{"id":"5ffe9142-37d6-4e89-b738-b98586c1fc4d","arxiv_id":"1908.05904","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"Cyclo[18]carbon was generated on a surface and imaged at atomic resolution, showing a polyynic ring with alternating single and triple bonds (D9h symmetry).","lead":"Researchers created a ring of 18 carbon atoms on a salt surface and used atomic force microscopy to show the ring alternates between single and triple bonds. The work settles a long-standing debate about the structure of cyclocarbons and opens a route to build new carbon materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The polyynic assignment depends on the bond-order-to-contrast mapping; the supporting AFM simulations rely on a nonstandard HSE exact-exchange mixing (α=0.8) that is the only tested functional yielding a polyynic input.","rationale":"I agree with the reader that this is a landmark experimental result and that the core observation is strong: the ninefold symmetry, the reproducible lobe-to-nonagon evolution with tip height, and the featureless PBE-based simulation as a negative control are all compelling evidence for a bond-order-modulated structure. The single most load-bearing soft spot is the connection between image features and bond order. The experimental image is interpreted through a contrast mechanism that is supported by a simulation whose only polyynic input comes from a nonstandard HSE mixing parameter (0.8). Because the simulation is a forward model, its agreement with experiment does not validate the input structure unless the model is independently calibrated for this system. The reader's conditional verdict is appropriate: the authors should justify the mixing parameter or provide an independent contrast calibration. A standard-functional or CCSD(T) geometry test would settle whether the HSE choice is critical. I therefore recommend no change to the reader's verdict.","tokens_in":22622,"tokens_out":11383,"duration_ms":123600,"concrete_test":"Recompute the gas-phase C18 geometry with a standard hybrid functional (HSE06, α=0.25) or with a high-level wavefunction method (CCSD(T)), then run the same probe-particle AFM simulation (same spring constant, charge, amplitude, and Δz values) with the resulting geometry and Hartree potential. If the simulated constant-height images still show nine lobes at the experimental tip heights, the nonstandard α=0.8 input is not load-bearing. If they are featureless (as in the PBE case), the authors would need to show that an alternative (e.g., surface-distorted cumulenic) structure cannot produce the observed ninefold pattern, or the experimental interpretation loses its computational support.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central inference is that the nine bright lobes in the constant-height AFM images of the decarbonylated product sit above the nine triple bonds of a polyynic C18 (D9h). This mapping is not read directly off the raw image; it is calibrated by the precursors (Fig. 3B–O) and by probe-particle simulations (Fig. 3S,T and fig. S15). Those simulations take as input a DFT geometry, and the DFT section of the SM states that the HSE mixing parameter was set to 0.8, with PBE converging to a cumulenic D18h structure (equal 1.284 Å bonds) and HSE(0.8) to a polyynic one (1.195/1.343 Å). The PBE-based simulation is the only negative control, and it produces a featureless ring. The load-bearing concern is therefore that the computational input that reproduces the experiment is generated by an atypical, hand-set functional, while the experimental image alone—a ninefold arrangement of lobes—does not independently prove bond-length alternation unless the lobe-to-triple-bond assignment is secure. The SM itself notes the molecule is non-planar (1–2° tilt, height differences 0.1–0.3 Å) and that nearby CO molecules can distort constant-height images; these effects are not in the gas-phase, flat-geometry simulations. If the true adsorbed structure were cumulenic but distorted by the surface, or if the lobe assignment failed, the D9h conclusion would not follow.