{"id":"6e42ee4a-b365-4692-89dc-97677cb23166","arxiv_id":"2411.15023","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Ta doping in SrTiO3 increases the optical band gap, saturates around 10-12.5% doping, and the material remains an indirect-gap n-type semiconductor.","lead":"This paper combines DFT calculations and spectroscopic ellipsometry to show that tantalum doping widens the optical band gap of strontium titanate while keeping it an indirect-gap conductor. The result matters for replacing indium-based transparent conducting oxides in displays and solar cells.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Experimental band-gap values are extracted by a Tauc n=1/2 extrapolation (Eq. 9) applied to a TL+Drude fit of degenerately doped metallic SrTiO3, although the text admits the TL model is insufficient for free carriers.","rationale":"Agree with the reader. The central argument is conditional: the qualitative DFT trend (HSE gap 3.16 eV for pristine STO, band unfolding preserving indirect character) is plausible and standard, and the qualitative ellipsometric peak shift is consistent, but the quantitative gap values and the specific saturation claim are not independently verifiable. The Tauc-Lorentz model is fitted to metallic films; a Tauc n=1/2 plot is not a valid measurement of an optical gap in a degenerately doped conductor. The manuscript itself flags the insufficiency of TL and the presence of FCA, so this is an internal inconsistency rather than merely an external disagreement with consensus. I also note that the ~10% saturation has no experimental point above 5% and only one theoretical point at 12.5%, so even a corrected experimental trend would need additional concentrations to establish the plateau. These concerns are addressable, not fatal: the paper's central claim may survive if the reanalysis confirms the trend. No code or raw data are shipped, so independent verification is currently impossible; conditional acceptance remains the appropriate outcome.","tokens_in":16241,"tokens_out":10542,"duration_ms":118466,"concrete_test":"Obtain the raw psi/delta spectra for the 0%, 2%, and 5% films and refit them with a degenerate-semiconductor model: an interband critical-point contribution multiplied by Fermi-Dirac occupation of the conduction-band final states (Fermi energy fixed by measured Hall carrier density) plus the same Drude term, with no a priori Tauc-Lorentz edge. Compare the resulting band-edge energies and their doping trend against the TL+Tauc values in Fig. 7. If the alternative fit has comparable MSE but yields different or non-monotonic onsets, the reported experimental gaps are model artifacts; if the onsets agree to within the claimed shifts, the Tauc concern is not decisive.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The weak load-bearing assumption is in Sec. III.C: the experimental Eg values (Fig. 7) are obtained by applying Tauc's formula (Eq. 9) with n=1/2 to absorption coefficients reconstructed from a Tauc-Lorentz+Drude ellipsometry model (Eqs. 6-8). The text itself states that free carriers make TL alone insufficient and that free-carrier absorption appears below the band edge. For a degenerate n-type conductor with E_F inside the conduction band, the optical absorption edge is a Pauli-blocked Fermi-edge transition contaminated by Drude/FCA, not the indirect semiconductor JDOS assumed by Tauc n=1/2; the persistence of an indirect ground-state band structure (Sec. III.B) does not validate that joint density of states. Moreover, TL already contains an Eg parameter and forces a (E-Eg)^2/E^2 edge, so a subsequent Tauc plot of the model's alpha is circular rather than an independent band-gap measurement. No raw psi/delta spectra, parameter uncertainties, or error bars on Eg are given (MSE <5% only constrains the overall fit), so the claimed monotonic increase and 'excellent agreement' with theory rest on this model-dependent reduction. If this extraction is biased, the experimental half of the central claim is not established, even though the qualitative epsilon1 peak shift and DFT results may still point in the right direction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript combines DFT (PBE and HSE06) supercell calculations with spectroscopic ellipsometry to study Ta-doped SrTiO3. The authors vary Ta concentration theoretically (0, 1.6%, 3.7%, 12.5%) and experimentally (0, 2%, 5%), and report structural relaxation, formation