{"id":"0be3d682-f3d7-4a74-b8e0-c4ae44ec6b09","arxiv_id":"2505.22462","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Time-resolved X-ray measurements show the Néel vector of TmFeO3 rotates from the c-axis to the a-axis, nearly 90 degrees, within about 20 ps after optical excitation.","lead":"Researchers used X-ray magnetic linear dichroism in reflection to watch the magnetic order of the antiferromagnet TmFeO3 rotate in real time after an ultrashort laser pulse. The spin axis completes about a 90 degree turn within 20 picoseconds, showing a direct X-ray route to tracking antiferromagnetic dynamics in thin films.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 7% transient signal is only ~85% of the 8.5% static XMLD asymmetry; under the usual cos² XMLD angular dependence that corresponds to roughly 65–70°, not the claimed full 90° rotation, so the headline angle overstates what the data establish.","rationale":"The most defensible reading is that the data show a large, fast rotation of the Néel vector within the a–c plane, with a temperature-like background of ~1% visible in the b-axis control. The reader already flagged the calibration/background issue and returned CONDITIONAL. My check is more direct: even granting the static calibration, the stated 7%/8.5% ratio is inconsistent with a full 90° rotation under the standard XMLD angular dependence. This is an internal-consistency concern, not a challenge to the consensus that TmFeO3 can be driven through its SRT. The static XMLD spectra are a genuine independent calibration, and the reflection-geometry concept is plausible; the weak point is the quantitative conversion of amplitude to angle. A line-shape fit with θ free, including the b-axis background, would settle whether the headline angle is defensible. If it is not, the appropriate change is to soften the abstract rather than reject the experiment, so the reader's CONDITIONAL verdict stands unchanged.","tokens_in":9075,"tokens_out":10957,"duration_ms":133722,"concrete_test":"Taking the spectra behind Fig. 2(c) and Fig. 3(c), fit the 170 ps transient asymmetry line shape as A(E;θ)=A_static(E)·[(3cos²θ−1)/2 −(3cos²(5°)−1)/2]/[(3cos²(85°)−1)/2 −(3cos²(5°)−1)/2] + Abg(E), with Abg(E) constrained by the b-axis 1% trace and errors propagated from the raw reflectivity. If the best-fit θ is consistent with 90° within one sigma, the abstract can stand; if θ<75°, the 90° claim should be replaced by a partial-rotation statement and the 20 ps timescale rephrased accordingly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Accepting the authors' own comparison, the ~7% transient signal at 170 ps is about 82–85% of the ~8.5% static XMLD asymmetry (Fig. 2c vs Fig. 3c). For Fe L2,3 XMLD the angular dependence is governed by (3cos²θ−1)/2, equivalently a cos²θ factor for the transition-matrix element. Starting from Ehor nearly parallel to c (θ≈5°) and rotating to Ehor nearly perpendicular to a (θ≈85°) gives essentially the full static contrast; a signal equal to only 82–85% of that contrast corresponds to a final angle of roughly 65–70° from c, not 90°. A true 90° rotation at the same normalization would produce a change close to the full 8.5%, not 7%. The authors offer no nonlinear angular calibration that would change this mapping. The b-axis control in Fig. 4(a) adds a ~1% pump-induced background over 60 ps attributed to thermal spin excitations; if any fraction of that background also contaminates the a-axis 7% signal, the rotation angle inferred from the magnetic part drops further, to ~60° or less. Thus the reported amplitudes quantitatively support a large in-plane rotation, but not specifically the 'full 90° rotation' asserted in the abstract; the 'complete within 20 ps' statement inherits this overreach.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports time-resolved X-ray magnetic linear dichroism (XMLD) measurements in reflection geometry on a TmFeO3 single crystal after near-infrared laser excitation. The static XMLD asymmetry is calibrated by two independent methods that agree, giving a maximum asymmetry of about 8.5% near the Fe L3 edge. At 170 ps after excitation, the pump-induced change for horizontal polarization reaches about 7%, which the authors compare with the static asymmetry to infer a reorientation of the Néel vector from the c-axis toward the a-axis. Delay scans with vertical polarization show a change that saturates around 20 ps and persists without recovery up to 2 ns. A control measurement with the b-axis vertical shows only a gradual ~1% signal, which the authors interpret as evidence that the rotation occurs primarily within the a-c plane. The central claim is that the Néel vector undergoes a full 90° rotation within about 20 ps.","tokens_in":9314,"tokens_out":4057,"duration_ms":49075,"significance":"If