{"id":"ec12d6f6-f6a1-4324-8676-99ecac0eff6a","arxiv_id":"1908.02872","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"The T-MOKE magnetic asymmetry at the M-edges of Fe and Ni is not generally proportional to the magnetization; the breakdown is strong for Fe during demagnetization and weak for Ni.","lead":"Ultrafast laser experiments and density functional theory calculations show that the magnetic signal read out from iron and nickel changes with probe energy in ways that rule out the usual assumption that the signal is simply proportional to the magnetization. The result warns that ultrafast demagnetization studies should measure several photon energies before trusting a single magnetic trace.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Fe MAT-ratio drift may stem from transient non-magnetic reflectivity changes rather than a nonlinear asymmetry–magnetization relation; the paper asserts but does not demonstrate that such changes are negligible for the chosen geometry.","rationale":"The paper does what it sets out to do: it proposes and applies a clean ratio test for proportionality between T-MOKE asymmetry and magnetization, and it substantiates the possibility of nonlinearity with constrained-moment DFT for Fe and Ni. The calculations also contain a useful distinction between nonlinearity in εxy(M) and nonlinearity in A(εxy). The weakest point is not the DFT itself but the experimental bridge: the MAT-ratio signature can be mimicked by pump-induced changes in the diagonal optical response, and the paper's one-sentence assertion of negligibility is not backed by a control measurement in this manuscript. This is a load-bearing concern because the stated conclusion specifically concerns 'asymmetry measured in pump-probe experiments,' not merely the existence of nonlinear model configurations. It does not warrant rejection: the central claim is a caution against a universal linear assumption, and the DFT calculations independently show that realistic excitation classes can break proportionality. But the empirical Fe evidence should be corrected for or protected against the denominator effect before quantitative conclusions about excitation mechanisms are accepted. The reader's CONDITIONAL verdict is therefore appropriate, and the proposed reflectivity check would settle whether the concern actually lands.","tokens_in":14228,"tokens_out":12953,"duration_ms":156563,"concrete_test":"Using the raw I±(E,t) traces, compute the transient total reflectivity R(E,t) = [I+(E,t)+I-(E,t)]/2 at each Fe harmonic (47, 51, 54, 57, 60, 63 eV). If R(E,t) shows an energy-dependent temporal variation, propagate it through the asymmetry formula A_raw(E,t) = ΔI/(2R(E,t)) and recompute the MAT ratios without assuming a constant denominator. If the corrected ratios become time-independent, the Fe drift is a non-magnetic optical artifact; if they remain time-dependent, the nonlinear asymmetry-magnetization interpretation is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central experimental evidence is the MAT-ratio test: if A(E,t)=K(E)M(t), then A(E,t)/A(E_ref,t) is time-independent. The paper's Introduction states that non-magnetic contributions from the pump-induced transient change of the refractive index are negligible for the chosen geometry, citing [30], but no measurement or calculation in the present paper supports this assertion for Fe and Ni over 40-72 eV. Since the measured asymmetry is (I+ - I-)/(I+ + I-) (Gamma only enters the theory), any delay-dependent change in the diagonal reflectivity R(E,t) alters the denominator by an energy-dependent amount because delta(E) and beta(E) vary steeply across the M-edge region. Each harmonic is independently normalized at t=0, so an energy-dependent R(t) produces precisely the time- and energy-dependent MAT-ratio signature reported for Fe. If this is the actual source of the drift, the experimental demonstration that Fe's asymmetry-magnetization relation is nonlinear during demagnetization is not established. The DFT calculations would remain as a model-based counterexample, but the paper's headline conclusion about pump-probe measurements would need to be softened.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates the relationship between the T-MOKE magnetic asymmetry and the magnetization in Fe and Ni across the M-edges between 40 and 72 eV, using pump-probe high-harmonic experiments and DFT-based modeling. The authors introduce the magnetisation-asymmetry test (MAT) ratio, A(E,t)/A(E_ref,t), which should be time-independent if A(E) is proportional to M(t). For Fe they observe strong energy- and time-dependent MAT ratios during demagnetization, whereas for Ni the ratios are nearly constant. DFT calculations with constrained magnetic configurations representing Stoner excitations, long-wavelength magnons, and short-wavelength magnons qualitatively reproduce this difference: Fe shows marked nonlinearity for Stoner and short-wavelength excitations, while Ni shows a more linear response. The paper concludes that a direct linear relationship between asymmetry and magnetization cannot be assumed in general, and that the MAT