{"id":"a04edae3-609b-4790-9216-f23221adfcbb","arxiv_id":"1908.06582","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A circular p-p junction in graphene confines Dirac fermions into whispering-gallery modes, shows a Berry-phase-induced level jump under magnetic field, and exhibits a split quasi-bound state near the Fermi level.","lead":"Using a scanning tunneling microscope, this paper shows that a circular graphene p-p junction traps electrons into whispering-gallery modes and exhibits a Berry-phase energy jump in a magnetic field. It also reports a quasi-bound state that splits into two peaks near the Fermi level, suggesting electron-electron interactions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 26 meV splitting near the Fermi level is the load-bearing novel claim, but the manuscript explicitly disclaims knowing its origin; the abstract upgrades an uncertain observation into a statement that strong electron-electron interactions are present.","rationale":"The reader's weakest assumption identifies exactly the same load-bearing concern: the attribution of the 26 meV splitting to electron-electron interactions. I agree that this is the weakest point because the authors explicitly disclaim knowledge of the splitting's origin and provide only one spectrum plus a crude energy-scale estimate. The WGM and Berry-phase results are supported by comparison with prior experiments and by lattice Green's function simulations, so they are not the principal vulnerability. I considered the absence of a magnetic-field sweep for the Berry-phase jump, which is a secondary weakness, but the interaction claim is more consequential because it is the only substantially new physics and its public statement exceeds the evidence. The proposed test—field/temperature dependence plus spatial maps—would distinguish an interaction-induced charging doublet from a single-particle near-degeneracy, and a non-interacting simulation with realistic disorder would directly test whether the splitting requires interactions at all. Since the reader already assigned a CONDITIONAL verdict based on this concern, my assessment does not change the verdict.","tokens_in":7583,"tokens_out":7358,"duration_ms":85352,"concrete_test":"Re-measure the quasi-bound state in Fig. 4(a) at B = 0, 2, 4, 6, and 8 T and at T = 4 K, and acquire dI/dV maps at the two peak energies. If the splitting is an interaction-induced charging effect, both peaks should share the same spatial LDOS pattern and their energy separation should remain approximately constant as B is varied; if the two peaks are distinct single-particle states, their spatial maps or B-dependence will differ. In parallel, run the lattice Green's function calculation with a slightly asymmetric or rough boundary and no interaction term: if it already produces two peaks separated by ~26 meV from a nominally degenerate level, the interaction interpretation is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most novel part of the paper is the reported splitting of a quasi-bound state into two peaks separated by ~26 meV when the Fermi level crosses it, interpreted as evidence for strong electron-electron interactions (Fig. 4 and abstract). This is load-bearing because the WGM confinement and Berry-phase jump in p-p junctions are incremental extensions of earlier p-n-junction studies (Refs. [7,10,14,15]), whereas the interaction-induced splitting is the genuinely new claim. The manuscript itself states: 'At present, we do not know the exact reason of the splitting.' The quantitative support is only an order-of-magnitude comparison with e2/(4πεR) ~ 30 meV. No gate, temperature, magnetic-field, or device-to-device dependence is shown, and no single-particle calculation with realistic boundary asymmetry or disorder is provided to exclude an accidental near-degeneracy of two distinct quasi-bound states, e.g., valley- or angular-momentum-split levels. Such a single-particle doublet would produce the same two-peak STS signature without invoking interactions. The abstract's wording 'indicating that there are strong electron-electron interactions' is also stronger than the main text's 'may play a vital role,' so the public claim exceeds the evidence presented.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports low-temperature STM/STS measurements of a circular graphene p-p junction resonator on a Cu substrate, with radius about 13 nm. The authors observe a series of quasi-bound states inside the junction, with an average energy spacing of about 48 meV and spatial distributions characteristic of whispering-gallery-mode (WGM) confinement. They support these observations with lattice Green's function simulations that use measured junction parameters. In an 8 T magnetic field, they report a jump in the quasi-bound-state energy of roughly half the level spacing, which they attribute to the magnetic-field-induced switching on of the π Berry phase. Finally, they report that a quasi-bound state lying near the Fermi level splits into two peaks separated by about 26 meV when partially filled, and they interpret this as evidence for strong electron-electron interactions in the quantum dot.","tokens_in":7736,"tokens_out":4070,"duration_ms":47431,"significance":"The WGM confinement and Berry-phase jump are incremental extensions of earlier studies of circular p-n junctions to p-p junctions, and the lattice Green's function simulation uses measured inputs rather than fitting parameters, which is a strength. The genuinely new claim is the observation of a ~26 meV splitting of a quasi-bound state near the Fermi level, presented as evidence for electron-electron interactions. If that claim were firmly established, it