{"id":"fe7506d1-b256-43d1-86c7-b7e58762ecb1","arxiv_id":"1908.07379","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Growing thin SrTiO3 films on Nb-doped SrTiO3 substrates tunes the surface 2DEG band filling, and for 0.5% Nb doping the Fermi level sits in the Zeeman gap, leaving a single spin-polarized Fermi surface.","lead":"A team used thin-film growth on doped strontium titanate to control the electron gas at the surface, pushing its Fermi level into a spin gap so that only a single spin-polarized Fermi surface remains. The result offers a simple, stable knob for tuning a 2D electron gas and, combined with strontium titanate's superconductivity, a possible platform for Majorana physics.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'EF inside the Zeeman gap' claim presumes an unmeasured spin gap in PLD films; the single-FS data could be masking an inner branch, and the paper itself calls for spin-splitting verification.","rationale":"I identified the same load-bearing concern as the Pith reader: the Zeeman gap is not measured in the PLD films, and its assumed size is essential to convert the observed single band and 80 meV band bottom into the statement that EF sits inside the gap. The paper's own SARPES data on one 10 u.c. film provide direct evidence of a single helical spin-polarized Fermi surface, which is a strong experimental result, but the energy and momentum resolution of the COPHEE endstation (70 meV, ~1.5°) are insufficient to exclude a weak inner branch with a small Fermi wavevector, especially given the intense central incoherent weight. The high-resolution spin-integrated data also show no obvious second band, but the 2D curvature analysis and the strong background leave room for a low-intensity inner band. The paper explicitly acknowledges the need for spin-splitting verification, which supports the conditionality of the claim. No other concern is more load-bearing: the doping-control claim is supported by two substrate concentrations but is a secondary conclusion, and the confinement inference is less central to the Majorana platform claim. Thus the reader's conditional verdict is appropriate, and no verdict change is needed. The proposed concrete test—high-resolution SARPES at the zone centre on representative films—would settle whether the gap is actually large enough for EF to lie inside it.","tokens_in":18023,"tokens_out":12643,"duration_ms":134030,"concrete_test":"Perform high-resolution spin-resolved ARPES on a 10 u.c. STO/0.5 wt% Nb:STO film at the SBZ centre (k=0) with energy resolution better than 10 meV and momentum resolution better than 0.02 Å^-1, acquiring spin-resolved EDCs and MDCs near EF. If a second spin-split dxy branch is resolved below EF, or an inner Fermi surface with kF > 0 is observed, the claim that EF is inside the Zeeman gap is falsified; if only one band is observed down to the band bottom and the spin gap is measured to be larger than the 80 meV band filling, the claim is confirmed. The same measurement should be repeated on 3 and 20 u.c. films to verify the thickness range of the claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that EF lies exactly in the Zeeman gap at the SBZ centre is load-bearing for the Majorana relevance, but it is not directly measured in these PLD-grown films. The observed single Fermi surface is consistent with a gap larger than the 80 meV band bottom, but it is also consistent with a smaller gap if an inner branch is hidden by the 70 meV energy resolution of the COPHEE SARPES, the modest angular resolution, and the strong incoherent central weight below 150 meV. For a rigidly shifted bulk band structure, a single outer FS requires the zone-centre gap Δ to exceed 80 meV; the paper does not provide any measurement of Δ in the films, only the bulk value from ref. [18], whose reproducibility is disputed (ref. [21]). The paper itself states (Sec. II A): 'a change in spin splitting cannot be excluded and requires experimental verification.' If Δ is in fact smaller, the upper spin-split branch would produce an inner Fermi surface, and the observed single band would not demonstrate a Zeeman-gap protected helical FS, undermining the Majorana platform conclusion. This is not an internal inconsistency, but an external parameter that the key interpretation rests on.