{"id":"c896b53f-0d75-413a-a1e5-24a093bb70d1","arxiv_id":"2608.06456","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Time-resolved resonant X-ray scattering separates europium and tellurium lattice dynamics in the charge density wave material EuTe4, revealing three coherent phonon modes and a previously unknown europium charge order.","lead":"Using X-rays tuned to the absorption edge of one element, the authors separately track vibrations of europium and tellurium atoms in a layered material where electric charge forms waves. This gives a way to identify which atoms move in each vibration, which could help in controlling materials with light.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Off-resonant Te weight may not be energy-independent; Eq. 1 decomposition could be underdetermined without error propagation.","rationale":"The reader's weakest_assumption already identifies Eq. 1 and the energy-independent Te weight as the key vulnerability, and I agree that the core observation (element-selective tr-RXS enhancement at Eu edges, three coherent modes with distinct resonant/off-resonant visibility, and a previously unreported Eu CDW component) is plausible and qualitatively well supported. My stress-test pass reinforces this concern with two sharper technical points: (1) the weights and the traces are not independent, since a_c(E) is extracted from the same energy-dependent I(E) that enters the global fit, so a contamination of the off-resonant baseline directly biases I_Te and I_Eu in a correlated way; (2) the 0.1 THz discrepancy between the Te and Eu channels near 1.3 THz is explicitly acknowledged but treated as a single mode, which weakens the mixed-character inference if the two features are distinct. The DFT comparison is appropriately hedged ('representative rather than unique'), so it does not compound the concern. The paper would be fully convincing with propagated uncertainties, a quantitative sensitivity test of the a_l assumption, and a two-Lorentzian test of the 1.3 THz feature; none of these would likely overturn the qualitative claim, so CONDITIONAL rather than REJECT is the right verdict. I agree with the reader's assessment; my additional specificity lies in the concrete tests and the framing of the off-resonant contamination as a correlated-bias problem rather than a generic fitting issue.","tokens_in":11155,"tokens_out":1762,"duration_ms":14593,"concrete_test":"Re-derive the decomposition using only the two most off-resonant energies (e.g., 1076 and 1106 eV) to fix a_l, then predict I(E,t) at the resonant energies without re-fitting; if the predicted resonant traces deviate by more than the shot noise, the energy-independent-a_l assumption fails. Additionally, refit the Fig. 3E/F spectra with two Lorentzians near 1.2-1.4 THz and report whether the 0.1 THz splitting is statistically required; if it is, the single 1.3 THz mixed-mode assignment in the central claim needs revision.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim is that I(E,t) = (a_l I_Te(t) + a_c(E) I_Eu(t))/(a_l + a_c(E)) can be globally inverted from the same energy-dependent diffraction data that determine a_l and a_c(E). The load-bearing assumption is that a_l is strictly energy-independent over the full 1076-1180 eV range, so that the off-resonant channel is a fixed Te-only reference. However, the CDW diffraction amplitude contains both Te and Eu form-factor/absorption contributions whose relative phases and magnitudes vary with energy; a fixed a_l is only justified if the off-resonant Eu contribution is negligible and if the Te non-resonant atomic form factor is truly constant across the Eu M4/M5 edges. Moreover, a_c(E) is read off the same equilibrium I(E) curve used to normalize I(E,t), so any contamination of the off-resonant baseline by Eu signal, or any transient change in the Eu resonant cross-section (Supplemental Note 4 rules out only the integrated XAS line shape, not the q-resolved resonant structure factor), would propagate directly into the extracted I_Te(t) and I_Eu(t), changing the apparent 1.3 THz mode's Te/Eu amplitude ratio. The 0.1 THz difference between the 1.3 THz features in Fig. 3E and F is acknowledged but not modeled; if the two features are actually distinct modes, the single-mode assignment in the band-pass filtering and the mixed-character claim are under-resolved. No propagated uncertainties are reported, so the significance of the Eu/Te amplitude separation is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The authors use time-resolved resonant X-ray scattering (tr-RXS) on the charge density wave material EuTe4, tuning the soft X-ray probe energy on and off the Eu M4/M5 absorption edges. They observe three coherent phonon modes near 0.85, 1.3, and 1.6 