would be of considerable interest for the physics of graphene quantum dots. However, as presented, the splitting evidence is limited to one device, one spectrum, and no control experiments, and the manuscript itself states that the exact origin of the splitting is unknown; the abstract's wording is therefore stronger than the data and analysis support.","major_comments":[{"comment":"The load-bearing novel claim is the ~26 meV splitting of a quasi-bound state near the Fermi level and its attribution to strong electron-electron interactions. The manuscript states, 'At present, we do not know the exact reason of the splitting,' and the quantitative support is only an order-of-magnitude comparison with e^2/(4πεR) ~ 30 meV. No gate-voltage dependence, temperature dependence, magnetic-field dependence, or device-to-device reproducibility is shown, and no single-particle calculation with realistic boundary asymmetry or disorder is provided to exclude an accidental near-degeneracy of two distinct quasi-bound states, such as valley- or angular-momentum-split levels. Such a single-particle doublet would produce the same two-peak STS signature without invoking interactions. The abstract's statement that the splitting 'indicat[es] that there are strong electron-electron interactions' is stronger than the main text's 'may play a vital role' and is not justified by the evidence presented. The authors should either supply additional measurements that distinguish interaction effects from single-particle splitting, or temper the abstract and conclusions accordingly.","section":"Section 4 (Fig. 4) and Abstract"},{"comment":"The Berry-phase-jump claim rests on a single pair of tunneling spectra measured at one location in one device, at 0 T and 8 T. The critical field B_C is estimated theoretically as about 4 T, but no field-dependent series is shown, so the assignment of the 0 T and 8 T peaks to the same quasi-bound states before and after the Berry-phase switch is not uniquely established. The 8 T spectrum also contains additional weak peaks attributed to Landau levels outside the GQD, which could complicate peak identification. Additional spectra at intermediate fields or spatial maps at 8 T would strengthen this central claim.","section":"Section 3 (Fig. 3)"},{"comment":"The experimental support for the WGM confinement and the Berry-phase jump comes from a single graphene quantum dot (one device). The paper says 'we systematically study a graphene quantum dot,' but the statistics are limited to one GQD, one magnetic-field pair, and one splitting observation. While single-device STM studies are common, the generality of the conclusions would be materially improved by showing data from at least one more device, or by explicitly stating the limitations of single-device statistics in the text.","section":"Section 2 (Fig. 1) and Section 3 (Fig. 3)"}],"minor_comments":[{"comment":"The affiliation contains a typo: 'Mocrosystem' should be 'Microsystem.'","section":"Affiliation line"},{"comment":"The abstract says the splitting indicates 'strong electron-electron interactions,' while the conclusion says electron-electron interactions 'may play an important role' and 'further experiments should be carried out.' These statements should be consistent; the more cautious wording is more appropriate to the evidence.","section":"Abstract and Conclusion"},{"comment":"The energy separation is described as ~26 meV with peaks at -13 mV and +13 mV. It would be helpful to state explicitly whether these energies are referenced to the Fermi level (zero bias) and whether the symmetric placement is a selection criterion or an outcome of the measurement.","section":"Section 4 (Fig. 4)"},{"comment":"Reference [27] is an arXiv preprint (arXiv:1904.06902); if the paper has been published by the time of revision, the published citation should be provided. Several other references are also preprints; the journal style should be followed.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a competent extension of the authors' and others' earlier work on p-n junction resonators, and the simulation approach is a strength. The main unresolved issue is that the most novel claim, the interaction-induced splitting, is explicitly acknowledged in the text to be of unknown origin, while the abstract presents it as established evidence for strong electron-electron interactions. This asymmetry should be addressed before publication. The editor may also wish to consider whether the incremental nature of the WGM and Berry-phase results, relative to Refs. [7,10,14,15], warrants the current framing."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the WGM confinement and Berry-phase jump in a p-p junction are real but incremental: they extend earlier p-n junction results to p-p, which is new but not conceptually surprising given the same Klein-tunneling mechanism. Second, the genuinely new claim is the 26 meV splitting of a quasi-bound state at the Fermi level, and that claim is not supported as strongly as the abstract suggests. The main text explicitly says \"we do not know the exact reason of the splitting,\" and the abstract upgrades that to \"strong electron-electron interactions.\" That gap is the paper's biggest problem.\n\nWhat the paper does well: the STM/STS data are consistent. The Dirac point offsets are measured via Landau levels, the quasi-bound-state spacing (~48 meV) matches hbar v_F/R, the spatial maps show ring structures, and the lattice Green's function simulation with measured parameters reproduces the main features without fitting to the target results. For the p-p WGM and Berry-phase parts, this is a competent, credible experimental paper.