{"id":"4119b074-8228-4635-a0d3-76ef55e79a12","arxiv_id":"2507.12991","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Planar tunneling into twisted bilayer graphene shows a tenfold conductance increase and a much narrower zero-bias suppression than Bernal bilayer graphene, attributed to low-energy layer-breathing phonons enabled by the small moiré Brillouin zone.","lead":"Experiments show that electrons tunnel much more easily into twisted bilayer graphene than into normal bilayer graphene through a tungsten diselenide barrier, thanks to low-energy lattice vibrations and a relaxed momentum-matching condition. The work introduces planar tunneling spectroscopy as a probe of electron-phonon coupling and moiré band structure in twisted van der Waals materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Device-to-device WSe2 barrier variation is not controlled; the comparative tBLG-vs-BLG enhancement could be a barrier-thickness artifact.","rationale":"The reader's weakest assumption identifies the absence of controlled barrier thickness and interface quality between the BLG and tBLG devices as the load-bearing uncertainty. I agree: the comparative claim is the heart of the paper, and tunneling conductance's exponential dependence on barrier thickness makes this the most plausible alternative explanation for the observed order-of-magnitude conductance increase. The phonon peak at ~8 meV in d2I/dV2 of tBLG, and its absence in BLG, is an important internal check, since barrier thickness alone cannot shift an inelastic peak energy; but this does not settle whether the magnitude of the enhancement is intrinsic. The paper's SI considers other physical mechanisms (ZA phonons, backfolding, three-terminal effects) but not device-to-device barrier variation, which is a notable omission. The proposed control experiment, using the same WSe2 flake for both BLG and tBLG junctions, would directly test whether the comparative changes survive with matched barriers. Because the current evidence is suggestive but not conclusive, the CONDITIONAL verdict from the reader is appropriate; my stress-test does not change it. I also credit the paper's independent phonon calculations and the reproducible 8 meV d2I/dV2 feature as genuine evidence that an interesting low-energy inelastic channel exists in tBLG, which is why the concern warrants a condition rather than rejection.","tokens_in":19029,"tokens_out":18093,"duration_ms":213336,"concrete_test":"Fabricate a control chip in which a single exfoliated WSe2 flake is transferred as a continuous barrier spanning adjacent BLG and tBLG regions, with junction areas set by identical electrode patterns; measure dI/dV and d2I/dV2 on at least three BLG and three tBLG junctions from the same flake. If tBLG still shows the ~5-10 mV zero-bias gap and an 8-10 meV d2I/dV2 peak while BLG shows the 30-40 mV gap and no low-energy peak, the twist-angle mechanism is confirmed; if the conductance enhancement or gap difference disappears, the original comparison was compromised by barrier variation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that tBLG shows enhanced phonon-assisted tunneling compared with BLG rests on comparing BLG device D1 with tBLG devices D2 and D3. Tunneling conductance is exponentially sensitive to barrier thickness, yet the paper reports no WSe2 thickness, junction area, or capacitance for any junction, and the devices were assembled in separate pickup-stack runs. If the tBLG barriers are even one monolayer thinner, their conductance would rise by roughly an order of magnitude with no role for moiré physics. The paper's own 'Discussion on other possibilities' (SI) addresses ZA modes, backfolding, and the three-terminal model but does not address this obvious sample-variation alternative. Reproducibility across multiple contacts within the same tBLG device (Figs. 2-3) does not constrain flake-to-flake barrier differences. The d2I/dV2 phonon peak near 8 meV in tBLG and its absence in BLG provide partial internal evidence that the inelastic threshold differs intrinsically, but the headline claims of an order-of-magnitude conductance enhancement and 'first clear deviation' from the 30-70 mV trend cannot be separated from barrier-thickness differences on the current data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports planar tunneling spectroscopy on metal/WSe2/BLG and metal/WSe2/tBLG (θ≈2°) junctions, measuring dI/dV and d2I/dV2 as functions of bias and top-gate voltage. The key observation is a much narrower zero-bias suppression in tBLG (≈5–10 meV) than in BLG (≈30–40 meV), an order-of-magnitude