Pith. sign in

REVIEW 4 major objections 5 minor 6 references

Revealing the innate sub-nanometer porous structure of carbon nanomembranes with molecular dynamics simulations and highly charged ion spectroscopy

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Carbon nanomembranes are open, sub-nanometer porous carbon networks, not dense films.

desk verdict Plausible but not proven: the new MD/HCI pipeline is a step forward, but the porosity claim rests on hole-enforced structures never validated against unconstrained pore formation. read the letter →

arxiv 2511.04266 v1 pith:AKPCDQFU submitted 2025-11-06 cond-mat.mtrl-sci

Filip Vuković , Anna Niggas , Levin Mihlan , Zhen Yao , Armin Gölzhäuser , Louise Fréville , Vladislav Stroganov , Andrey Turchanin
show 3 more authors
Jürgen Schnack Nigel A. Marks Richard A. Wilhelm
This is my paper
classification cond-mat.mtrl-sci
keywords carbonnanomembraneshighlychargedionspectroscopymoleculardynamicssub-nanometerporosityterphenylthioltensilemoduluschargeexchangeunder-coordinated
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to determine what a terphenylthiol-based carbon nanomembrane actually looks like at the atomic scale. By generating candidate structures with molecular dynamics—including an "exclusion cylinder" method that forces voids into the carbon network—and comparing simulated highly charged ion transmission spectra and tensile moduli to experiment, it argues that in-vacuum CNMs consist largely of under-coordinated carbon arranged around open sub-nanometer pores. The best-fitting model, a 150-cylinder structure annealed for 9 ps, reproduces the high-charge-state tail of the ion spectra and a tensile modulus in the measured 5–12 GPa range. If correct, the membranes are chemically reactive when removed from vacuum and are likely passivated by hydrogen and oxygen in air, with consequences for filtration, energy, and device applications.

What carries the argument

The work hinges on two matched computational tools. The first is "exclusion cylinder" molecular dynamics, which enforces a fixed set of cylindrical voids of radius about 2.5 A during high-temperature annealing to generate carbon-only candidate structures with controlled porosity; specular reflection keeps atoms out of the cylinders. The second is the time-dependent potential (TDPot) model, adapted from two-dimensional targets to finite-thickness membranes, which computes angle-resolved exit charge-state distributions for highly charged Xe ions by treating neutralization through interatomic Coulombic decay. The pore-detection image-processing routine and tensile-deformation simulations supply

What would settle it

A decisive check would be to measure the HCI transmission spectrum of a CNM whose hydrogen/oxygen content has been deliberately varied—for example, by controlled in-vacuum hydrogenation—and see whether the high-charge-state tail and bimodal distribution shift in the way the porous-network model predicts. Alternatively, atomically resolved imaging that revealed large graphitic domains or pore sizes far from the 5.5-A mean pore diameter of the best-fit model would falsify the specific structural claim.

Watch

Extended reading notes

Core claim

The paper's central claim is that in-vacuum terphenylthiol carbon nanomembranes consist largely of under-coordinated carbon arranged as an open sub-nanometer porous network, not a dense graphitic film. Exclusion-cylinder molecular dynamics produced candidate structures with enforced 2.5-A-radius voids; a time-dependent potential model of highly charged ion transmission converted them into angle-resolved exit charge-state spectra for comparison with Xe experiments. The 150-cylinder structure annealed for 9 ps best matches the measured high-charge-state tail and gives a tensile modulus in the 5–12 GPa experimental range. Since carbon loss during crosslinking is small, the authors infer that ev

Load-bearing premise

Everything rests on the assumption that the carbon-only, hole-enforced molecular dynamics structures are representative of real CNMs; the authors explicitly say the exclusion-cylinder simulations do not aim to model the SAM-to-CNM formation process, so if actual membranes contain hydrogen, oxygen, different pore geometries, or formation-induced artefacts, the match of the 150-cylinder model could be coincidental.