energies, Bader charges, band unfolding, and optical properties. Their central claims are that Ta doping increases the optical band gap of STO through the Moss-Burstein effect, that the band structure remains indirect at all studied concentrations, and that the gap increase saturates at roughly 10% doping, beyond which transparency in the visible is not significantly affected.","tokens_in":16525,"tokens_out":2643,"duration_ms":30301,"significance":"If established, the result would be practically relevant for designing Ta-doped SrTiO3 as a transparent conducting oxide and would resolve a discrepancy in earlier supercell DFT reports on the direct/indirect nature of the gap. The paper has clear strengths: the DFT calculations are not fitted to the ellipsometry data; the HSE06 gap for pristine STO is close to experiment; band unfolding is an appropriate tool for the supercell folding problem; and the comparison of formation energies across dopant concentrations is a useful addition. The qualitative trend that Ta doping shifts the interband absorption to higher energy is supported by both the calculated absorption and the measured epsilon1 peak shift. However, the quantitative band-gap values extracted from ellipsometry rest on a model-dependent Tauc analysis whose validity is not established for degenerately doped STO, and the saturation claim is inferred from a single high-concentration theoretical point with no experimental point above 5%.","major_comments":[{"comment":"The experimental Eg values in Fig. 7 are obtained by applying a Tauc plot with n=1/2 to absorption coefficients generated from a Tauc-Lorentz plus Drude model. This procedure is circular: the Tauc-Lorentz oscillator already contains a band-gap parameter Eg and enforces an (E-Eg)^2/E^2 absorption edge, so re-extracting Eg from a Tauc plot of the model's alpha essentially recovers the fit parameter rather than providing an independent measurement. For degenerately doped STO, where the Fermi level lies inside the conduction band, the optical edge is a Pauli-blocked Fermi-edge transition contaminated by free-carrier absorption; the persistence of an indirect ground-state band structure does not by itself validate the indirect-semiconductor joint density of states assumed by the n=1/2 Tauc form. The authors should justify the Tauc model for these metallic films, report the TL Eg parameter directly with uncertainties, or extract Eg from a critical-point analysis of the raw ellipsometric spectra.","section":"Sec. III.C, Eqs. (6)-(9)"},{"comment":"The central saturation claim rests on a single theoretical concentration point at 12.5% and on experimental data only up to 5% Ta. There is no experimental point above 5% and no second high-concentration theoretical point, so the statement that the band gap 'saturates at ~10% doping' is an extrapolation. The apparent saturation in Fig. 4(b) could also be a finite-supercell or band-filling artifact. The authors should either add intermediate high-concentration calculations (e.g., 6.25% and 9.375% with 16- and 8-atom supercells) or soften the saturation claim to a limited-concentration trend.","section":"Sec. III.C, Fig. 7 and Fig. 4(b)"},{"comment":"No numerical values, error bars, or parameter correlation information are given for the extracted Eg values; only an overall MSE below 5% is stated. Since the differences in Eg between samples are likely small, the claimed monotonic increase and 'excellent agreement' with theory cannot be assessed without the fitted TL parameters, their confidence limits, and preferably the raw psi/delta spectra for all samples. The authors should report the fitted parameter set and uncertainties for each film.","section":"Sec. III.C, Fig. 6 and Table S6"},{"comment":"The theoretical band-gap values underlying Fig. 4(b) are not tabulated or stated in the text, so a reader cannot reproduce the comparison or evaluate the claimed saturation quantitatively. Please provide the numerical HSE06 and PBE optical gaps, the Moss-Burstein shift ΔE, and the definition of the 'optical gap' used for the doped metallic systems.","section":"Sec. III.B and Fig. 4(b)"}],"minor_comments":[{"comment":"The word 'chemical' appears twice in the abstract ('structural, chemical, electronic, chemical, and optical properties'); one instance should be removed.","section":"Abstract and Sec. I"},{"comment":"The exponent in Eq. (9) is rendered as '$'5', which is unreadable; the text later says