{"id":"8677b822-20f1-4f33-9f54-6dc945b5f758","arxiv_id":"2505.07401","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Even when spin waves are launched along the reciprocal direction of a synthetic antiferromagnet stripe, the waves acquire large transverse wavevectors and radiate diagonally, so 1D propagating-wave analysis fails.","lead":"Experiments and simulations on a synthetic antiferromagnet stripe show that spin waves do not travel straight along the stripe even when the magnetic configuration is reciprocal along that direction. The work reveals non-reciprocal behavior that matters for how spin-wave data are interpreted in magnonic devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central mechanism relies on unmeasured edge coupling to high-k_x spin waves; the micromagnetic model that should establish it fails in the 8.4 GHz regime.","rationale":"The paper's central claim depends on the assertion that a line antenna, invariant along x, excites spin waves with large k_x via non-uniform regions near the stripe edges. This mechanism is not directly observed, and the micromagnetic model that includes the edge state fails in the 8.4 GHz regime, which is precisely where the same edge-seeding picture is invoked. The reader's weakest_assumption identifies this same link, so I agree with the conditional verdict. The experimental BLS patterns are strong evidence that something non-reciprocal and non-1D is happening, but the specific interpretation in terms of high-k_x waves seeded by edge non-uniformities needs a decisive test before the PSWS-invalidation claim can be considered fully established.","tokens_in":12910,"tokens_out":4372,"duration_ms":48819,"concrete_test":"Run a micromagnetic simulation of the same 5 µm SAF stripe and antenna, but with the equilibrium magnetization artificially forced to be perfectly uniform in the near-edge regions (e.g., by locally suppressing the edge demagnetization fields), and compare the predicted BLS images at 3.65, 4.5, and 5.2 GHz with the published results. If the transverse nodal pattern and lower-left-corner decay persist unchanged, edge non-uniformity is not the source of high-k_x waves; if they disappear or weaken, the edge-seeding mechanism is confirmed. A complementary experimental check is to fabricate stripes with intentionally modified edge profiles and measure whether the transverse node contrast and the 8.4 GHz bidirectional beating respond accordingly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's most consequential result is that a line-shaped antenna, whose rf field is translationally invariant along x, excites spin waves with large k_x components (up to about 25 rad/µm) whose interference creates the observed transverse nodal patterns. Because an x-uniform drive cannot directly source k_x≠0, this requires the x-translational invariance to be broken by the non-uniform equilibrium magnetization and demagnetizing fields near the stripe edges [Fig. 1(c)]. The paper asserts this coupling (\"likely the tiny regions near the stripe edges\") but neither measures it nor quantifies the transduction efficiency. The inferred k_x values come from applying the unbounded-film dispersion to real-space nodal spacings, not from a wavevector-resolved measurement, and some inferred values (e.g., −25 rad/µm) exceed the BLS collection limit of 18 rad/µm. The supporting micromagnetic model includes the edge state, yet at 8.4 GHz [Fig. 4(c)] it fails to reproduce the bidirectional beating that the same edge-seeding picture is invoked to explain, with the authors attributing the failure to incorrect modeling of the near-edge inhomogeneity. If edge non-uniformities do not efficiently convert the y-uniform antenna field into high-k_x spin waves, the transverse-node signature and the claimed invalidation of the 1D PSWS model lose their stated mechanism. This is the least secure link in the argument: the experimental patterns are convincing, but the physical source of the transverse wavevectors remains an assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates spin-wave propagation in narrow synthetic antiferromagnet (SAF) stripes, with the static field applied perpendicular to the stripe axis. The authors use Brillouin light scattering (BLS) microscopy and micromagnetic simulations to show that, even when the stripe (conduit) is aligned with the reciprocal direction of the material, non-reciprocal spin-wave dispersion produces unconventional propagation patterns: transverse nodal structures, tilted directional energy beams, and two-dimensional