{"id":"8dcd43ae-400c-435d-8a5d-fa230875718d","arxiv_id":"2501.05433","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The transient magneto-optical signal of Co and Ni at the M edge is strongly probe-energy dependent because of sign-canceling contributions from different core-state channels, so TMOKE changes cannot be read directly as magnetization changes.","lead":"The paper uses a fully quantum-mechanical simulation of cobalt and nickel to show that ultrafast laser experiments probing magnetism can give opposite signals at different probe energies. This matters because it explains why recent experiments disagreed about whether the materials demagnetize or not, and it tells experimentalists to measure at several well-chosen energies.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim about TR-TMOKE sign changes is supported only through Re eps_xy; the actual asymmetry depends on transient Fresnel factors and n, which are never computed, so the experimental implication lacks a direct test.","rationale":"After reading the paper in good faith, the strongest part is the direct TDDFT calculation of Re eps_xy and its mj decomposition, which gives a concrete mechanism for energy-dependent magneto-optical responses. The mj-resolved core-state picture is plausible because the 3p states are deep and the 1.55 eV pump cannot directly excite them; the reader's identified assumption is unlikely to fail in this regime. However, the paper's central message is aimed at TR-TMOKE experiments, and the experimental observable is the magnetic asymmetry A, not Re eps_xy. The SM provides the exact relation (S06-S07) with Fresnel coefficients and n, but the paper never evaluates A for the transient state. The denominator in the asymmetry is energy- and time-dependent, and in the overlap region (60-68 eV for Co/Ni) n has structure; sign changes in Re eps_xy do not guarantee sign changes in A. Without computing A, or comparing to experimental asymmetry traces, the claim that TMOKE measurements are 'muddled by artifacts arising from the choice of probe energy' is an unsupported extrapolation. This is conditional-worthy, not fatal: the paper can be accepted once the full asymmetry is computed and shown to reproduce the same energy-dependent sign pattern, or the conclusions are narrowed to the dielectric tensor. The reader's verdict of CONDITIONAL remains appropriate, so I keep the verdict unchanged.","tokens_in":12380,"tokens_out":9078,"duration_ms":90738,"concrete_test":"Compute the full TMOKE asymmetry A(omega) at t = 35 fs using the transient dielectric tensor components eps_xx(omega), eps_xy(omega) and the experimental geometry of Ref. [17] via Eqs. S06-S07, then compare the energy-resolved sign of delta-A(omega) with delta-Re eps_xy(omega) in the overlap region. If the sign reversals in delta-A coincide with those in delta-Re eps_xy, the central claim is supported; if not, the paper must qualify its conclusion to Re eps_xy only.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is about TR-TMOKE signals, but the only computed observable is the real part of the off-diagonal dielectric tensor, Re eps_xy. The actual TMOKE magnetic asymmetry A = (I+ - I-)/(I+ + I-) given in the SM (Eqs. S06-S07) also depends on the frequency-dependent refractive index n and the Fresnel coefficients I0 and Im, which are time-dependent and strongly structured in the M3/M2 overlap region (60-68 eV for Co and Ni). The paper never evaluates A for the transient state, nor does it compare delta-A to the experimental asymmetry traces cited as discrepancies (Refs. 15-17). Consequently, the load-bearing assertion that the sign of the TMOKE change does not track the magnetization change is an extrapolation from delta-Re eps_xy; it could fail if the denominator and the n-dependence in the Fresnel factors reverse or mask the sign pattern. This is not an internal inconsistency, but it is the weakest link between the ab initio calculation and the experimental conclusion.