THz whose FFT visibility depends strongly on probe energy, and they interpret this as sublattice selectivity: a Te-dominated mode, an Eu-dominated mode, and a mixed mode. To make this quantitative, they introduce a two-component decomposition of the normalized CDW diffraction intensity (Eq. 1) as a weighted sum of independent Te and Eu order-parameter dynamics, with weights a_l and a_c(E) obtained from the energy-dependent equilibrium diffraction intensity. They then extract decoupled Te and Eu time traces, band-pass filter them at the three phonon frequencies, and compare with DFT phonon eigenvectors that suggest modes at 0.70, 1.65, and 1.33 THz with corresponding Te, Eu, and mixed character. The paper also claims a previously unreported Eu-sublattice charge order component in EuTe4, supported by the resonant enhancement of the CDW peak at the Eu edges.","tokens_in":11438,"tokens_out":4906,"duration_ms":42622,"significance":"If the quantitative decomposition is valid, the work introduces a broadly applicable tr-RXS protocol for sublattice-resolved coherent phonon identification in multi-element CDW materials, circumventing the energy-resolution limits of RIXS. The qualitative contrast between resonant and off-resonant FFT spectra is compelling and the central frequency identification is not circular, since the phonon frequencies are obtained from Fourier transforms of the data rather than from the decomposition model. The paper also provides a useful demonstration that resonant enhancement can reveal an Eu charge-order component not previously reported. However, the quantitative disentangling step rests on assumptions about energy-independent Te weights and on the same energy-dependent data used to fix the weights, and the manuscript does not report propagated uncertainties; these issues must be addressed before the strength of the quantitative claims can be assessed.","major_comments":[{"comment":"The central quantitative claim is the decomposition I(E,t) = (a_l I_Te(t) + a_c(E) I_Eu(t))/(a_l + a_c(E)), with a_l assumed independent of energy over the full 1076-1180 eV range. This assumption is load-bearing: if the off-resonant signal retains a non-negligible Eu resonant contribution, or if the Te non-resonant form factor and absorption corrections vary appreciably across the Eu M4/M5 edges, then the extracted I_Te(t) and I_Eu(t) are contaminated and the Te/Eu amplitude ratio of the 1.3 THz mode is not reliable. The weights are also determined from the same energy-dependent equilibrium diffraction intensity that normalizes the data, and no error propagation is reported. Please provide a quantitative justification for the energy independence of a_l (for example, calculated energy-dependent structure factors or a control measurement at the Te edge) and report uncertainties on a_l, a_c(E), and the extracted time traces.","section":"Eq. (1), Fig. 3B"},{"comment":"The main text states that the penetration-depth-corrected time traces in Fig. 3A 'collapse onto a single curve,' yet the raw traces in Fig. 2A and their FFTs in Fig. 2B are clearly energy-dependent. If the full time traces, including oscillations, collapse after correction, then Eq. (1) cannot be inverted; if only the non-oscillatory melting background collapses, this must be stated explicitly. This ambiguity is central to the disentangling claim and should be resolved, for example by showing the corrected traces with the oscillatory components emphasized and by quantifying the residual spread among energies.","section":"Quantitative disentangling, Fig. 3A"},{"comment":"The acknowledged ~0.1 THz difference between the features near 1.3 THz in the Eu-decoupled and Te-decoupled FFT spectra is under-modeled. If these are two distinct phonon modes rather than a single mode, the single-mode band-pass filtering and the mixed-character assignment of the 1.3 THz mode become under-resolved. Please add a quantitative mode-discrimination analysis, such as two-oscillator fits with a statistical model comparison or a resolution-limited spectral decomposition, before treating this feature as a single mode.","section":"Fig. 3E,F"},{"comment":"The DFT comparison is explicitly hedged in the caption, where the calculated modes are called 'representative rather than unique assignment,' and the calculated 0.70 THz mode differs from the measured 0.85 THz mode by about 18%. Because the agreement with theory is used to corroborate the sublattice character of the modes, the manuscript should either provide a systematic assignment procedure (e.g., eigenvector overlap or displacement-participation ratios for nearby modes) or explicitly restrict the DFT claim to qualitative agreement only.","section":"Fig. 4, DFT comparison"}],"minor_comments":[{"comment":"The