{"id":"6e9f852f-2cb9-4fcd-b4f7-0e7e55fb78b1","arxiv_id":"1909.00990","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A proximity Eliashberg calculation reproduces the measured field-induced critical temperature shift in indium thin films, but only after fitting the surface layer thickness to the experimental data.","lead":"This paper computes how an electric field changes the superconducting critical temperature of thin indium films using a proximity Eliashberg model with density functional theory inputs, claiming to explain a 1960 experiment. Generalists may care because it suggests that extremely small charge densities can measurably tune superconductivity, a possibility recently questioned.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'no free parameters' claim is undercut because exact agreement requires ds=0.165 nm, 45% above the derived dTF=0.114 nm, and this sub-monolayer fitted thickness strains the bulk-phonon assumption for α2sF.","rationale":"The qualitative mechanism—extremely small doping changes λ and ωln enough to shift Tc by ~10^-4 K, with proximity dilution setting the final magnitude—is plausible and rests on real DFT inputs. The load-bearing weakness is precisely the surface-layer thickness ds: the derived dTF=0.114 nm is not what reproduces the experiment; the fitted 0.165 nm is. Because the fit is made to the target datum, the quantitative agreement cannot independently validate the model. The additional sub-monolayer concern reinforces this: a 0.165 nm slab is thinner than one In lattice constant, so the use of bulk α2F in that layer is itself an unverified assumption. None of this makes the framework internally inconsistent or the qualitative effect impossible; it means the central 'no free parameters' claim is not established. The reader's conditional verdict remains appropriate: a re-run at dTF or a first-principles extraction of ds would settle whether the parameter-free version actually works.","tokens_in":9169,"tokens_out":6957,"duration_ms":74503,"concrete_test":"Recompute the proximity-Eliashberg ΔTc for d=7 nm and x=3e-5 with ds=dTF=0.114 nm using the same code, and quote the resulting shift against the Glover-Sherrill value; if it deviates by more than the experimental uncertainty, the 'no free parameters' version of the model is falsified. To test the fitted value independently, repeat the DFT step in the field-effect slab geometry of Refs. [40,41], extract the induced-charge profile n(z), and derive ds and α2sF from that profile; if the derived ds is near 0.165 nm and the surface phonons are bulk-like, the fit is justified; if it is near 0.114 nm or the phonons shift, the quantitative agreement is an artifact of the free parameter.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4 computes dTF=0.114 nm, then states that 'in order to exactly reproduce the superconductive critical temperature shift measured in Ref. [27], ds=0.165 nm.' The parameter-free claim in the abstract therefore fails as written: the one number that controls the proximity-effect magnitude is adjusted upward by 45% to hit the target datum. For x=3e-5, the homogeneous calculation gives Tc,s=3.39683 K (Table 1), only 3.2 mK below Tc,b=3.4 K; the bilayer shift is produced by diluting this small difference through the surface/bulk thickness ratio, so ds sets the entire quantitative scale. Fig. 5 then uses ds=0.165 nm (not dTF=0.114 nm) for all film thicknesses, while the conclusion claims ds=dTF; this is circular. In addition, ds=0.165 nm is smaller than the In lattice constant a=3.25 Å, so the surface layer is sub-monolayer; computing α2sF from uniformly doped bulk In assumes bulk phonons exist in a region where they are unlikely to, and the monolayer caveat in Sec. 4 ([51]) is not applied here. The paper demonstrates that a one-parameter proximity model can fit one measured shift, but it does not demonstrate the advertised parameter-free reproduction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a theoretical explanation of the electric-field-induced superconducting critical-temperature shifts measured by Glover and Sherrill in 1960 for indium thin films. The authors model the field-perturbed surface layer and the unperturbed bulk as a superconductor/normal-metal junction described by proximity-effect Eliashberg equations. Inputs such as electron-phonon spectral functions, densities of states, and Fermi-energy shifts are computed with DFT for bulk and uniformly doped bulk indium. The paper's central claim is that in the weak-field limit the surface-layer thickness equals the Thomas-Fermi screening length, leaving no free parameters, and that the model then reproduces the