beating. The authors interpret these patterns using the unbounded-film dispersion relation and propose that non-uniform magnetizations near the stripe edges serve as sources of high-wavevector spin waves with wavevectors perpendicular to the conduit. They argue that these effects invalidate the one-dimensional analysis commonly used in propagating spin-wave spectroscopy (PSWS) of non-reciprocal materials.","tokens_in":13244,"tokens_out":3848,"duration_ms":35827,"significance":"If the interpretation holds, the results demonstrate a conceptual limitation of 1D PSWS models for non-reciprocal systems and reveal a new channeling behavior where line antennas can excite transverse high-k waves. The experimental dataset is substantial: BLS images at four applied fields and many frequencies, supplemented by micromagnetic simulations that reproduce the main features after optical convolution. Material parameters are taken from an independent VNA-FMR study, not fitted to the BLS images. The paper explicitly acknowledges uncertainties in the dispersion analysis for curved contours and the failure of the micromagnetic model at 8.4 GHz. The main weakness is that the proposed edge-seeding mechanism for high-kx waves is not directly measured, and the supporting simulation fails in a key frequency range.","major_comments":[{"comment":"The central claim that the x-uniform antenna field excites spin waves with large |kx| depends entirely on the assertion that \"tiny regions near the stripe edges ... are likely\" the sources (p.4). This is not directly demonstrated. The micromagnetic simulation, which includes these edge non-uniformities, fails at 8.4 GHz to reproduce the observed bidirectional beating (Fig. 4(c)), as stated in the Discussion. This inconsistency leaves the edge-seeding mechanism as a hypothesis rather than an established mechanism. To strengthen the claim, the authors should either measure the kx content directly (e.g., by wavevector-resolved BLS or by spatial Fourier analysis of the BLS images) or perform a control simulation with artificially uniform edges.","section":"Physical understanding / Interpretation guidelines"},{"comment":"The inference of large negative kx (e.g., -25 rad/µm at 4 GHz, p.5) from nodal spacings uses the unbounded-film dispersion relation, but the experimental BLS set-up has a collection limit of k_BLS_max = 18 rad/µm (Methods). Spatial frequencies above this limit cannot be imaged, so the observed BLS patterns cannot confirm modes with |kx| > 18 rad/µm. The authors should clarify how modes with -25 rad/µm contribute to the BLS image, and whether the apparent secondary maximum at the bottom right corner could be due to edge reflections or other lower-kx modes.","section":"Dispersion relations / Fig. 3"},{"comment":"The paper claims that the 1D PSWS model is invalid for non-reciprocal materials, but the evidence for the coexistence of multiple kx modes is qualitative. To make this claim quantitatively load-bearing, the authors should compare the measured spatial profiles with the predictions of the 1D model (which assumes only ky propagation) and show directly that the discrepancy arises from the transverse components. A quantitative measure, such as the spectral weight at kx versus ky in the spatial Fourier transform of the BLS images, would substantiate the central claim and would also help quantify the transduction efficiency of the proposed edge sources.","section":"Qualitative understanding / Discussion"}],"minor_comments":[{"comment":"The logarithmic color scale is not defined in the caption, and the stripe edges are not marked on all panels, making it difficult to assess the confinement of the BLS signal.","section":"Fig. 2"},{"comment":"The group-velocity arrows are drawn schematically; the caption should indicate whether they are computed from the dispersion relation or only illustrative.","section":"Fig. 3(b)-(e)"},{"comment":"The interval notation \"[-k_max_y,ant, -k_max_y,ant]\" appears to have a sign error; this should presumably be \"[-k_max_y,ant, +k_max_y,ant].