Experiments and non-equilibrium molecular dynamics (NEMD) simulations by Xu et al. (Nature Communications 5, 3689, 2014) showed that the thermal conductivity κ\kappa of suspended single-layer graphene at 300 K grows approximately as κ(L)∝log⁡L\kappa(L)\propto\log L with sample length LL up to about 9 µm, even for lengths well beyond the average phonon mean free path, and left open which microscopic mechanism dominates this behavior, whether the logarithmic growth continues to the 100 µm–millimeter scale or saturates, and how the conductivity decomposes over phonon branches and frequencies. A decade of subsequent work has split rather than settled the question: three-phonon Boltzmann treatments find a divergent (logarithmic) conductivity driven by momentum-conserving normal processes of flexural phonons, while first-principles calculations including four-phonon scattering and phonon renormalization find a convergent room-temperature value near 1300 W/(m·K), and other renormalization-based analyses argue the opposite. No suspended, length-controlled measurement at 100 µm–millimeter scales, and no mode-resolved experimental decomposition, exist. The problem — the mechanism, branch-resolved decomposition, and asymptotic L→∞L\to\infty behavior of κ\kappa of suspended defect-free single-layer graphene — remains open.

Background

Single-layer graphene (SLG) is an atomically thin carbon membrane whose thermal conductivity κ\kappa is extraordinarily high, making it both a technological heat-spreading material and the canonical experimental platform for heat conduction in two dimensions. In a crystal, heat is carried by phonons (quantized lattice vibrations); in graphene these separate into three acoustic branches — the in-plane longitudinal (LA) and transverse (TA) branches and the out-of-plane flexural branch (ZA) — plus optical branches. Phonons scatter off each other through anharmonic three-phonon and higher-order (four-phonon, ...) processes, classified as normal (momentum-conserving) or Umklapp (momentum-dissipating); only the latter directly relax the heat current. When normal processes dominate, the phonon gas flows like a fluid (phonon hydrodynamics). The phonon Boltzmann transport equation (BTE) computes κ\kappa from these scattering rates; NEMD (non-equilibrium molecular dynamics) drives a heat flux through an atomistic sample and extracts an effective, length-dependent κ\kappa.

For ideal low-dimensional nonlinear lattices (the Fermi–Pasta–Ulam–Tsingou setting), transport theory predicts that the thermal conductivity of a two-dimensional momentum-conserving system diverges with system size as κ∝log⁡L\kappa\propto\log L, instead of approaching a size-independent bulk value as in three dimensions. Whether a real, quasi-two-dimensional crystal such as graphene realizes this divergence has been contested: simulations gave contradictory results depending on technique, and experiments on nanotubes had verified the analogous one-dimensional scaling.

Xu et al. (2014) measured suspended SLG samples of controlled length between low-resistance contacts using Raman optothermal thermometry (inferring the temperature from the Raman shift while Joule-heating the sample), up to about 9 µm, and matched them with NEMD. They found κ(L)\kappa(L) increasing logarithmically even for LL an order of magnitude beyond the average phonon mean free path, and their NEMD attributed the divergence to the combination of reduced dimensionality (restricted three-phonon scattering phase space and selection rules, which particularly protect the flexural ZA modes) and a displacement of phonon populations out of equilibrium — specifically an augmented out-of-plane population under stationary non-equilibrium. They explicitly could not experimentally distinguish this from the alternative of ballistic, sample-length-limited long-wavelength acoustic phonons, and called for mode- and frequency-resolved measurements and for suspended measurements at 100 µm to millimeter scales, which did not exist.

Later theory sharpened rather than closed the question. Ab initio three-phonon BTE solutions with exact treatment of normal processes (Fugallo et al. 2014; Cepellotti et al. 2015; Lee et al. 2015) support a divergent, hydrodynamic-like conductivity of graphene at room temperature, with flexural modes carrying a large share (Fan et al. 2017 found roughly two thirds in unstrained graphene). But Feng and Ruan (2018) and, in particular, Han and Ruan (2023) showed that four-phonon scattering — to which the ZA selection rules do not apply — together with phonon renormalization (temperature-induced modification of the harmonic and anharmonic force constants) converges the room-temperature conductivity to about 1300 W/(m·K), below diamond; conversely, recent work on bending-rigidity renormalization (Ravichandran 2026) argues that elastic renormalization weakens Umklapp scattering and amplifies the conductivity and hydrodynamic behavior. Whether graphene's intrinsic κ\kappa is asymptotically finite, and which mechanism sets the length dependence under laboratory conditions, is listed as an open problem in current reviews (Li and Shi 2024; Di Lucente et al. 2026).