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports ARPES and spin-resolved ARPES studies of pulsed-laser-deposited SrTiO3 films grown on Nb-doped SrTiO3(001) substrates. The authors find that the surface 2DEG in these films has a much reduced dxy band filling compared with STO single crystals: the dxy band bottom lies at about 80 meV for films on 0.5 wt% Nb-doped substrates, and the heavy dxz/dyz bands are absent, resulting in a purely two-dimensional state. For these films the Fermi surface consists of a single band with a helical spin texture, which the authors interpret as the Fermi level lying inside the Zeeman gap at the surface Brillouin-zone centre. For films on 0.05 wt% Nb-doped substrates the band filling is larger (about 170 meV), and the authors report evidence for a second populated band. XPS data correlate the band filling with the amount of surface SrOx and suggest that the filling changes are related to the surface termination and possibly the dielectric response. The authors conclude that film growth on Nb-doped substrates provides a stable way to control the 2DEG filling and that the 0.5% films host a single spin-polarised Fermi surface relevant for Majorana physics.","tokens_in":18245,"tokens_out":8086,"duration_ms":85592,"significance":"The experimental data are of high quality and the paper introduces a promising growth-based route to tune the filling of the STO surface 2DEG. The thickness-independence study over 3–20 unit cells, the direct spin-resolved observation of a helical texture in the films, and the XPS correlation with surface SrOx are valuable additions to the field. The paper is also honest about its limitations, explicitly flagging the need for spin-splitting verification. However, the central claim that the Fermi level lies inside the Zeeman gap, and hence that the 0.5% films host a single helical Fermi surface relevant for Majorana physics, depends on an unmeasured parameter: the size (and even existence) of the Zeeman gap in these PLD-grown films is not directly established by the present data. This makes the headline result conditional on an external measurement whose reproducibility is debated.","major_comments":[{"comment":"The central conclusion that the Fermi level lies inside the Zeeman gap for the 0.5 wt% Nb-substrate films is not directly measured. The spin gap Δ is taken from Ref. [18] on STO single crystals and assumed to be unchanged in the PLD-grown films; the paper itself states in Section II.A that \"a change in spin splitting cannot be excluded and requires experimental verification.\" Because an inner spin-split branch would have a small Fermi wave vector (roughly 0.05 Å⁻¹ for Δ just below 80 meV) and could be masked by the 70 meV energy resolution of the COPHEE SARPES and the strong incoherent spectral weight below 150 meV, the observation of a single band in the spin-resolved MDC does not by itself place the Fermi level inside the gap. This assumption is load-bearing for the Majorana-platform claim in Section III and for the title/abstract statement of a single spin-polarised Fermi surface.","section":"Section II.A, Fig. 4"},{"comment":"The evidence that the 0.05 wt% Nb-substrate film hosts two occupied bands relies on a four-peak fit to low-resolution COPHEE data and on unresolved or broad spectral weight in the high-resolution ARPES data; the authors acknowledge in Section II.B that this \"requires further investigation.\" Since this film is one of only two substrate dopings used to demonstrate substrate-controlled band filling, the abstract's claim that band filling \"can be controlled\" by the substrate doping is stronger than the present evidence. A more systematic doping series or a clearly resolved second band in the 0.05% film would be needed to fully support the claim of controlled, monotonic tuning of the 2DEG filling.","section":"Section II.B and Appendix Sec. 5"}],"minor_comments":[{"comment":"The caption describes \"The 2DEG on 10 u.c. film grown on highly-doped substrate,\" but panels (a-c) are for 3, 5, and 20 u.c. films; this appears to be a typo and should be corrected.","section":"Fig. 3 caption"},{"comment":"The appendix contains leftover text from a thesis (\"6.2. Ba x Sr1°x TiO3 Thin Films\" and \"6.2.1 Film preparation\") that is not part of the paper and should be removed.","section":"Appendix Sec. 1"},{"comment":"The word \"spectromiscroscopy\" should be \"spectromicroscopy.\"","section":"Sec. II.C"},{"comment":"The sentence \"Along (a) ΓX and (b) ΓX\" should read \"Along (a) ΓX and (b) ΓM.