caption uses 'Tr-RXS' while the text uses 'tr-RXS'; please unify the abbreviation.","section":"Fig. 2 caption"},{"comment":"The normalization convention for I(E,t) and for the weights a_l and a_c(E) is not fully specified. Please state whether I(E,0)=1 at all energies and how the absolute scale of the weights is fixed in Fig. 3B.","section":"Eq. (1) and Fig. 3B"},{"comment":"The claim that the corrected time traces 'collapse onto a single curve' should be supported by a quantitative measure, such as the RMS deviation among traces, since the eye is not sufficient given the multiple overlapping curves.","section":"Fig. 3A"},{"comment":"Adding error bars or shaded confidence intervals to the FFT spectra would make the energy-dependent visibility contrast and the 0.1 THz difference more assessable.","section":"Figs. 2B and 3E,F"},{"comment":"The manuscript uses both 'space layer' and 'spacer layer' for the EuTe layers; please use one term consistently.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The experimental dataset and the qualitative sublattice-contrast observation are strong, and the paper is well within the scope of the journal. My main concern is that the quantitative decomposition is the key added value beyond the raw-data observation, and its validity currently rests on unquantified assumptions. I would support re-review after the authors add the requested error analysis and clarify the decomposition, rather than a rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something real: it uses energy-tuned time-resolved resonant X-ray scattering to separate coherent phonons on the Te and Eu sublattices of EuTe4, and in doing so uncovers a previously unreported Eu charge-order component. The qualitative evidence is convincing. The FFT spectra at resonant versus off-resonant probe energies show a clean contrast: 0.85 THz appears mainly off-resonance, 1.6 THz mainly at the Eu edges, and 1.3 THz shows up in both. That, by itself, is a nice result and a useful technique demonstration. The DFT-eigenvector comparison, while explicitly hedged, is consistent with those assignments.\n\nThe main soft spot is the quantitative decoupling in Eq. 1. The weights a_l and a_c(E) are determined from the same energy-dependent data used to extract I_Te(t) and I_Eu(t), and no uncertainties are propagated. The assumption that a_l is strictly energy-independent across the full range is plausible but not proven; off-resonant energies sit only about 20 eV from the Eu M edges, and tails from the resonant cross-section could bleed into the baseline. Similarly, the Te form factor is not truly constant in energy. This does not invalidate the qualitative story, but it means the reported Te/Eu amplitude ratios for the 1.3 THz mode should be treated as indicative, not quantitative. The roughly 0.1 THz difference between the filtered Te and Eu traces at 1.3 THz is acknowledged and set aside; that is fine for the main point, but it hints at more structure than the single-mode model captures.\n\nThe paper would be stronger if the authors reported error bars on the fitted weights, checked the sensitivity of the decomposition to the off-resonant baseline, and released the raw time traces and analysis scripts. As is, the central observation is sound and the technique advance is real. This deserves a serious referee, but a referee should push for the uncertainty analysis and the data. I would bring it to a reading group and would cite it if I worked on CDW phonon dynamics.","headline":"A genuinely useful demonstration that tr-RXS can separate coherent phonons by sublattice; the qualitative claims are solid, but the quantitative two-component decomposition needs error analysis before the amplitude ratios are trusted.","tokens_in":656,"tokens_out":630,"would_cite":true,"duration_ms":22944,"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":"By tuning a soft X-ray probe to the Eu M absorption edges, time-resolved resonant X-ray scattering resolves three coherent phonon modes in EuTe4 and reveals a previously unseen Eu-sublattice charge-order component.","keywords":["time-resolved resonant X-ray scattering","charge density wave","coherent phonons","sublattice-resolved dynamics","EuTe4","element-specific X-ray scattering","phonon eigenvectors","soft X-ray probe"],"falsifier":"Record the CDW diffraction intensity in a fully detuned probe (more than about 10 eV below the Eu M5 edge) after penetration-depth correction and Fourier-transform it: the paper's decomposition predicts that the decoupled Te trace contains no 1.6 THz component, so a robust 1.6 THz oscillation there would falsify the Eu-only assignment and the two-weight decomposition.","tokens_in":10940,"feed_emoji":"🔬","tokens_out":8346,"duration_ms":65308,"temperature":0.7,"pith_summary":"This