n=1/2, but the equation should be typeset correctly.","section":"Sec. III.C, Eq. (9)"},{"comment":"The sentence 'the band gap variation is not linear but follows Vegards law in a non-linear manner' is internally confusing, since Vegard's law is normally a linear interpolation; the authors likely mean a modified or nonlinear Vegard-type relation.","section":"Sec. III.C, text near Fig. 7"},{"comment":"Several supplementary figure labels are inconsistent with their captions (e.g., Figure S4 is labeled as DOS of Ta in the list but as a band structure in the caption, and Figure S5 is labeled as band structure but shows DOS). Please renumber and cross-check all SI figure references.","section":"Supplementary figures"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for a materials-science journal and the DFT part is competently executed. My main concern is that the experimental band-gap extraction is model-dependent to the point of circularity for the doped metallic films, and the headline saturation claim is not supported by the concentration range actually measured. These are fixable with a more careful analysis and more conservative claims, so I do not recommend rejection, but the revision needs to be substantive."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core result here is probably right, and it is worth a serious referee. What is new is the combination: band-unfolded HSE06/PBE calculations plus new spectroscopic ellipsometry on epitaxial films, used together to argue that Ta doping widens the STO optical gap while preserving its indirect character. That directly addresses a real discrepancy in the earlier literature, where Hou et al. reported a direct gap and Chen et al. an indirect one. The band-unfolding work is careful, the HSE06 gap for pure STO lands close to experiment, and the experimental epsilon1 peak shift toward higher energy with doping is a model-light piece of evidence that the trend is real. I give the authors credit for also flagging that the Tauc-Lorentz model alone is insufficient once free carriers are present, and for adding a Drude term.\n\nThe soft spots are real but not fatal. The stress-test concern about the Tauc n=1/2 extrapolation is legitimate: applying Tauc's formula to an absorption coefficient reconstructed from a TL+Drude fit is not an independent measurement of the band gap. For degenerate n-type STO with the Fermi level inside the conduction band, the absorption edge is a Pauli-blocked Moss-Burstein transition contaminated by free-carrier absorption, so the Tauc exponent and the resulting Eg values are model-dependent. Since the TL model already contains an Eg parameter, the subsequent Tauc plot is partly circular. The authors do not provide error bars on the extracted Eg, and the raw psi/delta spectra are not shipped, so the experimental half of the claim rests on that model. That said, the peak shift in epsilon1 and the DFT trends are consistent, so the qualitative conclusion is not a fitting artifact. What is weaker is the saturation claim: it is inferred from a single theoretical point at 12.5% and no experimental point above 5%. The text says the gap saturates at ~10%, but the evidence for a plateau is thin, and the practical TCO window should be stated more cautiously.\n\nThis paper is for people working on oxide TCOs and doped-STO optoelectronics. It deserves peer review; the reviewer should ask for a more rigorous treatment of the experimental gap extraction, ideally a direct fit with a justified model for degenerate carriers and uncertainty estimates, and a softened saturation claim. I would not cite the saturation number without checking it, but the main trend and the band-unfolding result are worth citing.","headline":"A worthwhile combined DFT+ellipsometry study that shows Ta-doped STO widens its optical gap and stays indirect; the saturation claim at ~10% is not as solid as the paper suggests.","tokens_in":17099,"tokens_out":1531,"would_cite":true,"duration_ms":17908,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["78.20.Ci","71.15.Mb","78.20.Bh"],"model":"deepseek-v4-flash","headline":"Ta doping widens SrTiO3's optical band gap through the Moss-Burstein effect, and the widening saturates near 10% doping while the band structure remains indirect.","keywords":["SrTiO3","transparent conducting oxide","Ta doping","Moss-Burstein effect","band gap renormalization","band unfolding","spectroscopic ellipsometry","density functional