{"id":"d46f2a7f-9b48-4a91-9c12-1bfcd349be1c","arxiv_id":"1908.01302","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"First-principles calculations predict that single-layer T-graphene and its potassium intercalate C4K are phonon-mediated superconductors with Tc of about 20.8 K and 30.4 K, respectively.","lead":"This paper predicts that a single-layer carbon sheet called T-graphene is an intrinsic superconductor with a transition temperature around 20.8 K, and proposes a high-pressure synthesis plus exfoliation route to make it. The same calculations give a potassium-intercalated version, C4K, a predicted transition temperature of 30.4 K, which would be the highest among layered carbon-based superconductors.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"T-graphene's λ=1.23 and Tc≈20.8 K rest on acoustic-phonon EPC computed with a single 8×8×1 q-mesh; no convergence check is reported for the dominant low-frequency contribution.","rationale":"The reader's weakest-assumption analysis targeted μ*=0.1, which is a general parametric uncertainty in Allen-Dynes McMillan theory. While valid, it is not the most load-bearing issue for this particular calculation: the paper's own Fig. 5(a) attributes 78.3% of λ to low-frequency acoustic modes with a soft out-of-plane branch near 160 cm−1, and the Methods section specifies only an 8×8×1 q-mesh for the phonon dynamical matrix. In 2D systems, the ZA branch's contribution to λ is concentrated near small q, and coarse q sampling can substantially misestimate the integral. This is a direct numerical-convergence concern rather than a modeling-assumption concern, and it can be settled definitively by denser q-mesh calculations. If the acoustic contribution changes with q-mesh, the headline Tc changes, regardless of μ*. The paper shows careful convergence checks for structural enthalpies (1050 eV cutoff, dense k-mesh) but none for the EPC λ/Tc, creating an internal asymmetry. I therefore view this as the single most load-bearing technical uncertainty. Since the reader's verdict was already CONDITIONAL with MODERATE confidence, my independent concern supports that verdict without moving it to ACCEPT or REJECT; the paper should add a q-mesh convergence check and ideally a μ* sensitivity scan. The synthesis-route claims are more speculative but are secondary to the superconductivity prediction and are already framed as proposals.","tokens_in":11526,"tokens_out":3775,"duration_ms":38273,"concrete_test":"Recompute electron-phonon coupling for single-layer T-graphene within DFPT using q meshes of 8×8×1, 16×16×1, 24×24×1, and 32×32×1 (with correspondingly scaled k meshes, e.g., 32×32×2 → 64×64×4), and compare the integrated λ, the low-frequency contribution from ω≤175 cm−1, and Allen-Dynes Tc at fixed μ*=0.1. If λ shifts by more than 10% or Tc by more than ~2 K between the two densest meshes, the claimed 20.8 K is not numerically converged; if λ stabilizes below ~1.1, the headline Tc should be revised downward accordingly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central Tc prediction for single-layer T-graphene depends on λ≈1.23, of which 78.3% comes from three acoustic modes below 175 cm−1 (Fig. 5(a)). In a free-standing 2D sheet the out-of-plane ZA branch has quadratic dispersion, and its EPC contribution to λ = 2∫α²F(ω)/ω dω is heavily weighted at small q. The DFPT calculation (Methods) uses only an 8×8×1 q-point mesh for the dynamical matrix and a 32×32×2 k mesh for linewidth integration. Such a coarse phonon grid is likely to undersample the low-frequency acoustic region where α²F is largest, so both λ and the resulting Allen-Dynes Tc could shift appreciably with denser q sampling. The paper reports no convergence test of λ or Tc against q-mesh size, unlike the careful convergence checks done for the formation enthalpies. Because the headline 'intrinsic 2D carbon superconductor with Tc≈20.8 K' relies directly on this λ value, the absence of a q-mesh convergence check is the most load-bearing uncertainty; μ*-sensitivity is secondary.