higher conductance in tBLG, and a low-energy d2I/dV2 peak at ≈8 meV in tBLG assigned to the layer breathing mode. The authors attribute the difference to a relaxed in-plane momentum-matching condition in the small moiré Brillouin zone: low-energy phonons can bridge the Fermi-surface mismatch in tBLG but not in BLG. Phonon dispersions for BLG and tBLG are computed with classical force fields (PARPHOM/TWISTER/LAMMPS/phonopy). The central claim is that this is the first clear deviation from the 30–70 meV zero-bias gap trend in graphitic tunnel junctions, establishing twist angle as a knob for phonon-assisted tunneling.","tokens_in":19198,"tokens_out":14045,"duration_ms":138055,"significance":"If the twist-induced enhancement is real, the paper introduces planar tunneling as a bulk probe of electron–phonon coupling in moiré materials and provides a new device-level control (twist angle) for tunnel junctions. The manuscript has notable strengths: reproducible data over multiple junctions in two tBLG devices (D2 and D3), d2I/dV2 spectra showing distinct phonon signatures, a comparative table of prior graphitic tunnel junctions, and phonon calculations based on standard, independently developed open-source codes. However, the significance is currently capped by two issues: the BLG/tBLG comparison is across separately fabricated devices without reported barrier thickness or area, and the central geometric statement about the moiré BZ 'encompassing' the Au Fermi-surface projection is numerically reversed. These issues need to be resolved before the main claim can be fully credited.","major_comments":[{"comment":"The central geometric argument contains a quantitative reversal. The text says that the moiré Brillouin zone (1.02×10^9 m^-1 for θ≈2°) 'encompasses the projection of the metal Fermi surface', but the reported k_F^Au = 1.20×10^10 m^-1 is an order of magnitude larger than the moiré BZ, so the projected Au Fermi disk can only encompass the moiré BZ, not the reverse. This matters because the relaxed momentum-matching claim in tBLG relies specifically on the Au projection filling the small moiré BZ. Please correct the wording, redraw the tBLG diagram in Fig. 4d with the Au disk and mBZ to scale or with an explicit scale, and restate the overlap criterion consistently with Eq. (3) and Fig. 4c.","section":"§2, Fig. 4c–d and Eq. (3)"},{"comment":"The BLG-vs-tBLG comparison is made across independently fabricated devices: BLG device D1 is compared with tBLG devices D2 and D3. The paper does not report the WSe2 barrier thickness, junction area, or capacitance for any junction. Since tunnel conductance is exponentially sensitive to barrier thickness and linearly proportional to area, the order-of-magnitude higher conductance in tBLG and even the narrower zero-bias feature could, in part, reflect sample-to-sample barrier differences. Reproducibility across contacts within D2/D3 does not constrain flake-to-flake variations. To support the headline claim, report the WSe2 thickness and junction dimensions, and provide a matched BLG control device fabricated in the same run with the same WSe2 flake (or an equivalent capacitance-based characterization of barrier thickness).","section":"§1 and §2, Figs. 1f–h vs 2a–c; Methods"},{"comment":"The paper does not demonstrate a quantitative link between the momentum-space picture and the measured bias gaps. The parameters in Eq. (3) (k_F^Au = 1.20×10^10 m^-1, k_F^BLG = 1.8×10^7 m^-1 at n = 1×10^12 cm^-2, and q_LBM = 0.25×10^10 m^-1) are quoted without a derivation, and the text does not explain how a momentum mismatch Δq translates into a bias gap of 30–40 meV or 5–10 meV, nor how that gap should depend on twist angle or carrier density. Since the model was constructed after the data, a sensitivity analysis over the plausible ranges of these inputs, or a prediction for the gap versus θ, would test whether the mechanism is actually responsible for the observed trend.","section":"§2, Eqs. (2)–(3) and Conclusion"}],"minor_comments":[{"comment":"The caption states the gate range as −5 V to 4 V, while the color plot axis spans −4 V to 4 V; please harmonize the text and axes.","section":"Fig. 2c"},{"comment":"The d2I/dV2 peak in tBLG is reported at ≈8 meV, while the layer breathing mode from the phonon calculation is quoted as ≈10 meV; the approximately 2 meV offset is not discussed and should be reconciled or explicitly attributed to model/measurement differences.","section":"Fig. S5 and main text"},{"comment":"The displayed inelastic current expression for the phonon-absorption term appears to contain a duplicated