theory"],"falsifier":"Measure the absorption edge of the same Ta-doped films by an independent method, such as direct transmission and reflection spectroscopy, and compare the edge position with Hall-measured carrier density. If the edge does not shift with carrier concentration, or if the Tauc-Lorentz-plus-Drude fit disagrees with the raw psi and delta spectra near the edge, the Moss-Burstein interpretation is not established.","tokens_in":16006,"feed_emoji":"🧪","tokens_out":7094,"duration_ms":63217,"temperature":0.7,"pith_summary":"Ta-doped strontium titanate (SrTiO3) is a candidate indium-free transparent conductor, but how doping changes its optical band gap has been unsettled. Combining density functional theory with spectroscopic ellipsometry on epitaxial films, this paper argues that substituting Ta for Ti widens the optical gap through the Moss-Burstein effect while the band structure stays indirect up to at least 12.5% doping. The widening is not open-ended: band-gap renormalization and free-carrier effects counteract the shift, so the gap saturates near 10% doping and further Ta concentration barely changes visible transparency. If correct, this identifies an optimal doping window and explains why earlier supercell calculations disagreed on the nature of the gap.","feed_headline":"Ta doping widens SrTiO3's optical band gap until about 10% doping","feed_subtitle":"Electrons filling the conduction band push up the absorption edge, but renormalization caps the gain near 10% Ta.","key_machinery":"The argument rests on two mechanisms and one methodological tool. The Moss-Burstein shift raises the absorption edge because added electrons fill the Ti-3d conduction-band bottom, so optical transitions need extra energy. Against that, band-gap renormalization from electron-electron and electron-impurity interactions narrows the gap, producing the saturation. The methodological tool is band unfolding, which projects a doped supercell's band structure back onto the primitive Brillouin zone to recover the true dispersion; it is what lets the paper identify the gap as indirect at every concentration studied. On the experimental side, a Tauc-Lorentz oscillator plus a Drude term is used to fit ellipsometric spectra and extract optical band gaps.","core_discovery":"The paper's central claim is that electron doping by Ta raises the optical band gap of SrTiO3 by filling the lower conduction band (the Moss-Burstein shift), but the increase saturates near 10% Ta because band-gap renormalization, electron-electron and electron-impurity interactions, and free-carrier absorption pull the edge back. The authors support this with hybrid-functional band structures, Bader-charge analysis showing partial reduction of Ti and Ta and rising carrier density, band-unfolded dispersions indicating the gap remains indirect at 1.6%, 3.7%, and 12.5% doping, and ellipsometric Tauc-Lorentz-plus-Drude fits on 0-5% epitaxial films. Theory and experiment agree on a nonlinear gap-versus-concentration curve that flattens at high doping. The paper also claims that the indirect-to-direct transition reported in some supercell calculations is an artifact of band folding, and that the true band structure matches angle-resolved photoemission observations on doped SrTiO3.","pith_inferences":["A direct test would grow a series of Ta-doped films and measure the absorption edge and Hall carrier density on the same samples; if the Moss-Burstein mechanism dominates, the edge shift should track the Fermi level rising into the conduction band.","At high doping the Tauc extrapolation with exponent n=1/2 may be questionable because the material is degenerately doped; a complementary extraction using a fixed absorption coefficient or a direct-transition model would clarify whether the reported saturation is physical or a fitting artifact.","The band-unfolding result suggests that other n-type dopants such as Nb or La in SrTiO3 may also preserve the indirect gap, which would unify angle-resolved photoemission observations with supercell calculations.","If transparency is the goal, the paper implies an optimal working window around 5-10% Ta; co-doping strategies should be evaluated against this baseline."],"forward_implications":["If the gap saturates near 10% Ta, then the best transparent-conductor performance for single-site Ta doping should be found at moderate doping, not at the highest achievable concentration.","Because the band structure