{"id":"462798e1-65b5-42ab-866e-f0aa0935ec3c","arxiv_id":"2506.02237","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Interfacial phonon modes in YBCO/LSAT thin films are identified by subtracting delocalized surface phonon contributions from atomic-scale EELS maps.","lead":"This paper uses atomic-scale electron microscopy to map phonon vibrations at the interface between the superconductor YBCO and its substrate LSAT. It identifies new interface-specific phonon modes and introduces a method to remove confusing surface signals.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The subtracted interface spectrum relies on the assumption that YBCO surface phonon scattering is identical at the YBCO/W and YBCO/LSAT edges; the two claddings differ dielectrically, so uncancelled surface polariton signal could masquerade as the claimed 40 and 75 meV interface peaks.","rationale":"The paper is carefully hedged and the experimental data include independent support: the atomically resolved EELS maps in Fig. 4 show 35–45 meV modulation tied to CuO2 planes and 70–80 meV signal localized at the interface, and the ELF calculation reproduces three broad features. Those pieces do not, however, establish the two cleaned peaks in Fig. 3(c), because those peaks are the output of the subtraction pipeline. The most load-bearing step in that pipeline is the W-side versus LSAT-side YBCO surface cancellation. The reader identified this exact assumption. The W layer is metallic and LSAT is a polar dielectric with strong Reststrahlen bands overlapping the claimed modes; surface-loss spectra at a dielectric interface and a metal interface are generically different. The paragraph dismissing W/YBCO phonon–plasmon coupling only addresses mode hybridization, not the difference in bare surface-loss functions, so the cancellation is unsupported. The alternative concern—the LSAT aloof-wedge background estimate—is also real, but the paper labels it an estimate and the YBCO subtraction is the first, qualitatively stronger assumption. A BEM simulation with the sample's actual geometry and published optical constants would settle the issue; if it shows large residuals, the interface peaks would not be robust, and if it confirms cancellation, the conditional acceptance should hold. Because the manuscript itself flags the estimate status, the reader's CONDITIONAL verdict is appropriate and no further adjustment is needed.","tokens_in":11007,"tokens_out":5911,"duration_ms":57193,"concrete_test":"Perform an independent boundary-element simulation of the exact cross-section—YBCO rod (~150 nm wide, 50 nm thick) with W on one side and LSAT wedge on the other—using the optical constants cited for YBCO and LSAT (Refs. [31,32]) and the MNPBEM toolbox cited in Ref. [33]. Compute the vibrational EELS loss probability for a 60 keV probe at the two equidistant positions used in the Fig. 3(a) subtraction. If the two simulated surface-loss spectra differ by more than the amplitude of the reported 40 or 75 meV peaks after the same subtraction, the cancellation assumption is falsified and the interface spectrum is not isolated. If they match, the YBCO subtraction step is validated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central result—the two broad interface peaks near 40 and 75 meV in Fig. 3(c)—is obtained by first subtracting a spectrum taken near the YBCO/LSAT interface from one at an equidistant position near the YBCO/W top interface (Section C). This step cancels YBCO surface contributions only if the surface-loss function at the two YBCO edges is identical. The surrounding media are not equivalent: one edge is adjacent to LSAT, a polar dielectric whose Reststrahlen bands lie in the 35–100 meV range, the other to a tungsten protective layer. Surface phonon polaritons of the YBCO rod are set by the dielectric boundary conditions of the whole cross-section, and a metal cladding modifies the surface-loss function even when plasmon–phonon coupling is detuned. The paper's dismissal of W/YBCO coupling (Section C) does not establish equality of the two bare surface contributions; it only excludes one hybridization channel. Consequently the subtraction leaves a residual difference of YBCO surface spectra that can be of the same order as the claimed interface signal. The manuscript itself states that the result is an 'estimate' and that short-wavelength interface contributions 'cannot be distinguished from remaining surface contributions' (Section C), so this is the load-bearing point rather than an internal inconsistency.