ratio can help identify the dominant excitation mechanism.","tokens_in":14391,"tokens_out":4591,"duration_ms":55179,"significance":"If the conclusions stand, the MAT ratio provides a parameter-free diagnostic of whether a measured T-MOKE asymmetry is proportional to the instantaneous magnetization, which is directly useful for interpreting ultrafast pump-probe experiments. The DFT modeling systematically compares three classes of magnetic excitations and identifies material-specific differences between Fe and Ni, including a nonlinearity between asymmetry and the off-diagonal dielectric tensor that is not of electronic-structure origin. The paper also gives credit for using constrained-moment DFT with fixed parameters (Γ, energy shift, Gaussian broadening) rather than tuning them per magnetization state. However, the experimental demonstration of the central claim currently rests on a stated but unsupported assumption about the negligibility of transient non-magnetic reflectivity changes.","major_comments":[{"comment":"The MAT-ratio test is the key experimental evidence that Fe's asymmetry is nonlinear in the magnetization. This test is valid only if the denominator I+(E)+I−(E) in Eq. (1) (or Eq. (4) with the constant Γ) is time-independent or, at least, changes in an energy-independent way. The paper states in the Introduction that non-magnetic contributions from the transient variation of the refractive index are negligible for the chosen geometry, citing Ref. [30], but no measurement or calculation for Fe and Ni in the 40–72 eV range is presented to support this. Because δ(E) and β(E) vary steeply across the M-edge, a delay-dependent refractive index would produce energy-dependent changes in the denominator; since each harmonic is independently normalized to its t=0 value, such changes would generate exactly the time- and energy-dependent MAT ratios reported for Fe. The authors should either provide an experimental or theoretical demonstration that the diagonal reflectivity is time-independent at these energies, or explicitly soften the conclusion that the measured Fe response is nonlinear during demagnetization.","section":"Introduction, Eq. (3) and Fig. 3"},{"comment":"The MAT-ratio data in Fig. 3 are presented without error bars or any estimate of the noise level. The Fe curves deviate from unity by a few percent, and some energy channels appear to deviate more than others. Without a quantitative uncertainty estimate, the reader cannot assess whether the observed time dependence is statistically significant or whether, for instance, the apparent early onset of the 54 eV curve in Fig. 2(a) is a real effect or a result of harmonic intensity normalization. Since the entire experimental case for nonlinearity rests on these deviations, the manuscript should include error bars or a noise analysis for the MAT ratios.","section":"Experimental results, Fig. 3"},{"comment":"The theoretical modeling maps the time-dependent demagnetization onto static DFT configurations through the quasi-static approximation of Ref. [31]. While this is a reasonable framework, its validity for the dielectric response at femtosecond timescales is not tested in the present work. In particular, the Stoner-like configuration is modeled as a collinear reduction of all magnetic moments, which is a highly constrained representation of the transient electronic state; the paper does not demonstrate that this captures the relevant transient modification of the band structure. This does not invalidate the model-based counterexample to a general linear relation, but it limits the strength of the conclusions drawn about the experimental data, especially the assignment of the remagnetization phase of Fe to magnons. A discussion of the expected size of corrections to the quasi-static approximation, or a comparison with a time-dependent calculation for at least one case, would strengthen the claim.","section":"Theory, quasi-static approximation"}],"minor_comments":[{"comment":"The caption labels the Ni asymmetry panel as '(b)', which duplicates the Fe panel label; the Ni asymmetry panel should be labeled '(d)'.","section":"Fig. 1 caption"},{"comment":"The sentence 'If the asymmetry were proportional to the magnetization, we would expect the asymmetry to be independent of the photon energy of the probe pulse during the demagnetization and remagnetization process' is imprecise: the asymmetry itself is energy-dependent in general; what is expected to be time-independent is the ratio of asymmetries at two fixed photon energies. The subsequent MAT-ratio discussion correctly implements the intended test, but the wording should be revised.","section":"Experimental results"},{"comment":"The paper states that the dielectric tensor is convoluted with a Gaussian of 1.2 eV, but it does not specify whether the convolution is applied separately to the real and imaginary parts of each tensor component or to the resulting asymmetry; this should be clarified, as it affects the comparison with experiment.","section":"Theory, computational details"},{"comment":"The procedure for normalizing each harmonic curve to its t=0 value is