{"id":"01ed07b4-75d7-46d9-a9c5-548d9b486e6e","arxiv_id":"2508.02374","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In Pt-WSe2-Pt FETs, transport crosses from thermionic emission to quantum tunneling as the channel shrinks below about 9 nm, and a 3 nm channel yields a subthreshold swing below Boltzmann's limit.","lead":"This paper simulates platinum-contacted WSe2 transistors with channels from 3 to 12 nanometers and shows that short devices switch by quantum tunneling rather than heat-driven emission. It introduces a competition parameter and predicts that a 3 nanometer channel can beat the usual 60 millivolts-per-decade subthreshold limit.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The gate-model alphas from Ref. 22 (with an h-AlN spacer) carry all Vg-dependent results, including the sub-60 mV/dec claim for 3 nm; a self-consistent gate sweep is needed before that claim can be trusted.","rationale":"The paper's actual content is a condensed-matter physics study, despite the mismatched Layout-HF metadata; treating the physics as the real claim, the strongest quantitative assertion is that the 3 nm device overcomes Boltzmann tyranny at 300 K. The weakest load-bearing element is the effective gate model: a rigid, piecewise-linear shift of the chemical potential with alpha_in = 0.83 and alpha_out = 0.33 imported from Ref. [22], a device with an h-AlN spacer, and applied unchanged to direct Pt-WSe2-Pt junctions of all channel lengths. Because the sub-60 mV/dec result is attributed to the steep slope of tau(E) near the chemical potential, any gate-induced modification of tau(E) or of gate efficiency could reshape or remove that window. The internal Top inconsistency (Abstract/Conclusion say Top < 300 K for Lch < 9 nm while Section III says Top > 300 K) further muddies the central claim. These issues do not make the physics impossible - tunneling-dominated sub-60 mV/dec is a recognized phenomenon - but they make the quantitative result conditional on an unverified electrostatic assumption. The proposed self-consistent NEGF-DFT gate sweep is a direct, feasible check. This is essentially the same weakest assumption the reader identified, so the CONDITIONAL verdict stands without change.","tokens_in":19323,"tokens_out":9795,"duration_ms":124380,"concrete_test":"Redo the gate electrostatics for Lch = 3 nm and 9 nm self-consistently: include an explicit gate electrode with EOT = 8 A, epsilon_r = 3.9 in the NEGF-DFT (NanoDCAL) calculation, sweep Vg from -1.5 to 1.5 V, recompute tau(E, Vg) and the Landauer current at T = 300 K, and extract the effective dmu/dVg (alpha_in, alpha_out) and SS(Vg) from the self-consistent results. If the SS/BT < 1 window for Lch = 3 nm persists with self-consistent tau(E, Vg), the sub-BT claim is robust; if it vanishes or shifts outside the reported Vg/T range, the fixed-alpha rigid-shift model is the source of the headline result. Also check whether Fig. 7(b) gives Top near or above 300 K for Lch = 3 and 6 nm to resolve the abstract/results contradiction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative conclusions (SS(Vg), Top, Tc, and the sub-Boltzmann swing window for Lch = 3 nm) are all computed from the rigid effective-gate model of Sec. II.C, in which mu(Vg) is shifted linearly with fixed efficiencies alpha_in = 0.83 and alpha_out = 0.33. These constants are taken from Ref. [22], a Pt-WSe2-Pt FET with an inserted h-AlN monolayer spacer, and are applied unchanged to the present direct Pt-WSe2-Pt junctions at Lch = 3, 6, 9, and 12 nm without re-verification. The model also keeps tau(E) fixed and shifts only the Fermi level; it neglects gate-induced changes to the channel band profile, tunnel barrier shape, and any in-gap resonances. Short-channel electrostatics (fringing fields, source-drain coupling, quantum capacitance) make this rigid shift least reliable precisely at Lch = 3 nm, where the beyond-Boltzmann swing is claimed. In addition, the paper is internally inconsistent about the threshold: the Abstract and Conclusion state that for Lch < 9 nm, Top < 300 K, while Section III reports Top > 300 K for those devices; one of these is a sign error in a central quantitative claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies monolayer Pt-WSe2-Pt field-effect transistors with channel lengths from 3 to 12 nm, combining NEGF-DFT transmission functions with an effective gate model. It derives an asymptotic