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports atomic-scale vibrational electron energy loss spectroscopy (EELS) mapping across an epitaxial YBa2Cu3O7-δ/(La,Sr)(Al,Ta)O3 (YBCO/LSAT) interface. The authors identify delocalized and localized phonon scattering contributions near the interface and propose a subtraction procedure to remove surface phonon polariton contributions from the surrounding environment. After subtracting a YBCO surface contribution via a spatial symmetry argument and an LSAT surface contribution estimated from an isolated wedge, the residual interface spectrum exhibits two broad peaks near 40 and 75 meV. The paper also presents atomically resolved maps showing distinct scattering modulations at CuO2 planes and CuO chains, and argues that these features are unique to the interface region.","tokens_in":11315,"tokens_out":3299,"duration_ms":31441,"significance":"If the subtraction assumptions are valid, the work would provide the first local vibrational spectroscopy of a YBCO/LSAT interface and a methodological template for isolating interfacial phonons in complex oxide heterostructures, with potential implications for understanding electron-phonon coupling at cuprate interfaces. The authors deserve credit for explicitly addressing delocalized surface phonon polariton contributions, which are often neglected in interface EELS studies, and for providing detailed experimental data along with a transparent statement of the method's limitations. However, the central result—the 40 and 75 meV interface peaks—rests on two unverified or approximate subtraction steps, and the manuscript itself acknowledges that short-wavelength interface contributions cannot be separated from remaining surface contributions. The significance is therefore conditional on additional validation.","major_comments":[{"comment":"The spatial subtraction step assumes that surface phonon scattering is identical at the YBCO/W and YBCO/LSAT edges so that subtracting an equidistant spectrum from the opposite side cancels YBCO surface contributions. The two interfaces are not physically symmetric: one edge is adjacent to a tungsten metal layer and the other to a polar dielectric LSAT substrate whose Reststrahlen bands lie in the 35–100 meV range. The surface-loss function of the YBCO rod depends on the full dielectric boundary condition of the cross-section, and a metal cladding modifies the surface-loss spectrum even when plasmon–phonon hybridization is detuned. The paper's dismissal of coupled phonon-plasmon modes due to energy detuning does not establish equality of the bare surface contributions; it only excludes one coupling channel. The residual difference between the two surface spectra can be of the same order as the claimed interface signal. This is the load-bearing assumption for the extracted 40 and 75 meV peaks, and it needs a direct test, for example by comparing subtraction of spectra taken at equidistant positions far from the interface, or by including simulations with the actual W and LSAT claddings.","section":"Section C, Figure 3(b)"},{"comment":"The LSAT surface contribution is estimated from an isolated LSAT wedge spectrum scaled by a screening factor to account for the YBCO presence. The scaling factor is effectively a free parameter, and no uncertainty analysis is provided. The manuscript states that the LSAT surface contributions account for up to 60–70% of the original scattering at the interface, so the final subtracted spectrum is highly sensitive to the choice of this scaling factor and to spectral alignment errors. The authors should provide a sensitivity analysis, such as varying the scaling factor by a reasonable range and showing that the positions and relative intensities of the 40 and 75 meV peaks remain within the experimental uncertainty, or else quantifying the uncertainty on the residual spectrum. Without such analysis, the existence of the two interface peaks is not quantitatively robust.","section":"Section C and Supplementary Material, Figure 3(c)"},{"comment":"The manuscript explicitly states that short-wavelength interface contributions 'cannot be distinguished from remaining surface contributions.' This acknowledged degeneracy means that the residual spectrum (green curve in Fig. 3(c)) cannot be uniquely assigned to interfacial phonon modes; it is at best a mixture of interface and uncancelled surface signals. Given that the subtraction steps involve an approximate LSAT estimate and an unverified spatial symmetry, the conclusions should be tempered accordingly. The claim of 'unveiling interfacial phonon modes' is overstated unless additional evidence—such as momentum-resolved measurements or full scattering calculations including impact scattering—is provided to break the degeneracy.","section":"Section C, final paragraph"}],"minor_comments":[{"comment":"The text contains a typo: '60-70%%' should be '60–70%'.","section":"Section C, Figure 3(c)"},{"comment":"The caption lists '(a)' twice; the schematic and the HAADF image appear to be labeled inconsistently with the in-text references.","section":"Figure 2 caption"},{"comment":"No error bars or confidence intervals are shown on any of the EELS spectra or maps. While the paper notes that noise is preserved, a quantitative statement of signal-to-noise ratio or repeated-measurement variability would help assess the significance of the reported peak positions and map contrasts.","section":"Methods and Figures 