Editorial extensions

If this is right

  • A correct CNM structure must be thought of as an open network: the membrane's low tensile modulus and its transmission signature both trace to sub-nanometer voids and a large fraction of under-coordinated carbon.
  • CNMs in vacuum are predicted to be chemically reactive; after exposure to air, dangling bonds should be stabilized by hydrogen, water, or oxygen groups, so ambient-condition CNM properties are those of a passivated network, not bare carbon.
  • Highly charged ion transmission spectroscopy can serve as a non-destructive structural probe for radiation-sensitive freestanding membranes, with the high-charge-state tail as a direct indicator of sub-nanometer porosity.
  • Multiple pore-formation mechanisms may operate during electron irradiation of the precursor SAM; momentum-transfer simulations produce broader pore-area distributions than enforced holes, implying that the real pore-size distribution may be complex.
  • Permeation and ion-selectivity behavior in filtration applications should be interpreted in terms of sub-nanometer channels and reactive under-coordinated sites rather than a dense graphitic layer.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If real membranes carry hydrogen or oxygen, the carbon-only models may be incomplete; a direct test would be to simulate H/O-passivated versions of the 150-cylinder structure and check whether the same ion spectra and modulus are reproduced.
  • The ratio between the incident-charge-state peak and the low-charge-state peak in HCI spectra could be developed into a quantitative porosity metric once the contribution from micrometer-scale cracks is subtracted or otherwise rejected.
  • The paper's two routes to low modulus—high under-coordinated fraction or high pore density—suggest a design principle: membranes with similar stiffness can be made with very different local chemistry, which should affect their reactivity and transport selectivity.
  • The momentum-transfer result that more irradiation events can produce fewer, larger pores is counterintuitive and testable: systematic permeation experiments with size-selected gases on membranes made with different electron doses could confirm whether pore coarsening occurs.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper combines molecular dynamics simulations with highly charged ion (HCI) transmission spectroscopy to infer the atomistic structure of terphenylthiol-derived carbon nanomembranes (CNMs). Two simulation approaches are used: 'exclusion cylinder' MD, which enforces a pre-defined distribution of cylindrical voids during high-temperature annealing, and 'momentum transfer' MD, which mimics electron-irradiation-induced crosslinking without enforced voids. The resulting structures are characterized by their tensile moduli and pore statistics, and a subset is fed into a time-dependent potential (TDPot) model to simulate angle-resolved HCI charge-exchange spectra. The authors report that a 150-exclusion-cylinder structure annealed for 9 ps best reproduces both the measured tensile modulus (~10 GPa) and the high-charge-state tail of the experimental transmission spectra, leading them to conclude that in-vacuum CNMs are open porous carbon networks with a significant fraction of under-coordinated carbon, likely stabilized by hydrogen/oxygen in ambient conditions.

Significance. If the conclusion holds, it would resolve a long-standing question about CNM structure and explain their high reactivity and gas/water permeation behavior. The work is methodologically valuable: it couples two state-of-the-art tools (MD structure generation and TDPot HCI spectroscopy) and makes the exclusion-cylinder simulation code openly available under GPL. The systematic sweep over 81 structures is a strength, and the direct comparison of simulated and experimental charge-state distributions is a novel, falsifiable test. However, the central claim rests on structures whose porosity was imposed a priori, and the key missing control—running TDPot on the unconstrained momentum-transfer structures—weakens the inference. The paper is therefore a promising proof-of-concept that requires additional validation before it can be considered a robust structural determination.