factor in the occupation-factor bracket; please check the algebraic expression and the signs.","section":"Supporting Information, inelastic current derivation"},{"comment":"The text contains several spacing and typographical errors (e.g., 'Avarietyofalternativetransporttechniquessuchas' and 'induce finer modifications'); a careful proofread is needed.","section":"Introduction and Conclusion"},{"comment":"For reproducibility, the paper would benefit from a statement of the moiré simulation cell size and number of atoms in the θ≈2° calculation, and from a data/code availability statement for the phonon calculations.","section":"Methods and Data Availability"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and the topic is timely. The main claim is plausible but is not yet separated from sample-to-sample barrier variations; the internal geometric inconsistency is fixable by rewording and a to-scale figure. If the authors supply the requested device characterization (WSe2 thickness, junction area, matched BLG control) and correct the geometric argument, the paper would be suitable for publication. I do not see grounds for rejection at this stage."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is the first planar tunneling spectroscopy on twisted bilayer graphene, and it reports a clean qualitative difference: a ~5-10 mV zero-bias gap in tBLG versus ~30-40 mV in BLG, with an order-of-magnitude higher conductance. The mechanism—that the small moiré BZ relaxes momentum conservation and lets the 10 meV layer breathing mode assist tunneling—is genuinely new and worth testing.\n\nWhat I like: the phonon calculations are independent, and the d2I/dV2 data in the SI show a sharp 8 meV feature in tBLG that is absent in BLG. That is strong internal evidence of a low-energy inelastic threshold only in tBLG. Reproducibility across multiple junctions in the same device is reassuring.\n\nThe soft spot is the uncontrolled device baseline. BLG and tBLG junctions come from separate fabrication runs, and the paper reports no WSe2 thickness, junction area, or capacitance. Tunneling conductance is exponentially sensitive to barrier thickness, so an order-of-magnitude difference could simply be a monolayer-thinner barrier in the tBLG samples. The paper's discussion of alternatives does not address this. The geometric model is also oversimplified: it takes a spherical Au Fermi surface, constant electron-phonon coupling, and the quoted k_BLG^F seems off by a factor of ten. The model was built to reproduce the observed gap sizes, which gives it some post-hoc flavor, though the phonon dispersions themselves are computed independently.\n\nThis is a paper for the moiré graphene and electron-phonon communities. It deserves serious peer review, but not acceptance on the current data. The authors should add barrier thickness characterization (AFM or capacitance), ideally a BLG device from the same stack, and a more quantitative tunneling model. The core idea is plausible and the spectroscopy is new; it just needs more evidence.\n\nI'd invite a revised version after those points are addressed.","headline":"First planar tunneling into tBLG: narrow gap and higher conductance, but the barrier-thickness confound makes the tBLG vs BLG comparison conditional.","tokens_in":19838,"tokens_out":4699,"would_cite":true,"duration_ms":46176,"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":"Planar tunneling into twisted bilayer graphene shows the moiré Brillouin zone relaxing in-plane momentum conservation, letting the ~10 meV layer breathing mode assist tunneling that is suppressed in Bernal bilayer graphene.","keywords":["twisted bilayer graphene","planar tunneling","phonon-assisted tunneling","moiré Brillouin zone","layer breathing mode","WSe2 tunnel barrier","Fermi surface momentum mismatch","zero-bias anomaly"],"falsifier":"Fabricate a single chip with adjacent Bernal and twisted junctions sharing the same WSe2 flake, then normalize conductances by junction area and barrier thickness; if the twisted conductance is not still about an order of magnitude higher, the twist-based explanation collapses. Alternatively, vary the twist angle continuously from zero to two degrees and check that the zero-bias gap interpolates monotonically from about 30 to 40 mV down to 5 to 10 mV as the moiré Brillouin zone shrinks.","tokens_in":1743,"feed_emoji":"⚛️","tokens_out":2181,"duration_ms":85192,"temperature":0.7,"pith_summary":"The paper sets out to show that planar tunneling, a bulk spectroscopic method, can reveal electron-phonon coupling in moiré materials. Its central experimental