{"id":"ab13aec8-4496-41d1-a33b-9e55bfeeeba1","arxiv_id":"2412.12029","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Epitaxial strain in PbTiO3 thin films tunes the Er3+ resonant fluorescence peak position, intensity, and linewidth, with a second peak set appearing in the most in-plane-polarized film.","lead":"Researchers grew erbium-doped lead titanate films on five substrates and found that the strain and ferroelectric domain orientation of the film change the erbium emission spectrum, including its brightness, color, and linewidth. The work points to ferroelectric materials as tunable hosts for telecom-wavelength quantum emitters.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Single-film-per-substrate design leaves the central strain/domain attribution confounded by uncontrolled Er incorporation and defect chemistry; repeated depositions and concentration checks are needed before the tuning claim is fully supported.","rationale":"Both the reader and I read this as a solid but conditional experimental characterization. The strongest quantitative support is the 7 K Y1→Z2 frequency shift of ~1.8 cm−1 between STO and DSO, which exceeds the 0.8 cm−1 spectrometer resolution; the intensity ordering across five substrates is also very large. Those data make it implausible that nothing structural is happening. The explicit sub-resolution linewidth statement (Sec. II C) is a real weakness, but it only removes one of the three sub-claims; energy and intensity would remain. The more load-bearing gap is that each substrate is a single film, with no measured Er concentration or defect density and no replicate samples. Substrate-dependent Er sticking, vacancy formation, or even different Er site populations (as evidenced by the NSO second peak set) could in principle generate the same trends. The manuscript itself acknowledges the linewidth limit but does not address the single-film issue, so the appropriate verdict remains conditional: the trends are suggestive and internally consistent, but a second deposition series and concentration characterization are required before the central strain/domain attribution is established. This does not move the verdict; it sharpens the condition.","tokens_in":17074,"tokens_out":5699,"duration_ms":56734,"concrete_test":"Deposit two additional Er:PTO films on STO and two on NSO under identical conditions (same PLD target and growth parameters), measure their emission under the same excitation, and quantify Er content by RBS or SIMS. If the within-substrate run-to-run spread in Y1→Z2 peak frequency, integrated intensity, and Gaussian width approaches or exceeds the reported LSAT-to-NSO differences (e.g., Δν up to ~1.8 cm−1 at 7 K), the strain/domain attribution fails; if the spread is much smaller and concentration is flat, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's causal claim—that epitaxial strain and c/a domain fraction tune Er3+ emission—requires that the five films differ only in strain/domain state. That condition is not secured. Each substrate condition is represented by one film; no RBS/SIMS or other measurement of Er concentration or defect density is reported, and PL intensity is not normalized to Er content. The intensity trend (LSAT>STO>DSO>GSO>NSO) is exactly the trend a systematic variation in Er incorporation or nonradiative recombination with substrate would also produce. Frequency shifts are less concentration-sensitive, but at 77 K they are 0.31–0.73 cm−1, comparable to the stated 0.8 cm−1 spectral resolution, and linewidth differences are explicitly admitted to be below resolution (Sec. II C). The NSO-specific second peak set (Sec. II D) shows that site occupancy or charge compensation changes between films, so 'domain fraction' is not the only structural variable. Without replicate films on the same substrate, the monotonic trends cannot be uniquely attributed to strain/domain rather than substrate-dependent chemistry or sample-to-sample variation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports resonant photoluminescence measurements of Er3+-doped PbTiO3 thin films grown on five substrates (LSAT, STO, DSO, GSO, NSO) with different epitaxial strain and ferroelectric domain configurations. The authors find that the 4I13/2 → 4I15/2 emission intensity, peak frequency, and linewidth vary systematically with the c/a domain fraction determined by XRD, with c-dominated films exhibiting brighter, lower-energy, narrower emission. An additional set of peaks in the NSO-grown film is attributed to a second Er3+ optical center, possibly a different substitution site or charge-compensated configuration, while similar peaks in GSO are traced to Er3+ impurities in the substrate. The paper concludes that epitaxial strain and ferroelectric domain engineering provide a tuning knob for Er3+ color centers.","tokens_in":17226,"tokens_out":3847,"duration_ms":32473,"significance":"If the claimed trends hold, this work introduces ferroelectric domain state as a practical control parameter for telecom-wavelength Er3+ emitters, a valuable addition to the quantum-defect toolkit. The paper's strengths include direct spectroscopic observation across a substrate series, excitation–emission spectral maps over a wide frequency range, and temperature-dependent control measurements (Appendix E) that argue against mounting artifacts. The identification of substrate-related Er3+ background in GSO is carefully done. However, the central claims rest on a single film per substrate with no measured Er concentration or defect density, and the reported linewidth and frequency shifts are partly at or below the stated spectrometer resolution.","major_comments":[{"comment":"The conclusion that strain/domain fraction tunes the Er3+ emission requires that the five films differ only in strain and domain state. Each condition is represented by a single film, the Er concentration is only nominal (0.01 at.