{"id":"4b83a2fe-8ff1-43fd-ab08-9711c66c5173","arxiv_id":"2607.10907","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"low","formal_verification":"none","parameter_count":6,"one_line_summary":"LPCVD grows controllable Ge-doped (010) β-Ga₂O₃ films with 62–105 cm²/V·s mobility, a 14 meV shallow donor, and functional Ni Schottky diodes with Ron,sp of 2.49 mΩ·cm².","lead":"Researchers grew Ge-doped β-Ga₂O₃ crystal films by low-pressure CVD with controllable n-type doping, high crystal quality, and working Schottky diodes. The result strengthens a simpler growth route for ultra-wide-bandgap layers used in high-voltage power electronics.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified beyond the reader's already-noted Ge-assignment caveat; the growth-and-diode claim holds on the reported metrology.","rationale":"The paper is a straightforward materials-and-device demonstration. Structural (XRD rocking curves, Raman Ag modes), morphological (SEM/AFM), chemical (XPS stoichiometry), transport (T-dependent Hall + charge-neutrality/Matthiessen fit), and diode (J–V thermionic emission with Wagner correction; C–V ND and φB) data are mutually consistent and improve on the cited prior LPCVD Ge mobility range. The only soft spot that could affect the strongest claim is chemical confirmation of the dopant identity; the reader already identified it correctly. That gap does not invalidate the reported electrical controllability, crystalline quality, or SBD metrics under the growth conditions used. No internal inconsistency, no hidden assumption that breaks the reported numbers, and no need to move the verdict from ACCEPT. A SIMS check would strengthen the Ge-specific language but is not required to accept the growth-and-diode package as written.","tokens_in":17059,"tokens_out":756,"duration_ms":10846,"concrete_test":"Obtain SIMS (or equivalent) Ge and Si depth profiles on the same Hall/SBD pieces used for the 7.4×10¹⁷ cm⁻³ film; if [Ge] is within ~factor of 2 of ND1 and [Si] ≪ ND1, the Ge-assignment concern is settled in the paper's favor; if [Si] dominates or [Ge] is far below Hall n, re-label the donor origin and soften the 'Ge doping' framing.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest assumption (that free electrons and the fitted ED ≈ 14 meV are primarily substitutional Ge, without SIMS or absolute [Ge] correlated to Hall n) is real but not load-bearing against the paper's actual central claim. The claim is that LPCVD with a metallic Ge source produces controllable n-type (010) β-Ga₂O₃ films of high structural/electronic quality, evidenced by multi-technique characterization, improved RT mobility vs prior LPCVD Ge reports, Hall/C–V agreement, and working vertical Ni SBDs. That package is supported by the data as stated: RMS 2.94–3.97 nm, (020) FWHM 97–124 arcsec, phase-pure Raman, near-stoichiometric XPS O/Ga ≈ 1.51, μ_RT = 105–62 cm²/V·s at n = 7.4×10¹⁷–2.57×10¹⁸ cm⁻³, peak μ = 234 cm²/V·s at 116 K, and SBD η = 1.32, φB,IV = 1.02 eV, φB,CV = 1.13 eV, Ron,sp = 2.49 mΩ·cm² with ND,CV ≈ Hall. Donor assignment rests on intentional Ge source + literature shallow-donor precedent + transport fit (Table 1: ND1 = 1.20×10¹⁸ cm⁻³, ED1 = 14 meV; low NA). Residual Si or other impurities cannot be rigorously excluded without chemical profiling, but this is a standard limitation of many doping papers and does not overturn the demonstrated controllability or device-grade quality under the stated growth conditions. High-voltage readiness remains prospective (doping still high for HV drifts; no breakdown data), which the reader already flags.