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the on-surface generation of cyclo[18]carbon (C18) from a C24O6 precursor on bilayer NaCl/Cu(111) by tip-induced decarbonylation, and its structural characterization using CO-functionalized atomic force microscopy (AFM). The AFM images show a ninefold symmetric pattern of bright lobes at moderate tip height and a nonagon with corners at the corresponding positions at smaller tip height. The authors interpret these images as direct evidence for a polyynic D9h structure with alternating triple and single bonds, and they support this assignment with probe-particle AFM simulations based on DFT geometries. The manuscript also demonstrates covalent fusion of cyclocarbon oxide intermediates by atom manipulation and reports charge bistability of the neutral and negatively charged molecule. The central claim is that cyclo[18]carbon on bilayer NaCl adopts a polyynic structure, resolving the long-standing polyynic versus cumulenic debate for this molecule.","tokens_in":22937,"tokens_out":6435,"duration_ms":65689,"significance":"If the central claim holds, this is the first structural characterization of a cyclocarbon and a landmark result in molecular carbon allotrope chemistry. The ninefold AFM pattern is a direct observation that distinguishes D9h from D18h symmetry, and the interpretation is supported by the evolution of contrast with tip height and by probe-particle simulations, including a falsifiable negative control in which the cumulenic PBE-based simulation yields a featureless ring. The paper also establishes a new on-surface synthetic route to cyclocarbons and demonstrates the reactivity of strained polyynic moieties, opening avenues for synthesizing other carbon-rich materials. The strength of the evidence is high, and the work is likely to become a reference for experimental cyclocarbon chemistry; however, the supporting simulations rely on a nonstandard DFT exact-exchange mixing parameter that requires scrutiny and justification.","major_comments":[{"comment":"The HSE exact-exchange mixing parameter is set to 0.8, with no justification in the manuscript. Standard HSE uses a mixing parameter of 0.25, and the choice of 0.8 is the only tested functional setting that yields a polyynic geometry for C18; PBE converges to a cumulenic structure with equal bond lengths. Because the AFM simulations that reproduce the experimental ninefold pattern are generated exclusively from the HSE(0.8) geometries, this hand-set parameter is load-bearing for the bond-order-to-contrast assignment. Please provide a rationale for α=0.8, for example by benchmarking against coupled-cluster calculations or experimental data, or show with a sensitivity study that the simulated AFM contrast is robust over a range of α values.","section":"SM1, 'Density functional theory calculations'"},{"comment":"The AFM simulations are performed for flat gas-phase geometries, whereas the experimental molecule is tilted by about 1–2 degrees with intramolecular height differences of 0.1–0.3 Å, as stated in SM1. The authors correctly note that the simulations do not reproduce brightness asymmetries, but the possibility that a surface-distorted cumulenic structure could also give rise to a ninefold lobe pattern is not explicitly ruled out by simulation. A short statement explaining why the fourfold-symmetric NaCl surface cannot generate a ninefold pattern from an 18-fold-symmetric cumulenic molecule, or a test simulation of a slightly distorted cumulenic ring, would close this gap and strengthen the assignment.","section":"Main text, Fig. 3 and SM1, fig. S15"}],"minor_comments":[{"comment":"The labels 'sim. far' and 'sim. close' in the caption are not defined; please specify the corresponding Δz values or state that they are the simulated analogues of the experimental 'AFM far' and 'AFM close' images.","section":"Fig. 3 caption"},{"comment":"The Δz values in the simulated AFM images are described as 'increase in tip-sample distance,' but the reference point (setpoint) is not given; please clarify the definition for consistency with the main text.","section":"SM1, fig. S15"},{"comment":"The reported 13% yield for generating C18 from C24O6 is based on 90 atom-manipulation events, but no statistical uncertainty is provided; please state the number of successful C18 formations or the standard error.","section":"Main text, 'On-surface synthesis' statistics"},{"comment":"The term 'molecular carbon allotrope' may be unfamiliar to some readers; a brief definition or explanatory phrase would improve accessibility, since 'allotrope' is often associated with bulk phases.","section":"Abstract and Introduction"}],"recommendation":"major_revision","confidential_remarks":"This is a landmark experimental paper with a strong direct observation at its core. The main concern is the unjustified HSE mixing parameter (α=0.8) used for the supporting simulations; if the authors can justify this choice or show robustness of the simulated contrast to α, the paper would be suitable for acceptance. In the absence of such justification, the simulation support should be downweighted in the conclusions, and the paper would still merit publication but with more cautious wording."