claim is that a metal/WSe2/twisted-bilayer-graphene junction with twist angle around 2 degrees conducts tunnel current about an order of magnitude better than an otherwise identical Bernal bilayer junction, while its zero-bias suppression shrinks from roughly 30 to 40 mV down to 5 to 10 mV. The paper attributes this to the moiré Brillouin zone being small enough that the ~10 meV layer breathing phonon can supply the in-plane momentum that separates the metal Fermi surface from the graphene Fermi pockets. If correct, this makes twist angle a continuous knob for tunnel currents and establishes planar tunneling as a bulk probe of electron-phonon coupling in twisted van der Waals materials.","feed_headline":"Twist shrinks graphene tunnel gap from 30 mV to 10 mV","feed_subtitle":"Planar tunneling into twisted bilayer graphene finds the layer-breathing phonon bridging the Fermi-surface mismatch.","key_machinery":"The central object is the in-plane momentum-conservation delta function $\\delta(k_T-k_B\\pm q)$ in the inelastic tunneling current, together with the phonon density of states $\\rho_{\\rm ph}(\\omega)$. The layer breathing mode, an optical phonon in which the two graphene layers vibrate out of phase, has a van Hove singularity in the phonon DOS near 10 meV, and the question is whether that phonon can supply the momentum needed to connect the metal Fermi surface to the graphene Fermi pockets. In Bernal bilayer graphene, the shortest required phonon wavevector exceeds what the mode provides, whereas in twisted bilayer graphene the folded moiré Brillouin zone relaxes the matching condition and lets the same phonon assist tunneling. This geometric criterion, expressed in the paper as $k_{BZ} = k_F^{Au} + k_F^{BLG} + \\Delta q$, is what carries the argument from measured dI/dV spectra to phonon-mediated mechanism.","core_discovery":"In a planar Au/WSe2/graphene tunnel junction, in-plane momentum must be conserved for elastic tunneling, but the Fermi-surface pockets of Bernal bilayer graphene around the K and K' points are disjoint from the metal's Fermi surface. The paper finds that inelastic tunneling must therefore supply a momentum mismatch of about $\\Delta q \\approx 0.5\\times10^{10}$ m$^{-1}$, which is larger than the maximum momentum $q_{\\rm LBM}\\approx0.25\\times10^{10}$ m$^{-1}$ available from the layer breathing mode near 10 meV. In twisted bilayer graphene, the moiré Brillouin zone is much smaller, roughly $1.02\\times10^9$ m$^{-1}$ at $\\theta=2^\\circ$, and it encompasses the metal Fermi surface projection, so the same breathing phonon bridges the mismatch and enhances tunneling. This produces an order-of-magnitude higher tunnel conductance and a zero-bias gate that is five to ten millivolts wide rather than thirty to forty millivolts.","pith_inferences":["Inference: if the geometric picture is correct, the enhancement should grow as the twist angle decreases toward the magic angle, where the moiré Brillouin zone is even smaller; this could be tested in a single fabrication run that varies only the twist angle.","Inference: the same mechanism predicts that tunneling into other moiré materials, such as twisted transition-metal dichalcogenides, may show phonon-assisted features at even lower biases limited only by the smallest phonon that bridges the momentum mismatch.","Inference: device-to-device normalization of barrier thickness, junction area, and capacitance is needed before the order-of-magnitude conductance claim can be made quantitative; without it, part of the observed enhancement could be barrier variation rather than twist physics."],"forward_implications":["Twist angle becomes a tunable parameter for van der Waals tunnel junctions: the zero-bias gap narrows and the conductance rises as the moiré Brillouin zone shrinks.","Planar tunneling can be used as a micrometer-scale bulk probe of electron-phonon coupling in moiré materials, complementing local STM probes.","The 10 meV layer breathing mode, largely invisible in earlier metal-graphene tunnel junctions with gaps of 30 to 70 mV, becomes visible in twisted bilayer graphene junctions in both dI/dV and d2I/dV2.","The dependence of tunnel conductance on both gate voltage and bias near the Dirac point indicates that planar tunneling also reads the moiré band density of states, not just phonon structure."],"supporting_citations":[{"why":"Supplies Bardeen's tunneling Hamiltonian that is the starting point for both the elastic and inelastic current expressions.","marker":"[48]"},{"why":"Provides the theory of phonon-mediated tunneling into graphene that gives the inelastic