pith. sign in

arxiv: 2605.22141 · v1 · pith:JDNT5IWSnew · submitted 2026-05-21 · ❄️ cond-mat.mtrl-sci

Theory-Guided, Machine-Learning-Accelerated Discovery of a 3D Carbon Nested Nodal-Surface Semimetal

Shuaihua Zhang , Silei Guo , Jingxiang Liu , Baoxin Hu , Yanling Wu , Jun Li This is my paper

Pith reviewed 2026-05-22 05:22 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci
keywords carbon allotropenodal surface semimetaltopological materialsmachine learning inverse designnon-symmorphic symmetryDirac crossingsdrumhead statesgraphene extension
0
0 comments X

The pith

A symmetry-engineering principle turns graphene Dirac cones into 3D nested nodal surfaces, yielding the stable carbon allotrope Netsene with drumhead states.

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

The paper seeks to extend two-dimensional Dirac physics from graphene into three-dimensional carbon allotropes that host higher-dimensional band degeneracies. It introduces a general principle of symmetry engineering that uses controlled layering and registry shifts to convert a Dirac cone into a three-dimensional nodal surface. This principle guides a machine-learning inverse-design workflow that combines a crystal diffusion variational autoencoder with a Crystal Transformer. The search produces Netsene, a body-centered tetragonal carbon structure that first-principles calculations show to be mechanically and dynamically stable. A reader would care because the resulting material combines ultrahigh carrier velocities, protected nodal surfaces, and surface states in a single, robust bulk platform suitable for topological and correlation studies.

Core claim

By applying a symmetry-engineering principle that converts graphene's Dirac cone into a three-dimensional nodal surface through controlled layering and registry shift, and accelerating the inverse search with a crystal diffusion variational autoencoder integrated with a Crystal Transformer, the authors identify a dynamically and mechanically stable carbon allotrope, Netsene (bct-C24), in space group I4/mcm. First-principles calculations establish that Netsene is a nested nodal-surface semimetal featuring a complex double-bowl-shaped nodal-surface system around the Fermi level protected by non-symmorphic symmetries, together with Dirac-like linear crossings whose Fermi velocities reach values

What carries the argument

The symmetry-engineering principle that systematically transforms graphene's Dirac cone into a 3D nodal surface via controlled layering and registry shift, used to guide the machine-learning inverse design of carbon allotropes.

If this is right

  • Netsene supplies a bulk, stable platform that unifies ultrahigh carrier mobility comparable to graphene with protected topological nodal surfaces.
  • The non-symmorphic symmetry protection enables complex nested nodal surfaces rather than isolated points or lines.
  • Drumhead surface states, including a nearly flat band, follow directly from the non-trivial bulk topology.
  • The material opens a route to study correlation physics on the flat surface band within an otherwise semimetallic carbon framework.
  • The same layering-and-shift principle can be reused to target additional three-dimensional topological phases in carbon.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach could be extended to other two-dimensional Dirac materials such as transition-metal dichalcogenides to generate new families of three-dimensional nodal-surface semimetals.
  • Doping or strain engineering of the nearly flat drumhead band might induce superconductivity or other ordered states without destroying the bulk nodal surfaces.
  • Experimental verification of the high Fermi velocities would require transport measurements on high-quality single crystals or thin films.
  • The registry-shift mechanism suggests a general design rule for engineering band degeneracies in layered van der Waals systems beyond carbon.

Load-bearing premise

The symmetry-engineering principle that transforms graphene's Dirac cone into a 3D nodal surface via controlled layering and registry shift is both valid and sufficient to guide discovery of a realizable, stable structure.

What would settle it

Synthesis of Netsene followed by angle-resolved photoemission spectroscopy that fails to detect the predicted double-bowl nodal surfaces or drumhead states would falsify the central topological claim.

Figures

Figures reproduced from arXiv: 2605.22141 by Baoxin Hu, Jingxiang Liu, Jun Li, Shuaihua Zhang, Silei Guo, Yanling Wu.