%), and no RBS/SIMS or other composition/defect characterization is reported. The observed intensity trend LSAT>STO>DSO>GSO>NSO is exactly what a systematic variation in Er incorporation or nonradiative recombination with substrate would produce. Replicate films on at least two substrates and a quantitative concentration measurement are needed to support the causal attribution.","section":"Section II C and Appendix A"},{"comment":"The linewidth changes between films are smaller than the stated spectrometer resolution of 0.8 cm−1, as the paper acknowledges. For example, at 77 K the Y1→Z1 linewidth varies from 2.36 to 2.57 cm−1, a spread comparable to the resolution, and the NSO sample at 2.43 cm−1 does not follow the monotonic trend. These sub-resolution differences cannot, by themselves, support the claim that c-domain films have narrower linewidths; a higher-resolution spectrometer or a statistical analysis over repeated measurements is required.","section":"Section II C, Fig. 4e–h"},{"comment":"At 77 K the reported emission frequency shifts are 0.31–0.73 cm−1, again comparable to the 0.8 cm−1 resolution. Although the consistency of the direction of the shifts across the three transitions is suggestive, no fit uncertainties or confidence intervals are given anywhere in the paper (Tables A1–A2 list fit parameters only). Without uncertainties it is impossible to assess whether the small shifts are significant.","section":"Section II C, Fig. 4g–h"},{"comment":"The additional set of peaks in the NSO film is attributed to a second Er3+ center (A-site vs B-site, charge compensation, or selection rules). The presence of a qualitatively different optical center in one of the five films means that the films are not simply strained versions of the same material—the local Er3+ environment changes across the series. The paper should explicitly address how this second center affects the comparison of the Y1→Z1/Z2/Z3 trends and whether the same center is being tracked on all five substrates.","section":"Section II D"}],"minor_comments":[{"comment":"The word 'ferroelectic' is a typo; 'ferroelectric' is intended.","section":"Abstract and Section I"},{"comment":"The caption reads 'Janus Cryostat'; this should be 'Janis cryostat' to match the text in Appendix B.","section":"Fig. A1 caption"},{"comment":"The phrase 'different optical selection thermodynamics' is vague; likely 'different optical selection rules' is intended.","section":"Section II D"},{"comment":"The NSO Y1→Z1 linewidth (2.43 cm−1) is smaller than the GSO value (2.57 cm−1), which breaks the monotonic trend claimed in the text; this should be discussed or at least noted.","section":"Table A2"},{"comment":"The nominal Er concentration of 0.01 at.% is stated, but no calibration or error bar is given; this should be mentioned as a nominal value in the main text when intensity comparisons are made.","section":"Appendix A"}],"recommendation":"major_revision","confidential_remarks":"The paper would be strengthened by a clear statement of the statistical significance of the sub-resolution shifts, and by replication of at least one sample to check sample-to-sample variability. The novelty is real, but the current evidence does not fully rule out substrate-dependent chemistry as the cause of the observed trends."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know right away: this is the first systematic five-substrate strain series for Er3+ in PbTiO3, and the monotonic intensity and peak-position trends are real enough to take seriously. But the paper's closing claim—narrower linewidths in c-domain films—rests on sub-resolution fitting differences, and every substrate condition is one film.\n\nThe work has real merits. The substrate series spans compressive to tensile strain, with c/a domain fractions independently quantified by XRD and corroborated by PFM. The temperature-dependent control measurements are a thoughtful check against mounting artifacts, and the bare-substrate controls convincingly assign the GSO features to Er contamination in the substrate, not the film. The NSO-specific second peak set is a genuinely new observation, and the authors are upfront about not being able to distinguish between site occupancy, charge compensation, and selection rules. They even flag that their linewidth differences are smaller than the 0.8 cm-1 spectrometer resolution. That honesty counts.