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"Elhanoty et al. report ab initio time-dependent density functional theory (TDDFT) calculations of the transient off-diagonal dielectric tensor component Re ε_xy at the M2,3 edges of fcc Ni and hcp Co, evaluated 35 fs after a 1.55 eV pump pulse. They decompose the equilibrium and transient M3 and M2 responses by the magnetic quantum number mj of the 3p3/2 and 3p1/2 core states and find that the sign and magnitude of ΔRe ε_xy vary strongly with probe energy, especially in the M3/M2 overlap region. The central claim is that the sign of a TR-TMOKE asymmetry change does not directly track the magnetization change, so the probe energy and the relevant matrix elements must be specified before interpreting ultrafast magnetization measurements.","tokens_in":12518,"tokens_out":6010,"duration_ms":63395,"significance":"The main strength is that the central spectral quantity, ΔRe ε_xy, is a direct output of a time-dependent linear-response calculation; the strong energy dependence and sign changes in Fig. 3 are not produced by fitting, and the mj decomposition is a post-hoc quantum-mechanical analysis of that output. If the conclusions hold, the paper provides a concrete and physically plausible mechanism—selection rules combined with M3/M2 overlap—for probe-energy-dependent TR-TMOKE responses, and it would strengthen the case for multi-energy probing in ultrafast magnetism experiments. However, the paper stops at Re ε_xy and does not evaluate the actual measured TMOKE asymmetry, so the experimental implications are not yet fully demonstrated.","major_comments":[{"comment":"The measured TMOKE magnetic asymmetry A = (I+ − I−)/(I+ + I−) is not determined by Re ε_xy alone: in Eqs. (S06)–(S07), the coefficients I0 and Im depend on the complex, transient refractive index n(ω,t), and the term I_m ε_xy brings the full complex ε_xy into the intensity. The manuscript computes only Re ε_xy at 35 fs and never evaluates A, n, or the Fresnel coefficients for the pumped state. The abstract's and conclusions' claim that TR-TMOKE signals can rise or fall independently of the magnetization change therefore goes beyond the computed quantity. The authors should either compute the transient A (or at least the transient Fresnel factors) or explicitly restrict the central claim to ΔRe ε_xy and flag the step from ΔRe ε_xy to the measured asymmetry as an untested extrapolation.","section":"SM, 'Transverse MOKE' (Eqs. S06–S07) and main Eq. (1)"},{"comment":"The selection-rule explanation assumes that the 3p core states remain pure |j,mj⟩ states with the same Clebsch–Gordan spin weights at 35 fs as in the ground state; however, Fig. S1 shows only ground-state spin projections. Since the pump modifies occupations and orbitals, the decomposition of the transient response into fixed mj channels may misattribute spectral weight if the transient core spinors are altered. The authors should verify from the time evolution that the core-state spin projections are unchanged at 35 fs, or quantify the uncertainty this assumption introduces in the mj-resolved Δε_xy of Figs. 3 and 4.","section":"SM, 'Spin Projections of 3p Core States'; Figs. 2 and 4"},{"comment":"The paper presents only a single nonequilibrium snapshot at t = 35 fs and does not directly compare any computed quantity with the experimental asymmetry traces of Refs. [15–17]. A single snapshot is sufficient to demonstrate energy-dependent sign changes at that instant, but it is not sufficient to resolve discrepancies that are defined by temporal traces over the first 100 fs. The authors should either add a time series of ΔRe ε_xy or soften the claim that these experimental discrepancies are addressed by the present calculation.","section":"Fig. 3; Refs. [15–17]"}],"minor_comments":[{"comment":"The phrase 'the the full potential ELK code' contains a duplicated article and should be corrected.","section":"Main text, numerical details section (SM)"},{"comment":"The symbol 'Im' is used both as a coefficient (I_m) and resembles the imaginary-part operator; using an explicit subscript, e.g. I_m, in Eqs. (S06)–(S07) and in the surrounding text would remove a real ambiguity.","section":"SM, Eq. (S07)"},{"comment":"The three Ni experiments are described as reporting 'starkly different responses,' but the manuscript does not list their probe energies or other experimental conditions; a brief table or sentence specifying the energy ranges would make the claimed discrepancy concrete and easier to compare with Fig. 3.","section":"Introduction, Refs. [15–17]"},{"comment":"The caption states that 'the energy scale is not the same for the valence levels and the core levels'; this is helpful, but the figure would be clearer if the two energy axes were labeled with their respective scales.","section":"Fig. 1 caption"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's core ab initio result is credible and the mj decomposition is an interesting analysis, but the gap between Re ε_xy and the experimentally measured TMOKE asymmetry is the main obstacle. The authors' own Eq. (S07) makes clear that A depends on transient Fresnel factors and the full complex ε_xy, so the central interpretive claim about measured signals is currently supported only by an extrapolation. I would encourage the editor to request the computation of the transient asymmetry (or a clearly stated restriction of the claim) and a verification of the mj purity assumption before publication. There is also a question of fit: the paper is positioned as resolving experimental discrepancies, but it does not yet quantitatively engage with the cited experimental traces."