Problem Statement

The problem, as posed by Xu et al. (2014; DOI: 10.1038/ncomms4689) and still unresolved, is to determine the origin and the ultimate limit of the length-dependent thermal conductivity of suspended, defect-free single-layer graphene at 300 K, where experiment and non-equilibrium molecular dynamics established κ(L)∝log⁡L\kappa(L)\propto\log L for lengths from below to about an order of magnitude above the average phonon mean free path. The single scientific objective — a validated microscopic account of the length dependence — comprises three linked questions. (1) Which mechanism dominates under laboratory conditions: (a) ballistic propagation of sample-length-limited long-wavelength low-frequency acoustic phonons, (b) intrinsic two-dimensional kinematic restrictions on anharmonic scattering (restricted three-phonon phase space and flexural selection rules), or (c) steady-state non-equilibrium redistribution of phonon populations (notably of flexural ZA modes) that alters effective scattering rates — and with what relative weights? (2) What is the asymptotic behavior of κ(L)\kappa(L) as LL grows to the tens-of-micrometers, 100 µm, and millimeter scales and beyond: does the logarithmic increase continue, or does κ\kappa saturate or cross over to different scaling, and at what length and to what value — equivalently, does suspended graphene possess a finite intrinsic thermal conductivity? (3) What is the quantitative, frequency- and polarization-resolved decomposition of κ(L)\kappa(L) (contributions of ZA, LA, TA, and optical branches as a function of length at the fixed 300 K condition) that discriminates among the candidate mechanisms? These three questions are posed jointly as the objective, and the relative weights asked in (1) follow from the branch- and frequency-resolved decomposition of (3).