remains indirect with near-CBM effective mass unchanged, carrier mobility should stay close to the pristine value at low-to-moderate doping.","Supercell band structures of doped SrTiO3 should be unfolded before assigning direct or indirect character; otherwise the nature of the gap can be misread.","Beyond the saturation point, further band-gap engineering requires another route, such as co-doping or a structural phase change, rather than simply adding more Ta.","The same Tauc-Lorentz-plus-Drude fitting strategy can be applied to other degenerately doped perovskite transparent conductors."],"supporting_citations":[{"why":"Provides the earlier hybrid-DFT supercell result claiming Ta does not modify the gap and gives a direct transition; the present paper's unfolding and density-of-states analysis directly address it.","marker":"[21]"},{"why":"Earlier supercell DFT finding of an indirect gap in electron-doped SrTiO3; one side of the direct/indirect discrepancy this paper resolves.","marker":"[28]"},{"why":"Angle-resolved photoemission study of Nb-doped SrTiO3 used as the experimental reference for the true band dispersion.","marker":"[29]"},{"why":"Supplies the effective-band-structure formalism that recovers the primitive-cell dispersion from supercell calculations.","marker":"[33]"},{"why":"The band-unfolding method for retaining an effective primitive-cell band structure, applied here to the Ta-doped supercells.","marker":"[46]"},{"why":"Extends the unfolding formalism to spectral weights; used to judge which bands are spurious supercell artifacts.","marker":"[47]"},{"why":"Documents electrostatically driven band-gap renormalization in a degenerate perovskite transparent conducting oxide, the counteracting effect invoked for the saturation.","marker":"[36]"},{"why":"Moss's original proposal that band filling widens the optical gap in degenerately doped semiconductors.","marker":"[39]"},{"why":"Burstein's formulation of the same shift in InSb, the mechanism the paper invokes for the gap increase.","marker":"[40]"},{"why":"Demonstrates the Tauc-Lorentz plus Drude oscillator fitting approach for doped oxide thin films, the model used here for ellipsometric data.","marker":"[70]"}],"fun_headline_variants":["Ta doping widens SrTiO3 gap, saturates near 10%","SrTiO3 band gap widens with Ta, then plateaus","Ta doping lifts SrTiO3 gap until 10% doping","Electron doping boosts SrTiO3 gap, caps at 10%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a Tauc plot with exponent n=1/2 applied to the ellipsometrically extracted absorption coefficient correctly gives the optical band gap even when the Fermi level lies inside the conduction band; if that semiconductor-style extrapolation is invalid for degenerately doped metallic SrTiO3, the reported experimental gap values and their saturation trend could be artifacts of the fitting model.","fun_headline_variants_meta":{"raw":{"variants":["Ta doping widens SrTiO3 gap, saturates near 10%","SrTiO3 band gap widens with Ta, then plateaus","Ta doping lifts SrTiO3 gap until 10% doping","Electron doping boosts SrTiO3 gap, caps at 10%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000432,"raw_usage":{"total_tokens":2225,"prompt_tokens":989,"completion_tokens":1236,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":605,"completion_tokens_details":{"reasoning_tokens":1158}},"tokens_in":605,"tokens_out":1236,"duration_ms":8828,"temperature":1.0,"reasoning_tokens":1158,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:35:37.291037+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the absorption edge of the same Ta-doped films by an independent method, such as direct transmission and reflection spectroscopy, and compare the edge position with Hall-measured carrier density. If the edge does not shift with carrier concentration, or if the Tauc-Lorentz-plus-Drude fit disagrees with the raw psi and delta spectra near the edge, the Moss-Burstein interpretation is not established.","supporting_citations":[{"cited_title":"Hybrid functionals based on a screened Coulomb potential","cited_arxiv_id":null,"evidence_quote":"Earlier supercell DFT finding of an indirect gap in electron-doped SrTiO3; one side of the direct/indirect discrepancy this paper resolves."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The band-unfolding method for retaining an effective primitive-cell band structure, applied here to the Ta-doped supercells."}],"review_version":1}