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"This manuscript reports LPCVD homoepitaxy of Ge-doped (010) β-Ga₂O₃ on native substrates, achieving room-temperature carrier concentrations of 7.4×10¹⁷–2.57×10¹⁸ cm⁻³ with Hall mobilities of 105–62 cm²/V·s. Multi-probe structural characterization (SEM/AFM RMS 2.94–3.97 nm, XRD (020) rocking FWHM 97–124 arcsec, full Ag Raman set, XPS O/Ga ≈ 1.51) supports phase-pure, near-stoichiometric films of good crystalline quality. Temperature-dependent Hall data on the lightest-doped film, fitted with a charge-neutrality and multi-scattering transport model (Eqs. 1–2, Table 1), yield a dominant shallow donor (ED1 = 14 meV, ND1 = 1.20×10¹⁸ cm⁻³), low compensation, and a peak mobility of 234 cm²/V·s at 116 K. Vertical Ni/β-Ga₂O₃ Schottky diodes on a ~2.54 µm Ge-doped drift layer show rectifying J–V behavior (η = 1.32, φB,IV = 1.02 eV, Ron,sp = 2.49 mΩ·cm²) and C–V ND = 7.7×10¹⁷ cm⁻³ with φB,CV = 1.13 eV, in agreement with Hall. The central claim is that LPCVD with a metallic Ge source enables controllable n-type doping while preserving device-grade structural and electronic quality.","tokens_in":17530,"tokens_out":1335,"duration_ms":11227,"significance":"If the results hold, the work strengthens LPCVD as a practical route for Ge-doped β-Ga₂O₃ homoepitaxy, with room-temperature mobilities clearly improved relative to earlier LPCVD Ge reports cited in the introduction and with Hall/C–V consistency plus working vertical SBDs that demonstrate electronic quality. The multi-technique structural package and the self-consistent transport fit (Table 1) are concrete strengths. High-voltage readiness remains prospective—doping is still high for thick HV drifts and no breakdown data are shown—but the manuscript positions the platform appropriately for future power-device development rather than claiming a finished HV device.","major_comments":[{"comment":"§III / Table 1 and the abstract claim of “efficient electrical activation of Ge donors” rest on intentional Ge-source growth plus literature precedent for shallow Ge donors, without SIMS or other absolute [Ge] profiles correlated to Hall n. Residual Si or other impurities cannot be rigorously excluded. This does not overturn the demonstrated n-type controllability or diode quality under the stated growth conditions, but a short, explicit caveat (or any available chemical evidence) should be added so the donor assignment is not overstated.","section":null},{"comment":"The abstract and conclusion frame the films as a platform for future high-voltage devices, yet the demonstrated doping (mid-10¹⁷ to low-10¹⁸ cm⁻³) and the absence of reverse-breakdown or high-voltage data leave that claim prospective. A brief, quantitative note on the doping range still needed for HV drift layers (and that lower doping is future work) would keep the device-level claim proportionate to the data.","section":null}],"minor_comments":[{"comment":"Throughout the manuscript and abstract, carrier-concentration exponents appear as corrupted placeholders (e.g., 7.4×10!\" , 2.57×10!#). These must be restored to proper scientific notation before publication.","section":null},{"comment":"Figure 2 caption refers to panel (c) (FWHM vs carrier concentration), but the figure description in the text should be checked for consistent panel labeling and axis units.","section":null},{"comment":"Eq. (5) barrier correction and the Richardson constant A* = 41.04 A cm⁻² K⁻² are standard; a one-line statement of the effective mass used for A* and NC would aid reproducibility.","section":null},{"comment":"Growth details (Ga–Ge spacing 33 cm, Ga–substrate 4.5 cm, 1.5 Torr, 1100 °C) are useful; if growth rate vs Ge source condition is known, a brief note would strengthen the “controllable doping” claim.","section":null},{"comment":"References include several arXiv preprints and in-press items; ensure final citations are updated and that prior LPCVD Ge work (e.g., Ranga et al.) is compared quantitatively on mobility at similar n.