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this is the first structural characterization of a cyclocarbon, and the evidence for the polyynic D9h form of cyclo[18]carbon is direct. The ninefold modulation in the AFM images rules out a cumulenic D18h ring, which would show no bond-order contrast. That settles a debate that has been running since Hoffmann's 1966 paper.\n\nThe genuinely new things are the structural assignment, the reversible charge bistability, and the covalent fusion between cyclocarbon oxides by atom manipulation. Prior gas-phase work (refs 4, 7, 15, 17) had generated C18 and measured electronic spectra, but never resolved the bond-length alternation question. The atom manipulation chemistry is careful: a 13% yield of C18 from 90 events, characterized intermediates, and a proposed mechanism for fusion.\n\nThe paper is honest about its assumptions. The assignment of the nine bright lobes to triple bonds rests on the established AFM contrast mechanism (bond-order discrimination) and is supported by probe-particle simulations. The authors also acknowledge the non-planar adsorption (1–2° tilt) and the proximity of CO molecules that can distort images.\n\nThe soft spot is the computational input to the simulations. HSE was used with an exact-exchange mixing of 0.8, which is not the standard value (usually 0.25) and is not justified in the text. That choice strongly favors the polyynic geometry, while PBE gives a cumulenic ring and, consequently, a featureless simulated image. So the only positive computational support for the ninefold pattern comes from a nonstandard functional. But the experimental observation itself does not depend on the simulation: the ninefold pattern is measured, not fitted. The simulation's role is to interpret the contrast, not to generate the symmetry. I see this as a genuine but minor issue that needs a sentence of justification or a sensitivity test, not a load-bearing flaw.\n\nThe paper deserves a serious referee. The result is important for the carbon allotrope and surface chemistry community, and the data are presented honestly and reproducibly. I would accept it with minor revision. Anyone working on cyclocarbons, polyynes, or on-surface synthesis should read it. I'd bring it to reading group.","headline":"First structural characterization of a cyclocarbon—C18 is polyynic on NaCl; the result is real, with a minor computational caveat.","tokens_in":23496,"tokens_out":2513,"would_cite":true,"duration_ms":23724,"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":"By removing six carbon monoxide groups from a C24O6 precursor with the tip of a scanning probe microscope, the authors produced cyclo[18]carbon on a cold salt surface and show with atomic force microscopy that it is polyynic: a ninefold…","keywords":["cyclo[18]carbon","cyclocarbons","polyynic structure","bond length alternation","atomic force microscopy","CO-functionalized tip","atom manipulation","carbon allotrope"],"falsifier":"Measure the carbon-carbon bond lengths of an isolated C18 molecule by a technique independent of AFM contrast, such as gas-phase electron diffraction or rotational spectroscopy: alternating bonds near 1.20 Å and 1.34 Å would confirm the polyynic assignment, while eighteen equal bonds near 1.28 Å would falsify it.","tokens_in":22437,"feed_emoji":"🔬","tokens_out":11938,"duration_ms":116012,"temperature":0.7,"pith_summary":"The paper reports the first isolation and structural characterization of a cyclocarbon, a long-sought family of carbon allotropes made of rings of two-coordinate carbon atoms. The authors generate cyclo[18]carbon on a bilayer sodium-chloride surface at 5 K by using voltage pulses from a scanning probe tip to strip six carbon monoxide groups from a precursor oxide, C24O6. Their central claim is that high-resolution atomic force microscopy with a carbon-monoxide-functionalized tip shows the ring in a polyynic form: nine alternating