2–4"},{"comment":"The discussion of phonon hardening in Y-124 structures cites reference [24] for YBa2Cu4O8, which is appropriate, but the connection to the observed spectral similarities between double CuO chains and CuO2 planes would benefit from a more explicit quantitative comparison.","section":"Section D, paragraph 7"}],"recommendation":"major_revision","confidential_remarks":"The manuscript relies heavily on the authors' own prior work for the methodological framework (refs. [4,11,14,20,22,23]), which is not inappropriate but should be balanced by independent validation. The paper fits the scope of the journal. The main concern is that the central interface-mode claim depends on a symmetry assumption that is not substantiated and an LSAT background estimate with a free scaling parameter; these points are fixable in revision but require substantial additional analysis or control experiments."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know: this is the first atomic-scale vibrational EELS study of a YBCO/LSAT interface, and the spatially resolved maps in Figure 4 are genuinely informative. The subtraction method used to isolate an 'interface spectrum' is the weak point; the symmetry assumption behind it doesn't hold up under scrutiny, but the paper's main claim doesn't rest on that subtraction alone.\n\nWhat's new: the authors map phonon scattering across a cuprate/substrate interface with sub-nanometer resolution, and they show energy-specific spatial modulations near the interface—e.g., the 35–45 meV signal oscillating with the Y-123 layer structure, the 70–80 meV enhancement at the interface, and the suppression at the first CuO chain. These are raw data, not subtraction artifacts. The paper also correctly emphasizes that long-range surface phonon polaritons from the surrounding environment can contaminate interface spectra, a point prior interface EELS papers mostly ignored. The authors are appropriately cautious in their claims, explicitly calling the subtracted spectrum an estimate and noting that short-wavelength contributions cannot be separated from residual surface signal.\n\nWhere it gets shaky: the surface-removal step in Section C assumes the YBCO surface phonon scattering is identical at the YBCO/W and YBCO/LSAT edges, so that spatial subtraction cancels it. That's not defensible. The two interfaces have different dielectric boundary conditions—metal cladding on one side, polar dielectric LSAT on the other—so the YBCO surface loss function will differ. Dismissing W/YBCO coupling due to energy detuning only rules out a hybridized mode; it doesn't establish equal bare surface contributions. The LSAT background estimate is also approximate, with no error bars anywhere. So the specific 40 and 75 meV peaks in the subtracted spectrum (Fig. 3c) should be treated with caution.\n\nThat said, the spatial maps in Fig. 4 show similar energies as local features, so the central conclusion—that interface-specific phonon modes exist—is not solely dependent on the flawed subtraction. I'd want the authors to either validate the subtraction with simulations of the full cross-section geometry or reframe the paper around the direct spatial evidence.\n\nWho's this for? Electron microscopists and anyone studying phonons at oxide interfaces. It deserves a serious referee, but the review should push on the subtraction assumption and demand error analysis. I'd send it out.\n\nBest.","headline":"First atomic-scale vibrational EELS of a YBCO/LSAT interface with real spatial evidence for interface phonons, but the surface-subtraction method rests on a questionable symmetry assumption.","tokens_in":11812,"tokens_out":3373,"would_cite":true,"duration_ms":31084,"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":"Vibrational electron spectroscopy isolates two interfacial phonon modes in an epitaxial YBCO/LSAT thin film after subtracting surface phonon contributions that otherwise dominate the signal.","keywords":["vibrational electron energy loss spectroscopy","interfacial phonons","surface phonon polaritons","YBCO thin films","LSAT substrate","cuprate superconductors","Begrenzung effect","atomic-scale EELS mapping"],"falsifier":"A decisive test would be to repeat the measurement on a YBCO film whose two faces are capped with the same material, making the surface subtraction symmetric; if the 40 and 75 meV peaks survive in that subtracted interface spectrum they are genuine interface modes, while their shift or disappearance would show the original subtraction leaked surface signal.","tokens_in":10841,"feed_emoji":"🔬","tokens_out":13585,"duration_ms":111252,"temperature":0.7,"pith_summary":"This paper aims to show that vibrational electron energy loss spectroscopy—probing lattice vibrations (phonons) with an atomic-scale electron beam—can resolve modes specific to the interface between a high-temperature superconducting YBa2Cu3O7−δ (YBCO) film and its (La,Sr)(Al,Ta)O3 (LSAT) substrate. The authors identify interfacial modes appearing as