the quantitative claim is upheld, the paper would demonstrate an element-specific, reflection-geometry X-ray probe of ultrafast antiferromagnetic order dynamics, with the static XMLD calibration carefully cross-checked by two methods and a useful control for out-of-plane rotation. The reflection geometry is a genuine methodological advance for accessing antiferromagnetic thin films. The paper is honest about several experimental limitations, such as the thermal background in the b-axis control and the absence of recovery within 2 ns. However, the headline quantitative conclusion of a full 90° rotation is not fully supported by the amplitude comparison, so the significance of the paper depends on the authors moderating or better supporting that claim. The reported qualitative picture—large in-plane rotation on a tens-of-picosecond timescale with a nonmagnetic background in the out-of-plane channel—is credible and worth publishing after revision.","major_comments":[{"comment":"The statement that the data show a 'full 90° rotation' is not quantitatively established. The transient pump-induced change at 170 ps is about 7%, whereas the static XMLD asymmetry between the c-axis and a-axis ground states is about 8.5%. The transient signal is therefore only about 82–85% of the full c-to-a contrast. Under the usual angular dependence of Fe L2,3 XMLD, reaching only 82–85% of the full contrast corresponds to a final Néel-vector angle of roughly 65–70° from c, not 90°, unless a nonlinear angular calibration is provided and justified. The authors should either report an inferred angle with an explicit angular calibration and uncertainty, or revise the abstract and conclusion to claim a large in-plane rotation rather than a complete 90° rotation.","section":"Results and Discussion, Figs. 2(c) and 3(c)"},{"comment":"The b-axis control signal of about 1% over 60 ps is attributed to thermally induced spin excitations, but it is not subtracted from or otherwise used to bound the a-plane transient signal. If even a fraction of this background contributes to the vertical-polarization signal in the a-axis geometry, the magnetic component of the 7% transient is smaller, further reducing the inferred rotation angle. The paper should quantify the nonmagnetic background contribution in the geometry of Fig. 3(c), or state the uncertainty that this background introduces into the angle estimate.","section":"Results and Discussion, Fig. 4(a)"},{"comment":"The transient delay scans are presented without error bars or a statistical uncertainty analysis, and the claim that the reorientation is 'complete within approximately 20 ps' is read from the shape of the curve without a fit. Given that the quantitative 90° angle claim depends on the plateau amplitude and the timescale claim depends on the rise time, the authors should provide error bars, a fit with confidence interval, or at minimum an explicit estimate of the systematic uncertainty in the 20 ps timescale.","section":"Results and Discussion, Figs. 4(a) and 4(b)"},{"comment":"The static XMLD spectrum used as a quantitative calibrant is measured at equilibrium temperatures of 50 K and 100 K, while the pump-induced transient is measured at 32 K under 45 mJ/cm² excitation. The paper does not assess whether pump-induced lattice expansion, electronic state filling, or transient changes in the crystal field could modify the XMLD cross-section itself, as opposed to only rotating the Néel vector. A control measurement of the nonmagnetic reflectivity change, for example at an off-resonance energy, would strengthen the calibration and the inferred angle.","section":"Experimental Methods and Results"}],"minor_comments":[{"comment":"The text 'on a ultrafast timescale' should read 'on an ultrafast timescale', and 'one of their most compelling feature' should be 'one of their most compelling features'.","section":"Abstract"},{"comment":"The notation 'Ehor ∥ N' in Fig. 2(a) is inconsistent with the text stating that Ehor forms a 5° angle with the c-axis; please clarify whether the parallel label is approximate or whether the figure geometry differs from the stated angle.","section":"Fig. 2(a) and Experimental Methods"},{"comment":"The phrase 'cf. inset of Fig. 3(b)' in the b-axis orientation description appears to refer to the inset of Fig. 4(b) or the b-axis geometry in Fig. 4(a); this cross-reference should be corrected.","section":"Results and Discussion, paragraph following Fig. 4"},{"comment":"The sentence 'the small negative XMLD signal observed may be attributed to a thermally induced spin-excitations' has a subject-verb agreement error and should be reworded, for example as 'may be attributed to thermally induced spin excitations'.","section":"Results and Discussion, Fig. 