\n\nSoft spots, in proportion. Statistics are thin: one quantum dot, one pair of spectra for the 8 T jump, and a single split peak. The splitting lacks magnetic-field, temperature, gate, or device-to-device dependence; without that, an accidental near-degeneracy of two quasi-bound states (valley or angular-momentum doublet, boundary asymmetry, disorder) is not excluded. The e^2/(4πεR) ~ 30 meV estimate is order-of-magnitude and not a test. No code or data are shipped, so the simulation cannot be checked independently. These issues are not fatal to the basic observations, but they are fatal to the interaction claim as stated. The citation pattern is fine; the self-cited Green's function method is used with measured inputs, not to fit the target results.\n\nWho this is for: anyone working on graphene quantum dots, Klein tunneling, or STM of confined Dirac states. It is a useful data point and deserves a serious referee. If I were the editor, I would send it out, with a clear request: soften the abstract and either provide more evidence for the interaction interpretation or present the splitting as an unexplained experimental observation.","headline":"Solid p-p WGM/Berry-phase extension, but the 26 meV splitting is overinterpreted as electron-electron interactions.","tokens_in":8355,"tokens_out":2660,"would_cite":true,"duration_ms":27204,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A circular graphene p-p junction confines Dirac fermions into whispering-gallery states, and a strong magnetic field flips their Berry phase.","keywords":["graphene quantum dot","p-p junction","whispering-gallery modes","Berry phase","electron-electron interaction","scanning tunneling microscopy","Dirac fermions","Klein tunneling"],"falsifier":"Measure the split-peak separation in circular p-p graphene quantum dots of different radii: if Coulomb interactions cause the 26 meV splitting, the separation should scale roughly as $1/R$, giving about 40 meV for $R=10$ nm and 20 meV for $R=20$ nm, whereas a constant or random separation would point to a tip artifact or accidental degeneracy.","tokens_in":7334,"feed_emoji":"🔬","tokens_out":9147,"duration_ms":90850,"temperature":0.7,"pith_summary":"This paper uses a low-temperature scanning tunneling microscope to study a graphene quantum dot formed by a circular p-p junction, a boundary where both sides are hole-doped but with different Dirac-point energies. It claims that this junction confines massless Dirac fermions into whispering-gallery quasi-bound states, exactly as p-n junction resonators do, and that these states can be imaged on the atomic scale. It further claims that a strong magnetic field switches on a $\\pi$ Berry phase for the confined fermions, producing a sudden energy jump of about half the level spacing. It also reports a 26 meV splitting of a quasi-bound state when the Fermi level crosses it, which it interprets as a sign of strong electron-electron interactions. If true, the result extends Klein-tunneling-based quantum confinement to same-polarity junctions and exposes interaction effects in a single graphene quantum dot.","feed_headline":"Graphene p-p junction traps electrons in whispering-gallery states","feed_subtitle":"Tunneling spectroscopy reveals a 48 meV level ladder, a Berry-phase jump, and a 26 meV split.","key_machinery":"The central object is a circular graphene p-p junction: a closed boundary between two hole-doped regions with different Dirac-point energies, which acts as a whispering-gallery resonator for massless Dirac fermions. The confinement mechanism is Klein-like anisotropic transmission, where glancing-angle trajectories are repeatedly reflected while near-normal trajectories transmit, so the circular interface traps quasi-bound states. The Berry-phase switch is the momentum-space mechanism: in a magnetic field, orbits with angular momentum antiparallel to the field bend into skipping orbits, so their closed momentum-space paths enclose the Dirac point and acquire a $\\pi$ Berry phase, jumping the level energies. The interaction claim is carried by a single spectral observation: a quasi-bound state at the Fermi level splits into two peaks separated by about 26 meV, comparable to the on-site Coulomb estimate $e^2/(4\\pi\\varepsilon R)\\sim30$ meV.","core_discovery":"The paper reports that a circular graphene p-p junction, a closed boundary between two hole-doped regions with different Dirac-point energies, can confine massless Dirac fermions into quasi-bound states by whispering-gallery-mode reflection. Scanning tunneling spectroscopy inside the dot reveals a ladder of resonances with an average spacing of about 48 meV, matching $\\hbar v_F/R$ for $R\\approx13$ nm, and the lowest state is centered while higher states form rings near the boundary. Applying an 8 T magnetic field shifts this ladder upward by roughly half the level spacing; the authors attribute the sudden jump to the magnetic field turning on a $\\pi$ Berry phase once the field exceeds a critical value of about 4 T. Finally, one quasi-bound state sitting near the Fermi level splits into two peaks separated by about 26 meV, which the authors interpret as evidence that electron-electron interactions are important in the dot.","pith_inferences":["A direct test left open by the paper: measure the splitting in dots with different radii; Coulomb scaling $e^2/(4\\pi\\varepsilon R)\\sim1/R$ would distinguish charging physics from tip artifacts or accidental degeneracy breaking.","The paper compares 0 T and 8 T only; a continuous field sweep across the predicted critical field of about 4 T would sharpen the Berry-phase jump and separate it from gradual orbital shifts.","If partial filling of the level