described in words but not shown quantitatively; providing the normalization factor for each harmonic would allow the reader to assess possible systematic offsets in the MAT ratios.","section":"Experimental results, Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"The central experimental claim depends on an assumption (negligible transient refractive-index changes) that is stated but not demonstrated for the specific energy range and materials. The authors should be encouraged to provide either a measurement of the pump-induced change in the sum I+(E)+I−(E) for Fe and Ni, or clearly recast the conclusion as a model-based prediction whose experimental verification remains conditional on this assumption. The paper's heavy reliance on the self-cited quasi-static approximation of Ref. [31] is also worth noting, although it is not by itself a reason to reject."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know. First, the paper's central claim is that T-MOKE asymmetry at the M-edge is not generically proportional to magnetization; it backs this with a new diagnostic (the MAT ratio) and with DFT modeling of three excitation families for Fe and Ni. Second, the strongest experimental evidence—time-dependent MAT ratios for Fe versus flat for Ni—is plausible but rests on an assumption about hot-electron reflectivity changes that the paper asserts but does not demonstrate. That does not sink the paper, but it needs handling.\n\nWhat is actually new: the MAT-ratio test is a clean way to check proportionality without absolute calibration, and the systematic comparison of Stoner, long-wavelength magnon, and short-wavelength magnon configurations for Fe and Ni at the M-edges is not in the prior literature. The experimental contrast between Fe and Ni is striking, and the DFT reproduces it. The distinction between nonlinearity coming from changes in epsilon_xy versus from the A–epsilon_xy relation in Ni is a useful contribution. The modeling is transparent: the three parameters (Gamma, energy shift, Gaussian broadening) are fixed on the ground state and held constant, not tuned to produce the nonlinearity.\n\nThe soft spots. The stress-test note is a fair concern. The asymmetry in Eq. 4 has the diagonal reflectivity in the denominator; if the pump transiently changes the refractive index, the MAT-ratio drift could come from a time-dependent denominator rather than a nonlinear A–M relation. The Introduction's claim that non-magnetic contributions are negligible cites [30], but that is not a measurement or calculation for Fe and Ni at 40–72 eV, and no control is shown. This is a genuine gap. That said, the DFT modeling stands independently as a counterexample showing that nonlinear A–M relations occur for plausible excited states, so even if the experimental demonstration is softened, the paper's core warning remains relevant. I would ask the authors to add error bars and to measure or calculate the transient reflectivity change for both elements.\n\nWorth noting: the theory–experiment comparison in Fig. 1 is qualitative, and the discussion is honest about the restricted supercell for short-wavelength magnons. The paper does not oversell its conclusions.\n\nWho it is for: anyone using T-MOKE or XUV magneto-optics in ultrafast magnetism, and people doing ab initio magneto-optical response. It deserves serious peer review; the experimental caveat should be addressed, but the work is solid enough for referees.","headline":"Useful, honest paper that gives the field a new diagnostic and a caution, but the experimental piece needs to rule out transient reflectivity changes before the Fe result is taken as proof.","tokens_in":15044,"tokens_out":3137,"would_cite":true,"duration_ms":30084,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper challenges the standard assumption that measured magnetic asymmetry tracks magnetization linearly, showing that Fe's asymmetry ratio varies with photon energy and pump-probe delay while Ni's stays nearly flat, and that…","keywords":["ultrafast demagnetization","transverse magneto-optical Kerr effect","T-MOKE","magnetic asymmetry","M-edge","3d transition metals","density functional theory","Stoner excitations"],"falsifier":"In the same pump-probe geometry, record the total reflected intensity at 51, 54, 60 and 63 eV as a function of delay without magnetic analysis, and compare the transient change in the denominator of Eq. 4 with the constant-background assumption. A measurable time-dependent diagonal reflectivity during the first picosecond would mean the Fe MAT-ratio drift could stem from the denominator, not from a nonlinear asymmetry–magnetization relation; a flat diagonal response would strengthen the authors' reading. A second check would be to compare the Fe MAT ratio with an independent magnetization probe, such as time-resolved X-ray magnetic circular dichroism or spin-resolved photoemission, at the same delays.","tokens_in":13993,"feed_emoji":"🧲","tokens_out":7618,"duration_ms":80911,"temperature":0.7,"pith_summary":"The paper sets out to test whether the magnetic asymmetry measured in transverse magneto-optical Kerr effect (T-MOKE) pump-probe