broad peaks near 40 and 75 meV that are distinct from the bulk phonon spectra of either material. They also demonstrate that long-range surface phonons from the surrounding specimen contribute up to 60–70% of the raw interface scattering, and they introduce a subtraction procedure that removes those delocalized contributions. This matters because local electron–lattice coupling at cuprate interfaces is thought to be relevant to high-temperature superconductivity, and conventional phonon probes lack the spatial resolution to see these modes.","feed_headline":"40 and 75 meV phonon modes isolated at cuprate interface","feed_subtitle":"Vibrational EELS subtraction reveals interface phonons hidden under surface noise in a cuprate superconductor.","key_machinery":"The engine of the argument is the two-step subtracted interface spectrum. The first step uses the spatial symmetry of the surface-scattering profile across the YBCO rod: by the Begrenzung effect—the depletion of bulk scattering near a surface in favor of surface modes—the spectrum very close to an edge is dominated by surface phonons, and the authors assume this surface contribution is the same at the W-coated side and the LSAT side, so subtraction of equidistant spectra cancels it. The second step estimates the LSAT surface phonon polariton contribution from an aloof-mode EELS measurement of an LSAT wedge with the same geometry, scaled by screening effects of the YBCO. The paper also computes an energy-loss function for long-wavelength LO interface phonons, which places three calculated peaks inside the experimental bands and supports the assignment of the measured features to interface bulk phonon modes.","core_discovery":"On the paper's own terms, the central discovery is that the YBCO/LSAT interface sustains vibrational modes that are neither bulk YBCO nor bulk LSAT modes, and that these modes become visible only after delocalized surface phonon scattering is removed. The subtraction proceeds in two steps: spectra at equidistant positions on the YBCO/W side and the YBCO/LSAT side are subtracted to cancel the YBCO surface contribution, then an aloof-mode spectrum of an isolated LSAT wedge, corrected for screening by YBCO, estimates the LSAT surface contribution to be subtracted. The resulting interface spectrum is dominated by two broad peaks centered near 40 and 75 meV, and the paper reports that the removed surface contributions accounted for up to 60–70% of the original signal. Atomic-scale EELS maps then show that the 35–45 meV scattering oscillates with the Y-123 layered structure, peaking at CuO2 planes and suppressed at CuO chains, while the 70–80 meV scattering is enhanced at the interface and suppressed specifically at the first CuO chain. These modes are presented as inaccessible to traditional phonon spectroscopy because of its limited spatial sensitivity.","pith_inferences":["Beyond the paper, the same subtraction logic could be applied to FeSe/SrTiO3 interface data, where interfacial phonons have been linked to enhanced superconductivity; comparing corrected and raw spectra would reveal whether surface polariton contamination was also present there.","Beyond the paper, a boundary-element simulation of the actual W/YBCO/LSAT wedge geometry would quantify how much residual surface signal survives the symmetric subtraction, testing the assumption that the two YBCO surfaces are equivalent scatterers.","Beyond the paper, the observed shift of the 40 meV band toward the interface and the suppression of the 75 meV band at the first CuO chain suggest these phonon signatures track local oxygen content or strain; a joint EELS measurement of oxygen K-edge and phonon maps on the same specimen could test that correlation.","Beyond the paper, because the EELS collection angle spans the full Brillouin zone, the mismatch between the measured subtracted spectrum and the q→0 energy-loss function implies that short-wavelength impact-scattering phonons contribute at the interface; computing those cross sections, as the authors suggest, is the natural next step."],"forward_implications":["Interface EELS studies of heterostructures should remove surface phonon polariton contributions before assigning interface modes; in this geometry they accounted for up to 60–70% of the raw interface scattering.","The two residual peaks near 40 and 75 meV constitute a spectral signature of the YBCO/LSAT interface that future electron–phonon coupling calculations for cuprate interfaces should reproduce.","Atomic-scale maps show that the first CuO2-plane and CuO-chain layers adjacent to the interface have vibrational responses distinct from the bulk, implying local lattice dynamics change within a few unit cells of the interface.","The same two-step subtraction procedure can be applied to other complex-oxide heterostructures to isolate interface phonons from surface background.","The delocalized scattering in the 80–90 meV range around the interface is consistent with an LO-type interface bulk phonon, giving