{"id":"1c23c1ac-69bf-4ac7-bf0e-cbb928e69633","arxiv_id":"2608.07823","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"LPCVD-grown tin-doped (010) beta-Ga2O3 films show record electron mobility for this growth method while keeping high growth rates and good crystal quality.","lead":"Researchers grew tin-doped beta-gallium oxide films with a low-pressure chemical vapor deposition system, controlling electron concentrations from 10^17 to 10^18 per cubic centimeter. The films combine high electron mobility with very fast growth rates, making this approach attractive for thick drift layers in high-voltage power devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'highest reported' LPCVD Sn-doped mobility claim lacks a direct comparison with prior LPCVD Sn-doped reports; the reader's thickness concern does not affect Hall mobility.","rationale":"I read the paper in good faith and find the growth, structural, and transport data plausible and internally consistent. The strongest claim is the record mobility. The reader's weakest assumption is the thickness inference from sapphire witness samples. That assumption affects carrier concentration (n = ns/t) and growth rate, but not Hall mobility, because mu = 1/(q*ns*Rs) is thickness-independent. Thus the reader's concern is not the most load-bearing for the headline mobility record. The actual load-bearing condition for 'highest reported values' is a complete comparison with all prior LPCVD-grown Sn-doped beta-Ga2O3 Hall data. The paper does not provide this comparison; it cites prior LPCVD Sn-doped work but omits those mobility values from Figure 6 and the text. This is a missing-support issue rather than an internal inconsistency. The concrete test is a targeted literature check of refs 72, 73, and 61. The reader's CONDITIONAL verdict remains appropriate; the condition should include explicit demonstration that the superlative is not contradicted by any prior LPCVD Sn-doped report.","tokens_in":15842,"tokens_out":10795,"duration_ms":102435,"concrete_test":"Retrieve refs 72, 73, and 61 and extract all reported Hall mobilities (room temperature and any sub-300 K values) for LPCVD-grown Sn-doped beta-Ga2O3. If any value is >=113 cm2/Vs at RT or >=380 cm2/Vs at 84 K, the record claim is false. If those papers report no Hall data, the authors should state that explicitly and qualify the superlative. This one literature check settles the central claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III and Figure 6 benchmark room-temperature mobility only against MOCVD/MBE data (refs 21, 22, 39, 40, 43, 44); no prior LPCVD-grown Sn-doped data points are shown. The text asserts 'highest reported value for LPCVD-grown Sn-doped beta-Ga2O3' without quoting Hall mobilities from refs 72 and 73 (LPCVD Sn-doped on sapphire) or from ref. 61 (the authors' own LPCVD Sn-doped Schottky diode work). The central superlative is therefore unsupported as written: if any prior LPCVD Sn-doped film had a room-temperature mobility >=113 cm2/Vs or an 84 K mobility >=380 cm2/Vs, the record claim is false. Separately, the reader's thickness-inference concern, while valid for absolute carrier concentrations and growth rates, does not undermine the Hall mobility, because van der Pauw mobility is computed from sheet carrier density and sheet resistance independently of film thickness.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a systematic study of Sn-doped (010) β-Ga2O3 homoepitaxial films grown by low-pressure chemical vapor deposition (LPCVD). The authors vary Sn source loading to achieve room-temperature carrier concentrations from 1.17×10^17 to 3.06×10^18 cm^-3, with Hall mobilities decreasing from 113 to 63 cm^2/V·s. Structural characterization (XRD, Raman, XPS, FESEM, AFM) shows phase-pure monoclinic films with step-flow morphology and rocking-curve FWHM as low as 68.4 arcsec. Temperature-dependent Hall measurements on two samples are analyzed with a two-donor charge-neutrality model and Matthiessen's-rule mobility model, yielding shallow donor energies of 32.7 and 26.7 meV, deeper donor levels near 95 and 80 meV, and compensating acceptor densities of 2.0×10^16 and 3.5×10^16 cm^-3. The central claim is that the sample with n=1.17×10^17 cm^-3 exhibits the highest room-temperature (113 cm^2/V·s) and low-temperature (380 