Pith. sign in

REVIEW 4 major objections 7 minor 2 cited by

Fluctuation-dissipation and virtual processes in interacting phonon systems

T0 review · 4 major / 7 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Delta-function phonon scattering breaks a key thermodynamic law.

desk verdict A coherent formal argument for FD-consistent phonon linewidths, with a BAs application that is suggestive but rests on an untested Markovian approximation in the very regime where it matters. read the letter →

arxiv 2502.03362 v2 pith:5MWKAYDG submitted 2025-02-05 cond-mat.mtrl-sci

Aloïs Castellano , J. P. Alvarinhas Batista , Matthieu J. Verstraete This is my paper · ORCID
classification cond-mat.mtrl-sci
keywords phonon-phononinteractionsfluctuation-dissipationtheoremself-consistentlinewidthsthermalconductivityboronarsenideMarkovianapproximationmemorykernelvirtualphonons
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 argues that the Dirac delta functions enforcing energy conservation at each microscopic phonon-phonon collision are an approximation that violates the fluctuation-dissipation theorem. It replaces them with convolutions of phonon spectral functions, solved self-consistently, so that energy is conserved only between initial and final states and intermediate phonons are virtual. In the Markovian limit this reduces to Lorentzian-broadened scattering kernels with linewidth increments given by sums of phonon linewidths. Applied to boron arsenide, the self-consistent three-phonon linewidths reproduce experimental thermal conductivity, undercutting the prior conclusion that four-phonon processes dominate. The framework is universal but its numerical impact is system-dependent: negligible in silicon, substantial in silver iodide.

What carries the argument

The memory kernel $K_\lambda(t)$ of the generalized Langevin equation is the central object; its real part $\Gamma_\lambda(\omega)$ is built from a convolution $S_{\lambda'\lambda''}(\omega)$ of two lesser phonon correlation functions. The spectral function $\chi''_\lambda(\omega)$ depends on $\Gamma_\lambda(\omega)$ and its Hilbert-transform partner $\Delta_\lambda(\omega)$. Replacing the memory-less delta approximation by Lorentzian convolutions with widths $\Gamma_{\lambda'}+\Gamma_{\lambda''}$ and iterating equations (2), (3), and (6) to convergence is the self-consistency mechanism that restores fluctuation-dissipation balance.

What would settle it

Compute the fully frequency-dependent self-consistent three-phonon linewidths in boron arsenide without the Markovian limit, keeping the full convolution and the real-part frequency shift $\Delta_\lambda(\omega)$, and compare the resulting thermal conductivity with experiments at 300 K and 900 K; a significant departure from the Lorentzian Markovian result would falsify the numerical claim.

Watch

Extended reading notes

Core claim

The central claim is that the energy-conserving delta functions in the phonon self-energy stem from neglecting the memory kernel, which treats phonons as infinitely long-lived quasiparticles, and that this neglect breaks the fluctuation-dissipation relation. The correct treatment keeps the convolution of lesser correlation functions and enforces self-consistency between the memory kernel and the spectral function; in the Markovian limit, the delta functions become Lorentzians with widths given by sums of phonon linewidths. For boron arsenide, this self-consistent three-phonon calculation yields thermal conductivities in line with experiments, and the agreement of energy-conserving four-phonon calculations is argued to be a coincidence of an over-high three-phonon value and an over-large four-phonon correction.

Load-bearing premise

The numerical boron-arsenide result rests on the Markovian approximation: the memory kernel is reduced to a single number $\Gamma_\lambda(\Omega_\lambda)$ and the spectral function to a sum of Lorentzians with no frequency shift, while the non-Markovian validity tests are performed without the fluctuation-dissipation condition and without collective effects; if memory effects matter in boron arsenide, the self-consistent linewidths and the three-phonon conclusion would change.

Editorial extensions

If this is right

  • In boron arsenide, the strict energy-conservation three-phonon thermal conductivity is too high, the strict energy-conservation four-phonon correction is too large, and the two errors roughly cancel; treating three-phonon scattering self-consistently with the fluctuation-dissipation condition reproduces the experimental values without invoking four-phonon dominance.
  • The apparent dominance of four-phonon processes in boron arsenide is overestimated; four-phonon interactions are not negligible, but they are not needed to explain the measured thermal conductivity.
  • Interactions between acoustic and optical phonons that are forbidden by selection rules under strict energy conservation are reopened through virtual-phonon-mediated processes, which expands the scattering phase space in large-band-gap materials.
  • The Markovian self-consistent scheme avoids the delicate numerical convergence parameters used to approximate delta functions, because the Lorentzian convolution provides a broader, physically motivated smearing.
  • The fluctuation-dissipation condition is universal, but its practical effect is system-dependent: it barely changes silicon, while it lowers the three-phonon thermal conductivity of silver iodide by 15% to 25% between 100 and 300 K.
  • The method extends beyond three-phonon processes: including four-phonon and isotope scattering in the self-consistent memory kernel preserves the fluctuation-dissipation structure while incorporating all scattering channels.