triple and single bonds with D9h symmetry, rather than the cumulenic form with eighteen equal bonds. If correct, this settles a long-standing debate in which most density-functional calculations predicted the cumulenic structure while high-level correlated calculations predicted the polyynic one, and it opens a surface route to other carbon allotropes through the demonstrated fusion of cyclocarbon molecules.","feed_headline":"C18 ring alternates triple and single bonds, AFM finds","feed_subtitle":"First direct images of cyclo[18]carbon settle the polyynic-versus-cumulenic debate.","key_machinery":"The load-bearing mechanism is the combination of atom manipulation and bond-order-sensitive AFM contrast. Voltage pulses delivered by the tip remove CO groups from C24O6 in pairs, and a CO-functionalized tip operated at constant height images the resulting ring; bright lobes at moderate tip height and nonagon corners at close approach are assigned to triple bonds because their high electron density shifts the tip's CO group in the probe-particle simulation. The simulation uses gas-phase density-functional geometries as input, and the assignment hinges on the contrast difference between a ninefold polyynic geometry (obtained with a hybrid functional using 80% exact exchange) and an eighteenfold cumulenic geometry (obtained with a semilocal functional), whose simulated image is featureless.","core_discovery":"The central discovery is that cyclo[18]carbon on bilayer NaCl/Cu(111), viewed with a CO-functionalized tip, displays a ninefold-symmetric pattern: at moderate tip height nine bright lobes mark the triple bonds, and at closer tip height the molecule appears as a nonagon whose corners sit at those same positions. Simulated AFM images built from a polyynic gas-phase geometry reproduce the experimental contrast, while simulations built from a cumulenic geometry give only a featureless ring. The authors therefore assign the neutral molecule a polyynic D9h structure with alternating short and long carbon-carbon bonds, and show that the negatively charged molecule distorts into a less symmetric, less planar geometry that can be switched back to the neutral form by changing the sample bias.","pith_inferences":["The same bright-lobe-to-triple-bond mapping should work for other strained sp-carbon scaffolds, so AFM with a CO tip could become a routine way to locate triple bonds in molecules that cannot be crystallized.","Because the hybrid-functional geometry was computed with a hand-set 80% exact-exchange fraction, systematically varying that fraction and re-simulating the AFM images would show how much of the conclusion rests on that parameter choice.","The experiment characterizes the molecule on a weakly interacting insulating film at 5 K; whether gas-phase cyclo[18]carbon is also polyynic, or whether the surface slightly biases the structure, is left open and could be tested by gas-phase spectroscopy or substrate-free calculations.","The demonstrated fusion of cyclocarbon oxides suggests that atom manipulation on insulating films could build larger all-carbon rings and networks one molecule at a time, a bottom-up route that does not rely on gas-phase coalescence."],"forward_implications":["If the assignment is correct, neutral cyclo[18]carbon on NaCl is the first structurally characterized cyclocarbon, and its alternating bond lengths show that electron correlation, not the 4n+2 aromaticity rule, fixes the ground-state geometry.","The tip-induced decarbonylation is a practical on-surface synthesis: C18 forms from C24O6 with 13% yield across the counted manipulation events, and the same tool can fuse nearby cyclocarbon oxides into larger covalently bonded carbon-rich structures.","The reversible charge-state switching between a ninefold neutral geometry and a distorted anionic geometry makes individual C18 molecules addressable single-molecule switches on this surface.","The close match between the measured AFM contrast and the polyynic geometry, together with the failure of the cumulenic simulation, provides a direct experimental test that favors high-level correlated theoretical predictions over conventional density-functional results for this molecule."],"supporting_citations":[{"why":"supplies the CO-tip AFM method that gives the submolecular resolution used to resolve the ring.","marker":"[18]"},{"why":"supplies the probe-particle simulation model used to