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core insight here is real and worth taking seriously: the transient M-edge response in Ni and Co, as captured by Re eps_xy, is not a monotonic function of the magnetization change. The mj-resolved decomposition of the M3/M2 edges shows that partial cancellation between mj channels can make the sign of the response flip with probe energy, which is the kind of thing that would explain conflicting TR-TMOKE results. As far as I can tell, the central number — the energy-dependent sign structure of Delta Re eps_xy at 35 fs — is a direct TDDFT output, not a fitted curve, and the mj decomposition is a standard post-hoc analysis. That is genuine evidence, and the paper deserves to be read by anyone using TR-TMOKE to infer demagnetization timescales.\n\nThe soft spots are not fatal, but they are real. The biggest one is the one the stress test flagged: the paper never computes the actual TMOKE asymmetry A, which depends on the refractive index n and the Fresnel coefficients, both of which are strongly structured in the M3/M2 overlap region. The authors show Eq. (1) and mention Ref. [17], but they do not evaluate A for the transient state or compare to the experimental traces they claim to address. As written, the conclusion that the TMOKE signal's sign does not track magnetization is an extrapolation from Delta Re eps_xy. It is probably right, but it is not demonstrated.\n\nThe single 35 fs snapshot is a second limitation. Dynamics are claimed, but only one time step is shown. That is a modest gap, since the paper's main point is about energy dependence rather than timescales, but it should be stated as a snapshot, not a dynamic conclusion. The assumption that the 3p core states remain pure jj eigenstates under the pump is reasonable in the small-perturbation limit, but the authors should either test it or explicitly justify it with the spin-projection data from Fig. S1. Finally, numerical details are thinner than I would like — more convergence information and xc-kernel specifics would help.\n\nThe citation pattern looks fair. The self-citations are to the group's own mixed TE+linear response method, which is appropriate here; the experimental references are the actual conflicting measurements.\n\nBottom line: this is a solid theoretical paper with a real interpretive advance. It needs a revision that closes the gap between eps_xy and the experimental observable, but it absolutely deserves peer review rather than a desk rejection.","headline":"A useful, mostly sound ab initio demonstration that M-edge TR-TMOKE responses in Ni and Co are strongly probe-energy dependent, though the leap from Re eps_xy to the actual measured asymmetry needs closure.","tokens_in":13157,"tokens_out":1607,"would_cite":true,"duration_ms":18816,"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":"This paper establishes that the transient TMOKE signal at the M edge of Ni and Co is a nontrivial function of probe energy, so a decrease in the measured signal does not necessarily mean a decrease in magnetization.","keywords":["ultrafast magnetization dynamics","magneto-optical Kerr effect","transverse MOKE","quantum selection rules","M-edge spectroscopy","3p core states","time-dependent density functional theory","dielectric tensor"],"falsifier":"A time-resolved scan of the TMOKE asymmetry (or of $\\Re \\epsilon_{xy}$) across the full M$_3$ and M$_2$ energy range in Ni or Co under one fixed pump would falsify the central claim if $\\Delta \\Re \\epsilon_{xy}$ had the same sign at every probe energy, or if the full energy dependence could be reproduced from ground-state rigid-band populations without including the transient $m_j$-resolved matrix elements.","tokens_in":12105,"feed_emoji":"🧲","tokens_out":12565,"duration_ms":104083,"temperature":0.7,"pith_summary":"The paper argues that the transient magneto-optical signal measured in pump-probe experiments—the time-resolved transverse magneto-optical Kerr effect, TR-TMOKE—at the $3p$ M edge of magnetic transition metals is not a direct readout of the instantaneous magnetization. Instead, the response is built from partial contributions of individual $m_j$ states of the $3p_{3/2}$ and $3p_{1/2}$ core manifolds, whose superposition can produce sign changes and oscillations at specific probe