Figure 1
Figure 1. Figure 1: FIG. 1: Schematic overview of the integrated research methodology. (a) Three-step symmetry-engineering design [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: (a) The conventional unit cell of [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: (b). The absence of any imaginary frequency modes confirms the dynamic stability of Netsene, indicating that the structure resides at a genuine local minimum on the potential energy surface. The highest phonon fre￾quency reaches approximately 1532 cm−1 , which is close to that of graphite (1610 cm−1 ) [23] and indicative of strong covalent bonding. Furthermore, ab initio molec￾ular dynamics (AIMD) simulati… view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: (a) Electronic band structure of Netsene along [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5: (a) 3D visualization of the nested nodal-surface [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6: Topological surface states of Netsene. (a) (110) [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
read the original abstract

Extending the Dirac physics of two-dimensional (2D) graphene into three dimensions (3D) carbon allotropes with higher-dimensional band degeneracies remains a central challenge in topological materials science. Here, we propose a general symmetry-engineering principle that systematically transforms graphene's Dirac cone into a 3D nodal surface via controlled layering and registry shift, and employ this principle to guide a machine-learning-accelerated inverse design. By integrating a crystal diffusion variational autoencoder(CDVAE) with a Crystal Transformer, we discover a novel, dynamically and mechanically stable carbon allotrope named \textbf{Netsene} (bct-C$_{24}$), which crystallizes in the body-centered tetragonal \textit{I4/mcm} space group. First-principles calculations confirm that Netsene is a unique nested nodal-surface semimetal: it hosts a complex, double-bowl-shaped nodal-surface system around the Fermi level, protected by non-symmorphic symmetries, alongside Dirac-like linear crossings with Fermi velocities comparable to graphene ($\sim 9 \times 10^5$~m/s). Its non-trivial bulk topology manifests in drumhead surface states, including a nearly flat band. Netsene provides a structurally robust, bulk platform that unifies ultrahigh carrier mobility, topological nodal surfaces, and potential correlation physics, demonstrating the power of theory-guided, machine-learning-accelerated discovery for engineering topological quantum phases.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript proposes a symmetry-engineering principle that converts graphene's 2D Dirac cones into 3D nested nodal surfaces through controlled layering and registry shifts. This principle is used to guide an inverse-design search combining a crystal diffusion variational autoencoder (CDVAE) with a Crystal Transformer, yielding a new dynamically and mechanically stable carbon allotrope, Netsene (bct-C24), in space group I4/mcm. First-principles calculations are reported to confirm a double-bowl-shaped nodal-surface system around the Fermi level protected by non-symmorphic symmetries, Dirac-like linear crossings with Fermi velocities ~9×10^5 m/s, and drumhead surface states including a nearly flat band.

Significance. If the central claims hold, the work would provide a bulk, structurally robust carbon platform that combines ultrahigh carrier mobility with topologically protected nodal surfaces and drumhead states, potentially enabling studies of correlation physics in a simple elemental system. The theory-guided ML inverse-design workflow represents a concrete strength, as does the identification of a realizable structure with graphene-comparable velocities.

major comments (3)
  1. [Symmetry-engineering principle and inverse design] The symmetry-engineering principle is load-bearing for the discovery claim, yet the manuscript provides no explicit symmetry analysis, model Hamiltonian, or tight-binding derivation demonstrating that the controlled layering and registry shift in I4/mcm enforces nodal-surface degeneracy rather than opening a gap or dispersing the cones (see the section introducing the principle and the subsequent inverse-design workflow).
  2. [Computational methods and results] First-principles confirmation of stability and band features is asserted without reported error bars, k-point convergence tests, or details on how the nodal surface was identified and protected (abstract and computational methods section). This weakens the claim that Netsene is a unique nested nodal-surface semimetal.
  3. [Machine-learning inverse design] The ML model (CDVAE integrated with Crystal Transformer) is central to the discovery, but training data, validation metrics, and how the symmetry principle was encoded as a constraint are not described, making it impossible to assess whether the search was truly guided or merely sampled.
minor comments (2)
  1. [Abstract] Notation for the space group (I4/mcm) and structure name (Netsene, bct-C24) should be introduced consistently in the abstract and main text.
  2. [Electronic structure results] The Fermi velocity value (~9×10^5 m/s) is given without comparison to the exact graphene reference value used or the direction of the linear crossing.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed review. We address each major comment point by point below, providing clarifications and indicating revisions made to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Symmetry-engineering principle and inverse design] The symmetry-engineering principle is load-bearing for the discovery claim, yet the manuscript provides no explicit symmetry analysis, model Hamiltonian, or tight-binding derivation demonstrating that the controlled layering and registry shift in I4/mcm enforces nodal-surface degeneracy rather than opening a gap or dispersing the cones (see the section introducing the principle and the subsequent inverse-design workflow).