convert AFM contrast into bond positions.","marker":"[25]"},{"why":"shows that AFM brightness can discriminate bond order, grounding the assignment of bright lobes to triple bonds.","marker":"[27]"},{"why":"gives the earlier assignment of bright AFM features to triple bonds in polyyne chains, which the paper extends to the cyclic molecule.","marker":"[22]"},{"why":"documents direct AFM imaging of covalent bond structure in surface reactions, supporting the bond-level interpretation of the images.","marker":"[28]"},{"why":"reports the cyclocarbon oxide C24O6 and related precursors that are decarbonylated on the surface to make the ring.","marker":"[15]"},{"why":"provides the high-level coupled-cluster prediction of a polyynic ground state that the experiment supports.","marker":"[12]"},{"why":"provides the density-functional prediction of a cumulenic structure that the experiment and simulations rule out.","marker":"[9]"},{"why":"supplies the semilocal functional whose cumulenic geometry yields a featureless simulated AFM ring, the key negative control.","marker":"[48]"},{"why":"supplies the hybrid functional with high exact-exchange mixing whose polyynic geometry reproduces the measured ninefold AFM contrast.","marker":"[49]"}],"fun_headline_variants":["AFM settles cyclo[18]carbon bond-order debate","Cyclo[18]carbon shows alternating bonds, AFM reveals","First images of C18 confirm polyynic structure","Ninefold pattern in C18 identifies triple bonds","Direct AFM evidence for polyynic cyclo[18]carbon"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion rests on the assumption that the bright lobes and nonagon corners in the AFM images sit exactly above triple bonds, so that the images directly map bond order; this mapping is supported by simulations, but the simulations themselves assume one particular density-functional geometry.","fun_headline_variants_meta":{"raw":{"variants":["AFM settles cyclo[18]carbon bond-order debate","Cyclo[18]carbon shows alternating bonds, AFM reveals","First images of C18 confirm polyynic structure","Ninefold pattern in C18 identifies triple bonds","Direct AFM evidence for polyynic cyclo[18]carbon"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000197,"raw_usage":{"total_tokens":1311,"prompt_tokens":836,"completion_tokens":475,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":452,"completion_tokens_details":{"reasoning_tokens":393}},"tokens_in":452,"tokens_out":475,"duration_ms":4592,"temperature":1.0,"reasoning_tokens":393,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:00:51.824524+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the carbon-carbon bond lengths of an isolated C18 molecule by a technique independent of AFM contrast, such as gas-phase electron diffraction or rotational spectroscopy: alternating bonds near 1.20 Å and 1.34 Å would confirm the polyynic assignment, while eighteen equal bonds near 1.28 Å would falsify it.","supporting_citations":[{"cited_title":"Gross, F","cited_arxiv_id":null,"evidence_quote":"supplies the CO-tip AFM method that gives the submolecular resolution used to resolve the ring."},{"cited_title":"Hapala, G","cited_arxiv_id":null,"evidence_quote":"supplies the probe-particle simulation model used to convert AFM contrast into bond positions."},{"cited_title":"Gross, F","cited_arxiv_id":null,"evidence_quote":"shows that AFM brightness can discriminate bond order, grounding the assignment of bright lobes to triple bonds."},{"cited_title":"Pavliček, P","cited_arxiv_id":null,"evidence_quote":"gives the earlier assignment of bright AFM features to triple bonds in polyyne chains, which the paper extends to the cyclic molecule."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"documents direct AFM imaging of covalent bond structure in surface reactions, supporting the bond-level interpretation of the images."},{"cited_title":"Rubin, M","cited_arxiv_id":null,"evidence_quote":"reports the cyclocarbon oxide C24O6 and related precursors that are decarbonylated on the surface to make the ring."},{"cited_title":"Arulmozhiraja, T","cited_arxiv_id":null,"evidence_quote":"provides the high-level coupled-cluster prediction of a polyynic ground state that the experiment supports."},{"cited_title":"Parasuk, J","cited_arxiv_id":null,"evidence_quote":"provides the density-functional prediction of a cumulenic structure that the experiment and simulations rule out."}],"review_version":1}