paper shows that time-resolved resonant X-ray scattering can be tuned to a specific element's absorption edge to tell which atoms are moving in each coherent phonon mode, without needing to track many Bragg peaks or rely on energy-resolution-limited spectroscopies. Applying the method to the charge density wave material EuTe4, the authors find three coherent phonon modes at about 0.85, 1.3, and 1.6 THz with distinct sublattice character: the lowest is Te-dominated, the highest is Eu-dominated, and the middle one is mixed. Along the way, they obtain evidence that the CDW is not confined to the Te monolayers: an additional, previously unreported charge-order component develops on the Eu sublattice with the same wavevector. Because the approach works in the time domain and scans X-ray energy, it sidesteps the energy-resolution ceiling of inelastic X-ray scattering and should generalize to other multi-element materials.","feed_headline":"Element-tuned X-rays split phonon modes by atom type","feed_subtitle":"Three coherent modes in EuTe4 traced to Te and Eu sublattices, exposing a hidden Eu charge order","key_machinery":"The load-bearing element is equation (1): the normalized CDW intensity at X-ray energy $E$ and delay $t$ is written as a weighted average $I(E,t) = (a_l I_{\\mathrm{Te}}(t) + a_c(E) I_{\\mathrm{Eu}}(t))/(a_l + a_c(E))$, where $I_{\\mathrm{Te}}$ and $I_{\\mathrm{Eu}}$ are the Te and Eu sublattice order-parameter dynamics, $a_l$ is an energy-independent Te weight, and $a_c(E)$ is an energy-dependent Eu weight whose line shape is read off the resonant enhancement of the static CDW peak across the Eu M edges. This identity turns a set of time traces measured at several probe energies into a determined system from which the two sublattice dynamics can be globally fitted, after normalizing for probe penetration depth. The second ingredient is the time-domain readout: Fourier-transforming the oscillatory part of the decoupled traces separates the modes by frequency without the energy-resolution limits of inelastic scattering.","core_discovery":"On its own terms, the paper claims the following: when the incident X-ray energy is placed on the Eu M4/M5 absorption edges, the CDW superlattice diffraction peak gains a resonant enhancement that can only come from charge modulation on the Eu sublattice, establishing that EuTe4's CDW has an Eu component in addition to the known Te-lattice distortion. Using the energy dependence of that enhancement to decompose the time-dependent diffraction intensity into separate Te and Eu order parameters, the authors extract two sublattice-resolved time traces. Fourier analysis of these traces yields three coherent phonon modes at approximately 0.85, 1.3, and 1.6 THz: the first involves mainly Te monolayer motion, the third mainly EuTe spacer-layer motion, and the middle involves both. Density functional theory phonon eigenvectors at 0.70, 1.33, and 1.65 THz match these assignments. The claim is therefore that energy-dependent time-resolved resonant X-ray scattering provides element-level resolution of coherent phonon eigenvectors in a multi-element CDW material.","pith_inferences":["Applying the same energy-scanning decomposition at the Te M edges would provide a consistency check: if the extracted Te dynamics reproduce the off-resonance trace, the model's energy-independent $a_l$ assumption is validated.","The roughly 0.1 THz difference between the Te and Eu FFT peaks near 1.3 THz hints that two closely spaced phonon modes, rather than a single mixed mode, may be present; longer time windows or higher fluence could split them.","Extending the scan across momentum transfer $q$ could map how the sublattice character of each mode disperses, distinguishing amplitude-type from phase-type CDW dynamics.","If the Eu charge component is intrinsic rather than a hybridization artifact, resonant pumping at the Eu M edge may allow direct optical addressing of the Eu sublattice order, a control channel not accessible through the Te sublattice."],"forward_implications":["Targeted THz driving can selectively excite the 0.85 THz Te-dominated mode or the 1.6 THz Eu-dominated mode in EuTe4, providing a route to control the CDW sublattice by sublattice.","The decomposition method transfers to any multi-element material with element-specific absorption edges, including CDW superstructures and van der Waals heterostructures.","Models and calculations of EuTe4's electronic structure must include an Eu-sublattice charge modulation at the same wavevector as the Te CDW, not only the Te monolayer distortion.","Soft phonons below roughly 10 THz, which energy-domain inelastic scattering cannot resolve, become addressable in frequency by this time-domain