    Authors: We appreciate this observation. The introduction outlines the symmetry-engineering principle by describing how controlled layering and registry shifts extend graphene's 2D Dirac cones into 3D nodal surfaces protected by non-symmorphic symmetries in I4/mcm. To provide a more rigorous foundation, the revised manuscript adds a dedicated subsection with a simplified model Hamiltonian and tight-binding derivation. This explicitly shows that the registry shift and non-symmorphic operations enforce the nodal-surface degeneracy without gapping the bands, consistent with the first-principles results. revision: yes

  2. Referee: [Computational methods and results] First-principles confirmation of stability and band features is asserted without reported error bars, k-point convergence tests, or details on how the nodal surface was identified and protected (abstract and computational methods section). This weakens the claim that Netsene is a unique nested nodal-surface semimetal.

    Authors: We agree that additional computational details would enhance rigor. The original methods section used standard DFT convergence settings, but the revised version now includes explicit k-point convergence tests (energy converged to <1 meV), error estimates on Fermi velocities from multiple samplings, and a step-by-step description of nodal-surface identification via dense k-grids combined with symmetry analysis confirming protection by the non-symmorphic elements of I4/mcm. These additions support the uniqueness of the nested nodal-surface features. revision: yes

  3. Referee: [Machine-learning inverse design] The ML model (CDVAE integrated with Crystal Transformer) is central to the discovery, but training data, validation metrics, and how the symmetry principle was encoded as a constraint are not described, making it impossible to assess whether the search was truly guided or merely sampled.

    Authors: We thank the referee for this point. The methods section has been expanded in the revision to detail the training dataset (carbon allotropes drawn from the Materials Project and related databases), validation metrics (reconstruction accuracy and property-prediction errors), and the encoding of the symmetry principle via conditional generation with space-group filters and a symmetry-aware term in the Crystal Transformer loss function. This demonstrates that the search was guided by the principle rather than random sampling. revision: yes

Circularity Check

0 steps flagged

No significant circularity: principle proposed then verified independently via DFT

full rationale

The paper proposes a symmetry-engineering principle for converting graphene Dirac cones into 3D nodal surfaces through layering and registry shift, then applies it to guide a CDVAE+Crystal Transformer inverse design that yields the Netsene structure. Topological features including the double-bowl nodal surfaces, linear crossings, and drumhead states are subsequently confirmed by independent first-principles DFT calculations on the discovered allotrope in space group I4/mcm. No equation or step reduces a reported prediction to a fitted parameter defined by the same calculation, nor does any load-bearing claim collapse to a self-citation chain or ansatz smuggled from prior work by the same authors. The derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The abstract relies on an unproven symmetry-engineering principle and on the assumption that the ML model can generate dynamically stable structures whose DFT bands match the design target; no numerical free parameters are stated.

axioms (1)
  • domain assumption Symmetry-engineering principle that transforms 2D Dirac cones into 3D nodal surfaces via layering and registry shift
    Invoked to guide the inverse design; stated in the opening sentence of the abstract.
invented entities (1)
  • Netsene (bct-C24) no independent evidence
    purpose: New carbon allotrope realizing the nested nodal-surface semimetal
    Postulated as the output of the ML search; independent evidence would require experimental synthesis or external verification not mentioned in the abstract.

pith-pipeline@v0.9.0 · 5810 in / 1477 out tokens · 49189 ms · 2026-05-22T05:22:47.366035+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages

  1. [1]

    K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov, Science306, 666 (2004)

    work page 2004

  2. [2]