4(a)"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the experimental work is technically solid and the static calibration is well cross-checked, but the quantitative 90° claim in the abstract and conclusion is not supported by the amplitude comparison alone. I recommend major revision rather than rejection, since the qualitative finding of a large in-plane rotation on a ~20 ps timescale is credible and the reflection-geometry XMLD method is a valuable contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's the short version: this is a genuinely new experiment—time-resolved XMLD in reflection on the Fe L edges of TmFeO3 tracking the pump-induced Néel vector rotation. The static calibration is done properly, with two independent methods agreeing to about 0.5%. That gives the result real legs. But the headline claim of a full 90° rotation is overreach based on the numbers they report. The transient signal at 170 ps is roughly 7%, which is about 85% of the 8.5% static asymmetry. Unless there is a nonlinear angular mapping that changes the relationship, a cos²-like angular dependence puts the final angle at roughly 65–70°, not 90°. They do not give error bars on the transient data, so we cannot tell how significant that difference is. The 1% background in the b-axis orientation, which they attribute to thermal spin excitations, could also eat into the 7%, pushing the inferred angle even lower. So the data support a large in-plane rotation—and the b-axis control supports the in-plane character—but not the specific full 90° within 20 ps claim.\n\nWhat the paper does well: the static XMLD asymmetry is cross-checked by polarization switching and by temperature-driven spin reorientation, and both agree. The reflection geometry at 5° grazing incidence gives elemental and depth sensitivity, and the comparison to Kimel's earlier transmission result is a sensible way to explain the difference. The b-axis control is the right experiment to show the rotation is in-plane. The 20 ps timescale is plausible for a thermal or anisotropy-driven reorientation.\n\nWhere it falls short: no statistical uncertainty or error bars on any of the transient reflectivity points. Given the low flux at slicing mode, this is a real omission. The 90° conclusion is inferred from an 85% amplitude comparison. If the pump also reduces the XMLD contrast via demagnetization or state filling, then 7% could still be consistent with full rotation, but the authors would need to say that. As written, the amplitude read is ambiguous and the abstract overstates it. The 2 ns lack of recovery is interesting but probably just a thermal effect; no quantitative thermal modeling is offered.\n\nMy take: this deserves a serious referee. The experimental demonstration is valuable and the static work is careful. But the authors need to either dial back the full 90° language to a large rotation, likely close to 90°, or provide error analysis and a proper angular mapping. The core technique is solid and the paper will be read, but the current claims outrun the evidence.\n\nRecommendation: send it to review, but expect heavy revision on the quantitative claims.","headline":"New and careful static XMLD calibration supports a large ultrafast in-plane Néel rotation in TmFeO3, but the reported 7% transient signal does not quantitatively establish the claimed full 90° rotation.","tokens_in":9953,"tokens_out":3839,"would_cite":true,"duration_ms":44161,"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":"Using time-resolved X-ray magnetic linear dichroism in reflection, this paper reports that a near-infrared pulse rotates the Néel vector of the antiferromagnet TmFeO3 by a full 90° within about 20 picoseconds, with no recovery within 2…","keywords":["time-resolved X-ray magnetic linear dichroism","Néel vector dynamics","spin reorientation transition","antiferromagnetic spintronics","rare-earth orthoferrite","reflection geometry","femtosecond X-ray probe","TmFeO3"],"falsifier":"Measure the same pump-induced reflectivity change at the Fe L3 edge on a TmFeO3 crystal whose Néel vector is prevented from rotating, for example by pinning it with a strong magnetic field along the c-axis or by staying far below the spin-reorientation temperature with the same excitation fluence; if a comparable 7 percent spectral change at the XMLD asymmetry energy still appears, the signal is not purely a magnetic rotation.","tokens_in":8864,"feed_emoji":"🧲","tokens_out":10462,"duration_ms":90751,"temperature":0.7,"pith_summary":"By combining a near-infrared pump with time-resolved X-ray magnetic linear dichroism (XMLD) measured in reflection at the Fe L edges, the paper reports a direct, element-specific view of laser-driven spin reorientation in the antiferromagnet TmFeO3. The central observation is that after optical