measured Tc shift for electron doping x=3e-5.","tokens_in":9433,"tokens_out":7777,"duration_ms":82437,"significance":"If the central claim were established, the paper would resolve a long-standing puzzle by connecting ab initio DFT inputs with a proximity-Eliashberg mechanism, and it would provide a falsifiable prediction of the film-thickness dependence of the Tc shift. The computational pipeline is a strength: the DFT-derived spectral function is benchmarked against tunneling data, and the same framework has previously been applied to Pb and MgB2. The paper also gives explicit credit to the limitations of the bulk-phonon assumption for monolayer films. However, the advertised parameter-free reproduction of the data is not supported by the manuscript as written, because the surface-layer thickness is adjusted to the target experimental shift and at least one other input is tuned or borrowed; the thickness-dependence prediction is therefore conditional on a fitted parameter.","major_comments":[{"comment":"The 'no free parameters' claim is contradicted by the fitting of ds. The paper computes dTF=0.114 nm and then states that 'in order to exactly reproduce the superconductive critical temperature shift measured in Ref. [27], ds = 0.165 nm', a 45% adjustment of the parameter that controls the overall magnitude of the proximity effect. At x=3e-5 the homogeneous Tc,s is 3.39683 K (Table 1), only 3.2 mK below Tc,b, so the bilayer shift is essentially this small difference multiplied by a surface/bulk dilution factor controlled by ds. The manuscript does not report the predicted shift at ds=dTF, so the reader cannot judge whether the disagreement is within experimental uncertainty. The abstract's and conclusion's parameter-free claim therefore fails as written; at best the paper demonstrates a one-parameter fit.","section":"Section 4, Fig. 4, Abstract, Conclusions"},{"comment":"The bulk Coulomb pseudopotential mu*_b is tuned to reproduce the experimental Tc,b=3.4 K ('forcing Tc,b to its experimental value', Section 4), which is an additional fitted parameter beyond ds. In addition, the junction area A=10^-6 m^2 is taken from Ref. [29] rather than from the Glover-Sherrill experiment; since Gamma_s and Gamma_b in Eq. (7) are proportional to A, the predicted Delta-Tc depends on this choice unless an explicit cancellation is demonstrated. The statement in Sec. 2 that d and A are experimental data is not accurate for A in this context.","section":"Section 4, Eq. (7), Sec. 2"},{"comment":"The surface layer is treated as uniformly doped bulk indium, but the fitted ds=0.165 nm is smaller than the indium lattice constant (a=3.25 Å), and even dTF=0.114 nm is sub-unit-cell. Using the bulk alpha^2F for such a sub-monolayer region assumes bulk phonons in a region where they are unlikely to exist. The paper itself notes that monolayer films would require a different spectral function (Sec. 4, Ref. [51]), yet applies the bulk alpha^2F to ds=0.165 nm in all of Fig. 5. This approximation is load-bearing because the sign and magnitude of the homogeneous Tc change are controlled by a delicate competition between lambda and omega_ln (Table 1).","section":"Section 4, Fig. 5, monolayer caveat"},{"comment":"Fig. 5 uses ds=0.165 nm for all film thicknesses, while the conclusion states that in the weak-field limit ds equals the Thomas-Fermi screening length dTF and thus no free parameters remain. These two statements are mutually inconsistent. The thickness-dependence prediction of Fig. 5 is therefore not a parameter-free prediction; it is conditioned on the fitted ds. The authors should either renounce the parameter-free claim or provide an independent determination of ds and demonstrate that the experimental shift is reproduced without adjustment.","section":"Fig. 5 and Conclusions"}],"minor_comments":[{"comment":"There are several typos: 'shoed' should be 'showed', 'Paneal' should be 'Panel', 'Ernxerhof' should be 'Ernzerhof', and 'emplyed' should be 'employed'.","section":"Introduction, Fig. 3, Sec. 3, Conclusions"},{"comment":"Units are missing or inconsistent in a few places: 'aexp = 3.25[46]' and 'ath = 3.30 A' should include angstroms, and the text alternates between 'unitary cell' and 'unit cell' for the same quantity.","section":"Sec. 3, Sec. 4"},{"comment":"The caption states 'The red line is the experimental data. [27]', but the original experiment reported five films with thicknesses in the 60–120 Å range; the figure should identify which film or film thickness the comparison refers to, and ideally show the experimental uncertainty.","section":"Fig. 4 caption"},{"comment":"The phrase 'total density of space per