\"","section":"Appendix Sec. 3"},{"comment":"The labels \"b) SrTiO3 thin film\" and \"b) SrTiO3 single crystal\" appear to be interchanged; the schematic in panel (b) shows the film structure and should be labelled accordingly.","section":"Fig. 1"}],"recommendation":"major_revision","confidential_remarks":"The experimental work is solid and the manuscript is generally well written, but the headline claim goes beyond the data: the Zeeman gap is not measured in the films, and the only direct measurement of such a gap in crystals (Ref. [18]) has a disputed reproducibility record (Ref. [21]). The authors' own caveat in Section II.A should be treated as a central limitation, and the manuscript would be substantially strengthened by additional spin-resolved measurements that directly probe the spin splitting in these films, or by a rephrasing that clearly separates the observed facts (reduced filling, single band in the resolved window, helical spin texture) from the inferred placement of the Fermi level inside the gap. The appendix also contains leftover thesis text that indicates the manuscript was not fully cleaned. I recommend major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth your time. The genuinely new thing here is a growth-based tuning knob for the STO surface 2DEG: homoepitaxial PLD films on Nb-doped substrates reproducibly lower the band filling, and on 0.5% Nb-doped material the filling drops to a point where only one dxy band crosses EF, with spin-resolved ARPES showing a single helical Fermi surface. That is the first time anyone has gotten a stable single-spin-channel surface state in this system, and it is a plausible route to a Majorana platform in one material.\n\nWhat is done well: the thickness series (3, 5, 20 u.c.) shows no finite-size shift, so the filling change is not quantum confinement; the absence of the heavy dxz/dyz bands is clean; and the low-doping substrate (0.05%) gives two bands, which is a nice control showing the filling really moves with substrate doping. The XPS correlation with surface SrOx is suggestive rather than proven, but honestly framed. They also engage the conflicting SARPES result from Walker et al. in an appendix, which is the right way to handle a disputed observation.\n\nThe soft spot is the headline conclusion. The claim that EF sits inside the Zeeman gap is not directly measured in these films; it is inherited from the Zeeman gap size seen in single crystals by Santander-Syro et al. The gap could in principle shrink or change in the films, and the paper says exactly that in Sec. II A: 'a change in spin splitting cannot be excluded and requires experimental verification.' Given the 70 meV energy resolution of the spin-resolved data and the strong incoherent weight near the zone centre, a small inner branch could hide. The 0.05% data reduce that worry somewhat, because there the rigid shift is large enough to resolve two bands. But the single-band observation in the 0.5% films is consistent with, not proof of, a Zeeman-gap-protected single Fermi surface.\n\nThe two dopant concentrations are enough to show the effect exists, not to map it. The 2 u.c. confinement bound is an inference from thickness independence, fine but indirect.\n\nNet: this deserves a serious referee. The control experiment is credible, the data quality looks adequate, and the limitation statements are honest. I would engage with it and probably cite it. My recommendation: send it to peer review rather than desk reject; the Zeeman-gap point needs a direct spin-resolved measurement in films before the Majorana claim hardens, but the paper does not overclaim beyond what it shows.","headline":"A credible growth-based route to a single helical Fermi surface on STO, with the Zeeman-gap interpretation still resting on crystal data rather than film data.","tokens_in":18855,"tokens_out":2631,"would_cite":true,"duration_ms":27729,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["73.20.At","71.70.Ej","79.60.-i"],"model":"deepseek-v4-flash","headline":"Thin SrTiO3 films grown on niobium-doped substrates reduce the surface 2D electron gas to a single spin-polarised Fermi surface, a configuration that could host Majorana bound states.","keywords":["strontium titanate","two-dimensional electron gas","spin-polarised Fermi surface","Rashba spin-orbit coupling","Zeeman gap","Majorana bound states","angle-resolved photoemission","pulsed laser deposition"],"falsifier":"A decisive test is to measure the spin-resolved dispersion of the 10 u.c. film across a range of binding energies down to the band bottom. If the spin splitting between the two $d_{xy}$ branches is smaller than 80 meV, or if a second spin-polarised branch is found crossing $E_F$, then the Fermi level is not in the Zeeman gap and the central claim is refuted.","tokens_in":17810,"feed_emoji":"🧲","tokens_out":9959,"duration_ms":89805,"temperature":0.7,"pith_summary":"The paper reports a way to control the band filling of the two-dimensional electron gas (2DEG) that forms at the surface of SrTiO3(001). By growing thin SrTiO3 films on niobium-doped SrTiO3 substrates, the authors lower the bottom of the main $d_{xy}$ band from the usual 230 meV to 80 meV below the Fermi level, so the Fermi level sits inside the spin gap at the surface Brillouin zone centre. Spin-resolved photoemission then shows exactly one helical, spin-polarised band. If correct, this gives a stable single-spiral Fermi surface in a superconducting material, which