cm^2/V·s at 84 K) Hall mobilities reported for LPCVD-grown Sn-doped β-Ga2O3, and that LPCVD growth rates of 6.4–16.6 μm/h enable thick drift layers.","tokens_in":16077,"tokens_out":3007,"duration_ms":26046,"significance":"If the superlative mobility claims are adequately supported, this work is a valuable contribution to the β-Ga2O3 power-device materials literature: it demonstrates controlled Sn doping by LPCVD over a wide range, shows good structural quality and high growth rates, and provides temperature-dependent Hall data with a scattering analysis. The paper also usefully extends the comparison of LPCVD Sn-doped films against MOCVD/MBE Sn-doped layers. The main significance hinges on the record-mobility claim, which is not currently substantiated because no quantitative comparison with prior LPCVD Sn-doped work is provided. The experimental dataset (XRD, Raman, XPS, AFM, Hall) is internally consistent, and the growth-rate/thickness claims are plausible, although the thickness-inference method introduces a systematic uncertainty that propagates to reported carrier concentrations.","major_comments":[{"comment":"The claim that the 113 cm^2/V·s room-temperature mobility and the 380 cm^2/V·s at 84 K are 'the highest reported values for LPCVD-grown Sn-doped β-Ga2O3' is unsupported as written. Figure 6 benchmarks only against MOCVD and MBE data (refs 21, 22, 39, 40, 43, 44). No mobility values from prior LPCVD-grown Sn-doped β-Ga2O3 reports are quoted, including refs 72 and 73 (LPCVD Sn-doped on sapphire) and ref. 61 (the authors' own LPCVD Sn-doped Schottky diode work with presumably similar drift layers). To sustain the superlative, the authors must either add those prior LPCVD data points to the benchmarking figure and explicitly compare the mobility values, or temper the claim to 'among the highest' with a clear statement of the comparison basis.","section":"Section III, Figure 6; Abstract"},{"comment":"Film thicknesses of the homoepitaxial layers were not measured directly; they were estimated from cross-sectional FESEM images of co-loaded β-Ga2O3 films grown on sapphire, assuming identical growth rates on sapphire and on the (010) β-Ga2O3 substrate. This assumption is not validated, and because carrier concentrations in Table 1 are computed as sheet density divided by thickness, the reported values of 1.17×10^17–3.06×10^18 cm^-3, as well as the growth rates of 6.4–16.6 μm/h, carry an unquantified systematic error. The Hall mobility is not affected by this thickness uncertainty because the van der Pauw mobility is obtained from sheet carrier density and sheet resistance, as noted in the paper's own data flow. The authors should either measure the thickness of the actual homoepitaxial films (e.g., by step profilometry, ellipsometry, or TEM) or explicitly quantify the expected difference in growth rate between sapphire and (010) β-Ga2O3 and provide an uncertainty estimate for the carrier concentrations.","section":"Section II, Experimental Details; Table 1"},{"comment":"The two-donor charge-neutrality model is fitted to the very same temperature-dependent Hall carrier-concentration data from which the donor parameters are then 'extracted,' and Table 3 sets the shallow donor concentration ND1 equal to the room-temperature carrier concentration by construction (1.17×10^17 and 3.28×10^17 cm^-3). The 'excellent agreement' between measured and fitted curves is therefore a property of the fit, not an independent validation of the two-donor hypothesis. The extracted values of ND2, ED2, and NA should be presented as model-dependent estimates rather than as directly determined materials parameters. A concrete test would be to compare the fitted donor activation energies with values obtained from a separate technique (e.g., capacitance-voltage profiling or admittance spectroscopy) or to state explicitly that the two-donor model is one plausible parametrization and not a unique solution.","section":"Section III, Eqs. (1)–(2); Table 3"}],"minor_comments":[{"comment":"The Raman axis label in Figure 3 is garbled ('Raman Shi9'); the intended text is likely 'Raman Shift (cm^-1)'.","section":"Section III, Figure 3"},{"comment":"The legend entries in Figure 6 mix growth-method labels (e.g., 'MBE MOCATAXY', 'MOVPE') with an inconsistent naming scheme; for clarity, use a uniform