spin at the Fermi energy' should be 'density of states per spin at the Fermi energy'.","section":"Eq. (11)"}],"recommendation":"major_revision","confidential_remarks":"The modeling framework is plausible and within the journal's scope, and the DFT pipeline is a genuine strength. The main problem is that the abstract and conclusion overclaim parameter-free status: ds is fitted to the target datum, mu*_b is tuned to Tc,b, and A is borrowed from another context. I recommend major revision rather than rejection because the qualitative mechanism is sound and the thickness-dependence prediction could become a valuable falsifiable statement if the claims are reframed or if an independent determination of ds is supplied."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know about arXiv:1909.00990 is that it is a real attempt to explain the 1960 Glover–Sherrill data on field-induced Tc shifts in indium, and it brings genuinely new material-specific input: doped electron-phonon spectral functions, λ, ω_ln, density of states, Fermi-energy shifts, and a proximity Eliashberg calculation for a surface/bulk bilayer. That is the first time this model, previously applied to Pb and MgB2, has been tested on a third material, and the qualitative takeaway—that a doping of 3×10^-5 electrons/atom can measurably lower Tc—is plausible and worth taking seriously.\n\nWhat the paper does well is the ab initio machinery and the honesty in the middle sections: the authors show the computed spectra, tabulate the input parameters, and tell you exactly where they adjusted things. The citations to their own prior work are legitimate, since those papers introduced the method.\n\nThe soft spot is the headline claim. The abstract says that in the weak-field limit the Thomas-Fermi approximation leaves no free parameters. In Section 4, however, the value that reproduces the measured shift is ds = 0.165 nm, while the calculated d_TF is 0.114 nm. The text openly states that 0.165 nm is chosen \"in order to exactly reproduce\" the experimental shift. So the central quantitative result is a fit, not a prediction. The conclusion then claims ds = d_TF, which is not consistent with the calculation. This matters because the bilayer shift is obtained by diluting the small homogeneous Tc difference through the surface/bulk thickness ratio; ds sets the entire scale.\n\nA second issue: 0.165 nm is below the indium lattice constant, so the \"surface layer\" is sub-monolayer. Using a bulk-like α2F for a region thinner than a unit cell is a stretch. The authors acknowledge the monolayer caveat in Section 4, but do not apply it to their own ds. Also, mu*_b is tuned to reproduce Tc,b = 3.4 K, which is standard practice but is another adjusted parameter, and the junction area A is taken from prior work. These are minor compared to the ds problem.\n\nOverall, the paper demonstrates that a one-parameter proximity model can fit a single measured shift. It does not demonstrate the advertised parameter-free reproduction. The qualitative framework and the new ab initio data are still useful, but the quantitative validation is overclaimed.\n\nThis is a paper for specialists in field-effect superconductivity, especially anyone working on the Glover–Sherrill puzzle or the recent debate about whether tiny induced charge densities matter. It should go to peer review rather than desk reject, but it needs major revision: either derive ds honestly with an uncertainty, or present the d_TF = 0.114 nm result as the prediction and propose a thickness-dependence experiment that could test it. As written, the central claim would not survive refereeing.","headline":"New ab initio results for indium, but the 'no free parameters' claim fails: the surface-layer thickness is tuned to hit the target shift, so the paper is a conditional validation rather than a parameter-free explanation.","tokens_in":769,"tokens_out":1079,"would_cite":false,"duration_ms":33684,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["74.45.+c","74.62.