resonant scattering approach.","The mixed character of the 1.3 THz mode shows dynamic coupling between Te monolayers and EuTe spacer layers, relevant to interlayer order."],"supporting_citations":[{"why":"Establishes resonant elastic soft X-ray scattering as an element-selective probe of order parameters, giving the physical basis for tuning to Eu edges.","marker":"[16]"},{"why":"Shows resonant enhancement of CDW diffraction in rare-earth tritellurides and supplies the basis for attributing intensity changes at absorption edges to charge components.","marker":"[33]"},{"why":"Provides the resonant X-ray decomposition of a spatial modulation into separate atomic (Cu 3d and O 2p) contributions, the template for equation (1).","marker":"[43]"},{"why":"Extends the weighted-sublattice decomposition to bond and ligand order in IrTe2, supporting the same functional form.","marker":"[44]"},{"why":"Characterizes EuTe4 and establishes the Te-monolayer CDW at $q = (0, 0.644 b, 0)$, the material and wavevector probed here.","marker":"[27]"},{"why":"Supplies the DFT functional used to compute the phonon eigenvectors compared with the experimental modes.","marker":"[45]"},{"why":"Demonstrates tr-RXS disentangling of transient charge order from structural dynamics in manganites, the direct methodological precedent for separating sublattice contributions in the time domain.","marker":"[21]"}],"fun_headline_variants":["X-ray energy peels phonon modes by atomic sublattice","Hidden Eu order surfaces in Te CDW phonon map","Three coherent modes, two sublattices: EuTe4 phonons resolved","Resonant X-rays trace each atom's role in CDW's phonons","Element-specific phonons in EuTe4 exposed by X-ray timing"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole separation into Te and Eu dynamics rests on assuming the measured intensity at each X-ray energy is a clean, weighted sum of a pure Te signal and a pure Eu signal, with the Te weight independent of energy; if the off-resonant X-rays also see Eu, or the weights do not cleanly separate the sublattices, the assignment of each mode to a sublattice breaks down.","fun_headline_variants_meta":{"raw":{"variants":["X-ray energy peels phonon modes by atomic sublattice","Hidden Eu order surfaces in Te CDW phonon map","Three coherent modes, two sublattices: EuTe4 phonons resolved","Resonant X-rays trace each atom's role in CDW's phonons","Element-specific phonons in EuTe4 exposed by X-ray timing"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000255,"raw_usage":{"total_tokens":1569,"prompt_tokens":938,"completion_tokens":631,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":554,"completion_tokens_details":{"reasoning_tokens":538}},"tokens_in":554,"tokens_out":631,"duration_ms":5994,"temperature":1.0,"reasoning_tokens":538,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:33:16.292861+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Record the CDW diffraction intensity in a fully detuned probe (more than about 10 eV below the Eu M5 edge) after penetration-depth correction and Fourier-transform it: the paper's decomposition predicts that the decoupled Te trace contains no 1.6 THz component, so a robust 1.6 THz oscillation there would falsify the Eu-only assignment and the two-weight decomposition.","supporting_citations":[{"cited_title":"Burkel, J","cited_arxiv_id":null,"evidence_quote":"Establishes resonant elastic soft X-ray scattering as an element-selective probe of order parameters, giving the physical basis for tuning to Eu edges."},{"cited_title":"Xiao, W.-H","cited_arxiv_id":null,"evidence_quote":"Shows resonant enhancement of CDW diffraction in rare-earth tritellurides and supplies the basis for attributing intensity changes at absorption edges to charge components."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the resonant X-ray decomposition of a spatial modulation into separate atomic (Cu 3d and O 2p) contributions, the template for equation (1)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Extends the weighted-sublattice decomposition to bond and ligand order in IrTe2, supporting the same functional form."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Characterizes EuTe4 and establishes the Te-monolayer CDW at $q = (0, 0.644 b, 0)$, the material and wavevector probed here."},{"cited_title":"Takubo, R","cited_arxiv_id":null,"evidence_quote":"Supplies the DFT functional used to compute the phonon eigenvectors compared with the experimental modes."},{"cited_title":"Rettig, C","cited_arxiv_id":null,"evidence_quote":"Demonstrates tr-RXS disentangling of transient charge order from structural dynamics in manganites, the direct methodological precedent for separating sublattice contributions in the time domain."}],"review_version":2}