\n\nThe biggest problem is the design. One film per substrate, no measurement of Er concentration in each film, no RBS/SIMS, and PL intensity normalized without reference to Er content. The intensity ordering (LSAT > STO > DSO > GSO > NSO) is exactly what you would expect from uncontrolled Er incorporation or non-radiative recombination differences. The frequency shifts at 77 K are 0.31–0.73 cm-1, comparable to the stated 0.8 cm-1 resolution, so the 7 K data (0.58–1.78 cm-1) carry the argument. The linewidth variation is explicitly below resolution and fits have no error bars, so I'd treat that part of the conclusion as not demonstrated. The NSO second peak set shows that site chemistry changes between films, which is another reason to be cautious about attributing all trends to domain fraction.\n\nWho is this for? Experimental researchers working on Er-based quantum emitters in ferroelectrics and perovskite oxides. They will cite it for the substrate series and the NSO observation. It deserves a serious referee; the core intensity/frequency trend is likely real, but the paper needs replicate samples, concentration checks, and a more careful spectral resolution statement before the linewidth narrative is accepted.\n\nSend it to peer review. Not because it is flawless, but because it is a solid, honest experimental study that will improve with referee demands for replication and characterization. A desk reject would lose a useful dataset.","headline":"First systematic strain series for Er3+ in PbTiO3 with a convincing intensity/frequency trend, but the linewidth claim and single-film design need more support before the full tuning story stands.","tokens_in":17839,"tokens_out":2524,"would_cite":true,"duration_ms":23200,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In Er3+-doped ferroelectric PbTiO3 thin films, the brightness, peak energy, and linewidth of telecom emission all track the epitaxial strain and ferroelectric domain configuration imposed by the substrate.","keywords":["Er3+ color centers","PbTiO3 thin films","ferroelectric domains","epitaxial strain","resonant fluorescence","telecom emission","rare-earth dopants","crystal field tuning"],"falsifier":"Measure the three transitions on the same set of films with a spectrometer resolution better than 0.8 cm−1; if the fitted linewidth ordering among substrates does not persist, the narrowing claim is refuted. Alternatively, grow multiple films per substrate or verify Er concentration and defect density to test whether the intensity trend tracks strain or instead tracks sample-to-sample variations.","tokens_in":16839,"feed_emoji":"🔬","tokens_out":6084,"duration_ms":50751,"temperature":0.7,"pith_summary":"The paper sets out to show that the optical emission of Er3+ ions in a ferroelectric host can be tuned by the host's structural state. By growing PbTiO3 films on substrates with different lattice constants, the authors vary the epitaxial strain and therefore the ferroelectric domain orientation, and they report that the Er3+ telecom emission follows this structural knob. Films dominated by c-domains (out-of-plane polarization) are brighter, emit at lower energy, and have narrower lines than films dominated by a-domains. A second set of emission peaks appears in films with in-plane polarization; the authors attribute it to a different Er3+ lattice site, charge compensation, or selection rules. If the claim holds, strain engineering of a ferroelectric matrix becomes a practical tuning parameter for quantum-relevant rare-earth color centers.","feed_headline":"Epitaxial strain tunes erbium emission in ferroelectric films","feed_subtitle":"In PbTiO3, c-domain films emit brighter, lower-energy Er3+ telecom light than a-domain films.","key_machinery":"The central object is the Er3+ 4I13/2→4I15/2 transition at about 6500 cm−1 (1.55 eV, telecom band), probed by resonant fluorescence spectroscopy. The tuning mechanism is the tetragonal distortion of PbTiO3 (c:a ≈ 4.11 Å : 3.91 Å) and the domain configuration it adopts under epitaxial strain: compressive substrates favor c-domains with out-of-plane polarization, while tensile substrates favor a-domains with in-plane polarization. The c:a domain fraction, quantified from X-ray diffraction peak ratios, serves as the structural proxy that correlates with emission intensity, peak position, and linewidth. A secondary feature is the second set of peaks observed only in predominantly a-domain films, which the paper hypothesizes to arise from Er3+ substituting different perovskite cation sites, different charge compensation, or altered selection rules.","core_discovery":"The central claim is that the Er3+ emission spectrum in PbTiO3 is controlled by the epitaxial strain and resulting ferroelectric domain configuration of the host. Across five films of identical nominal composition on LSAT, STO, DSO, GSO, and NSO substrates, the authors observe a systematic trend: as the c-domain fraction decreases, photoluminescence intensity drops, emission frequencies shift to higher energy, and linewidths broaden. For the Y1→Z2 transition at 7 K the shift reaches 1.78 cm−1, and at 77 K the integrated counts fall by 57% for STO, 86% for DSO, and about 92% for GSO and NSO relative to LSAT. The paper also reports an extra set of Er3+ peaks in the NSO sample that is absent from the bare substrate and not explained by Er contamination in the GSO substrate case, and proposes that it arises from Er3+ occupying a different cation site or from