major comments (4)
  1. [Methods, 'Momentum transfer simulations' and Results, 'Pore Detection'] The central conclusion—that CNMs are open sub-nanometer porous networks—is tested by HCI–TDPot only for the exclusion-cylinder family (Figs. 5 and 6). The momentum-transfer structures, which form pores without pre-imposed geometry, are analyzed only for pore-area statistics (Fig. 7). Since the exclusion-cylinder method explicitly 'do[es] not aim to model the SAM to CNM formation process,' the match of the 150-cylinder/9 ps model could be an artifact of the imposed hole distribution. Please run TDPot transmission simulations on the momentum-transfer structures and compare to experiment. This is the missing critical test for the porosity claim.
  2. [Results, Fig. 6 and surrounding text] The identification of the 150-cylinder, 9 ps structure as 'best fitting' is not supported by a quantitative fit metric. The comparison is qualitative and the manuscript itself lists discrepancies (less overall neutralization, different angle scales). Provide a quantitative goodness-of-fit measure (e.g., χ² or Kolmogorov–Smirnov statistic) evaluated across all 81 structures, and describe the angle-rescaling procedure alluded to in the Fig. 6 caption but not given in the text.
  3. [Methods, 'Highly-charged ion transmission simulations'] TDPot ICD parameters are calibrated against single-, bi-, and tri-layer graphene benchmark data. Disordered, porous, under-coordinated carbon networks present a different electronic environment, and no sensitivity analysis is provided. Please discuss the transferability of the ICD parameters or quantify the sensitivity of the predicted spectra to reasonable variations in the ICD rate/prefactor, otherwise the quantitative match in Fig. 6 may be parameter-dependent.
  4. [Results, 'Tensile moduli' and Fig. 5] The 9 ps annealing time is highlighted because bimodality in the exit charge-state distribution appears at that time and longer annealing gives moduli above the experimental range. This is a post hoc selection criterion. The manuscript should either provide a more principled basis for choosing the annealing time or demonstrate that the qualitative conclusion (high-charge tail, modulus) is robust across the 4–16 ps window for the 150-cylinder family, rather than singling out one frame.
minor comments (5)
  1. [Fig. 7 and text] Inconsistency: the text refers to 'the exclusion cylinder 10 ps annealed CNMs' while the Fig. 7(c) caption says '4 ps annealed region restricted simulations.' Please reconcile.
  2. [Equation (1)] The rotation matrices in Eq. (1) are not fully explained; state explicitly that θ is the angle between the atom's velocity vector and the x-axis, and define the intermediate vectors v2–v4.
  3. [Fig. 6 caption] The caption states 'the scattering angles are different in experiment and simulation, details are given in the text' but no such details appear in the main text. Either describe the calibration in the main text or move it to the SI.
  4. [Acknowledgment] Typographical error: 'Deutche Forschungsgemeinschaft' should be 'Deutsche Forschungsgemeinschaft.'
  5. [References] References 15 and 42 are duplicate citations of the same paper (Angelova et al., ACS Nano 2013). Consider consolidating.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central inference is model selection against independent HCI, tensile-modulus, and graphene-calibration benchmarks.

full rationale

The paper generates candidate CNM structures by two MD routes (exclusion-cylinder and momentum-transfer), computes HCI transmission spectra with the TDPot method calibrated on multilayer graphene data (Refs. 36 and 73), and compares with measured angle-resolved charge states and literature tensile moduli. The porosity of the selected 150-cylinder/9-ps model was indeed an input to the generation procedure, so the conclusion is not a first-principles prediction of porosity; but it is also not a circular reduction in the sense defined here. The HCI spectra and tensile modulus are external benchmarks not used to construct the TDPot parameters or the MD potential, and the 0- and low-hole-count models provide non-porous alternatives that fail to reproduce the high-charge tail. The paper explicitly disclaims that the hole-enforced simulations model SAM-to-CNM formation ('It should be emphasized that these simulations do not aim to model the SAM to CNM formation process'), and it admits pore-size distributions cannot currently be extracted from HCI charge exchange data; these are limitations of the inference, not evidence of circularity. The absence of a quantitative goodness-of-fit metric for the 'best fitting' label and the incomplete angle calibration in Fig. 6 weaken the model-selection argument but do not make it definitional. I therefore find no circular step requiring a non-zero score.

Assumptions & free parameters 5 free parameters · 6 assumptions · 1 invented entities

The central structural conclusion depends on several modeling choices: an unphysical high-temperature annealing protocol, a carbon-only description, a hole-enforcement scheme, a potential model, and a charge-exchange model calibrated on graphene. The paper is transparent about many of these, but they constitute the bulk of the prior content the reader must accept.