\"","section":"Eq. (1) and surrounding text"},{"comment":"The term \"scissors state\" is used without definition; a brief description of the equilibrium configuration would help readers unfamiliar with SAFs.","section":"Introduction / Fig. 1(c)"},{"comment":"The statement that the BLS amplitude is maximal near the left edge of the antenna is not clearly illustrated because the antenna position is not drawn on the images; adding a schematic of the antenna would improve clarity.","section":"Fig. 2(a)"}],"recommendation":"major_revision","confidential_remarks":"The paper reports novel and well-presented experimental observations, and the material parameters are anchored to an independent study. The main risk is the edge-seeding mechanism, which is not directly evidenced and is contradicted by the simulation failure at 8.4 GHz. A revision that adds a direct test (e.g., wavevector-resolved BLS or a control simulation with modified edges) or at least a quantitative analysis of the transverse spectral content would make the paper suitable for publication. I would not reject the manuscript; the observations are timely and the interpretation is plausible, but the load-bearing mechanism needs stronger support."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear X,\n\nQuick take: this paper earns its place. The experiment is clear: BLS images of a SAF stripe show that even when the antenna and conduit are aligned with the reciprocal direction, the spin-wave response is full of non-reciprocal signatures—transverse nodal patterns, corner emission, and caustic beams that ignore the conduit. That is a direct, useful warning to anyone doing propagating spin-wave spectroscopy on non-reciprocal materials: the usual 1D assumption is not safe. The authors deserve credit for the honesty of their analysis; they spell out where the plane-wave description breaks and where the micromagnetic model fails.\n\nWhat is new is the emphasis on propagation rather than eigenmode frequencies. Prior work on confined DMI systems looked at frequency shifts and sometimes claimed confinement restores reciprocity. This paper shows that the propagative pattern itself remains strongly non-reciprocal even when the intended channeling direction is the reciprocal one. That is a genuine step.\n\nThe soft spots are in the mechanism section. The paper attributes the large transverse wavevectors to the tiny non-uniform edge regions that break the translational symmetry of the antenna drive. That is plausible, but it is inferred, not measured. The micromagnetic simulations, which include the edge state, reproduce most patterns after optical convolution—good—but they fail at 8.4 GHz to reproduce the bidirectional beating, and the authors say so themselves. So the edge-seeding picture is not fully closed. Also, the inferred high-k values (up to 25 rad/µm) exceed the BLS collection limit, so they are extrapolations from nodal spacings using the unbounded-film dispersion. That is fine in context, but it is indirect.\n\nNone of that undermines the core conclusion. The BLS images are direct evidence that the response is not 1D, and the simulation agreement at the other frequencies supports the general narrative. The 8.4 GHz blemish is a limitation to fix in a follow-up, not a reason to stop the paper.\n\nMy recommendation: send this to peer review. It deserves a careful referee—someone who can check the edge-coupling story and who understands PSWS. I would bring it to the reading group and cite it if I worked on that kind of spectroscopy.\n\nBest,","headline":"A solid experimental caution for the PSWS community: non-reciprocity leaks into supposedly reciprocal channeling, and the 1D analysis deserves suspicion.","tokens_in":13721,"tokens_out":3327,"would_cite":true,"duration_ms":30604,"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","85.75.-d"],"model":"deepseek-v4-flash","headline":"Non-reciprocity defeats spin-wave channeling even when the stripe is aligned with the material's reciprocal direction.","keywords":["spin waves","non-reciprocity","synthetic antiferromagnet","Brillouin light scattering","spin-wave channeling","propagating spin wave spectroscopy","isofrequency contours","micromagnetic simulation"],"falsifier":"Measure the wavevector distribution across the stripe width in the low-frequency regime (e.g., by wavevector-resolved or phase-resolved Brillouin light scattering): if no spin-wave components with $|k_x|\\gtrsim 10$ rad/µm are detected, the edge-seeding mechanism is falsified. Alternatively, fabricate a stripe with deliberately modified edges (smoothed or ion-damaged) and check whether the transverse nodal pattern disappears.","tokens_in":12752,"feed_emoji":"🧲","tokens_out":7695,"duration_ms":63206,"temperature":0.7,"pith_summary":"This paper asks whether non-reciprocal waves can be guided in arbitrary directions. Using Brillouin light scattering microscopy and modeling, the authors study acoustic spin waves in a narrow stripe of a synthetic antiferromagnet, with the static field applied perpendicular to