excitation the Néel vector rotates within the a–c crystal plane, moving from the c-axis to the a-axis, and that this full $90^\\circ$ rotation is complete within about 20 picoseconds. The rotated state persists for more than 2 nanoseconds, which the authors attribute to slow heat dissipation in the insulating crystal. A sympathetic reader cares because the probe is sensitive to the local Fe magnetic moment rather than to charge or transport effects, and because the reflection geometry removes the need for X-ray-transparent substrates, making the method applicable to antiferromagnetic thin films.","feed_headline":"X-rays clock a 90-degree Néel flip in 20 picoseconds","feed_subtitle":"Direct XMLD in reflection tracks antiferromagnetic order without optical or transport artifacts.","key_machinery":"The carrying object is the XMLD asymmetry, defined as the normalized difference between X-ray reflectivities measured with linear polarization parallel and perpendicular to the spin axis; because XMLD scales with the square of the sublattice magnetization, it exists in antiferromagnets with no net moment. In this experiment the static asymmetry measured below (c-axis) and above (a-axis) the spin-reorientation transition serves as the quantitative calibrant: the pump-induced change in reflectivity at the Fe L3 edge is compared with that static asymmetry to infer a near-$90^\\circ$ rotation. The time-resolved path of the Néel vector is then read from the pump-induced reflectivity transient at the maximum-asymmetry photon energy, and the a–c plane confinement is established by comparing signals with the a-axis versus the b-axis aligned vertically.","core_discovery":"On its own terms, the paper establishes that transient XMLD signals track the Néel vector of TmFeO3 directly through the spin-reorientation transition. The static XMLD asymmetry between the low-temperature c-axis state and the high-temperature a-axis state is roughly 8.5–9 percent at the Fe L3 edge, and the pump-induced reflectivity change at a delay of 170 ps reaches about 7 percent, approximately 85 percent of that static asymmetry. Time-resolved traces at the maximum-asymmetry photon energy show the signal settling to a plateau within about 20 ps, while measurements with the b-axis aligned vertically show only a gradual about 1 percent change, indicating that the rotation stays within the a–c plane rather than tilting out of it. The authors conclude that the surface region probed by grazing-incidence reflection undergoes nearly the full $90^\\circ$ reorientation, with no recovery up to 2 ns, and that this is an element-specific and largely artifact-free measure of the Néel vector dynamics.","pith_inferences":["A systematic fluence and base-temperature series could reveal whether the $90^\\circ$ rotation is threshold-like, which would support a thermally driven anisotropy-switch mechanism and predict a minimum energy cost per bit.","The small roughly 1 percent signal in the b-axis orientation suggests a non-rotational magnetic background; a two-color pump-probe scheme or polarization-dependent analysis could separate this thermal spin-disorder contribution and sharpen the rotation-angle estimate.","If the rotation is complete only in the roughly 10 nm surface region probed at grazing incidence, bulk-sensitive probes should see a smaller or slower rotation; testing this depth dependence would clarify how well the result transfers to thin-film devices.","The same reflection-XMLD approach could be applied to other rare-earth orthoferrites and compensated insulators, where the absence of net magnetization makes transport-based readout unreliable."],"forward_implications":["A single near-infrared pulse can rotate the Néel vector of TmFeO3 by nearly $90^\\circ$ within 20 ps, establishing a fast optical route toward writing antiferromagnetic order.","The rotated state is stable beyond 2 ns in this sample, meaning the written state persists on device-relevant timescales even though it is not permanent.","Reflection-geometry XMLD can resolve ultrafast antiferromagnetic dynamics at the Fe sites without requiring X-ray-transparent substrates, opening the method to thin films on application-relevant substrates.","The comparison with the earlier transmission study indicates that the measured rotation angle depends on probe depth; surface regions can rotate fully while transmission averages over a depth-dependent temperature gradient.","Because the b-axis signal remains small, any out-of-plane component of the Néel vector trajectory during the reorientation is negligible on the measured timescale."],"supporting_citations":[{"why":"Reports the earlier laser-induced spin reorientation in TmFeO3 with a $30^\\circ$ rotation; the present paper contrasts its near-$90^\\circ$ surface result