is what triggers the splitting, gating the dot through the Fermi energy should turn the split on and off, turning the resonator into a tunable single-electron interaction device.","The p-p junction geometry, with no carrier-type inversion, may be easier to model with lattice Green's functions and could give cleaner interaction parameters than p-n junction dots."],"forward_implications":["Circular p-p junctions can act as graphene quantum dots even though they do not invert carrier type, extending Klein-tunneling-based confinement beyond p-n junctions.","A magnetic field can abruptly switch the energy ladder of the confined states by about half a level spacing, so the dot behaves as a tunable Berry-phase resonator.","The observed quasi-bound-state spacing of about 48 meV matches $\\hbar v_F/R$, providing a straightforward spectroscopic ruler for the dot radius and Fermi velocity.","A quasi-bound state at the Fermi level splits into two peaks about 26 meV apart, indicating that electron-electron interactions can dominate in these small graphene dots."],"supporting_citations":[{"why":"It demonstrates whispering-gallery-mode confinement in a circular graphene p-n junction, the phenomenon this paper extends to p-p junctions.","marker":"[7]"},{"why":"It provides the imaging method and expected spatial distribution for electrostatically confined Dirac fermions in graphene quantum dots.","marker":"[9]"},{"why":"It establishes the STM/STS platform for atomically sharp graphene junctions and the doping control used to form p-p junctions.","marker":"[12]"},{"why":"It shows the on/off $\\pi$ Berry-phase switch in circular graphene resonators, the effect the paper reproduces in p-p geometry.","marker":"[14]"},{"why":"It reports interaction-driven splitting-like features in graphene quantum dots and motivates the electron-electron interaction interpretation.","marker":"[15]"},{"why":"It provides the lattice Green's function calculation method used to simulate the quasi-bound states and the local density of states.","marker":"[20]"},{"why":"It gives the theoretical Berry-phase jump mechanism for Dirac quantum dots that explains the magnetic-field energy jump.","marker":"[22]"},{"why":"It gives the effective dielectric constant of graphene used to estimate the on-site Coulomb repulsion of about 30 meV.","marker":"[36]"}],"fun_headline_variants":["Graphene p-p junction confines electrons into whispering-gallery modes","Berry phase jump detected in graphene quantum dot at 4 T","Graphene dot reveals interacting electrons via 26 meV splitting","Whispering-gallery resonances and Berry phase in graphene dot","STM maps quasi-bound states in circular graphene p-p junction"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument's weakest link is the assumption that the observed splitting of the confined level is caused by electron-electron repulsion inside the dot, rather than by the microscope tip or some accidental double-peak effect; the paper explicitly says the exact reason is not known.","fun_headline_variants_meta":{"raw":{"variants":["Graphene p-p junction confines electrons into whispering-gallery modes","Berry phase jump detected in graphene quantum dot at 4 T","Graphene dot reveals interacting electrons via 26 meV splitting","Whispering-gallery resonances and Berry phase in graphene dot","STM maps quasi-bound states in circular graphene p-p junction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000197,"raw_usage":{"total_tokens":1335,"prompt_tokens":889,"completion_tokens":446,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":505,"completion_tokens_details":{"reasoning_tokens":358}},"tokens_in":505,"tokens_out":446,"duration_ms":5168,"temperature":1.0,"reasoning_tokens":358,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:39:27.676882+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the split-peak separation in circular p-p graphene quantum dots of different radii: if Coulomb interactions cause the 26 meV splitting, the separation should scale roughly as $1/R$, giving about 40 meV for $R=10$ nm and 20 meV for $R=20$ nm, whereas a constant or random separation would point to a tip artifact or accidental degeneracy.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It demonstrates whispering-gallery-mode confinement in a circular graphene p-n junction, the phenomenon this paper extends to p-p junctions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the imaging method and expected spatial distribution for electrostatically confined Dirac fermions in graphene quantum dots."},{"cited_title":"Bai, J.-J","cited_arxiv_id":null,"evidence_quote":"It establishes the STM/STS platform for atomically sharp graphene junctions and the doping control used to form p-p junctions."},{"cited_title":"Ghahari, D","cited_arxiv_id":null,"evidence_quote":"It shows the on/off $\\pi$ Berry-phase switch in circular graphene resonators, the effect the paper reproduces in p-p geometry."},{"cited_title":"Gutié rrez, D","cited_arxiv_id":null,"evidence_quote":"It reports interaction-driven splitting-like features in graphene quantum dots and motivates the electron-electron interaction interpretation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the lattice Green's function calculation method used to simulate the quasi-bound states and the local density of states."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It gives the theoretical Berry-phase jump mechanism for Dirac quantum dots that explains the magnetic-field energy jump."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It gives the effective dielectric constant of graphene used to estimate the on-site Coulomb repulsion of about 30 meV."}],"review_version":1}