charge-compensation differences. The authors argue that temperature-induced mounting artifacts cannot explain the trends by comparing with temperature-dependent measurements.","pith_inferences":["The paper leaves open whether the correlation is causal through strain directly or through the domain fraction; a test on a single film with electric-field-induced domain switching would separate the two and is an obvious next step.","The linewidth claim is the least secure because the fitted differences are below the 0.8 cm−1 spectrometer resolution; higher-resolution spectroscopy or single-ion measurements could confirm whether c-domains genuinely narrow the inhomogeneous line.","If Er3+ occupies both A and B sites with strain-dependent occupancy, then temperature- or strain-dependent site-selective excitation could turn PbTiO3 into a tunable multi-species emitter; the paper's excitation-emission maps already show the pathways needed to test this.","The systematic strain dependence, if confirmed, could be used in reverse as a non-invasive local strain probe, reading out lattice distortions through the Er3+ emission energy and linewidth."],"forward_implications":["Choosing or switching the ferroelectric domain orientation gives a direct handle on Er3+ emission energy, linewidth, and brightness without changing the dopant or its concentration.","The demonstrated coupling implies that other ferroic order parameters of the host, such as polarization direction or electric-field-induced domain reorientation, could modulate the defect spectrum dynamically, enabling strain- or field-driven control of rare-earth qubits.","The appearance of a second emission set in predominantly a-domain films suggests that strain may select which Er3+ substitutional site dominates, potentially enabling site-selective doping.","Resonant fluorescence at 7 K resolves shifts as small as about 0.6–1.8 cm−1, establishing PbTiO3 as a sensitive platform for probing crystal-field changes in epitaxial heterostructures."],"supporting_citations":[{"why":"Supplies the reference Er3+:PbTiO3 spectrum (6512, 6496, 6398 cm−1) that the new peak positions are compared against.","marker":"1"},{"why":"Provides the epitaxial strain engineering and XRD c:a domain fraction quantification method for PbTiO3 thin films.","marker":"27"},{"why":"Gives evidence that lanthanides can occupy A or B cation sites in perovskite titanates, underpinning the second-site hypothesis.","marker":"32"},{"why":"Demonstrates narrow optical linewidths in Er3+:TiO2, supporting the promise of rare-earth ions in oxide hosts and the site-sensitive emission picture.","marker":"17"},{"why":"Documents parasitic erbium photoluminescence, which the paper uses to identify Er contamination in the GSO substrate.","marker":"33"},{"why":"Shows ferroelectric polarization in PbTiO3 can control dopant spin anisotropy, motivating coupling between ferroic order and defect states.","marker":"19"}],"fun_headline_variants":["Strain twists erbium light in ferroelectric films","Ferroelectric strain bends erbium emission","Domain strain reshapes erbium spectra","Strain tunes erbium photons in PbTiO3","In-plane strain reveals extra erbium lines"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The narrower-linewidth claim rests on fitted linewidth differences smaller than the spectrometer resolution (0.8 cm−1), and each substrate condition is represented by a single film, so any uncontrolled film-to-film variation in Er concentration or defect density could account for the observed trends.","fun_headline_variants_meta":{"raw":{"variants":["Strain twists erbium light in ferroelectric films","Ferroelectric strain bends erbium emission","Domain strain reshapes erbium spectra","Strain tunes erbium photons in PbTiO3","In-plane strain reveals extra erbium lines"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000278,"raw_usage":{"total_tokens":1683,"prompt_tokens":1003,"completion_tokens":680,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":619,"completion_tokens_details":{"reasoning_tokens":605}},"tokens_in":619,"tokens_out":680,"duration_ms":5720,"temperature":1.0,"reasoning_tokens":605,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T14:20:55.546225+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the three transitions on the same set of films with a spectrometer resolution better than 0.8 cm−1; if the fitted linewidth ordering among substrates does not persist, the narrowing claim is refuted. Alternatively, grow multiple films per substrate or verify Er concentration and defect density to test whether the intensity trend tracks strain or instead tracks sample-to-sample variations.","supporting_citations":[{"cited_title":"Ma \\ and\\ author A","cited_arxiv_id":null,"evidence_quote":"Provides the epitaxial strain engineering and XRD c:a domain fraction quantification method for PbTiO3 thin films."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates narrow optical linewidths in Er3+:TiO2, supporting the promise of rare-earth ions in oxide hosts and the site-sensitive emission picture."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents parasitic erbium photoluminescence, which the paper uses to identify Er contamination in the GSO substrate."}],"review_version":1}