channel and the role of the phonon momentum q.","marker":"[30]"},{"why":"Demonstrates inelastic electron tunneling into graphene nanostructures on metal surfaces, establishing the momentum-matching role of phonons.","marker":"[29]"},{"why":"Reports planar metal-hBN-graphene junctions with 30 to 50 mV zero-bias gaps, the baseline trend from which the twisted bilayer data deviate.","marker":"[47]"},{"why":"Shows phonon-mediated tunneling into graphene on metal surfaces and is used to discuss why ZA phonons are not responsible for the observed features.","marker":"[27]"},{"why":"Provides the BLG-on-Ru comparison used to argue that ZA modes do not explain the low-energy tunneling in free-standing bilayer graphene.","marker":"[57]"},{"why":"Supplies the phonon dispersions and density of states used to identify the 10 meV layer breathing mode in both Bernal and twisted bilayers.","marker":"[56]"}],"fun_headline_variants":["Twist enhances graphene tunnel current via breathing phonon","Moiré twist lets breathing phonon boost tunneling","Twist enables phonon-assisted tunneling boost in graphene","Twisted bilayer graphene shows stronger phonon tunneling","Layer-breathing phonon bridges gap in twisted graphene"],"cache_read_input_tokens":21888,"weakest_assumption_plain":"The central comparison assumes that the WSe2 barrier thickness and interface quality are essentially identical across separately fabricated Bernal and twisted bilayer devices, since barrier thickness, junction area, and capacitance are not reported and the conclusion rests on one Bernal device and two twisted devices.","fun_headline_variants_meta":{"raw":{"variants":["Twist enhances graphene tunnel current via breathing phonon","Moiré twist lets breathing phonon boost tunneling","Twist enables phonon-assisted tunneling boost in graphene","Twisted bilayer graphene shows stronger phonon tunneling","Layer-breathing phonon bridges gap in twisted graphene"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000647,"raw_usage":{"total_tokens":2944,"prompt_tokens":887,"completion_tokens":2057,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":503,"completion_tokens_details":{"reasoning_tokens":1982}},"tokens_in":503,"tokens_out":2057,"duration_ms":16760,"temperature":1.0,"reasoning_tokens":1982,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:34:22.649675+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fabricate a single chip with adjacent Bernal and twisted junctions sharing the same WSe2 flake, then normalize conductances by junction area and barrier thickness; if the twisted conductance is not still about an order of magnitude higher, the twist-based explanation collapses. Alternatively, vary the twist angle continuously from zero to two degrees and check that the zero-bias gap interpolates monotonically from about 30 to 40 mV down to 5 to 10 mV as the moiré Brillouin zone shrinks.","supporting_citations":[{"cited_title":"Tunnelling from a Many-Particle Point of View","cited_arxiv_id":null,"evidence_quote":"Supplies Bardeen's tunneling Hamiltonian that is the starting point for both the elastic and inelastic current expressions."},{"cited_title":"O.; Grigorenko, I.; Lichtenstein, A","cited_arxiv_id":null,"evidence_quote":"Provides the theory of phonon-mediated tunneling into graphene that gives the inelastic channel and the role of the phonon momentum q."},{"cited_title":"O.; Kr\\\"oger, J","cited_arxiv_id":null,"evidence_quote":"Demonstrates inelastic electron tunneling into graphene nanostructures on metal surfaces, establishing the momentum-matching role of phonons."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports planar metal-hBN-graphene junctions with 30 to 50 mV zero-bias gaps, the baseline trend from which the twisted bilayer data deviate."},{"cited_title":"Understanding and Engineering Phonon-Mediated Tunneling into Graphene on Metal Surfaces","cited_arxiv_id":null,"evidence_quote":"Shows phonon-mediated tunneling into graphene on metal surfaces and is used to discuss why ZA phonons are not responsible for the observed features."},{"cited_title":"Monolayer and Bilayer Graphene on Ru(0001): Layer-Specific and Moiré-Site-Dependent Phonon Excitations","cited_arxiv_id":null,"evidence_quote":"Provides the BLG-on-Ru comparison used to argue that ZA modes do not explain the low-energy tunneling in free-standing bilayer graphene."},{"cited_title":"R.; Jain, M","cited_arxiv_id":null,"evidence_quote":"Supplies the phonon dispersions and density of states used to identify the 10 meV layer breathing mode in both Bernal and twisted bilayers."}],"review_version":1}