equivalence between the Landauer formula and Richardson's law in the classical limit, introduces a competition parameter zeta to separate quantum tunneling from thermionic emission, and computes the subthreshold swing, OFF current, and two characteristic temperatures Top and Tc. The paper claims that the 3 nm device remains tunneling-dominated up to 500 K and achieves a sub-Boltzmann subthreshold swing, while Lch near 10 nm is identified as the optimal channel length.","tokens_in":19618,"tokens_out":12651,"duration_ms":126302,"significance":"If the device-level predictions are correct, the paper offers a concrete route to sub-60 mV/dec switching in 2D FETs and a useful scaling criterion near Lch = 10 nm. The first-principles NEGF-DFT transmission calculations and the explicit analytic connection between the Landauer formula and Richardson's law are valuable contributions, and the zeta parameter is a crisp diagnostic for the tunneling-to-thermionic crossover. However, the quantitative device predictions rest on an effective gate model whose parameters are imported from a prior self-cited study without re-verification, and the manuscript contains an internal contradiction about Top that must be resolved before the central claim is reliable.","major_comments":[{"comment":"The threshold behavior of Top is stated inconsistently. The Abstract and the Conclusion state 'For Lch < 9 nm, Top < 300 K', while Section III, in the paragraph discussing Fig. 7(b), states 'short-channel FETs (Lch < 9 nm) exhibit Top > 300 K'. Since Top defines the tunneling-dominated versus thermionic-dominated regimes and supports the optimal-length claim, this is a load-bearing inconsistency. Please correct the sign and ensure the abstract, results, and conclusion all agree.","section":"Abstract; Section III (Fig. 7(b)); Section IV"},{"comment":"The effective gate model in Section II.C adopts alpha_in = 0.83 and alpha_out = 0.33 from Ref. [22], which treats a Pt-WSe2-Pt junction with an h-AlN monolayer spacer, and applies them unchanged to the present direct Pt-WSe2-Pt junctions at all channel lengths. All Vg-dependent quantities (SS, JOFF, Top, Tc) are computed from this linear shift, and the sub-Boltzmann claim for the 3 nm device depends on the resulting position of mu(Vg) in the transmission gap. The manuscript itself notes in Section III that low gate voltages 'may be insufficient to generate substantial effects on charge screening, interface state filling, and gate leakage.' Because short-channel electrostatics make the rigid shift least reliable at 3 nm, please either (i) perform self-consistent NEGF-DFT gate sweeps for at least the 3 nm device, or (ii) provide a sensitivity analysis showing the results are robust to variations in alpha_in and alpha_out within a physically plausible range.","section":"Section II.C; Section III"},{"comment":"The closed-form expression for SS in Eq. (16) is missing a numerical factor. Direct differentiation of the Landauer current with respect to Vg gives dI/dVg proportional to (1/(4kBT)) times an integral over [sech^2((E-mu_L)/(2kBT)) - sech^2((E-mu_R)/(2kBT))] tau(E), so the overall prefactor in Eq. (16) should be 4 ln(10) kBT/e rather than ln(10) kBT/e. With the printed factor, the thermionic limit of Eq. (16) becomes SS -> BT/(4 alpha_in), but the correct limit is BT/alpha_in, and the values reported in Fig. 4(c) (SS approximately 72.6 mV/dec) are consistent with BT/alpha_in. Please correct Eq. (16) and confirm that the numerical implementation uses the corrected expression.","section":"Section II.E, Eq. (16)"},{"comment":"The correspondence-principle derivation is carried out for a parabolic single-band model with a sharp affinity step, not for the actual DFT transmission function tau(E) shown in Fig. 1(b). The statement that 'the Landauer formula asymptotically converges to Richardson's law' is therefore proven for an idealized model rather than for the computed junctions. Please either soften the universality claim or demonstrate the asymptotic behavior on the actual tau(E), for example by checking the slope of log10(J/T) versus 1/T for the 12 nm device in the classical regime.","section":"Section II.D"}],"minor_comments":[{"comment":"The caption states 'the tunneling-dominated triangular region ultimately vanishes at temperatures exceeding approximately 250, 325, and 500 K for the 12, 9, and 12 nm FETs', listing 12 nm twice and omitting the 6 nm