled film, no fits to multiple modes that could pin Ms. The absence of a same-experiment YIG control is a lesser issue; the 5× comparison relies on literature values, which is acceptable but not ideal. There is also minor overreach in the conclusion: 'isotropic behavior' is inferred from two widths, not really established.\n\nWho is this for? Magnonics experimentalists and anyone thinking about nanoscale spin-wave logic. It deserves a serious referee, but conditional: the authors should be asked to either measure Ms on patterned films or argue convincingly why 17.51 is the right value, and to show how the k-axis error bar propagates into the claimed vg comparison.\n\nMy recommendation: peer review yes, with revision required. This is a solid, reproducible-looking experimental result with a calibration gap that needs closing, not a paper with no merit.","headline":"Solid experimental demonstration of fast exchange-dominated spin waves in Ga:YIG nanowires, with a genuine calibration caveat in the adjusted Ms that should be fixed but doesn't sink the main time-of-flight result.","tokens_in":11003,"tokens_out":1043,"would_cite":true,"duration_ms":13664,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Gallium-substituted YIG nanowaveguides carry exchange-dominated spin waves at up to 600 m/s, several times faster than pure YIG, with speed nearly independent of waveguide width.","keywords":["magnonics","Ga:YIG","gallium-substituted yttrium iron garnet","spin-wave propagation","exchange-dominated spin waves","group velocity","nanowaveguides","Brillouin light scattering"],"falsifier":"Measure the saturation magnetization or FMR spectrum of the actual patterned 145-nm Ga:YIG waveguide rather than the unpatterned film. If Ms comes out near 22.76 kA/m instead of 17.51 kA/m, recompute the dispersion: the wavevector assignments in the measured group-velocity plot move, and the reported 600 m/s value and width-independence likely fail to match simulation. Alternatively, measure group velocity at two well-separated wavevectors and check whether it stays linear through the origin.","tokens_in":10069,"feed_emoji":"🧲","tokens_out":8307,"duration_ms":79679,"temperature":0.7,"pith_summary":"This paper argues that gallium-substituted yttrium iron garnet (Ga:YIG) solves a scaling problem in magnonic devices: as waveguides shrink, conventional YIG spin waves slow down and decay quickly, but Ga:YIG's reduced saturation magnetization makes exchange interactions dominate even at long wavelengths. That turns the dispersion into a nearly parabolic curve, so the group velocity grows almost linearly with wavevector and barely changes with waveguide width. Combining micro-focused Brillouin light scattering on fabricated nanowaveguides (145–530 nm wide, 73 nm thick) with analytical dispersion calculations and finite-element micromagnetic simulations, the paper reports group velocities up to 600 m/s and decay lengths of several micrometers, several times better than comparable pure-YIG conduits. The practical stake: if true, Ga:YIG offers a path to nanoscale spin-wave circuits with predictable, geometry-tolerant transport.","feed_headline":"Ga:YIG nanowires carry spin waves at 600 m/s","feed_subtitle":"Reduced magnetization makes exchange dominate, so speed no longer depends on wire width—a plus for tiny magnonic circuits.","key_machinery":"The load-bearing object is the exchange-dominated dispersion relation for the fundamental mode in backward-volume geometry (magnetic field parallel to propagation), built from dynamic demagnetization factors for a rectangular waveguide plus an exchange term and the material's uniaxial anisotropy. With the reduced Ms of Ga:YIG, the exchange term dominates, making the dispersion near-parabolic in kx and giving group velocity proportional to kx regardless of width. The same model, using an adjusted Ms of 17.51 kA/m, matches both finite-element micromagnetic simulations and the measured BLS group velocities.","core_discovery":"The central claim is that Ga:YIG waveguides host exchange-dominated spin waves whose dispersion stays nearly parabolic in wavevector even at long wavelengths, because gallium substitution cuts the saturation magnetization by about an order of magnitude relative to YIG. As a result, group velocity grows almost linearly with kx and barely changes with waveguide width, reaching about 600 m/s at kx≈11.25 rad/µm in a 145 nm × 73 nm guide. The paper reports at least a fivefold advantage over pure YIG at the same wavevector and dimensions, with decay lengths of 7.43 µm in the narrowest guide and up to 10.2 µm in wider ones.","pith_inferences":["If the adjusted Ms is close to the true patterned-film value, the width-independence means designers can fix spin-wave velocity by frequency or wavevector rather than by controlling waveguide width, simplifying inverse design of magnonic circuits.","A direct test of the exchange-dominated picture would be to measure vg over a wider kx range and check strict linearity; any downward bend would signal the return of dipolar effects at long wavelengths.","Because the reduced Ms also changes the demagnetizing-field landscape, the width-independence may not survive in Damon–Eshbach (transverse-field) geometry; measuring group velocity under transverse fields would map the boundaries of the claimed isotropy.","The same mechanism suggests a material-design lever: tuning gallium content tunes Ms and hence the exchange length, so even faster or slower waves could be engineered for impedance matching between waveguide sections."],"forward_implications":["At equal wavevector (≈11.25 rad/µm) and comparable dimensions, Ga:YIG nanowaveguides give measured group velocities of 529–600 m/s versus simulated 97–106 m/s for pure YIG: at least a fivefold