    A. H. Castro Neto, F. Guinea, N. M. R. Peres, K. S. Novoselov, and A. K. Geim, Reviews of Modern Physics 81, 109 (2009)

    work page 2009

  3. [3]

    A. A. Burkov and L. Balents, Physical Review Letters 107, 127205 (2011)

    work page 2011

  4. [4]

    N. P. Armitage, E. J. Mele, and A. Vishwanath, Reviews of Modern Physics90, 015001 (2018)

    work page 2018

  5. [5]

    W. Wu, Y. Liu, S. Li, C. Zhong, Z.-M. Yu, X.-L. Sheng, Y. X. Zhao, and S. A. Yang, Physical Review B97, 115125 (2018)

    work page 2018

  6. [6]

    Bzduˇ sek and M

    T. Bzduˇ sek and M. Sigrist, Physical Review B96, 155105 (2017)

    work page 2017

  7. [7]

    J.-T. Wang, S. Nie, H. Weng, Y. Kawazoe, and C. Chen, Physical Review Letters120, 026402 (2018)

    work page 2018

  8. [8]

    J.-T. Wang, H. Weng, and C. Chen, Advances in Physics: X4, 1625724 (2019)

    work page 2019

  9. [9]

    T. Xie, X. Fu, O.-E. Ganea, R. Barzilay, and T. Jaakkola, International Conference on Learning Representations (ICLR) (2021), arXiv:2110.06197

    work page arXiv 2021

  10. [10]

    Z. Cao, X. Luo, J. Lv, and L. Wang, Science Bulletin70, 3522 (2025)

    work page 2025

  11. [11]

    J. O. Sofo, A. S. Chaudhari, and G. D. Barber, Physical Review B75, 153401 (2007)

    work page 2007

  12. [12]

    X. Yuan, J. Liu, J. Liu, J. Wei, L. Chen, W. Wang, and C. Li, Diamond and Related Materials149, 111625 (2024)

    work page 2024

  13. [13]

    A. Jain, S. P. Ong, G. Hautier, W. Chen, W. D. Richards, S. Dacek, S. Cholia, D. Gunter, D. Skinner, G. Ceder, and K. A. Persson, APL Materials1, 011002 (2013)

    work page 2013

  14. [14]

    Hoffmann, A

    R. Hoffmann, A. A. Kabanov, A. A. Golov, and D. M. Proserpio, Angewandte Chemie International Edition55, 10962 (2016)

    work page 2016

  15. [15]

    Giannozzi, O

    P. Giannozzi, O. Andreussi, T. Brumme, O. Bunau, M. B. Nardelli, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, M. Cococcioni, N. Colonna, I. Carnimeo, A. D. Corso, S. d. Gironcoli, P. Delugas, R. A. DiStasio, A. Ferretti, A. Floris, G. Fratesi, G. Fugallo, R. Gebauer, U. Gerstmann, F. Giustino, T. Gorni, J. Jia, M. Kawa- mura, H.-Y. Ko, A. Kokalj, E. K¨...

    work page 2017

  16. [16]

    J. P. Perdew, K. Burke, and M. Ernzerhof, Physical Re- view Letters77, 3865 (1996). 8

    work page 1996

  17. [17]

    Q. Wu, S. Zhang, H.-F. Song, M. Troyer, and A. A. Soluyanov, Computer Physics Communications224, 405 (2018)

    work page 2018

  18. [18]

    A. A. Mostofi, J. R. Yates, Y.-S. Lee, I. Souza, D. Vander- bilt, and N. Marzari, Computer Physics Communications 178, 685 (2008)

    work page 2008

  19. [19]

    Z. Z. Li, C. S. Lian, J. Xu, L. F. Xu, J. T. Wang, and C. F. Chen, Phys. Rev. B91, 214106 (2015)

    work page 2015

  20. [20]

    J. T. Wang, H. Weng, S. Nie, Z. Fang, Y. Kawazoe, and C. F. Chen, Phys. Rev. Lett.116, 195501 (2016)

    work page 2016

  21. [21]