-c","74.20.Fg"],"model":"deepseek-v4-flash","headline":"A nanometre-thin charged surface layer is enough to explain the measured electric-field shifts in superconducting indium films.","keywords":["electric field effect","superconductivity","indium thin films","proximity effect","Eliashberg theory","Thomas-Fermi screening","critical temperature shift","ab initio calculation"],"falsifier":"Measure the field-induced $T_c$ shift on indium films of several well-characterized thicknesses while independently extracting the surface charge profile; the model fails if a single $d_s$ cannot fit all thicknesses, if the shift vanishes when the induced charge is below a few times $10^{-5}$ electrons per atom, or if hole doping does not produce the opposite shift.","tokens_in":8943,"feed_emoji":"⚡","tokens_out":9031,"duration_ms":85018,"temperature":0.7,"pith_summary":"The paper aims to explain a decades-old puzzle: why a static electric field shifts the superconducting transition temperature of indium films by about $10^{-4}$ K. It argues that the field dopes a thin surface layer, changing its local critical temperature, and that proximity coupling between that layer and the untouched bulk controls the film's overall $T_c$. All microscopic inputs, including electron-phonon spectral functions, Fermi-energy shifts, and densities of states, are computed from first principles, leaving only the surface-layer thickness as an adjustable parameter. In the weak-field limit the Thomas-Fermi screening length fixes that thickness, and with it the calculation reproduces the experimental sign, magnitude, and thickness dependence of the shifts, even for induced charge densities of about $3\\times10^{-5}$ electrons per atom that are usually dismissed as negligible.","feed_headline":"Small surface charges explain indium film Tc shifts","feed_subtitle":"Proximity-Eliashberg theory with a Thomas-Fermi surface layer reproduces the 1960 field-effect data with one adjustable thickness.","key_machinery":"The working engine is a set of four coupled imaginary-axis Eliashberg equations for two subsystems: a field-perturbed surface layer of thickness $d_s$ with its own lower critical temperature $T_{c,s}$, and an unperturbed bulk of thickness $d-d_s$ with bulk $T_{c,b}$. A proximity parameter couples the two, so the surface gap and the bulk gap renormalize each other, and the equations are solved together for the effective $T_c$ of the film. The surface layer is treated as uniformly doped bulk indium, with its doping level set by the induced sheet charge divided by $d_s$, and in the weak-field limit $d_s$ is identified with the Thomas-Fermi screening length, about $0.114$ nm. This machinery converts a small change in surface carrier density into a measurable shift of the film's critical temperature, and it lets the paper compute the shift from first principles rather than fit it.","core_discovery":"The paper's central claim is that a static electric field does not dope the entire indium film uniformly; it changes the carrier density only in a surface layer roughly a tenth of a nanometre thick, and the coupling between this layer and the untouched bulk produces the observed shift of the superconducting critical temperature. Solving the one-band s-wave Eliashberg equations with proximity effect, with all microscopic inputs for doped bulk indium, gives a decrease of $T_c$ with electron accumulation that matches the old measurements in sign and size. The induced charge involved is tiny, about $3\\times10^{-5}$ electrons per atom, and the paper's point is that such small doping is still enough to matter because the proximity effect concentrates its influence. The predicted shift also shrinks as the total film thickness grows, matching the thickness trend in the data. The one fitted parameter, the surface-layer thickness, comes out at $0.165$ nm, close to the Thomas-Fermi screening length of $0.114$ nm, which the paper reads as validating the parameter-free weak-field limit.","pith_inferences":["I would infer a sharp test the paper does not perform: positive charging (hole doping) should shift $T_c$ in the opposite direction from electron doping if the mechanism is purely electrostatic, and measuring that sign reversal would separate this model from heating or strain artifacts.","Because the exact fit requires $d_s=0.165$ nm rather than the derived $0.114$ nm, an independent measurement of the near-surface charge profile would convert the calculation from one-parameter to zero-parameter; in the meantime, $0.165$ nm is best read as an effective thickness that absorbs phonon and interface effects.","The model keeps bulk phonons in the thin surface layer; a fully self-consistent treatment with surface-modified phonons might absorb the remaining difference between $d_s$ and $d_{TF}$, which is a natural theoretical extension."],"forward_implications":["If the model is correct, electric-field control of metallic superconductivity is a genuine surface-doping effect: induced charge densities around $3\\times10^{-5}$ electrons per atom measurably lower $T_c$ in thin films.","Thinner films should show larger field-induced shifts; the calculation gives a clear hierarchy from $d=3$ nm to $d=7$ nm at fixed doping, so thickness can be used to amplify or suppress the effect.","In the weak-field