Reading between the lines

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

  • If the violation is general, perturbative quasiparticle calculations that enforce on-shell energy conservation for intermediate states, such as electron-phonon and magnon-phonon problems, may need the same replacement; the paper notes that self-consistent electronic linewidths already resolve divergences in piezoelectric materials.
  • The Markovian assumption is the main limiting step for the quantitative boron-arsenide claim; a fully frequency-dependent self-consistent calculation with the Kramers-Kronig shift would determine whether the Lorentzian tails are quantitatively accurate.
  • A testable extension would apply the same self-consistent convolution to materials with very large acoustic-optical gaps other than boron arsenide, where strict energy-conservation selection rules are expected to overestimate thermal conductivity.
  • The paper's argument implies that energy-conservation tests in strongly anharmonic materials should be revisited with frequency-dependent spectral functions rather than delta function smearing, because the choice of Gaussian versus Lorentzian broadening can change the outcome.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. This Letter develops a self-consistent, fluctuation-dissipation (FD) framework for phonon linewidths within the authors' Mori-Zwanzig/mode-coupling theory of anharmonic lattice dynamics. The central formal result is that the standard delta-function expression for the three-phonon self-energy (Eq. 5) follows from inserting non-interacting spectral functions into the exact spectral convolution (Eq. 4), and that replacing the deltas by Lorentzian convolutions with self-consistently determined linewidths (Eq. 6 in the Markovian limit) restores consistency with the fluctuation-dissipation theorem at the level of individual scattering events. Applying this scheme to boron arsenide with an MLIP-based ab initio force-constant pipeline, the authors find that self-consistent three-phonon linewidths alone bring the thermal conductivity into the experimental range, and argue that the agreement of the energy-conserving (EC) three-phonon-plus-four-phonon calculation with experiment is coincidental. Silicon and silver iodide are presented as control systems with, respectively, negligible and moderate FD corrections.

Significance. If correct, the paper makes three contributions: a formal justification for previously proposed self-consistent phonon broadening schemes; the identification of long Lorentzian tails (rather than self-consistency per se) as the mechanism that activates acoustic-optical scattering across the BAs band gap; and a sharp, falsifiable prediction that four-phonon scattering is not required to explain the measured BAs conductivity. The calculations are parameter-free in the sense that matters: the linewidths are fixed points of a self-consistent integral equation, not fits to experiment, and the experimental values enter only as benchmarks. The MLIP is validated against DFT (force RMSE 0.016 eV/Å), the EC results reproduce published calculations, and the Si control behaves as expected for a gapless harmonic system, which strengthens the causal argument that the BAs effect stems from the gap. The main risk is that the numerical BAs claim rests on the Markovian and small-linewidth reductions in exactly the regime (large self-consistent linewidths) where those reductions are least controlled, as detailed in the major comments.