device; the intended sequence is likely 12, 9, and 6 nm.","section":"Figure 3 caption"},{"comment":"The slope of the Richardson plot is given as m = (log10 e)/kB W, but the standard expression is negative, m = -(log10 e) W/kB, when plotting log10(J/T) versus 1/T; please include the minus sign or define the axes consistently.","section":"Section III, text near Eq. (14)"},{"comment":"The caption for Fig. 7 states 'Vds = 50 V', but the text and all other figures use Vds = 50 mV; this is a typo.","section":"Figure 7 caption"}],"recommendation":"major_revision","confidential_remarks":"The provided full text is a condensed-matter paper on 2D TMD FETs, while the metadata abstract describes a computer-vision paper titled 'Uni-Layout'. The referee report above evaluates the full-text manuscript; the editorial office should confirm which manuscript is actually under consideration. Beyond that, the main technical issues are the internal Top inconsistency, the unverified gate-model parameters, and an apparent missing factor of 4 in Eq. (16)."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the provenance: this arXiv listing is for a cs.CV layout-generation paper, but the full text is a condensed-matter transport study. That mismatch alone should stop the review process. Ignoring that, the physics is a mixed bag.\n\nWhat's new: a quantitative NEGF-DFT study of Pt-WSe2-Pt FETs at 3, 6, 9, and 12 nm, with a competition parameter zeta separating tunneling and thermionic regimes, and Top/Tc criteria for operating windows. The Landauer-to-Richardson reduction is a textbook derivation but cleanly executed. The zeta maps in Fig. 3 are a useful way to visualize the crossover.\n\nThe soft spots, in order. (1) The gate model uses alpha_in=0.83 and alpha_out=0.33 from Ref. [22], which was a device with an h-AlN spacer, applied here to direct Pt-WSe2-Pt junctions at all lengths. The paper does not re-verify these alphas for the shorter channels. This is load-bearing because the sub-60 mV/dec claim for the 3 nm device comes entirely from this rigid shift. (2) The abstract and conclusion say Top < 300 K for Lch < 9 nm, but Section III clearly states the opposite ('short-channel FETs (Lch < 9 nm) exhibit Top > 300 K'). One of these is wrong, and it is a central quantitative claim. (3) The beyond-Boltzmann swing via tunneling is already established in TFETs, so the novelty is incremental. (4) No code or data is shipped; 'available from the corresponding author' is weak for a computational study.\n\nThe stress-test note is accurate: the gate model is the Achilles heel. The paper does not self-consistently recompute transmission as the gate shifts the potential, so any gate-dependent result beyond the rigid shift is unverified. That includes the main sub-Boltzmann claim.\n\nFor peer review: this should not be accepted as is. The metadata mismatch is a desk-reject reason. If the authors resubmit the physics paper under its own heading, fix the Top contradiction, and provide a self-consistent gate sweep (or at least justify transferring alphas from Ref. 22), it could be a modest contribution to the 2D-FET scaling literature. As presented, it's not there.","headline":"Provenance mismatch aside, this is a device-physics paper with a coherent Landauer-to-Richardson reduction, but its sub-Boltzmann claim rests on an unverified gate model and the text contradicts itself on Top.","tokens_in":20193,"tokens_out":4609,"would_cite":false,"duration_ms":54749,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["73.40.Qv","73.63.-b","85.30.Tv"],"model":"deepseek-v4-flash","headline":"Scaling a monolayer Pt–WSe2–Pt transistor from 12 nm to 3 nm shifts transport from thermionic emission to quantum tunneling, and at 3 nm the subthreshold swing drops below the 60 mV/dec Boltzmann limit at room temperature.","keywords":["2D TMD FETs","quantum tunneling","thermionic emission","subthreshold swing","Boltzmann tyranny","Landauer formula","NEGF-DFT","channel scaling"],"falsifier":"Measure the transfer characteristic (drain current versus gate voltage) of a 3 nm channel monolayer Pt–WSe2–Pt FET at 300 K with a thin EOT gate dielectric: if the subthreshold swing stays above about 60 mV/dec, or if the effective chemical-potential shift deviates significantly from 0.83 eV per volt of gate bias, then the tunneling-dominated beyond-Boltzmann claim fails.","tokens_in":19119,"feed_emoji":"⚡","tokens_out":5914,"duration_ms":64229,"temperature":0.7,"pith_summary":"This