gain.","Group velocity stays essentially flat across widths from 145 nm to 530 nm (529±56 m/s and 524±41 m/s at the same kx), so transport speed can be set by excitation frequency rather than lithography.","Decay lengths of 7.43 µm (145 nm wide) up to 10.2 µm (530 nm wide) exceed those reported for non-substituted YIG nanowaveguides of similar size.","The exchange-dominated, monotonically rising dispersion excites a single wavevector per frequency with well-separated width modes over 6.45–8.14 GHz, enabling selective single-mode operation.","Numerical predictions for 50×50 nm cross-sections (613 m/s Ga:YIG vs 70 m/s YIG) indicate the advantage persists at smaller scales."],"supporting_citations":[{"why":"Supplies the experimental comparison baseline: group velocities and decay lengths in non-substituted YIG nanowaveguides that Ga:YIG is claimed to outperform, plus the BLS normalization used here.","marker":"[19]"},{"why":"Defines the Ga:YIG material: liquid-phase-epitaxy growth, VSM saturation magnetization, and stress-induced uniaxial anisotropy parameters entering the model.","marker":"[25]"},{"why":"Prior demonstration of fast long-wavelength exchange spin waves in Ga:YIG; provides the exchange stiffness Aex adopted for the dispersion calculations.","marker":"[27]"},{"why":"Supports the claim that group velocity in Ga:YIG has minimal dependence on waveguide width.","marker":"[31]"},{"why":"Supplies the analytical dispersion framework for nanoscopic waveguides, including dynamic demagnetization factors and the unpinned-edge assumption used in Eq. (3)-(5).","marker":"[40]"},{"why":"Finite-element micromagnetic package used for numerical dispersion, lifetime, and antenna absorption simulations compared with experiment.","marker":"[42]"},{"why":"Dynamic-matrix method underlying the finite-element spin-wave dispersion and lifetime calculations.","marker":"[43]"},{"why":"Micro-focused Brillouin light scattering technique that enables the time-resolved group-velocity and decay-length measurements.","marker":"[37]"}],"fun_headline_variants":["Ga:YIG nanowires boost spin-wave speed to 600 m/s","Narrow Ga:YIG guides keep spin waves fast and long-lived","Exchange spin waves in Ga:YIG: uniform speed down to 145 nm","Ga:YIG nanowaveguides: spin waves at 600 m/s, width-independent","Spin waves in Ga:YIG nanowires: faster, longer, width-agnostic"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The entire wavevector-versus-velocity interpretation rests on the choice of saturation magnetization Ms = 17.51 kA/m, adjusted downward from the measured film value 22.76 kA/m to account for nanofabrication, without an independent measurement of the patterned film; if the true Ms differs, the dispersion curves, extracted wavevectors, and claimed agreements shift.","fun_headline_variants_meta":{"raw":{"variants":["Ga:YIG nanowires boost spin-wave speed to 600 m/s","Narrow Ga:YIG guides keep spin waves fast and long-lived","Exchange spin waves in Ga:YIG: uniform speed down to 145 nm","Ga:YIG nanowaveguides: spin waves at 600 m/s, width-independent","Spin waves in Ga:YIG nanowires: faster, longer, width-agnostic"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000154,"raw_usage":{"total_tokens":1050,"prompt_tokens":751,"completion_tokens":299,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":495,"completion_tokens_details":{"reasoning_tokens":192}},"tokens_in":495,"tokens_out":299,"duration_ms":3239,"temperature":1.0,"reasoning_tokens":192,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T05:39:12.581615+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the saturation magnetization or FMR spectrum of the actual patterned 145-nm Ga:YIG waveguide rather than the unpatterned film. If Ms comes out near 22.76 kA/m instead of 17.51 kA/m, recompute the dispersion: the wavevector assignments in the measured group-velocity plot move, and the reported 600 m/s value and width-independence likely fail to match simulation. Alternatively, measure group velocity at two well-separated wavevectors and check whether it stays linear through the origin.","supporting_citations":[{"cited_title":"Heinz, T","cited_arxiv_id":null,"evidence_quote":"Supplies the experimental comparison baseline: group velocities and decay lengths in non-substituted YIG nanowaveguides that Ga:YIG is claimed to outperform, plus the BLS normalization used here."},{"cited_title":"Dubs and O","cited_arxiv_id":null,"evidence_quote":"Defines the Ga:YIG material: liquid-phase-epitaxy growth, VSM saturation magnetization, and stress-induced uniaxial anisotropy parameters entering the model."},{"cited_title":"B¨ ottcher, M","cited_arxiv_id":null,"evidence_quote":"Prior demonstration of fast long-wavelength exchange spin waves in Ga:YIG; provides the exchange stiffness Aex adopted for the dispersion calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supports the claim that group velocity in Ga:YIG has minimal dependence on waveguide width."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the analytical dispersion framework for nanoscopic waveguides, including dynamic demagnetization factors and the unpinned-edge assumption used in Eq. (3)-(5)."},{"cited_title":"K¨ orber, G","cited_arxiv_id":null,"evidence_quote":"Finite-element micromagnetic package used for numerical dispersion, lifetime, and antenna absorption simulations compared with experiment."},{"cited_title":"K¨ orber, G","cited_arxiv_id":null,"evidence_quote":"Dynamic-matrix method underlying the finite-element spin-wave dispersion and lifetime calculations."},{"cited_title":"Sebastian, K","cited_arxiv_id":null,"evidence_quote":"Micro-focused Brillouin light scattering technique that enables the time-resolved group-velocity and decay-length measurements."}],"review_version":1}