experiments is proportional to the instantaneous magnetization. For Fe and Ni probed at the M-absorption edges between 40 and 72 eV, it compares measured asymmetry ratios at several photon energies with density-functional-theory calculations for three classes of magnetic excitations. It finds that Fe's asymmetry–magnetization relation is strongly nonlinear and energy-dependent during demagnetization, while Ni's is nearly linear. The authors conclude that a direct linear relationship between asymmetry and magnetization cannot be assumed in general, and that the type of magnetic excitation determines the coupling. The stakes are practical: ultrafast demagnetization studies routinely convert asymmetry into a magnetization trace, and this conversion needs material- and energy-specific justification.","feed_headline":"Fe's magneto-optical asymmetry is not proportional to magnetization","feed_subtitle":"During ultrafast demagnetization, Fe's asymmetry ratios drift with energy and time; Ni's stay flat.","key_machinery":"The key diagnostic is the magnetisation-asymmetry test ratio (MAT ratio): the ratio of T-MOKE asymmetries at two photon energies plotted against pump-probe delay, which should be time-independent if asymmetry is proportional to magnetization. The supporting machinery is a quasi-static constrained-moment DFT scheme in which the sample's magnetization is fixed at a reduced value in three distinct ways: collinearly shrunk moments for Stoner-like excitations, uniformly tilted moments averaged over azimuth for long-wavelength magnons, and randomly tilted moments in 16-atom supercells for short-wavelength magnons. The dielectric tensor from each constrained state is put through Fresnel equations with a constant background term to produce theoretical asymmetry spectra, and those spectra are compared with the measured MAT-ratio behaviour to identify which excitation family is active.","core_discovery":"The central claim is that A(E)=K(E)M, the working equation behind most T-MOKE demagnetization studies, fails for Fe but holds approximately for Ni at the M-edge. The evidence is the magnetisation-asymmetry test ratio: if the linear relation held, the ratio A(E)/A(E') would be constant in time. In Fe the measured ratio drifts markedly with delay during the first sub-picosecond, especially near 51, 57, 60 and 63 eV relative to 54 eV, while in Ni the ratios stay nearly flat. Constrained-moment DFT reproduces this difference: Fe's calculated asymmetry depends strongly on whether magnetization is reduced by Stoner-like moment collapse, long-wavelength magnons, or short-wavelength magnons, whereas Ni's asymmetry is similar across all three. The modelling also separates two sources of nonlinearity in Ni, a nearly linear off-diagonal dielectric tensor versus magnetization but a nonlinear asymmetry versus that tensor near the main peak, and shows that Fe's remagnetization is consistent with magnon-dominated response.","pith_inferences":["If the Fe result holds generally for M-edge probing, previously published single-harmonic demagnetization curves for Fe may have folded energy-dependent optical artefacts into the magnetization dynamics; a multi-energy re-analysis could shift reported demagnetization time constants.","The clean separation between nonlinearity in the off-diagonal dielectric tensor versus magnetization and nonlinearity in asymmetry versus that tensor suggests an experimental route: measuring a broadband asymmetry spectrum at each delay would allow one to invert the Fresnel expression and extract both diagonal and off-diagonal dielectric response, separating electronic-structure dynamics from geom","Extending the constrained-moment comparison to Co and to alloys across a range of pump fluences could turn the three excitation families into a phase diagram for which microscopic mechanism controls the asymmetry–magnetization relationship."],"forward_implications":["For Fe, single-energy asymmetry curves cannot be read directly as magnetization traces during demagnetization; the time- and energy-dependent MAT ratio shows the out-of-equilibrium state has a more complex magneto-optical response.","For Fe at delays beyond about 2 ps, the near-constant MAT ratios indicate magnons dominate remagnetization, with only small Stoner-like contributions.","For Ni, the asymmetry–magnetization relation at the M-edge is close to linear for all modelled excitations, so single-energy Ni traces are more trustworthy; the residual nonlinearity between asymmetry and the off-diagonal dielectric tensor near 65 eV is a Fresnel-level effect, not a band-structure change.","Any T-MOKE demagnetization study should record at least two probe energies to verify the proportionality before converting asymmetry to magnetization."],"supporting_citations":[{"why":"Provides the T-MOKE geometry and the stated assumption that transient hot-electron refractive-index changes are negligible, supporting the asymmetry expression used in Eq. 4.","marker":"[30]"},{"why":"Supplies the quasi-static constrained-moment method used to compute dielectric tensors for partially demagnetized states.","marker":"[31]"},{"why":"Exemplifies