with that study.","marker":"[22]"},{"why":"Shows that resonant X-ray magneto-optical reflection spectroscopy can probe antiferromagnetic films, supporting the reflection geometry used here.","marker":"[29]"},{"why":"Examines the applicability and breakdown of transient magnetic linear dichroism, motivating the care taken in interpreting the ultrafast XMLD signal.","marker":"[31]"},{"why":"Gives the static XMLD spectrum and the Fe and Tm moment evolution across the spin-reorientation transition in TmFeO3, used as the spectral reference.","marker":"[36]"},{"why":"Describes the femtosecond X-ray slicing source used to generate the about 100 fs probe pulses that resolve the 20 ps dynamics.","marker":"[38]"},{"why":"Provides the optical absorption coefficient of TmFeO3 used to estimate the surface thermal gradient and probe depth.","marker":"[39]"},{"why":"Demonstrates XMLD contrast in reflection geometry, supporting the claim that reflection yields strong magnetic signals.","marker":"[37]"}],"fun_headline_variants":["X-rays watch Néel vector spin 90° in 20 ps","Direct X-ray probe tracks 90° spin reorientation in 20 ps","Ultrafast X-rays capture 90° antiferromagnetic flip","20 ps: X-rays see Néel vector rotate 90°","X-ray reflection reveals 90° Néel rotation in 20 ps"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the static XMLD spectrum measured at equilibrium temperatures remains a valid quantitative calibrant for the pump-induced reflectivity change at 32 K, so a roughly 7 percent transient signal can be read as about 85 percent of the static asymmetry and therefore a near-$90^\\circ$ rotation.","fun_headline_variants_meta":{"raw":{"variants":["X-rays watch Néel vector spin 90° in 20 ps","Direct X-ray probe tracks 90° spin reorientation in 20 ps","Ultrafast X-rays capture 90° antiferromagnetic flip","20 ps: X-rays see Néel vector rotate 90°","X-ray reflection reveals 90° Néel rotation in 20 ps"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000628,"raw_usage":{"total_tokens":2864,"prompt_tokens":868,"completion_tokens":1996,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":1901}},"tokens_in":484,"tokens_out":1996,"duration_ms":15090,"temperature":1.0,"reasoning_tokens":1901,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:06:42.004342+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the same pump-induced reflectivity change at the Fe L3 edge on a TmFeO3 crystal whose Néel vector is prevented from rotating, for example by pinning it with a strong magnetic field along the c-axis or by staying far below the spin-reorientation temperature with the same excitation fluence; if a comparable 7 percent spectral change at the XMLD asymmetry energy still appears, the signal is not purely a magnetic rotation.","supporting_citations":[{"cited_title":"Laser-induced ultrafast spin reorienta- tion in the antiferromagnet TmFeO3,","cited_arxiv_id":null,"evidence_quote":"Reports the earlier laser-induced spin reorientation in TmFeO3 with a $30^\\circ$ rotation; the present paper contrasts its near-$90^\\circ$ surface result with that study."},{"cited_title":"Buried antiferromagnetic films investigated by x- ray magneto-optical reflection spectroscopy,","cited_arxiv_id":null,"evidence_quote":"Shows that resonant X-ray magneto-optical reflection spectroscopy can probe antiferromagnetic films, supporting the reflection geometry used here."},{"cited_title":"Applicability and breakdown of transient magnetic linear dichroism,","cited_arxiv_id":null,"evidence_quote":"Examines the applicability and breakdown of transient magnetic linear dichroism, motivating the care taken in interpreting the ultrafast XMLD signal."},{"cited_title":"Interplay of Fe and Tm moments through the spin-reorientation transition in TmFeO3,","cited_arxiv_id":null,"evidence_quote":"Gives the static XMLD spectrum and the Fe and Tm moment evolution across the spin-reorientation transition in TmFeO3, used as the spectral reference."},{"cited_title":"Fem- toSpeX: a versatile optical pump–soft X-ray probe fa- cility with 100fs X-ray pulses of variable polarization,","cited_arxiv_id":null,"evidence_quote":"Describes the femtosecond X-ray slicing source used to generate the about 100 fs probe pulses that resolve the 20 ps dynamics."},{"cited_title":"Optical properties of thulium orthoferrite TmFeO3,","cited_arxiv_id":null,"evidence_quote":"Provides the optical absorption coefficient of TmFeO3 used to estimate the surface thermal gradient and probe depth."},{"cited_title":"Magnetocrystalline anisotropy in x-ray magnetic linear dichroism at the 3 p edges of crystalline Fe thin films,","cited_arxiv_id":null,"evidence_quote":"Demonstrates XMLD contrast in reflection geometry, supporting the claim that reflection yields strong magnetic signals."}],"review_version":1}