the stripe so that the stripe axis is the material's reciprocal direction. They find that the spin-wave response is nevertheless dominated by non-reciprocity: a line-shaped antenna excites waves with wavevectors perpendicular to the stripe, and energy flows along a tilted, caustic-like beam that does not follow the conduit. The explanation traces these effects to the shape of the acoustic-branch isofrequency contours and to non-uniform magnetization near the stripe edges that seeds high transverse wavevectors. If correct, the results invalidate the standard one-dimensional analysis of propagating spin wave spectroscopy for non-reciprocal materials.","feed_headline":"Spin waves dodge the guide, even on the reciprocal axis","feed_subtitle":"A line antenna excites sideways wavevectors and tilted beams that break 1D spectroscopy models.","key_machinery":"The load-bearing object is the isofrequency contour $\\omega_{\\mathrm{ac}}(k_x,k_y)=\\omega_{\\mathrm{applied}}$ of the acoustic spin-wave branch, computed with a dynamical-matrix formalism that includes interlayer dipole-dipole coupling. At low frequencies the contour is an egg shape extending to large negative $k_x$; near the uniform-mode frequency it collapses to a line along the $k_x=-k_y$ diagonal; at higher frequencies it develops a protuberance that yields two distinct group-velocity directions. The analysis selects the wavevector space reachable by the antenna ($|k_y|<3$ rad/µm) and with group velocity directed away from the antenna ($\\vec{\\nabla}_{\\vec{k}}\\omega\\cdot\\hat{y}>0$). Within this window the direction of the group velocity determines where energy flows, while the large transverse wavevector components determine the nodal spacing of the interference pattern. The non-uniform equilibrium magnetization near the stripe edges is proposed as the source that seeds those high-$|k_x|$ waves.","core_discovery":"The central discovery is that non-reciprocity does not disappear when spin waves are guided along a direction in which the dispersion is symmetric. In a synthetic antiferromagnet stripe, the antenna excites acoustic spin waves whose wavevectors can run parallel to the antenna—perpendicular to the intended guiding direction—because the isofrequency contour at the driving frequency crosses the antenna's coupling window at large $|k_x|$. These waves share a common group velocity, leading to interference patterns with nodes transverse to the stripe at low frequencies and to a single tilted 'caustic-like' beam near the uniform-mode frequency, where the contour collapses to a line along the $k_x=-k_y$ diagonal. The authors show that a qualitative analysis based on the unbounded-film dispersion relation reproduces the main experimental features, and micromagnetic simulations—which include the non-uniform edge magnetization—match the images after convolution with the optical resolution, although discrepancies remain at intermediate frequencies. The paper concludes that wavevectors and group velocities are generally not collinear with the conduit, so models of spin-wave transport that assume a one-dimensional flow fail for strongly non-reciprocal materials.","pith_inferences":["The transverse nodal patterns imply that in non-reciprocal waveguides, a single 'wavelength' measured along the conduit may actually be a superposition of the antenna's $k_y$ window and large $k_x$ components; phase-resolved imaging could disentangle these.","The same isofrequency-contour reasoning could predict off-axis emission in other non-reciprocal wave systems, such as DMI films or magnonic crystals, where the dispersion surface is tilted.","A direct experimental test of the edge-seeding hypothesis would be to modify the stripe edges (e.g., by smoothing, ion irradiation, or exchange-biasing) and observe whether the transverse nodal pattern disappears or shifts in frequency.","The caustic beam near the uniform-mode frequency is an energy-focusing effect that might be exploitable for microwave signal routing, but its direction depends on field and frequency, suggesting tunable beam steering in SAF conduits."],"forward_implications":["Propagating spin wave spectroscopy on non-reciprocal materials cannot be reduced to a 1D model: wavevectors and group velocities are generally not collinear with the conduit, so spatial imaging is required to interpret the electrical response.","The standard method of extracting spin-wave attenuation lengths from the exponential decay of a signal along the stripe becomes unreliable whenever the energy flow is tilted away from