free parameters (5)
  • Exclusion cylinder count = 150 cylinders (0.84 nm^-2) for the best-fit structure; scanned 0-150
    The number of pre-enforced holes is a discrete model parameter; choosing 150 as 'best fitting' effectively fits the pore density to the HCI and tensile data.
  • Exclusion cylinder radius = 2.5 Å with Gaussian sigma 0.5 Å
    Chosen from gas permeation and ion selectivity literature rather than from the target HCI data, but it controls the pore size in all generated porous models.
  • High-temperature annealing time = 9 ps for best fit; scanned 0-64 ps
    Annealing at 3000 K is not physical; the branching time controls sp2/sp ratio, modulus, and HCI spectra, so selecting 9 ps is a fitted choice.
  • TDPot ICD parameters = Tuned to SLG/BLG/TLG graphene benchmarks
    The parameters controlling interatomic Coulombic decay rate are calibrated on graphene and then applied to CNM structures; numerical values are not reported.
  • Momentum-transfer simulation parameters = Secondary force 300-550 eV/Å; 30-50 impact events
    These are scanned in the alternative formation model to explore pore formation, but they are not used to define the final best-fit structure.
assumptions (6)
  • domain assumption CNMs can be represented as carbon-only networks in vacuum; all H, S, and O atoms are omitted from the MD models.
    All MD starts with 'all other elements omitted' (Methodology), so the under-coordinated carbon conclusion assumes the real UHV membrane is effectively carbon-only after crosslinking.
  • domain assumption The carbon EDIP potential accurately describes bonding, energetics, and elastic response of disordered carbon networks.
    MD structure generation and tensile predictions rely on Marks EDIP; no CNM-specific recalibration is provided.
  • domain assumption TDPot parameters tuned on crystalline graphene transfer to disordered, under-coordinated CNMs.
    ICD rate parameters are tuned to SLG/BLG/TLG benchmarks (Methods) and then used unchanged for CNM structures without uncertainty propagation.
  • domain assumption Only a negligible amount of carbon is lost during CNM crosslinking (<=5%).
    Stated in the Conclusion as an assumption; it is needed to interpret density inhomogeneities as intrinsic sub-nanometer porosity rather than carbon loss or etch artifacts.
  • ad hoc to paper Annealing at 3000 K explores carbon configurations relevant to real electron-irradiation crosslinking.
    The authors note the annealing 'does not have any physical analogue to the experimental preparation of CNMs' and is used for computational tractability.
  • ad hoc to paper Exclusion-cylinder enforced voids survive relaxation and represent physical sub-nanometer pores.
    The algorithm prevents atoms from entering cylinder regions during dynamics; after constraints are removed, pores remain, but the method guarantees at least some porosity in many generated structures.
invented entities (1)
  • Exclusion cylinder (computational constraint)
    purpose: To enforce cylindrical void regions during MD structure generation while relaxing the carbon network.
    The exclusion cylinder is a modeling device, not a claimed physical object. Its radius is loosely tied to permeation experiments, but it has no falsifiable handle of its own.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Revealing the innate sub-nanometer porous structure of carbon nanomembranes with molecular dynamics simulations and highly charged ion spectroscopy." pith.science (2026). https://pith.science/paper/AKPCDQFU

@misc{pith2026251104266,
  author       = {Pith},
  title        = {Pith review of: Revealing the innate sub-nanometer porous structure of carbon nanomembranes with molecular dynamics simulations and highly charged ion spectroscopy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AKPCDQFU}},
  note         = {Machine review of arXiv:2511.04266}
}
read the original abstract

Carbon nanomembranes (CNMs) are nanometer-thin disordered carbon materials that are suitable for a range of applications, from energy generation and storage, through to water filtration. The structure-property relationships of these nanomembranes are challenging to study using traditional experimental characterization techniques, primarily due to the radiation-sensitivity of the free-standing membrane. Highly charged ion spectroscopy is a novel characterization method that is able to infer structural details of the carbon nanomembrane without concern of induced damage affecting the measurements. Here we employ molecular dynamics simulations to produce candidate structural models of terphenylthiol-based CNMs with varying degrees of nanoscale porosity, and compare predicted ion charge exchange data and tensile moduli to experiment. The results suggest that the in-vacuum CNM composition likely comprises a significant fraction of under-coordinated carbon, with an open sub-nanometer porous structure. Such a carbon network would be reactive in atmosphere and would be presumably stabilized by hydrogen and oxygen groups under atmospheric conditions.

Figures

Figures reproduced from arXiv: 2511.04266 by the authors.