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The Ge-assignment caveat is real but standard for many doping papers and is not load-bearing against the growth-and-diode claim as written. I agree with the reader that the package is sound enough for minor revision rather than major revision or reject. Scope fits a materials/device-oriented condensed-matter or applied-physics journal."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This is a clean experimental materials paper. What is new is not Ge doping itself—MBE, MOCVD, and earlier LPCVD already did that—but LPCVD Ge on (010) with room-temperature mobility of 62–105 cm²/V·s (vs the 20–43 they cite for prior LPCVD Ge), a full structural/chemical suite, temperature-dependent transport fitting that pulls a dominant 14 meV donor, and a vertical Ni SBD on that drift layer with Hall and C–V doping in agreement.\n\nThey do the work carefully. Surfaces are smooth (RMS ~3–4 nm) with the usual [001] striations. XRD shows only (020), rocking FWHM 97–124 arcsec that only modestly widens with doping. Raman has the full Ag set; XPS O/Ga ≈ 1.51. Hall n = 7.4×10¹⁷–2.57×10¹⁸ cm⁻³, peak μ = 234 at 116 K. The charge-neutrality + Matthiessen fit (Table 1: ND1 = 1.2×10¹⁸, ED1 = 14 meV, low NA, Nline) is standard parameter extraction, not circular. The diode is honest: Von = 0.74 V, η = 1.32, φB,IV = 1.02 eV / φB,CV = 1.13 eV, Ron,sp = 2.49 mΩ·cm², ND,CV = 7.7×10¹⁷ matching Hall. Citations cover the right prior Ge and LPCVD work.\n\nSoft spots, in proportion: no SIMS or absolute [Ge] correlated to Hall n, so the “Ge donors” assignment rests on intentional Ge source + literature shallow-donor precedent. That is a real caveat, not a load-bearing hole—the central claim is controllable n-type films of good quality under the stated LPCVD conditions, and that package holds. Doping is still mid-10¹⁷ for a “high-voltage platform,” and there is no breakdown data; that framing is prospective, which they mostly acknowledge in the conclusion. Transport parameters are fits; free parameters are the usual ones.\n\nWho it is for: people growing or doping β-Ga₂O₃ by LPCVD/CVD who need a process benchmark and diode-level check. Math and metrology look solid. I would send it to peer review; it is the kind of growth-and-device note that belongs in the literature with the usual requests for chemical profiling and clearer HV caveats.","headline":"Solid LPCVD Ge-doping process paper: better mobility than prior LPCVD Ge, multi-probe film quality, and a working vertical SBD with Hall/C–V agreement; high-voltage framing is still prospective.","tokens_in":18245,"tokens_out":639,"would_cite":true,"duration_ms":6822,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"LPCVD grows Ge-doped (010) β-Ga₂O₃ films with controllable n-type doping and working Schottky diodes.","keywords":["β-Ga2O3","LPCVD","Ge doping","Schottky barrier diode","homoepitaxy","Hall mobility","ultra-wide bandgap"],"falsifier":"Secondary-ion mass spectrometry (or equivalent) depth profiles showing Ge atomic density that does not track the Hall and C–V donor densities across the reported doping series would falsify the claim that Ge is the electrically active dopant.","tokens_in":17941,"feed_emoji":"⚡","tokens_out":1187,"duration_ms":10915,"temperature":0.7,"pith_summary":"This paper shows that low-pressure chemical vapor deposition can put germanium into (010) β-Ga₂O₃ layers grown on native substrates and turn that germanium into free electrons at useful concentrations. Room-temperature carrier densities span 7.4×10¹⁷ to 2.57×10¹⁸ cm⁻³ with mobilities of 105–62 cm²/V·s, while XRD, Raman and XPS confirm the films stay phase-pure, near-stoichiometric and only modestly rougher as doping rises. Temperature-dependent Hall data fit a dominant shallow donor at 14 meV, and vertical Ni Schottky diodes made on a 2.54 µm Ge-doped drift layer show rectifying behavior with turn-on 0.74 V, ideality factor 1.32, barrier height ~1.0–1.1 eV and specific on-resistance 2.49 mΩ·cm². The authors argue that LPCVD therefore supplies both doping control and device-grade material