the conduit axis.","A line-shaped antenna in a non-reciprocal medium can radiate energy in a narrow, caustic-like beam whose direction is set by the isofrequency contour, not by the antenna orientation—opening a design route for angle-selective spin-wave emitters.","For very narrow stripes (near 1 µm), the non-reciprocity is progressively lost and the response crosses over to the familiar reciprocal Damon-Eshbach regime with mode interference."],"supporting_citations":[{"why":"Introduces propagating spin wave spectroscopy, whose 1D analysis the paper challenges.","marker":"[20]"},{"why":"Gives the reciprocal-mode interference picture (periodic self-focusing) that serves as the baseline contrast for the non-reciprocal nodal patterns.","marker":"[21]"},{"why":"Derives the conditions under which inductive-antenna PSWS yields unidirectional energy flow, the 1D framework the paper shows to be invalid for non-reciprocal materials.","marker":"[22]"},{"why":"Reports unidirectional spin waves in synthetic antiferromagnets measured by PSWS, providing the experimental prior that motivates studying the role of non-reciprocity.","marker":"[5]"},{"why":"Establishes the unidirectionality of spin waves in SAFs and the acoustic-branch dispersion properties on which the isofrequency-contour analysis is built.","marker":"[36]"},{"why":"Supplies the plane-wave demagnetizing tensors used in the dynamical-matrix calculation of the dispersion, including interlayer dipolar interactions.","marker":"[35]"}],"fun_headline_variants":["Spin waves leak sideways even on reciprocal axis","Guide can't tame non-reciprocal spin waves","1D spin-wave models fail: non-reciprocity guides","Reciprocal? Not for guided spin waves"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's explanation relies on the untested claim that tiny regions near the stripe edges, where the equilibrium magnetization and demagnetizing fields are non-uniform, efficiently couple the antenna field to spin waves with transverse wavevectors up to roughly 25 rad/µm; the micromagnetic model itself fails at 8.4 GHz to reproduce the observed bidirectional beating, indicating that this edge coupling is not fully captured.","fun_headline_variants_meta":{"raw":{"variants":["Spin waves leak sideways even on reciprocal axis","Guide can't tame non-reciprocal spin waves","1D spin-wave models fail: non-reciprocity guides","Reciprocal? Not for guided spin waves"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000893,"raw_usage":{"total_tokens":3827,"prompt_tokens":899,"completion_tokens":2928,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":515,"completion_tokens_details":{"reasoning_tokens":2865}},"tokens_in":515,"tokens_out":2928,"duration_ms":21491,"temperature":1.0,"reasoning_tokens":2865,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:16:54.524308+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the wavevector distribution across the stripe width in the low-frequency regime (e.g., by wavevector-resolved or phase-resolved Brillouin light scattering): if no spin-wave components with $|k_x|\\gtrsim 10$ rad/µm are detected, the edge-seeding mechanism is falsified. Alternatively, fabricate a stripe with deliberately modified edges (smoothed or ion-damaged) and check whether the transverse nodal pattern disappears.","supporting_citations":[{"cited_title":"Bailleul, D","cited_arxiv_id":null,"evidence_quote":"Introduces propagating spin wave spectroscopy, whose 1D analysis the paper challenges."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the reciprocal-mode interference picture (periodic self-focusing) that serves as the baseline contrast for the non-reciprocal nodal patterns."},{"cited_title":"Devolder, Propagating-spin-wave spectroscopy using induc- tive antennas: Conditions for unidirectional energy flow, Phys- ical Review Applied20, 054057 (2023)","cited_arxiv_id":null,"evidence_quote":"Derives the conditions under which inductive-antenna PSWS yields unidirectional energy flow, the 1D framework the paper shows to be invalid for non-reciprocal materials."},{"cited_title":"Thiancourt, S","cited_arxiv_id":null,"evidence_quote":"Reports unidirectional spin waves in synthetic antiferromagnets measured by PSWS, providing the experimental prior that motivates studying the role of non-reciprocity."},{"cited_title":"Millo, J.-P","cited_arxiv_id":null,"evidence_quote":"Establishes the unidirectionality of spin waves in SAFs and the acoustic-branch dispersion properties on which the isofrequency-contour analysis is built."}],"review_version":1}