energies even while the magnetization simply decreases. Using ab initio time-dependent density functional theory for hcp Co and fcc Ni, the authors show that $\\Delta \\Re \\epsilon_{xy}$ in the overlapping M$_3$/M$_2$ region varies dramatically with probe energy and does not track the overall demagnetization. The practical consequence is that TR-TMOKE demagnetization claims are trustworthy only when the probe energy and the matrix elements at that energy are specified.","feed_headline":"Selection rules set the sign of ultrafast magneto-optical response","feed_subtitle":"For Ni and Co at the M edge, a dip in the transient Kerr signal need not mean the magnetization dropped.","key_machinery":"The central object is the $m_j$-resolved decomposition of the real part of the off-diagonal dielectric tensor, $\\Re \\epsilon_{xy}$, at the M$_3$ and M$_2$ absorption edges. The authors disentangle the total static and 35 fs transient signals into contributions from each $|3p_{3/2}, m_j\\rangle$ state ($m_j = -3/2, -1/2, 1/2, 3/2$) and each $|3p_{1/2}, m_j\\rangle$ state ($m_j = -1/2, 1/2$), using dipole selection rules and Clebsch-Gordan coefficients to connect core states to the spin character of available conduction states. This decomposition shows that the M$_3$ edge receives a large negative $m_j = -3/2$ contribution and a smaller positive $m_j = -1/2$ contribution, the M$_2$ edge receives a large positive $m_j = -1/2$ and a smaller negative $m_j = 1/2$ contribution, and the superposition of all channels in the overlap region is what makes the transient response a nontrivial function of probe energy.","core_discovery":"The central claim is that the energy-dependent transient response of the off-diagonal dielectric tensor component $\\Re \\epsilon_{xy}$ at the M edge is governed by the superposition of partial contributions from the individual $m_j$ states of the $3p_{3/2}$ (M$_3$) and $3p_{1/2}$ (M$_2$) manifolds. Each $m_j$ channel has its own sign and magnitude, fixed by dipole selection rules and Clebsch-Gordan coefficients, and the pump laser modifies these channels differently. In the energy range where M$_3$ and M$_2$ overlap, the constituent changes can cancel or add so that $\\Delta \\Re \\epsilon_{xy}$ oscillates and changes sign in a way that does not directly relate to the overall decrease in magnetization. The authors therefore conclude that interpreting a TMOKE asymmetry change solely as a change in magnetization is invalid unless the probe energy is specified.","pith_inferences":["A direct experimental check would be to probe the same pump-driven Ni or Co sample at two energies on opposite sides of the M$_3$/M$_2$ crossing and look for opposite signs in $\\Delta \\Re \\epsilon_{xy}$; the paper's mechanism predicts such a sign flip.","The same $m_j$-decomposition logic should apply to other $3d$ elements and alloys with overlapping $3p$ edges; extending it to Fe or permalloy would show whether the energy-dependent artifacts are generic.","The paper's use of full transient Kohn-Sham states, rather than ground-state rigid bands, implies that population-only models may miss the main source of the energy dependence; a population-only calculation would give a different and testable $\\Delta \\Re \\epsilon_{xy}$ line shape."],"forward_implications":["A single TR-TMOKE measurement at one probe energy cannot by itself establish demagnetization or enhancement in Ni and Co; the sign of the measured asymmetry change depends on energy.","Measurements in the M$_3$/M$_2$ overlap region (roughly 65--68 eV for Ni and 60--63 eV for Co) are the most prone to artifacts and should be interpreted with the $m_j$-resolved dielectric response.","Experiments should report or scan multiple probe energies, especially at the M$_3$ edge, and pair measurements with theory that resolves individual core-state contributions.","For magnetic alloys, where optical intersite spin transfer is expected, the probe-energy dependence complicates attribution of asymmetry changes to intersite spin transfer.","The overlap between M$_3$ and M$_2$ also means ground-state XMCD sum rules cannot be applied directly at the M edge of these metals."],"supporting_citations":[{"why":"supplies the experimental observation of energy-dependent asymmetry across the Co M edge that motivates the probe-energy question.","marker":"[8]"},{"why":"provides the photoabsorption and atomic scattering-factor data against which the ground-state $\\Re \\epsilon_{xy}$ is validated.","marker":"[14]"},{"why":"reports time-resolved Ni M$_{2,3}$ absorption and magnetic circular dichroism results with distinct responses at different