limit the model has no free parameters once the Thomas-Fermi length is used, meaning the same framework can predict field-effect $T_c$ shifts for other elemental superconductors from first-principles inputs alone.","The successful reproduction of the six-decade-old indium data supplies a concrete case where small charge doping, previously dismissed as negligible, quantitatively controls superconductivity, reinforcing the physical picture behind recent field-effect experiments on metallic superconductors."],"supporting_citations":[{"why":"supplies the 1960 measurements of field-induced $T_c$ shifts in indium films that the model must reproduce.","marker":"[27]"},{"why":"provides the proximity-effect Eliashberg treatment of a field-doped surface layer, previously used for lead.","marker":"[28]"},{"why":"introduces the Thomas-Fermi layer application and the zero-free-parameter weak-field limit used in this paper.","marker":"[29]"},{"why":"supports the ideal-interface assumption through experimental doping-dependent $T_c$ data on niobium nitride.","marker":"[3]"},{"why":"supplies the one-band s-wave Eliashberg equations and the bulk-$T_c$ calibration fixing the Coulomb pseudopotential.","marker":"[31]"},{"why":"gives the ab initio electron-phonon spectral function for bulk indium used as an input and comparison.","marker":"[44]"},{"why":"provides the Thomas-Fermi screening length estimate that fixes the surface-layer thickness in the weak-field limit.","marker":"[33]"}],"fun_headline_variants":["Surface charges, not bulk doping, shift indium Tc","Nanoscale surface layer explains indium Tc shifts","Tiny induced charges move indium film Tc","Proximity effect turns small charge into Tc shift"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the electric field acts only by converting the surface of the film into a uniformly doped slab whose thickness is a free parameter and whose electron-phonon spectrum is the same as bulk indium; the exact match requires $d_s=0.165$ nm rather than the derived Thomas-Fermi value of $0.114$ nm, so this slab picture, not the measured data, carries the weight.","fun_headline_variants_meta":{"raw":{"variants":["Surface charges, not bulk doping, shift indium Tc","Nanoscale surface layer explains indium Tc shifts","Tiny induced charges move indium film Tc","Proximity effect turns small charge into Tc shift"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000573,"raw_usage":{"total_tokens":2692,"prompt_tokens":912,"completion_tokens":1780,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":528,"completion_tokens_details":{"reasoning_tokens":1718}},"tokens_in":528,"tokens_out":1780,"duration_ms":12686,"temperature":1.0,"reasoning_tokens":1718,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:29:28.852919+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the field-induced $T_c$ shift on indium films of several well-characterized thicknesses while independently extracting the surface charge profile; the model fails if a single $d_s$ cannot fit all thicknesses, if the shift vanishes when the induced charge is below a few times $10^{-5}$ electrons per atom, or if hole doping does not produce the opposite shift.","supporting_citations":[{"cited_title":"Ghinovker, V.B","cited_arxiv_id":null,"evidence_quote":"supplies the 1960 measurements of field-induced $T_c$ shifts in indium films that the model must reproduce."},{"cited_title":"Walter, H","cited_arxiv_id":null,"evidence_quote":"provides the proximity-effect Eliashberg treatment of a field-doped surface layer, previously used for lead."},{"cited_title":"De Simoni, F","cited_arxiv_id":null,"evidence_quote":"introduces the Thomas-Fermi layer application and the zero-free-parameter weak-field limit used in this paper."},{"cited_title":"Electronic and vibrational properties for both sub- systems are computed for bulk Indium in its body-centered tetragonal structure, which has one atom per unit cell (see Fig","cited_arxiv_id":null,"evidence_quote":"supports the ideal-interface assumption through experimental doping-dependent $T_c$ data on niobium nitride."},{"cited_title":"Paolucci, G","cited_arxiv_id":null,"evidence_quote":"supplies the one-band s-wave Eliashberg equations and the bulk-$T_c$ calibration fixing the Coulomb pseudopotential."},{"cited_title":"Pines and P","cited_arxiv_id":null,"evidence_quote":"gives the ab initio electron-phonon spectral function for bulk indium used as an input and comparison."},{"cited_title":"Ummarino, E","cited_arxiv_id":null,"evidence_quote":"provides the Thomas-Fermi screening length estimate that fixes the surface-layer thickness in the weak-field limit."}],"review_version":1}