major comments (4)
  1. [SM S5, Eqs. (S38)-(S40), Figs. S8-S9] The validity test of the Markovian approximation is performed only in the energy-conserving (EC) regime; SM S5 states explicitly that the non-Markovian thermal conductivity is computed 'without the fluctuation dissipation theorem' and 'maintaining strict energy conservation'. Under EC the BAs three-phonon linewidths are small because the acoustic-optical gap forbids the relevant processes, so the small-linewidth conditions behind the Markovian reduction (Eqs. S29 and S32) hold by construction. The FD calculation is precisely the regime in which the linewidths become substantially larger (main text Fig. 2c) and in which the self-consistent feedback of Eq. (6) can broaden the spectral functions further. The central numerical claim of the Letter therefore rests on an approximation whose validity is untested in the regime that matters most. I ask that the authors either perform a non-Markovian FD test, for example by solving the frequency-dependent convolution in Eqs. (3) and (4) with G< built from the full susceptibility of Eq. (2) on a subset of modes or a coarser q-grid, or provide a quantitative bound on the Markovian error at the self-consistent FD linewidths.
  2. [SM S2.B, Eqs. (S29)-(S32); main text Eq. (6)] Independent of the memory-kernel question, the reduction to Eq. (6) involves two further approximations that are uncontrolled in the FD regime. First, the real part Δλ(ω) of the memory kernel is dropped: the full susceptibility of Eq. (2) contains the 2ωΔλ(ω) term in the denominator, whereas Eq. (S29) and the iteration cycle retain only Γλ = Γλ(Ωλ); since Γλ and Δλ are Kramers-Kronig partners, a substantially broadened FD linewidth implies a non-negligible shift. Second, the Bose factors n(±Ω')n(±Ω'')/n(ω) in Eq. (S32) are replaced by their values at the Lorentzian peaks under the 'small enough linewidth' assumption; at 100-300 K the scale kBT over which n(ω) varies can be comparable to the detunings and widths that control the acoustic-optical unblocking. These errors enter directly into the self-consistent linewidths that produce the BAs agreement, so they should be quantified, for example by comparing Eq. (6) with the direct convolution of Bose-weighted Lorentzian spectral functions over the full frequency range for a sample of representative modes.
  3. [Fig. 2b and SM Fig. S2] The complete FD treatment including four-phonon interactions appears as the 'fd 3ph+4ph' curve in SM Fig. S2, but the corresponding numbers are never quoted or discussed in the main text. The reader therefore cannot determine whether the fully converged FD 3ph+4ph result agrees with experiment or falls below it; should it fall below, the relationship between the virtual-phonon processes generated by the self-consistent three-phonon channel (the diagram of Fig. 1a is itself an effective four-phonon process) and the explicit fourth-order vertex Γ4ph of Eq. (S36) would need to be examined for double counting. In addition, the claim that the EC 3ph+4ph agreement with experiment is 'actually a coincidence' requires uncertainty estimates for the EC calculations, which are not provided; the Monte-Carlo noise visible in the FD convergence curve (Fig. S10) suggests that the EC numbers carry comparable scatter. Please report the numerical values of all six curves of Fig. S2 at 300 K with their variances and discuss the FD 3ph+4ph outcome explicitly.
  4. [Eqs. (4) and (5), SM S2.A] The headline statement that the delta functions of Eq. (5) 'violate the fluctuation-dissipation theorem' is stronger than the derivation supports. SM S2.A obtains Eq. (5) by inserting delta-function spectral functions into the formally exact convolution (Eq. 4), and the Bose factors of Eq. (5) (via the identity S26) preserve the detailed-balance structure of the rates; this is the standard lowest-order self-energy evaluated on free propagators, which is FDT-consistent at that order for an equilibrium system. What the manuscript actually establishes is the need for self-consistency between the spectral functions used to evaluate the memory kernel and the spectral function that the kernel generates, with the delta functions as the zero-broadening limit of that cycle. I recommend qualifying the FDT-violation wording, or demonstrating a concrete failure of the detailed-balance condition for the delta-function S.
minor comments (7)
  1. [SM S5, Eq. (S41)] The Markovian-limit substitution in Eq. (S41) appears to have a prefactor inconsistency: with χ''λ normalized as in Eq. (S29), one has ∫ dω χ''λ(ω)χ''λ(ω) = 1/(2πΓλ), so the replacement should read c(Ωλ)/(2πΓλ). Please check the constant, as it underlies the heuristic argument that strongly non-Markovian modes contribute little to κ.
  2. [Acknowledgements] The sentence 'Simulation time was awarded by by PRACE on Discoverer' contains a duplicated 'by'.
  3. [SM S5, Figs. S8-S9] The non-Markovian test is performed in the single-mode approximation and omits the collective (off-diagonal) transport contribution, which the text identifies as important in BAs; the 'we anticipate' statement should be replaced by an explicit estimate or a clearly stated limitation of the test.
  4. [SM S6, Fig. S10] The convergence procedure (ten iterations, mean over the last five) has no stated tolerance, and the iteration-to-iteration evolution in Fig. S10 shows visible oscillation; please report the numerical variance over the averaged iterations and define in the captions what the shaded areas in Figs. 2b and S2-S5 represent.
  5. [Ref. [7]] The supplemental material is cited with the placeholder 'URL-will-be-inserted-by-publisher'; a stable identifier is needed for review.
  6. [SM S6] The trained BAs MLIP and the self-consistent iteration script are not made available; releasing the potential and the numerical κ(T) data plotted in Fig. S2 would permit independent verification of the headline BAs result.
  7. [Main text, after Eq. (5)] The claims that the reasoning 'holds identically with higher orders' and that 'similar arguments will hold for all of the commonly used (perturbative) theories' are broader than the demonstrated derivation, which uses the mode-coupling decoupling of four-point correlations (SM Eq. S18); this generality should be substantiated or hedged.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the FD linewidths are a self-consistent fixed point of the derived GLE equations, and experimental conductivity is used only as a benchmark.