    Z. Z. Li, J. Chen, S. Nie, L. F. Xu, H. Mizuseki, H. Weng, and J. T. Wang, Carbon133, 39 (2018)

    work page 2018

  22. [22]

    W. I. F. David, R. M. Ibberson, J. C. Matthewman, K. Prassides, T. J. S. Dennis, J. P. Hare, H. W. Kroto, R. Taylor, and D. R. M. Walton, Nature353, 147 (1991)

    work page 1991

  23. [23]

    Maultzsch, S

    J. Maultzsch, S. Reich, C. Thomsen, H. Requardt, and P. Ordejon, Phys. Rev. Lett.92, 075501 (2004)

    work page 2004

  24. [24]

    H. J. McSkimin and P. Andreatch, Journal of Applied Physics43, 2944 (1972)

    work page 1972

  25. [25]

    Bosak, M

    A. Bosak, M. Krisch, M. Mohr, J. Maultzsch, and C. Thomsen, Physical Review B75, 153408 (2007)

    work page 2007

  26. [26]

    R. S. Deacon, K.-C. Chuang, R. J. Nicholas, K. S. Novoselov, and A. K. Geim, Physical Review B76, 081406 (2007)

    work page 2007

  27. [27]

    Y. Cao, V. Fatemi, S. Fang, K. Watanabe, T. Taniguchi, E. Kaxiras, and P. Jarillo-Herrero, Nature556, 43 (2018)

    work page 2018

  28. [28]

    Bistritzer and A

    R. Bistritzer and A. H. MacDonald, Proceedings of the National Academy of Sciences108, 12233 (2011)

    work page 2011

  29. [29]

    J. Li, S. Jin, F. Yang, and D.-X. Yao, Physical Review B 102, 174509 (2020)

    work page 2020

  30. [30]

    Malko, C

    D. Malko, C. Neiss, F. Vines, and A. Gorling, Physical Review Letters108, 086804 (2012)

    work page 2012

  31. [31]

    Zou and B.-G

    J.-Y. Zou and B.-G. Liu, Physical Review B93, 205435 (2016)

    work page 2016

  32. [32]

    Meunier, A

    V. Meunier, A. Souza Filho, E. Barros, and M. Dressel- haus, Reviews of Modern Physics88, 025005 (2016)

    work page 2016

  33. [33]

    Q. Fan, L. Yan, M. W. Tripp, O. Krejc´ ı, S. Dimos- thenous, S. R. Kachel, M. Chen, A. S. Foster, U. Ko- ert, P. Liljeroth, and J. M. Gottfried, Science372, 852 (2021)

    work page 2021

  34. [34]

    Z. Liu, Y. Li, and Y.-F. Yang, Chinese Physics B28, 077103 (2019)

    work page 2019

  35. [35]

    Y. Yao, J. Li, J. Ye, F. Yang, and D.-X. Yao, Chinese Physics Letters42, 067504 (2025)

    work page 2025

  36. [36]

    Cheng, X

    C. Cheng, X. Zhang, S. Ma, and S. Wang, Physical Chemistry Chemical Physics25, 13116 (2023)

    work page 2023

  37. [37]

    J.-T. Wang, K. Bu, Y. Qian, H. Weng, and C. Chen, Physical Review B104, 245143 (2021)

    work page 2021

  38. [38]

    K. Bu, Y. Qian, J.-T. Wang, and H. Weng, Physical Re- view B103, L081108 (2021)

    work page 2021

  39. [39]

    Bu, J.-T

    K. Bu, J.-T. Wang, H. Weng, and C. Chen, Physical Review B101, 205104 (2020)

    work page 2020

  40. [40]

    Jiang, J

    W. Jiang, J. Jiang, Z. Zhang, W. Wu, L.-C. Zhang, Y. Xie, and Y. Chen, Nanoscale16, 13543 (2024)

    work page 2024

  41. [41]

    Farid and S

    B. Farid and S. Kose, Physica B: Condensed Matter369, 185 (2005)

    work page 2005

  42. [42]

    Barhoumi, D

    M. Barhoumi, D. Rocca, M. Said, and S. Leb` egue, Solid State Communications324, 114136 (2021)

    work page 2021