naming convention and define the symbols in the caption.","section":"Section III, Figure 6"},{"comment":"Table 1 lists sample numbers but omits growth duration and measured thickness repeatability. Adding these parameters would aid reproducibility, especially given the thickness-inference approach.","section":"Section III, Table 1"},{"comment":"The Hall measurement temperature range is stated as 80 to 350 K, but the low-temperature peak mobility is reported at 84 K; the authors should clarify whether 84 K is the lowest measured temperature or the temperature of the observed peak.","section":"Section II, Experimental Details"},{"comment":"Reference 4 contains a typo ('1202A1202' should be '1202A2'), and several arXiv preprints (refs 53, 63, 64) are cited without archival validation; please check the final published versions or include appropriate caveats.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of cond-mat.mtrl-sci and reports potentially useful growth data, but the central superlative claim is currently not backed by a direct comparison with the prior LPCVD Sn-doped literature. The thickness-inference issue is important for quantitative carrier concentrations, while the transport-model circularity is a common but still nontrivial modeling concern. I would advise the editor to request a revision that addresses the three major points, with particular emphasis on either adding the missing LPCVD comparisons or softening the record claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper delivers the first systematic study of Sn-doped homoepitaxial (010) beta-Ga2O3 grown by LPCVD, and the core result is believable: controlled n-type doping from 1.17e17 to 3.06e18 cm-3, Hall mobilities that track doping in a sensible way (113 to 63 cm2/Vs at room temperature), and a low-temperature peak of 380 cm2/Vs at 84 K for the lowest-doped sample. Growth rates of 6.4-16.6 um/h with thicknesses up to 11.3 um are genuinely impressive for this technique. The structural work is solid and internally consistent: XRD rocking curves widen gradually with doping, Raman shows the expected monoclinic modes, XPS gives near-stoichiometric O/Ga, and AFM shows step-flow morphology throughout. The transport modeling is standard and gives reasonable donor energies and low compensation.\n\nThe soft spots are real but not fatal. The \"highest reported value for LPCVD-grown Sn-doped beta-Ga2O3\" claim is not actually supported by the comparison shown. Figure 6 benchmarks only against MOCVD and MBE; refs 72 and 73 (LPCVD Sn-doped on sapphire) and ref. 61 (their own LPCVD Sn-doped diodes) are cited but never quoted. If any of those prior films had a room-temperature mobility above 113 or an 84 K mobility above 380, the record claim is wrong. The claim should be qualified to homoepitaxial LPCVD and directly compared with those numbers.\n\nThe thickness issue is real but narrower than the reader suggests. The films' thicknesses come from co-loaded sapphire witnesses, not from the homoepitaxial layers themselves. That affects absolute carrier concentrations, growth rates, and derived donor densities, because all of them scale with the assumed thickness. It does not affect the Hall mobilities, since van der Pauw mobility is thickness-independent. The authors should measure thickness on the actual films or, at minimum, report the uncertainty introduced by the witness assumption.\n\nMissing error bars are a broader problem. Hall data, rocking-curve widths, and XPS compositions are all presented without uncertainty estimates, so it is hard to say whether sample-to-sample differences are meaningful. The two-donor and Matthiessen fits are performed on the same temperature-dependent data they are said to confirm; the agreement is a property of the fit, not an independent check. That is common practice, but the language \"excellent agreement confirms\" oversells it.