the standard use of T-MOKE asymmetry as approximately proportional to magnetization, the assumption this paper tests.","marker":"[32]"},{"why":"A second standard T-MOKE analysis that reads asymmetry as local magnetization, defining the proportionality the paper challenges.","marker":"[33]"},{"why":"Earlier Co M-edge study attributing distinct spectral fingerprints to Stoner and spin-wave excitations, which this paper extends to Fe and Ni.","marker":"[35]"},{"why":"Follow-up Co M-edge work that motivates comparing calculated asymmetry spectra for different magnetic configurations.","marker":"[36]"},{"why":"Attributes the DFT-versus-experiment energy shift to many-body and local-field effects, justifying the rigid shifts used in the comparison.","marker":"[55]"},{"why":"Long-standing theoretical prediction of nonlinear relationships between magnetization and the off-diagonal dielectric tensor, corroborated here.","marker":"[58]"},{"why":"Computational study of temperature-dependent magnetism in Fe and Ni that supports the different dominant excitation channels.","marker":"[59]"}],"fun_headline_variants":["Fe's M-edge asymmetry breaks linear magnetization rule","Ultrafast demagnetization: Fe nonlinear, Ni linear at M-edge","T-MOKE linearity fails for Fe, holds for Ni","Magnetic asymmetry: Fe defies linear law, Ni obeys","Stoner vs magnon: Fe's nonlinear asymmetry explained"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The interpretation assumes that non-magnetic transient changes in the refractive index caused by the hot-electron distribution are negligible in the experimental geometry; if those changes are not negligible, the Fe MAT-ratio drift could come from a time-dependent denominator in Eq. 4 instead of from an intrinsically nonlinear asymmetry–magnetization relation.","fun_headline_variants_meta":{"raw":{"variants":["Fe's M-edge asymmetry breaks linear magnetization rule","Ultrafast demagnetization: Fe nonlinear, Ni linear at M-edge","T-MOKE linearity fails for Fe, holds for Ni","Magnetic asymmetry: Fe defies linear law, Ni obeys","Stoner vs magnon: Fe's nonlinear asymmetry explained"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000969,"raw_usage":{"total_tokens":4167,"prompt_tokens":1034,"completion_tokens":3133,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":650,"completion_tokens_details":{"reasoning_tokens":3046}},"tokens_in":650,"tokens_out":3133,"duration_ms":21958,"temperature":1.0,"reasoning_tokens":3046,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:31:11.893071+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In the same pump-probe geometry, record the total reflected intensity at 51, 54, 60 and 63 eV as a function of delay without magnetic analysis, and compare the transient change in the denominator of Eq. 4 with the constant-background assumption. A measurable time-dependent diagonal reflectivity during the first picosecond would mean the Fe MAT-ratio drift could stem from the denominator, not from a nonlinear asymmetry–magnetization relation; a flat diagonal response would strengthen the authors' reading. A second check would be to compare the Fe MAT ratio with an independent magnetization probe, such as time-resolved X-ray magnetic circular dichroism or spin-resolved photoemission, at the same delays.","supporting_citations":[{"cited_title":"La-O-Vorakiat, E","cited_arxiv_id":null,"evidence_quote":"Provides the T-MOKE geometry and the stated assumption that transient hot-electron refractive-index changes are negligible, supporting the asymmetry expression used in Eq. 4."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the quasi-static constrained-moment method used to compute dielectric tensors for partially demagnetized states."},{"cited_title":"Mathias, C","cited_arxiv_id":null,"evidence_quote":"Exemplifies the standard use of T-MOKE asymmetry as approximately proportional to magnetization, the assumption this paper tests."},{"cited_title":"Mathias, C","cited_arxiv_id":null,"evidence_quote":"A second standard T-MOKE analysis that reads asymmetry as local magnetization, defining the proportionality the paper challenges."},{"cited_title":"Turgut, D","cited_arxiv_id":null,"evidence_quote":"Earlier Co M-edge study attributing distinct spectral fingerprints to Stoner and spin-wave excitations, which this paper extends to Fe and Ni."},{"cited_title":"Zusin, P","cited_arxiv_id":null,"evidence_quote":"Follow-up Co M-edge work that motivates comparing calculated asymmetry spectra for different magnetic configurations."},{"cited_title":"Willems, S","cited_arxiv_id":null,"evidence_quote":"Attributes the DFT-versus-experiment energy shift to many-body and local-field effects, justifying the rigid shifts used in the comparison."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Long-standing theoretical prediction of nonlinear relationships between magnetization and the off-diagonal dielectric tensor, corroborated here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Computational study of temperature-dependent magnetism in Fe and Ni that supports the different dominant excitation channels."}],"review_version":1}