paper tries to establish how the transport mechanism in two-dimensional transition-metal-dichalcogenide transistors changes as the channel is scaled from 12 nm down to 3 nm, and what that means for device performance. It argues that the quantum-mechanical Landauer current asymptotically reduces to classical thermionic emission in the long-channel, high-temperature limit, and uses a competition parameter to draw the boundary between the two regimes. The central result is that a 3 nm Pt–WSe2–Pt device stays tunneling-dominated up to 500 K and achieves a subthreshold swing below the classical 60 mV/dec Boltzmann limit at room temperature, while channels around 10 nm represent the best compromise between leakage and gate control. If correct, these results give concrete guidance for choosing operating temperature and gate-voltage windows in sub-10 nm 2D FETs.","feed_headline":"3 nm 2D transistor beats the 60 mV/dec limit","feed_subtitle":"Tunneling-dominated switching breaks Boltzmann tyranny, but ~10 nm channels are the practical sweet spot.","key_machinery":"The central objects are the transmission coefficient τ(E) obtained from NEGF-DFT calculations, the Landauer–Büttiker formula for gate-modulated current, and an effective gate model in which the applied voltage shifts the chemical potential linearly with coefficients αin = 0.83 inside the band gap and αout = 0.33 outside it. From the Landauer current, the paper separates the tunneling contribution IQM (energies inside the band gap) from the thermionic contribution ISC (energies above the conduction minimum or below the valence maximum) and defines the competition parameter ζ. Two characteristic temperatures, Top (where OFF current is minimal) and Tc (where thermionic emission takes over), quantify the crossover and define the operating window for each channel length.","core_discovery":"The paper claims that for monolayer Pt–WSe2–Pt FETs the OFF current and subthreshold swing are governed by two competing mechanisms: quantum tunneling, which is temperature-insensitive and grows exponentially as the channel shortens, and semiclassical thermionic emission, which is length-insensitive and follows Richardson's law. Using transmission coefficients computed from NEGF-DFT and a Landauer formula with an effective gate model, the authors define a competition parameter ζ = (ISC − IQM)/(ISC + IQM) that cleanly separates the regimes. They find that for Lch < 9 nm the OFF current is tunneling-dominated with an optimal temperature Top below 300 K, whereas for Lch ≥ 9 nm thermionic emission dominates and the subthreshold swing approaches the Boltzmann tyranny divided by the gate-control efficiency αin. The 3 nm device remains tunneling-dominated up to 500 K and its subthreshold swing falls below Boltzmann tyranny because the steep slope of the transmission function τ(E) steepens the log-current versus gate-voltage curve.","pith_inferences":["The steep-τ(E) mechanism that produces sub-Boltzmann switching at 3 nm suggests that engineering the transmission edge—through spacer layers, band alignment, or dielectric screening—might extend the sub-60 mV/dec benefit to longer channels or higher temperatures.","The competition parameter ζ could serve as a practical design metric for other sub-10 nm devices, including tunnel FETs, because it directly quantifies how much of the current is quantum versus thermionic at a given temperature and gate voltage.","Because the authors keep the gate voltage below 1.5 V and note that screening, interface-state filling, and gate-leakage effects are not substantially probed, an experimental test at higher gate bias would clarify the validity range of the effective gate model.","A low-temperature transport experiment that separates temperature-insensitive tunneling from Arrhenius-type thermionic emission could directly verify the predicted Top and Tc for each channel length."],"forward_implications":["Sub-10 nm 2D FETs can switch below the 60 mV/dec Boltzmann limit when transport is ballistic and tunneling-dominated, at least for a 3 nm Pt–WSe2–Pt channel at room temperature.","A channel length near 10 nm offers the best trade-off between leakage current and gate control for 2D FETs, because longer channels lose some gate efficiency while shorter channels leak more.","The two characteristic temperatures Top and Tc give device designers concrete