\n\nFor the Ga2O3 epitaxy community this is a useful data point and a candidate benchmark for LPCVD Sn doping. It deserves a serious referee, not a desk reject. My recommendation: send it to review, and require the authors to substantiate the record claim with direct LPCVD comparisons, address the thickness determination, and add uncertainty estimates. The central findings will likely survive those revisions.","headline":"Useful first systematic LPCVD Sn-doping study with likely record mobilities, but the superlative is not directly benchmarked against prior LPCVD work and the thickness inference muddies the carrier-density numbers.","tokens_in":856,"tokens_out":1269,"would_cite":true,"duration_ms":31952,"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":"Tin-doped β-Ga₂O₃ films grown by LPCVD can be doped controllably to $1.17\\times10^{17}$–$3.06\\times10^{18}$ cm⁻³ and reach Hall mobilities of 113 cm²/V·s at room temperature and 380 cm²/V·s at 84 K, the highest reported for this growth…","keywords":["β-Ga2O3","LPCVD","Sn doping","homoepitaxial growth","Hall mobility","power electronics","wide-bandgap semiconductor","donor activation"],"falsifier":"Cut a cross-section of one of the $\\beta$-Ga$_2$O$_3$ homoepilayers and measure its thickness directly with electron microscopy, or profile Sn by secondary-ion mass spectrometry (SIMS) through the layer. If those numbers disagree with the sapphire-based thickness estimate, the reported carrier concentrations and growth rates would need revision; the mobility values themselves are largely unaffected, but the doping calibration and speed advantage rest on this measurement.","tokens_in":15680,"feed_emoji":"⚡","tokens_out":8865,"duration_ms":73484,"temperature":0.7,"pith_summary":"Using solid gallium and tin sources in low-pressure chemical vapor deposition, this paper shows that tin can dope homoepitaxial (010) $\\beta$-Ga$_2$O$_3$ films controllably across $1.17\\times10^{17}$ to $3.06\\times10^{18}$ cm$^{-3}$ while preserving monoclinic phase, step-flow morphology, and narrow X-ray rocking curves. The best film has room-temperature Hall mobility 113 cm$^2$ V$^{-1}$ s$^{-1}$ and 380 cm$^2$ V$^{-1}$ s$^{-1}$ at 84 K; the authors report these as the highest values for LPCVD-grown Sn-doped $\\beta$-Ga$_2$O$_3$. Because growth runs at 6.4–16.6 $\\mu$m/h, the process makes 1.7–11.3 $\\mu$m thick layers far faster than metal-organic CVD or molecular-beam epitaxy, which is what matters for high-voltage power-device drift layers. The paper argues this combination makes LPCVD a viable and scalable route for thick $\\beta$-Ga$_2$O$_3$ drift layers.","feed_headline":"LPCVD tin-doped Ga2O3 hits its highest reported mobility","feed_subtitle":"Thick drift layers grow at 6–16 µm/h with mobilities matching MOCVD and MBE epilayers.","key_machinery":"The mechanism is the solid-source Sn/Ga crucible arrangement: tin content in the source charge sets the electron concentration, while substrate temperature, pressure, and gas flows are fixed. The paper then uses a two-donor charge-neutrality model and Matthiessen's-rule mobility decomposition—ionized-impurity, neutral-impurity, polar-optical-phonon, and acoustic deformation-potential scattering—to extract donor energies and compensation from temperature-dependent Hall data. Growth thickness, estimated from cross-sectional FESEM of co-loaded sapphire films, converts sheet Hall data into bulk carrier concentration and growth rates.","core_discovery":"Central claim: LPCVD is not just a fast growth method but also a clean doping environment for Sn in $\\beta$-Ga$_2$O$_3$, with electrical quality close to metal-organic CVD and molecular-beam epitaxy. The evidence is a doping series grown at 1000 °C and ~1.5 Torr from Ga and Sn metal, with Sn loading setting the donor density; structural characterization (XRD, Raman, XPS) shows phase-pure, near-stoichiometric films; temperature-dependent Hall measurements fitted with a two-donor model give low shallow-donor activation energies (32.7 and 26.7 meV), a second deeper donor, and compensating acceptor concentrations about an order of magnitude below the donors. The standout sample, at $1.17\\times10^{17}$ cm$^{-3}$, reaches 113 cm$^2$ V$^{-1}$ s$^{-1}$ (room temperature) and 380 cm$^2$ V$^{-1}$ s$^{-1}$ (84 K).","pith_inferences":["Editorial extension: a direct thickness measurement on the homoepitaxial films themselves would settle whether the growth-rate advantage is as large as claimed; that is the one experiment this paper leaves undone.","A natural follow-up is to place a vertical Schottky diode or transistor on the 11.3 $\\mu$m layer and compare blocking voltage with Si-doped MOCVD drift layers; the paper's earlier Sn-doped LPCVD diode work makes that test