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript predicts, using a machine-learning-accelerated crystal-structure search and first-principles density-functional calculations, that a planar single-layer carbon sheet with four- and eight-membered rings (T-graphene) is an intrinsic elemental superconductor with a critical temperature of about 20.8 K at 0 GPa. It further predicts a potassium-intercalated compound C4K (P4/mmm) that becomes thermodynamically stable above roughly 11.5 GPa, is quenchable to ambient pressure, and has a calculated Tc of about 30.4 K, which would be the highest among known layered carbon-based superconductors if confirmed. The authors propose two synthesis routes to single-layer T-graphene: electrochemical exfoliation from C4K, or evaporation of potassium from C4K to form bulk T-graphite C4 followed by mechanical exfoliation. The stability claims are supported by formation-enthalpy calculations with two vdW functionals benchmarked against graphite and potassium, phonon spectra at 0 and 20 GPa, AIMD simulations up to 1000 K, and exfoliation-energy calculations benchmarked against graphene. Superconducting parameters are obtained from DFPT electron-phonon coupling and the Allen-Dynes modified McMillan equation with a fixed screened Coulomb parameter μ* = 0.1.","tokens_in":11725,"tokens_out":13174,"duration_ms":115346,"significance":"If the predictions hold, this would be the first intrinsic single-layer elemental carbon superconductor and would add a concrete high-pressure synthesis route to a 2D superconductor, making the paper of substantial interest to the 2D-materials and superconductivity communities. The workflow has genuine strengths: the formation enthalpies are computed with two independent vdW functionals that are calibrated against measured lattice constants and phase stability of graphite and potassium; the exfoliation energy of graphene (24.1 meV/Ų) reproduces the experimental value; the C4K compound is checked by phonons at two pressures and by AIMD at 300 and 1000 K; and the Tc values are derived from first-principles λ and ω_log rather than fitted to target data, with μ* = 0.1 as the only semi-empirical input. The principal weaknesses are the absence of convergence tests for the electron-phonon quantities and the lack of sensitivity analysis for μ*, both of which bear directly on the headline Tc values.","major_comments":[{"comment":"The electron-phonon calculation for single-layer T-graphene uses an 8×8×1 q-point mesh for the dynamical matrix and a 32×32×2 k-mesh for the linewidth integration, and no convergence test of λ, ω_log, or Tc versus q-mesh is reported. The paper states that 78.3% of λ ≈ 1.23 comes from three acoustic modes below 175 cm⁻¹, and since λ = 2∫α²F(ω)/ω dω and the Allen-Dynes formula weight the low-frequency region heavily, the result is particularly sensitive to the q-point sampling near the zone center. In a free-standing 2D sheet the out-of-plane ZA branch has quadratic dispersion, so its α²F contribution is concentrated at small q and can be severely undersampled by an 8×8×1 mesh. This is the most load-bearing uncertainty for the headline claim Tc ≈ 20.8 K; the authors should report λ, ω_log, and Tc for at least two finer q-meshes (for example 16×16×1 and 24×24×1, with correspondingly denser k-meshes) and demonstrate that the acoustic contribution is converged. The same request applies, with less urgency, to C4K, for which only an 8×8×8 q-mesh is used.","section":"Methods (EPC paragraph), Fig. 5(a)"},{"comment":"The Tc values of both T-graphene (around 20.8 K) and C4K (around 30.4 K) are quoted for a single value of the screened Coulomb parameter μ* = 0.1, with no sensitivity analysis. For λ ≈ 1.23 the Allen-Dynes formula has a strong nonlinear dependence on μ*, and modest changes in μ* translate into shifts of several kelvin; the abstract's phrasing 'Tc up to around 20.8 K' implies an upper bound that the present data do not establish. The authors should report Tc for a small set of μ* values (for example 0.10, 0.12, and 0.15) for both systems and discuss the resulting uncertainty, or justify a narrower range from the materials in question.","section":"Page 4, Allen-Dynes modified McMillan equation paragraph"},{"comment":"The proposed electrochemical exfoliation route from C4K rests on an extrapolation from graphite intercalation compounds, but the computed exfoliation energy of T-graphene from C4K is 85.8 meV/Ų, which is above the 30–35 meV/Ų 'easily exfoliable' range cited by the authors themselves and 3.5 times the computed graphene value of 24.1 meV/Ų. While 85.8 meV/Ų is below the generic 130 meV/Ų threshold for 'potentially exfoliable', the statement that T-graphene 'can very likely be exfoliated from bulk C4K' is stronger than this number supports. The authors should discuss what the higher exfoliation energy implies for the electrochemical route and temper the claim accordingly, or provide additional evidence (for example, the effect of intercalation on the interlayer binding).","section":"Page 3, exfoliation paragraph"}],"minor_comments":[{"comment":"The claim that C4K 'sets a new record for layered carbon-based superconductors' refers to a calculated value; it should be rephrased as 'the highest predicted Tc among layered carbon-based superconductors' to avoid confusion with experimental records.","section":"Abstract and conclusions"},{"comment":"The phrases 'haven been widely used' and 'monolayerT -graphite' are typos for 'have been widely used' and 'monolayer T-graphene'.","section":"Pages 2 and 7"},{"comment":"The phrase 'The calculate electronic structures' should read 'The calculated electronic structures'.","section":"Page 3"},{"comment":"The phrase 'Figis. 