References

  1. Xiangfan Xu, Luiz F. C. Pereira, Yu Wang, Jing Wu, Kaiwen Zhang, Xiangming Zhao, Sukang Bae, Cong Tinh Bui, Rongguo Xie, John T. L. Thong, Byung Hee Hong, Kian Ping Loh, Davide Donadio, Baowen Li, Barbaros Özyilmaz, 'Length-dependent thermal conductivity in suspended single-layer graphene', Nature Communications 5, 3689 (2014), DOI: 10.1038/ncomms4689, https://doi.org/10.1038/ncomms4689 (source of the open question; also arXiv:1404.5379)
  2. Giorgia Fugallo, Andrea Cepellotti, Lorenzo Paulatto, Michele Lazzeri, Nicola Marzari, Francesco Mauri, 'Thermal Conductivity of Graphene and Graphite: Collective Excitations and Mean Free Paths', Nano Letters 14, 6109–6114 (2014), DOI: 10.1021/nl502059f, https://doi.org/10.1021/nl502059f
  3. Andrea Cepellotti, Giorgia Fugallo, Lorenzo Paulatto, Michele Lazzeri, Francesco Mauri, Nicola Marzari, 'Phonon hydrodynamics in two-dimensional materials', Nature Communications 6, 6400 (2015), DOI: 10.1038/ncomms7400, https://doi.org/10.1038/ncomms7400
  4. Sangyeop Lee, David Broido, Keivan Esfarjani, Gang Chen, 'Hydrodynamic phonon transport in suspended graphene', Nature Communications 6, 6290 (2015), DOI: 10.1038/ncomms7290, https://doi.org/10.1038/ncomms7290
  5. Youdi Kuang, Lucas Lindsay, Sanqiang Shi, Xinjiang Wang, Ruiqiang Guo, Baoling Huang, 'Thermal conductivity of graphene mediated by strain and size', preprint (2015), arXiv:1506.08380, https://arxiv.org/abs/1506.08380
  6. Zheyong Fan, Luiz Felipe C. Pereira, Petri Hirvonen, Mikko M. Ervasti, Ken R. Elder, Davide Donadio, Tapio Ala-Nissila, Ari Harju, 'Thermal conductivity decomposition in two-dimensional materials: Application to graphene', Physical Review B 95, 144309 (2017), DOI: 10.1103/PhysRevB.95.144309, https://doi.org/10.1103/PhysRevB.95.144309
  7. Tianli Feng, Xiulin Ruan, 'Four-phonon scattering reduces intrinsic thermal conductivity of graphene and the contributions from flexural phonons', Physical Review B 97, 045202 (2018), DOI: 10.1103/PhysRevB.97.045202, https://doi.org/10.1103/PhysRevB.97.045202
  8. Gwangseok Hwang, O. Kwon, 'Measuring the size dependence of thermal conductivity of suspended graphene disks using null-point scanning thermal microscopy', Nanoscale 8, 5280–5290 (2016), DOI: 10.1039/c5nr08097e, https://doi.org/10.1039/c5nr08097e
  9. A. Majee, Z. Akšamija, 'Dynamical thermal conductivity of suspended graphene ribbons in the hydrodynamic regime', Physical Review B 98, 024303 (2018), DOI: 10.1103/PhysRevB.98.024303, https://doi.org/10.1103/PhysRevB.98.024303
  10. Ruixia Su, Xing Zhang, 'Size effect of thermal conductivity in monolayer graphene', Applied Thermal Engineering 144, 488–494 (2018), DOI: 10.1016/j.applthermaleng.2018.08.062, https://doi.org/10.1016/j.applthermaleng.2018.08.062
  11. E. Mercado, J. Anaya, M. Kuball, 'Impact of Polymer Residue Level on the In-Plane Thermal Conductivity of Suspended Large-Area Graphene Sheets', ACS Applied Materials & Interfaces 13, 17910–17919 (2021), DOI: 10.1021/acsami.1c00365, https://doi.org/10.1021/acsami.1c00365
  12. Cuiqian Yu, Yulou Ouyang, Jie Chen, 'A perspective on the hydrodynamic phonon transport in two-dimensional materials', Journal of Applied Physics 130, 010902 (2021), DOI: 10.1063/5.0056315, https://doi.org/10.1063/5.0056315
  13. Jie Wang, Chao Li, Yunhe Sheng, Yu Su, Lina Yang, 'Cross-scale investigation of the size effect on thermal conductivity of monolayer graphene', Applied Physics Letters 121, 042202 (2022), DOI: 10.1063/5.0096870, https://doi.org/10.1063/5.0096870
  14. Zherui Han, Xiulin Ruan, 'Thermal conductivity of monolayer graphene: Convergent and lower than diamond', Physical Review B 108, L121412 (2023), DOI: 10.1103/PhysRevB.108.L121412, https://doi.org/10.1103/PhysRevB.108.L121412 (arXiv:2302.12216, posted under the title 'Is Thermal Conductivity of Graphene Divergent and Higher Than Diamond?')
  15. Xun Li, Li Shi, 'Open problems in transport physics of ultrahigh-thermal conductivity materials', Journal of Materials Research 39, 2757–2774 (2024), DOI: 10.1557/s43578-024-01441-2, https://doi.org/10.1557/s43578-024-01441-2
  16. Chuang Zhang, H. Rezgui, Meng Lian, Hong Liang, 'Non-equilibrium transport and phonon branch-resolved size effects based on a multi-temperature kinetic model', Materials Today Physics 59, 101893 (2025), DOI: 10.1016/j.mtphys.2025.101893, https://doi.org/10.1016/j.mtphys.2025.101893 (arXiv:2505.02411)
  17. J. Tur-Prats, Zherui Han, A. Beardo, Xiulin Ruan, F. X. Alvarez, 'High-Order Anharmonicities Shape Phonon Hydrodynamic Effects in Graphene', Nano Letters 25, 11203–11209 (2025), DOI: 10.1021/acs.nanolett.5c00855, https://doi.org/10.1021/acs.nanolett.5c00855
  18. Enrico Di Lucente, Michele Simoncelli, Nicola Marzari, 'Thermal transport in crystals: from the quantum Dyson equation to mesoscopic phonon hydrodynamics', Advances in Physics 75, 83–222 (2026), DOI: 10.1080/00018732.2026.2685376, https://doi.org/10.1080/00018732.2026.2685376 (arXiv:2608.13339)
  19. Navaneetha K. Ravichandran, 'Elasticity reshapes heat flow in graphene', preprint (2026), arXiv:2604.03910, https://arxiv.org/abs/2604.03910