is the ingredient combination needed for Majorana bound states, and it also implies the 2DEG is confined within the top two unit cells of the film.","feed_headline":"Thin SrTiO3 films yield a single spin-polarised Fermi surface","feed_subtitle":"By growing SrTiO3 on niobium-doped substrates, the Fermi level lands in a spin gap, leaving one helical band.","key_machinery":"The argument rests on two ingredients. The first is a growth-controlled chemical-potential shift: pulsed-laser-deposited homoepitaxial SrTiO3 films on Nb-doped substrates fix the $d_{xy}$ band bottom at 80 meV, far below the 230 meV filling of cleaved or annealed crystals. The second is the Rashba-Zeeman spin gap at the surface Brillouin zone centre, a spin gap that opens at $\\bar{\\Gamma}$ when spin-orbit and magnetic interactions combine, as previously measured for STO crystal surfaces. Placing the Fermi level inside that gap yields exactly one helical Fermi surface; the simultaneous fit of spin-integrated intensity and the three spin-polarisation components of the momentum distribution curve is the tool that establishes this single band. The Nb-doping level of the substrate moves the band bottom and thus selects how many bands cross the Fermi level.","core_discovery":"For pulsed-laser-deposited SrTiO3 films between 3 and 20 unit cells (u.c.) thick on 0.5 wt% Nb-doped SrTiO3(001), the surface 2DEG consists of a single $d_{xy}$-derived parabolic band with a band bottom at 80 meV binding energy and an effective mass of $0.74\\,m_e$; the heavier $d_{xz}$ and $d_{yz}$ bands are absent, and the state is purely two-dimensional in photon-energy scans. Spin-resolved angle-resolved photoemission on a 10 u.c. film resolves one band whose in-plane spin polarisation is perpendicular to the crystal momentum and reverses sign across the zone centre, matching a helical Rashba-like texture. Interpreting this through the Zeeman gap previously reported at the STO surface Brillouin zone centre, the authors conclude that the Fermi level for these films lies inside the gap, leaving a single spin-polarised Fermi surface. On 0.05 wt% Nb-doped substrates the band bottom moves back up to roughly 170 meV and a second inner band appears to be populated, showing that substrate doping tunes the filling. The authors further infer, from the thickness independence and from XPS of Sr 3d and Ti 3p core levels, that the 2DEG is confined to no more than two unit cells beneath the top TiO2 layer and that the amount of surface SrOx, rather than the Ti3+ fraction, tracks the band filling.","pith_inferences":["If the Zeeman gap in the films has the same size as on crystals, there should be an intermediate Nb-doping level at which the Fermi level crosses the gap edge; mapping that crossover would turn the single-vs-double Fermi surface transition into a controlled phase boundary.","Direct SARPES of the full band dispersion, not just the Fermi level, could measure the film's spin splitting directly and test whether the gap is actually larger than the 80 meV band bottom.","The SrOx correlation suggests a testable extension: preparing adjacent regions of the same film with different SrOx coverage (e.g., by localised annealing or patterned growth) should produce regions with different band fillings on one substrate.","If the confinement to two unit cells is right, then the single-spiral regime should be independent of film thickness beyond 3 u.c. but disappear when the surface termination is changed to TiO2-rich; this is checkable with existing PLD and ARPES tools."],"forward_implications":["Stable control of band filling: unlike the universal 230 meV filling seen on cleaved and annealed SrTiO3 crystals, films on Nb-doped substrates put the Fermi level where the experimenter chooses within the spin gap, with no gate voltage or surface contamination required.","A single helical Fermi surface exists for films of 3 to at least 20 u.c. on 0.5 wt% Nb-doped substrates; combined with the superconductivity of STO, this is a single-material platform for Majorana bound states.","The 2DEG does not extend beyond the top two unit cells, so its properties are governed by the surface termination and the film's dielectric response rather than the bulk vacancy concentration.","Changing the substrate Nb doping from 0.5 to 0.05 wt% shifts the band bottom from 80 meV to about 170 meV, showing the filling can be tuned across the spin-gap window.","The correlation between surface SrOx content and band filling suggests surface chemistry is an additional, possibly independent, tuning knob."],"supporting_citations":[{"why":"Provides the helical Rashba spin texture and the Zeeman gap at the zone centre for STO crystal surfaces, the basis for interpreting the single band as a single spin-polarised