energies.","marker":"[15]"},{"why":"gives few-femtosecond Ni core-level transient absorption data showing a response that differs from other Ni studies.","marker":"[16]"},{"why":"provides TR-MOKE measurements and the Fresnel analysis connecting magnetic asymmetry to magnetization.","marker":"[17]"},{"why":"supplies the magneto-optical formalism relating the Kerr effect to $\\epsilon_{xy}$.","marker":"[20]"},{"why":"derives the relationship between magnetic asymmetry and magnetization that the paper argues is incomplete without probe-energy information.","marker":"[21]"},{"why":"is the DFT code used for the TDDFT time evolution and linear response calculations.","marker":"[22]"},{"why":"supplies the angular-momentum coupling coefficients used to decompose the $m_j$ contributions.","marker":"[23]"},{"why":"is the theoretical basis for optical intersite spin transfer whose interpretation is affected by the probe-energy effects.","marker":"[24]"}],"fun_headline_variants":["Probe energy controls the sign of ultrafast magneto-optical response","A Kerr dip may not mean the magnetization dropped","Selection rules, not just spins, shape ultrafast Kerr signals","Magnetization readouts depend on probe energy in ultrafast experiments","Why probe energy matters in ultrafast magneto-optics"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The explanation assumes the 3p core states retain their fixed spin and angular-momentum character, with the same mixture of spin-up and spin-down components, while the pump laser is active; if the pump distorts or mixes those core levels, the sign-cancellation picture would not describe the total response.","fun_headline_variants_meta":{"raw":{"variants":["Probe energy controls the sign of ultrafast magneto-optical response","A Kerr dip may not mean the magnetization dropped","Selection rules, not just spins, shape ultrafast Kerr signals","Magnetization readouts depend on probe energy in ultrafast experiments","Why probe energy matters in ultrafast magneto-optics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000769,"raw_usage":{"total_tokens":3396,"prompt_tokens":925,"completion_tokens":2471,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":541,"completion_tokens_details":{"reasoning_tokens":2385}},"tokens_in":541,"tokens_out":2471,"duration_ms":17752,"temperature":1.0,"reasoning_tokens":2385,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:14:36.756554+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A time-resolved scan of the TMOKE asymmetry (or of $\\Re \\epsilon_{xy}$) across the full M$_3$ and M$_2$ energy range in Ni or Co under one fixed pump would falsify the central claim if $\\Delta \\Re \\epsilon_{xy}$ had the same sign at every probe energy, or if the full energy dependence could be reproduced from ground-state rigid-band populations without including the transient $m_j$-resolved matrix elements.","supporting_citations":[{"cited_title":"The 2022 magneto-optics roadmap,","cited_arxiv_id":null,"evidence_quote":"supplies the experimental observation of energy-dependent asymmetry across the Co M edge that motivates the probe-energy question."},{"cited_title":"X-ray interactions: Pho- toabsorption, scattering, transmission, and reflection at e = 50- 30,000 ev, z = 1-92,","cited_arxiv_id":null,"evidence_quote":"reports time-resolved Ni M$_{2,3}$ absorption and magnetic circular dichroism results with distinct responses at different energies."},{"cited_title":"Elec- tron thermalization and relaxation in laser-heated nickel by few-femtosecond core-level transient absorption spectroscopy,","cited_arxiv_id":null,"evidence_quote":"provides TR-MOKE measurements and the Fresnel analysis connecting magnetic asymmetry to magnetization."},{"cited_title":"Verification of ultrafast spin transfer effects in iron-nickel alloys,","cited_arxiv_id":null,"evidence_quote":"supplies the magneto-optical formalism relating the Kerr effect to $\\epsilon_{xy}$."},{"cited_title":"Magneto-optical kerr spectra,","cited_arxiv_id":null,"evidence_quote":"derives the relationship between magnetic asymmetry and magnetization that the paper argues is incomplete without probe-energy information."},{"cited_title":"Relationship between magnetic 6 asymmetry and magnetization in ultrafast transverse magneto- optical kerr effect spectroscopy in the extreme ultraviolet spec- tral range,","cited_arxiv_id":null,"evidence_quote":"is the DFT code used for the TDDFT time evolution and linear response calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"is the theoretical basis for optical intersite spin transfer whose interpretation is affected by the probe-energy effects."}],"review_version":1}