Figure 1
Figure 1. Exit charge state spectra of 90 keV Xe10+ after transmission through the CNM before (a) and after (b) contaminant removal through ion irradiation. (c) A fluence of ∼ 3 × 1011 ion impacts per cm2 is necessary to reach an equilibrium mean charge state of qout ∼ 2 [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Schematic diagram of the specular reflection imposed on atoms that enter the [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. Planar top-down view of a selection of exclusion cylinder enforced carbon [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Predicted average tensile moduli of the CNM models as a function of the number [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]
Figure 5
Figure 5. Figure 5: Simulated exit charge state for Xe ions incident on various hole-enforced nanomem [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]
Figure 6
Figure 6. Figure 6: Experimental (left half) and simulated (right half) charge exchange spectra through [PITH_FULL_IMAGE:figures/full_fig_p021_6.png]
Figure 7
Figure 7. Figure 7: Pore area distributions shown in blue shaded regions (left [PITH_FULL_IMAGE:figures/full_fig_p022_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

6 extracted references

  1. [1]

    (1) Griffin, E. et al. Proton and Li-Ion Permeation through Graphene with Eight-Atom- Ring Defects.ACS Nano2020,14, 7280–7286. (2) Rajendran, S.; Tang, Z.; George, A.; Cannon, A.; Neumann, C.; Sawas, A.; Ryan, E.; Turchanin, A.; Arava, L. M. R. Inhibition of Lithium Dendrite Formation in Lithium Metal Batteries via Regulated Cation Transport through Ultra...

    2021

  2. [50]

    Compact electron beam ion sources/traps: Review and prospects (invited).Rev

    (39) Zschornack, G.; Kreller, M.; Ovsyannikov, V.; Grossman, F.; Kentsch, U.; Schmidt, M.; Ullmann, F.; Heller, R. Compact electron beam ion sources/traps: Review and prospects (invited).Rev. Sci. Instrum.2008,79, 02A703–02A703. (40) Kentsch, U.; Landgraf, S.; Zschornack, G.; Grossmann, F.; Ovsyannikov, V. P.; Ull- mann, F. Dresden EBIT: Results and persp...

    2008

  3. [89]

    A versatile ion beam spectrometer for studies of ion interaction with 2D materials.Rev

    (37) Schwestka, J.; Melinc, D.; Heller, R.; Niggas, A.; Leonhartsberger, L.; Winter, H.; Facsko, S.; Aumayr, F.; Wilhelm, R. A versatile ion beam spectrometer for studies of ion interaction with 2D materials.Rev. Sci. Instrum.2018,89, 085101–085101. (38) Niggas, A.; Schwestka, J.; Weichselbaum, D.; Heller, R.; Aumayr, F.; Wilhelm, R. A. Coincidence techni...

    2018

  4. [180]

    Elastic properties of crystals of single-walled carbon nanotubes.Solid State Commun.2000,114, 395–399

    (74) Popov, V.; Van Doren, V.; Balkanski, M. Elastic properties of crystals of single-walled carbon nanotubes.Solid State Commun.2000,114, 395–399. 34 (75) Zhang, X.; Mainka, M.; Paneff, F.; Hachmeister, H.; Beyer, A.; Gölzhäuser, A.; Huser, T. Surface-Enhanced Raman Spectroscopy of Carbon Nanomembranes from Aromatic Self-Assembled Monolayers.Langmuir2018...

    2000

  5. [662]

    Genera- tion of Surface Amino Groups on Aromatic Self-Assembled Monolayers by Low Energy Electron Beams—A First Step Towards Chemical Lithography.Adv

    (41) Eck, W.; Stadler, V.; Geyer, W.; Zharnikov, M.; Gölzhäuser, A.; Grunze, M. Genera- tion of Surface Amino Groups on Aromatic Self-Assembled Monolayers by Low Energy Electron Beams—A First Step Towards Chemical Lithography.Adv. Mater.2000,12, 805–808. (42) Angelova, P.; Vieker, H.; Weber, N.-E.; Matei, D.; Reimer, O.; Meier, I.; Kurasch, S.; Biskupek, ...

    2000

  6. [6103]

    Graphene Growth by Conversion of Aromatic Self-Assembled Monolay- ers.Ann

    27 (17) Turchanin, A. Graphene Growth by Conversion of Aromatic Self-Assembled Monolay- ers.Ann. Phys. (Berl.)2017,529, 1700168. (18) Dementyev, P.; Wilke, T.; Naberezhnyi, D.; Emmrich, D.; Gölzhäuser, A. Vapour per- meation measurements with free-standing nanomembranes.Phys. Chem. Chem. Phys. 2019,21, 15471–15477. (19) Winter, A.; Ekinci, Y.; Gölzhäuser,...

    2017

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.