criteria for selecting operating temperature and gate-voltage windows when scaling 2D FETs.","In the thermionic regime the normalized subthreshold swing SS/αin approaches the fundamental Boltzmann limit, implying that thinner gate dielectrics and better gate control directly improve the subthreshold swing.","The quantum-to-classical crossover picture should generalize to other TMD channel materials, with barrier height and chemical-potential alignment shifting the quantitative predictions but not the qualitative behavior."],"supporting_citations":[{"why":"Supplies the effective gate model with the linear chemical-potential shift parameters αin = 0.83 and αout = 0.33 used for all channel lengths.","marker":"[22]"},{"why":"Provides the transmission formalism τ(E) and the derivation connecting the Landauer formula to classical current, underpinning the correspondence-principle argument.","marker":"[23]"},{"why":"Defines the earlier quantum-classical competition parameter that this paper adapts and generalizes as ζ.","marker":"[36]"},{"why":"Establishes the 'Boltzmann tyranny' limit for subthreshold swing that the 3 nm device claims to overcome.","marker":"[21]"},{"why":"These works provide the NEGF-DFT method (NanoDCAL) used to compute the transmission coefficients for each junction.","marker":"[31-33]"},{"why":"Reports near-ideal subthreshold swing in an h-BN encapsulated MoS2 FET, serving as an experimental benchmark that the ballistic-regime predictions extend beyond.","marker":"[41]"}],"fun_headline_variants":["Uni-Layout: Unified layout generation with human-mimicking evaluation","One framework to generate and judge layouts, guided by human feedback","Layout generator and evaluator aligned via dynamic-margin preference optimization","Uni-Layout lifts layout tasks with 100k expert-annotated feedback","From generation to evaluation: Uni-Layout integrates human judgments"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Every gate-dependent prediction assumes a fixed linear relation between applied gate voltage and chemical-potential shift (83% inside the band gap, 33% outside), parameters taken from prior work and applied unchanged to all channel lengths from 3 to 12 nm.","fun_headline_variants_meta":{"raw":{"variants":["Uni-Layout: Unified layout generation with human-mimicking evaluation","One framework to generate and judge layouts, guided by human feedback","Layout generator and evaluator aligned via dynamic-margin preference optimization","Uni-Layout lifts layout tasks with 100k expert-annotated feedback","From generation to evaluation: Uni-Layout integrates human judgments"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000273,"raw_usage":{"total_tokens":1666,"prompt_tokens":1004,"completion_tokens":662,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":620,"completion_tokens_details":{"reasoning_tokens":571}},"tokens_in":620,"tokens_out":662,"duration_ms":7850,"temperature":1.0,"reasoning_tokens":571,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T04:58:17.335277+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the transfer characteristic (drain current versus gate voltage) of a 3 nm channel monolayer Pt–WSe2–Pt FET at 300 K with a thin EOT gate dielectric: if the subthreshold swing stays above about 60 mV/dec, or if the effective chemical-potential shift deviates significantly from 0.83 eV per volt of gate bias, then the tunneling-dominated beyond-Boltzmann claim fails.","supporting_citations":[{"cited_title":"Jiang, L","cited_arxiv_id":null,"evidence_quote":"Supplies the effective gate model with the linear chemical-potential shift parameters αin = 0.83 and αout = 0.33 used for all channel lengths."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the transmission formalism τ(E) and the derivation connecting the Landauer formula to classical current, underpinning the correspondence-principle argument."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the earlier quantum-classical competition parameter that this paper adapts and generalizes as ζ."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the 'Boltzmann tyranny' limit for subthreshold swing that the 3 nm device claims to overcome."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports near-ideal subthreshold swing in an h-BN encapsulated MoS2 FET, serving as an experimental benchmark that the ballistic-regime predictions extend beyond."}],"review_version":1}