a concrete prediction for future scattering-cross-section calculations."],"supporting_citations":[{"why":"It establishes the surface-versus-bulk vibrational EELS framework and the Begrenzung effect used in the surface-subtraction procedure.","marker":"[11]"},{"why":"It supplies the YBCO phonon density-of-states mode assignments used to attribute spectral bands to CuO2 planes, CuO chains, and BaO.","marker":"[19]"},{"why":"It provides the LSAT infrared-active phonon energies used to interpret and remove the substrate-side surface contribution.","marker":"[21]"},{"why":"It documents localized surface phonon polariton modes in rod/wedge geometries and the detuning argument used to exclude coupled modes.","marker":"[22]"},{"why":"It provides the detailed-balancing-based zero-loss background subtraction used to extract inelastic EELS signal.","marker":"[14]"},{"why":"It frames the sample context by documenting epitaxial strain and CuO intergrowths in YBCO heterostructures.","marker":"[3]"},{"why":"It characterizes the atomic structure of epitaxial YBCO/LSAT interfaces and planar defects used to interpret the HAADF images.","marker":"[16]"},{"why":"It supplies optical phonon data for double-chain YBCO structures used to interpret the double CuO chain's resemblance to CuO2 planes.","marker":"[24]"},{"why":"It establishes the approximation of meV-resolution phonon EELS by the phonon density of states, used to compare measured spectra with expected bulk modes.","marker":"[20]"}],"fun_headline_variants":["Interface phonons in cuprate revealed after surface noise removal","40 and 75 meV modes pinned to cuprate interface","Subtracting surface noise uncovers hidden interface phonons","Atomic-scale map shows interfacial phonons in YBCO/LSAT","New method isolates interface phonons in high-Tc superconductor"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the surface phonon scattering from YBCO is the same at its tungsten-coated side and its LSAT side, so subtracting equidistant spectra from the two sides cancels all YBCO surface contributions; if the two interfaces scatter differently, the residual peaks at 40 and 75 meV contain uncancelled surface signal.","fun_headline_variants_meta":{"raw":{"variants":["Interface phonons in cuprate revealed after surface noise removal","40 and 75 meV modes pinned to cuprate interface","Subtracting surface noise uncovers hidden interface phonons","Atomic-scale map shows interfacial phonons in YBCO/LSAT","New method isolates interface phonons in high-Tc superconductor"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000197,"raw_usage":{"total_tokens":1363,"prompt_tokens":942,"completion_tokens":421,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":558,"completion_tokens_details":{"reasoning_tokens":334}},"tokens_in":558,"tokens_out":421,"duration_ms":4153,"temperature":1.0,"reasoning_tokens":334,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:26:57.981805+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test would be to repeat the measurement on a YBCO film whose two faces are capped with the same material, making the surface subtraction symmetric; if the 40 and 75 meV peaks survive in that subtracted interface spectrum they are genuine interface modes, while their shift or disappearance would show the original subtraction leaked surface signal.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It establishes the surface-versus-bulk vibrational EELS framework and the Begrenzung effect used in the surface-subtraction procedure."},{"cited_title":"Renker, F","cited_arxiv_id":null,"evidence_quote":"It supplies the YBCO phonon density-of-states mode assignments used to attribute spectral bands to CuO2 planes, CuO chains, and BaO."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the LSAT infrared-active phonon energies used to interpret and remove the substrate-side surface contribution."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It documents localized surface phonon polariton modes in rod/wedge geometries and the detuning argument used to exclude coupled modes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the detailed-balancing-based zero-loss background subtraction used to extract inelastic EELS signal."},{"cited_title":"Zhang, N","cited_arxiv_id":null,"evidence_quote":"It frames the sample context by documenting epitaxial strain and CuO intergrowths in YBCO heterostructures."},{"cited_title":"Gauquelin, H","cited_arxiv_id":null,"evidence_quote":"It characterizes the atomic structure of epitaxial YBCO/LSAT interfaces and planar defects used to interpret the HAADF images."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies optical phonon data for double-chain YBCO structures used to interpret the double CuO chain's resemblance to CuO2 planes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It establishes the approximation of meV-resolution phonon EELS by the phonon density of states, used to compare measured spectra with expected bulk modes."}],"review_version":1}