for high-voltage β-Ga₂O₃ power electronics.","feed_headline":"LPCVD Ge-dopes β-Ga₂O₃ with 14 meV donors and working diodes","feed_subtitle":"Controllable n-type films and vertical Ni Schottky rectifiers point to a path for high-voltage power devices","key_machinery":"Self-consistent charge-neutrality plus multi-scattering transport modeling of temperature-dependent Hall data, which isolates a dominant 14 meV shallow donor (assigned to substitutional Ge) and quantifies residual line-defect and acceptor densities.","core_discovery":"LPCVD of Ge-doped (010) β-Ga₂O₃ on native substrates yields controllable n-type doping (7.4×10¹⁷–2.57×10¹⁸ cm⁻³, µ = 105–62 cm²/V·s), a dominant shallow donor of 14 meV, and vertical Ni Schottky diodes whose Hall, C–V and I–V parameters agree, establishing the layers as a platform for high-voltage power devices.","pith_inferences":["Because earlier LPCVD Ge reports showed substantially lower mobilities, the present growth geometry (source spacing, 1100 °C, ~1.5 Torr) is likely the practical lever that recovered device-grade transport.","If residual Si is ruled out by SIMS, Ge becomes a competitive alternative to Si for LPCVD doping windows where Si incorporation is hard to throttle.","The modest FWHM and roughness rise with doping suggests Ge lattice strain remains tolerable up to at least mid-10¹⁸ cm⁻³, so thicker multi-micron drifts should remain structurally viable."],"forward_implications":["LPCVD can supply Ge-doped drift layers for vertical β-Ga₂O₃ Schottky and power transistors without metal-organic or chloride precursors.","The 14 meV donor level implies near-complete room-temperature ionization, so doping set-points can be used directly for on-resistance and breakdown design.","Further reduction of growth-related line-defect density should raise low-temperature mobility above the observed 234 cm²/V·s peak.","Extension of the same LPCVD process toward lower 10¹⁵–10¹⁶ cm⁻³ doping would open high-voltage drift layers for multi-kV devices."],"fun_headline_variants":["LPCVD Ge-dopes (010) β-Ga₂O₃ with 14 meV donors and Ni diodes","Ge-doped β-Ga₂O₃ via LPCVD: 7e17–2e18 cm⁻³, 14 meV donor, rectifiers","LPCVD homoepitaxy yields Ge n-type β-Ga₂O₃ and vertical Schottky diodes","Controllable LPCVD Ge doping of β-Ga₂O₃ gives 14 meV donors and diodes","LPCVD Ge:β-Ga₂O₃ films show shallow donors and matching Hall-CV-IV diodes"],"cache_read_input_tokens":128,"weakest_assumption_plain":"That the free electrons and the fitted 14 meV donor come mainly from the intentionally added germanium rather than residual silicon or other impurities, even though no absolute Ge concentration profile is measured.","fun_headline_variants_meta":{"raw":{"variants":["LPCVD Ge-dopes (010) β-Ga₂O₃ with 14 meV donors and Ni diodes","Ge-doped β-Ga₂O₃ via LPCVD: 7e17–2e18 cm⁻³, 14 meV donor, rectifiers","LPCVD homoepitaxy yields Ge n-type β-Ga₂O₃ and vertical Schottky diodes","Controllable LPCVD Ge doping of β-Ga₂O₃ gives 14 meV donors and diodes","LPCVD Ge:β-Ga₂O₃ films show shallow donors and matching Hall-CV-IV diodes"]},"model":"grok-4.5","effort":"low","cost_usd":0.005962,"raw_usage":{"total_tokens":1765,"prompt_tokens":1061,"num_sources_used":0,"completion_tokens":131,"cost_in_usd_ticks":59620000,"prompt_tokens_details":{"text_tokens":1061,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":573,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":1061,"tokens_out":131,"duration_ms":5583,"temperature":1.0,"reasoning_tokens":573,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-14T08:22:45.295399+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Secondary-ion mass spectrometry (or equivalent) depth profiles showing Ge atomic density that does not track the Hall and C–V donor densities across the reported doping series would falsify the claim that Ge is the electrically active dopant.","supporting_citations":[],"review_version":1}