immediate.","The transport model assumes only four scattering mechanisms; a temperature-dependent Hall study down to lower temperatures or a mobility-versus-thickness series could reveal whether defects at the film/substrate interface contribute at low temperature.","Sn's octahedral-site preference and low activation energy suggest that co-doping or multi-layer stacks with Si might be feasible, but abrupt doping-transition behavior in LPCVD is not addressed and would need calibration for device structures."],"forward_implications":["A growth rate of 6.4–16.6 $\\mu$m/h means a 10 $\\mu$m drift layer, the kind needed for multi-kilovolt vertical devices, can be grown in roughly an hour rather than a day.","Sn doping with LPCVD covers the $10^{17}$–$10^{18}$ cm$^{-3}$ range with mobilities comparable to MOCVD and MBE, giving device designers a second donor species beyond Si in a fast growth platform.","The measured activation energies (26.7–32.7 meV) and low compensation imply nearly complete donor ionization at room temperature, a prerequisite for low on-resistance drift layers.","Above roughly $3\\times10^{18}$ cm$^{-3}$, surface roughness and rocking-curve width degrade, which defines a practical doping ceiling if smooth step-flow morphology is needed."],"supporting_citations":[{"why":"MOCVD-grown Sn-doped Ga₂O₃ with record electrical properties; the baseline the mobility comparison must match.","marker":"[21]"},{"why":"MOVPE Si- and Sn-doped homoepitaxial layers on (010) substrates; defines donor behavior and the comparison range.","marker":"[22]"},{"why":"MBE-grown Sn-doped Ga₂O₃ films; supplies the transport and electronic-structure benchmark for Hall mobility.","marker":"[39]"},{"why":"Sn doping of (010) β-Ga₂O₃ by plasma-assisted MBE; the benchmark for low-temperature mobility and compensation.","marker":"[44]"},{"why":"Reports high-temperature LPCVD of β-Ga₂O₃ and provides the growth-mode background (step-flow morphology) used here.","marker":"[54]"},{"why":"The authors' earlier LPCVD Si-doped films showing high growth rates and mobilities; the precedent this Sn study extends.","marker":"[59]"},{"why":"Their demonstration of Sn-doped LPCVD Schottky diodes; the device motivation and prior evidence that LPCVD drift layers are usable.","marker":"[61]"},{"why":"LPCVD Sn-doped β-Ga₂O₃ grown on sapphire from Ga-Sn alloy; earlier LPCVD Sn-doping work this paper goes beyond with homoepitaxy.","marker":"[72]"}],"fun_headline_variants":["Record LPCVD mobility for Sn-doped β-Ga2O3","LPCVD-grown Sn:Ga2O3 hits 113 cm²/V·s mobility","Thick Sn-doped Ga2O3 drift layers via fast LPCVD","Highest mobility yet in LPCVD Sn-doped β-Ga2O3","Sn doping in Ga2O3: LPCVD matches MOCVD quality"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the homoepitaxial films grown on $\\beta$-Ga$_2$O$_3$ substrates have the same thickness as simultaneously grown films on sapphire, because the film thickness itself was never measured on the actual samples; every carrier concentration and growth-rate number inherits that inference.","fun_headline_variants_meta":{"raw":{"variants":["Record LPCVD mobility for Sn-doped β-Ga2O3","LPCVD-grown Sn:Ga2O3 hits 113 cm²/V·s mobility","Thick Sn-doped Ga2O3 drift layers via fast LPCVD","Highest mobility yet in LPCVD Sn-doped β-Ga2O3","Sn doping in Ga2O3: LPCVD matches MOCVD quality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000288,"raw_usage":{"total_tokens":1811,"prompt_tokens":1192,"completion_tokens":619,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":808,"completion_tokens_details":{"reasoning_tokens":514}},"tokens_in":808,"tokens_out":619,"duration_ms":5900,"temperature":1.0,"reasoning_tokens":514,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T04:12:22.904178+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Cut a cross-section of one of the $\\beta$-Ga$_2$O$_3$ homoepilayers and measure its thickness directly with electron microscopy, or profile Sn by secondary-ion mass spectrometry (SIMS) through the layer. If those numbers disagree with the sapphire-based thickness estimate, the reported carrier concentrations and growth rates would need revision; the mobility values themselves are largely unaffected, but the doping calibration and speed advantage rest on this measurement.","supporting_citations":[],"review_version":1}