Fermi surface.","marker":"[18]"},{"why":"Reports the universal ~230 meV band filling of the 2DEG on cleaved STO crystals, the baseline the films are compared against.","marker":"[10]"},{"why":"Supplies the reference 2DEG electronic structure ($d_{xy}$/$d_{xz}$/$d_{yz}$ bands, $m^*=0.65\\,m_e$, inner potential $V_0=14.5$ eV) used to identify the films' purely $d_{xy}$-derived 2D state.","marker":"[12]"},{"why":"Reports MBE-grown STO films on 0.05% Nb-doped substrates with a ~170 meV band bottom, corroborating the substrate-doping dependence of band filling.","marker":"[14]"},{"why":"The conflicting SARPES study that failed to see spin polarisation; the paper argues its photon energy and surface state mask the signal, defending the spin interpretation.","marker":"[21]"},{"why":"Establishes the Majorana-bound-state context: a single helical Fermi surface plus superconductivity is the required platform.","marker":"[22]"},{"why":"Gives the simultaneous fitting routine for total intensity and spin polarisation used to determine the number of bands and their spin texture.","marker":"[36]"},{"why":"Explains the out-of-plane spin polarisation as a photoemission interference effect, allowing the in-plane helical signal to be isolated.","marker":"[35]"}],"fun_headline_variants":["SrTiO3 films show a single spin-polarised Fermi surface","Substrate doping gives SrTiO3 films one helical band","Thin SrTiO3 films: a single spin-polarised 2DEG for Majorana physics","Spin-polarised single band in SrTiO3 films via Nb doping"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper assumes that the magnetic spin gap seen at the surface of SrTiO3 crystals is still present in the thin films and is larger than 80 meV; the authors do not measure this gap in the films and explicitly write that \"a change in spin splitting cannot be excluded and requires experimental verification.\"","fun_headline_variants_meta":{"raw":{"variants":["SrTiO3 films show a single spin-polarised Fermi surface","Substrate doping gives SrTiO3 films one helical band","Thin SrTiO3 films: a single spin-polarised 2DEG for Majorana physics","Spin-polarised single band in SrTiO3 films via Nb doping"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000634,"raw_usage":{"total_tokens":2958,"prompt_tokens":1009,"completion_tokens":1949,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":625,"completion_tokens_details":{"reasoning_tokens":1870}},"tokens_in":625,"tokens_out":1949,"duration_ms":16053,"temperature":1.0,"reasoning_tokens":1870,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:19:16.324844+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is to measure the spin-resolved dispersion of the 10 u.c. film across a range of binding energies down to the band bottom. If the spin splitting between the two $d_{xy}$ branches is smaller than 80 meV, or if a second spin-polarised branch is found crossing $E_F$, then the Fermi level is not in the Zeeman gap and the central claim is refuted.","supporting_citations":[{"cited_title":"Meevasana, P","cited_arxiv_id":null,"evidence_quote":"Provides the helical Rashba spin texture and the Zeeman gap at the zone centre for STO crystal surfaces, the basis for interpreting the single band as a single spin-polarised Fermi surface."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the universal ~230 meV band filling of the 2DEG on cleaved STO crystals, the baseline the films are compared against."},{"cited_title":"Kalisky, J","cited_arxiv_id":null,"evidence_quote":"Supplies the reference 2DEG electronic structure ($d_{xy}$/$d_{xz}$/$d_{yz}$ bands, $m^*=0.65\\,m_e$, inner potential $V_0=14.5$ eV) used to identify the films' purely $d_{xy}$-derived 2D state."},{"cited_title":"Cancellieri, M","cited_arxiv_id":null,"evidence_quote":"Reports MBE-grown STO films on 0.05% Nb-doped substrates with a ~170 meV band bottom, corroborating the substrate-doping dependence of band filling."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The conflicting SARPES study that failed to see spin polarisation; the paper argues its photon energy and surface state mask the signal, defending the spin interpretation."},{"cited_title":"Delugas, V","cited_arxiv_id":null,"evidence_quote":"Establishes the Majorana-bound-state context: a single helical Fermi surface plus superconductivity is the required platform."},{"cited_title":"Zhang, P","cited_arxiv_id":null,"evidence_quote":"Gives the simultaneous fitting routine for total intensity and spin polarisation used to determine the number of bands and their spin texture."},{"cited_title":"Ogawa, K","cited_arxiv_id":null,"evidence_quote":"Explains the out-of-plane spin polarisation as a photoemission interference effect, allowing the in-plane helical signal to be isolated."}],"review_version":1}