4 (c) and (d)' should read 'Figs. 4(c) and (d)'.","section":"Page 4"},{"comment":"The word 'perturbated' appears three times and should be 'perturbed'.","section":"Page 4"},{"comment":"The name 'NoseHoover' should be 'Nosé–Hoover', and 'core radium' should be 'core radius'.","section":"Methods"},{"comment":"The method used to compute the charge transfer (presumably Bader analysis) is not stated and should be specified in the caption.","section":"Table I caption"},{"comment":"The manuscript references many Supplemental Material items (Fig. S1–S10 and Table S1) that were not available in the reviewed version; the authors should ensure the SM is complete and explicitly include the EPC convergence tests requested above in the main text or SM.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The q-mesh convergence issue raised by the stress-test is, on my reading, real and is the primary reason for requesting major revision; the μ*-sensitivity issue is secondary but should also be addressed. I would note for the editor that the structure-search method is cited as a self-reference [42] and is treated as a black box, but the central claims are validated by independent formation-enthalpy and dynamics calculations, so I do not see a circularity problem. The version I reviewed (arXiv v1) does not include the Supplemental Material, so the requested convergence checks may already exist there; if so, they should be reported in the main text explicitly, since the current manuscript gives no indication of them. Finally, the paper is prediction-only, and the 'new record' language for an unmeasured Tc may need toning down for the final version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know: this is a credible, mostly well-executed first-principles prediction of intrinsic superconductivity in single-layer T-graphene (Tc ~ 20.8 K) and in the potassium-intercalated C4K (Tc ~ 30.4 K). The new physics is real: prior work on T-graphene considered structure, Dirac fermions, mechanics, and hydrogen storage, not superconductivity; the C4K compound and the two-step synthesis route are also new. The authors did the standard stability homework — formation enthalpies with two vdW functionals, phonons at 0 and 20 GPa, AIMD at 300 and 1000 K, and exfoliation energies benchmarked against graphene. The EPC analysis traces the large lambda to soft out-of-plane acoustic modes, and the frozen-phonon band comparison shows they couple to the Fermi surface. No fitted parameters beyond mu* = 0.1, and the structure-search tool is used as a generator, not as an input to the result. I don't see circularity.\n\nThe main soft spot is the one the stress-test flags: the lambda = 1.23 for T-graphene gets 78% of its weight from acoustic modes below 175 cm^-1, and the DFPT QE calculation uses only an 8x8x1 q-mesh for the phonons. No convergence test of lambda or Tc against q-mesh is reported. For a 2D sheet with a quadratic ZA branch, that small-q region is exactly where alpha2F/omega can be under-resolved. This is a legitimate concern and the single most load-bearing uncertainty in the paper. The fixed mu* = 0.1 is secondary but worth a sensitivity table; going to 0.15 can cut Allen-Dynes Tc by a meaningful amount. The 'record' claim for C4K should be hedged to 'among computed layered carbon superconductors' or similar, since it is a prediction, not a measurement. The electrochemical exfoliation of T-graphene from C4K is speculative, but the exfoliation energies are in the exfoliable range and the authors cite analogous GIC exfoliation, so I don't think it's fatal. I also couldn't find the SM or code; if the journal can't make them available, that's a minor reproducibility issue.\n\nIf I were editing: send to a serious referee, but ask for a q-mesh convergence check and mu* sensitivity before acceptance. The paper is for the 2D superconductivity and computational materials community, and it will be a useful data point even if the Tc shifts. I'd bring it to reading group as a 'maybe', and I wouldn't cite the headline number in my own work until the convergence is shown.","headline":"Solid computational prediction of a new 2D carbon superconductor, but the headline Tc needs a q-mesh convergence check before I'd trust the number.","tokens_in":12303,"tokens_out":2962,"would_cite":false,"duration_ms":28290,"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":"The paper predicts that a single carbon sheet called T-graphene is an intrinsic superconductor at about 20.8 K.","keywords":["T-graphene","single-layer superconductor","carbon allotrope","electron-phonon coupling","high-pressure synthesis","graphite intercalation compound","Allen-Dynes McMillan equation","2D materials"],"falsifier":"Recompute the electron-phonon coupling for T-graphene and C$_4$K with $\\mu^*$ swept over the range 0.1 to 0.15, or solve the full Migdal-Eliashberg equations, and check whether $T_c$ stays in the 20-30 K range; alternatively, synthesize C$_4$K above 11.5 GPa, quench it, and look for a superconducting transition near 30 K, then exfoliate and look for the monolayer transition near 20 K.","tokens_in":11277,"feed_emoji":"⚛️","tokens_out":8137,"duration_ms":67707,"temperature":0.7,"pith_summary":"The paper predicts that a single flat layer of carbon arranged in four- and eight-membered rings, named T-graphene, is an intrinsic elemental superconductor with $T_c$ around 20.8 K at ambient pressure, needing no doping, strain, or twisting. It further predicts that the layered potassium compound C$_4$K has $T_c$ near 30.4 K, and lays out a synthesis route: make C$_4$K above 11.5 GPa, quench it to ambient conditions, then obtain T-graphene by electrochemical exfoliation or by peeling from bulk C$_4$ after evaporating the potassium. If these predictions hold, this would be the first intrinsic single-layer elemental superconductor made of carbon and a new record $T_c$ among layered carbon-based superconductors.","feed_headline":"Pure carbon sheet predicted to superconduct at 20.8 K","feed_subtitle":"Calculations identify flat T-graphene as an intrinsic 2D superconductor and map a high-pressure route to make it.","key_machinery":"The load-bearing mechanism is the electron-phonon coupling of the low-frequency out-of-plane carbon vibrations, computed with density-functional perturbation theory and converted to a $T_c$ estimate through the Allen-Dynes modified McMillan equation with a screened Coulomb parameter $\\mu^*=0.1$. In T-graphene, almost 78% of the total coupling constant $\\lambda$ comes from the three low-frequency acoustic modes with $\\omega \\le 175$ cm$^{-1}$, and the softest of these has the largest phonon linewidth and the highest peak in the Eliashberg spectral function. In C$_4$K, the analogous out-of-plane $E_g$ mode is the key pairing channel, and the cylinder-like, nested Fermi surfaces of both systems give the 2D electronic character that sustains the coupling.","core_discovery":"On the paper's own terms, the central discovery is that T-graphene, the planar carbon sheet with 4- and 8-membered rings, is a phonon-mediated superconductor in its undoped elemental form, with electron-phonon coupling constant $\\lambda\\approx 1.23$ and an Allen-Dynes $T_c$ of about 20.8 K. In C$_4$K, potassium intercalation pushes the Fermi level up, increases occupation of the $\\pi$ bands, and raises the calculated $T_c$ to about 30.4 K. The pairing is carried by low-frequency out-of-plane carbon vibrations: in T-graphene the softest acoustic mode (the $A_1$ mode at about 160 cm$^{-1}$) dominates the electron-phonon spectral function, while in C$_4$K the out-of-plane $E_g$ mode and potassium-related vibrations produce the main coupling peaks near 300 and 450 cm$^{-1}$.","pith_inferences":["A direct testable extension would be a systematic sweep of electron filling per carbon atom, which the paper's mechanism suggests should tune $T_c$ smoothly; identifying the filling that maximizes $\\lambda$ would give a simpler synthesis target than the full C$_4$K route.","Substrate effects could cut either way: a substrate that softens the out-of-plane acoustic mode would raise $T_c$, while one that stiffens it would suppress superconductivity, so the measured value on a given support may differ substantially from the freestanding prediction.","The specific 20.8 K and 30.4 K numbers are less secure than the qualitative claim that this carbon lattice superconducts, since $\\mu^*$ is fixed at 0.1 without a sensitivity check; the discovery would still stand if a fuller Eliashberg treatment changed the numbers by a few kelvin."],"forward_implications":["A freestanding T-graphene monolayer would join borophene as one of very few intrinsic elemental single-layer superconductors, without charge doping, strain, or a twist between layers.","The layered compound C$_4$K would set a new ambient-pressure record for carbon-based layered superconductors, with $T_c$ more than an order of magnitude above that of C$_8$K.","The proposed route makes T-graphene a practical building block: high-pressure synthesis of C$_4$K, quenching, potassium evaporation to C$_4$, then exfoliation, with exfoliation energies below the thresholds used for known 2D materials.","Because higher $\\pi$-band occupation correlates with higher $T_c$, charge doping of T-graphene by metallic atoms or gate voltage is expected to raise its transition temperature beyond 20.8 K."],"supporting_citations":[{"why":"supplies the machine-learning-accelerated crystal structure search that identifies P4/mmm C4K as the stable phase in the C-K system at 20 GPa.","marker":"[42]"},{"why":"provides the Allen-Dynes modified McMillan equation used to turn the computed electron-phonon coupling into the Tc estimates.","marker":"[55]"},{"why":"is the density-functional perturbation theory implementation used for phonon dispersions, Eliashberg functions, and coupling constants.","marker":"[66]"},{"why":"the optB88-vdW functional that gives lattice constants and formation enthalpies used to establish C4K stability against C+K and C8K+K.","marker":"[40]"},{"why":"the optPBE-vdW functional used as a second vdW treatment confirming the stability conclusions.","marker":"[41]"},{"why":"provides the Na4Si24-to-Si24 potassium-evaporation analog used to argue bulk T-graphite C4 can be obtained from C4K.","marker":"[46]"},{"why":"supplies the exfoliation-energy thresholds used to judge that T-graphene can be peeled from C4K and from bulk C4.","marker":"[49]"},{"why":"shows electrochemical exfoliation works for alkali graphite intercalation compounds, the basis of the proposed C4K exfoliation route.","marker":"[50]"},{"why":"establishes borophene as the prior intrinsic single-layer elemental superconductor that T-graphene would join or surpass.","marker":"[8]"}],"fun_headline_variants":["Pure carbon T-graphene superconducts at 20.8 K","20.8 K Tc predicted in single-layer carbon T-graphene","Synthesis route found for superconducting T-graphene","T-graphene: 20.8 K superconductor with a make-it route"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The headline $T_c$ values rest on setting the screened Coulomb parameter $\\mu^*=0.1$ in the Allen-Dynes equation, and the paper gives no sensitivity analysis; if the true $\\mu^*$ for these 2D carbon systems is higher, the estimated $T_c$ would fall, possibly below the claimed regime.","fun_headline_variants_meta":{"raw":{"variants":["Pure carbon T-graphene superconducts at 20.8 K","20.8 K Tc predicted in single-layer carbon T-graphene","Synthesis route found for superconducting T-graphene","T-graphene: 20.8 K superconductor with a make-it route"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000837,"raw_usage":{"total_tokens":3696,"prompt_tokens":1035,"completion_tokens":2661,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":651,"completion_tokens_details":{"reasoning_tokens":2582}},"tokens_in":651,"tokens_out":2661,"duration_ms":21049,"temperature":1.0,"reasoning_tokens":2582,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:17:23.703969+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the electron-phonon coupling for T-graphene and C$_4$K with $\\mu^*$ swept over the range 0.1 to 0.15, or solve the full Migdal-Eliashberg equations, and check whether $T_c$ stays in the 20-30 K range; alternatively, synthesize C$_4$K above 11.5 GPa, quench it, and look for a superconducting transition near 30 K, then exfoliate and look for the monolayer transition near 20 K.","supporting_citations":[{"cited_title":"Klimeˇ s, D","cited_arxiv_id":null,"evidence_quote":"supplies the machine-learning-accelerated crystal structure search that identifies P4/mmm C4K as the stable phase in the C-K system at 20 GPa."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"is the density-functional perturbation theory implementation used for phonon dispersions, Eliashberg functions, and coupling constants."},{"cited_title":"Zhang, A","cited_arxiv_id":null,"evidence_quote":"the optB88-vdW functional that gives lattice constants and formation enthalpies used to establish C4K stability against C+K and C8K+K."},{"cited_title":"Zhang, Y","cited_arxiv_id":null,"evidence_quote":"the optPBE-vdW functional used as a second vdW treatment confirming the stability conclusions."},{"cited_title":"Zacharia, H","cited_arxiv_id":null,"evidence_quote":"supplies the exfoliation-energy thresholds used to judge that T-graphene can be peeled from C4K and from bulk C4."},{"cited_title":"Ziambaras, J","cited_arxiv_id":null,"evidence_quote":"shows electrochemical exfoliation works for alkali graphite intercalation compounds, the basis of the proposed C4K exfoliation route."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"establishes borophene as the prior intrinsic single-layer elemental superconductor that T-graphene would join or surpass."}],"review_version":1}