full rationale

The paper's central equations (2)-(6) form a closed self-consistency problem: Γλ is the fixed point of Γλ = (1/16)Σ|Ψ|²S(Ωλ; Γ), with S built from Lorentzian spectral functions whose widths are the Γλ themselves. The linewidths are not fitted to experimental thermal conductivity; the experimental values cited (Kang, Li, Tian) appear only as comparison points in Fig. 2b. The replacement of delta functions by Lorentzian convolutions is derived in the SM (S2.B) from the generalized Langevin equation and the fluctuation-dissipation relation G< = n(ω)χ'', not assumed as an input. The Markovian form (6) is attributed to the authors' prior work [7], but the derivation is reproduced in the manuscript (SM eqs S28-S33), so the step does not reduce to a bare self-citation. The self-consistency iteration is a numerical fixed-point scheme with stated details (10 iterations, last 5 averaged), not a tautological input-output relation. The SM S5 limitation that Markovian validity tests are performed only under strict energy conservation is an approximation-validity concern, not a circularity: it does not make the FD prediction equal to an input. No fitted parameter is renamed as a prediction, no uniqueness theorem is imported from the authors, and the qualitative role of Lorentzian tails is explicitly attributed to prior independent work (Leggett and d'Haar, ref [12]). The derivation chain is therefore self-contained against external benchmarks and warrants a non-circularity finding.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No parameter is fitted to experimental thermal conductivity. The phonon linewidths Gamma_lambda are self-consistently computed from the mode-coupling memory kernel; numerical choices such as grids, cutoffs, and 10 iterations are convergence parameters, not fitted constants. No new particles, forces, or conserved quantities are introduced; 'virtual phonons' are standard off-shell intermediate states used to interpret the convolution.

assumptions (5)
  • standard math Quantum equilibrium lattice dynamics is governed by the Mori-Zwanzig generalized Langevin equation with a memory kernel satisfying the fluctuation-dissipation theorem.
    Invoked at main text eq (1) and SM eqs (S5)-(S13); standard projection operator formalism.
  • domain assumption Four-point Kubo correlations can be decoupled into products of two-point correlations (mode-coupling approximation).
    Used in deriving the memory kernel expression eq (3) via SM eq (S18); uncontrolled beyond the stated mode-coupling order.
  • domain assumption Phonon dynamics in the systems studied are Markovian, so the memory kernel reduces to Gamma_lambda(Omega_lambda) and spectral functions are Lorentzian with negligible frequency shift Delta_lambda.
    Used in main text after eq (5) and in SM S2.B and S5; tested only outside the full FD self-consistent calculation.
  • domain assumption Linewidths are small enough that Bose-Einstein prefactors can be evaluated at the bare frequencies in eq (6) and eq (S33).
    Used to pass from the convolution of Lorentzians to a single Lorentzian with width Gamma_lambda' + Gamma_lambda''; could fail for very anharmonic or high-temperature cases.
  • domain assumption The MLIP trained on 122 DFT configurations accurately represents the Born-Oppenheimer surface for BAs at the relevant temperatures.
    Validated by RMSE and phonon dispersion comparison in SM figs S6 and S7, but the potential is not made available.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fluctuation-dissipation and virtual processes in interacting phonon systems." pith.science (2026). https://pith.science/paper/5MWKAYDG

@misc{pith2026250203362,
  author       = {Pith},
  title        = {Pith review of: Fluctuation-dissipation and virtual processes in interacting phonon systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5MWKAYDG}},
  note         = {Machine review of arXiv:2502.03362}
}
read the original abstract

Phonon-phonon interactions are fundamental to understanding a wide range of material properties, including thermal transport and vibrational spectra. In conventional perturbative approaches, energy conservation during each microscopic phonon interaction is enforced using delta functions. We demonstrate that these delta functions stem from an incomplete treatment, that violates the fluctuation-dissipation theorem governing systems at equilibrium. By replacing delta functions with convolutions and introducing a self-consistency condition for the phonon spectral function, we provide a more accurate and physically consistent framework. For systems where phonon dynamics can be approximated as Markovian, we simplify this approach, reducing the dissipative component to a single parameter tied to phonon lifetimes. Applying this method to boron arsenide, we find that self-consistent linewidths better capture the phonon scattering processes, significantly improving agreement with experimental thermal conductivity values. These results also challenge the conventional view of four-phonon processes as dominant in BAs, demonstrating the adequacy of a three-phonon description, provided it is self-consistent. With this method we address critical limitations of perturbative approaches, offering new insights into dissipation and phonon-mediated processes, and enabling more accurate modeling of anharmonic materials.

Figures

Figures reproduced from arXiv: 2502.03362 by the authors.

Figure 1
Figure 1. FIG. 1. Examples of second-order three-phonon Feynman di [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Self-Consistent Phonon Spectral Functions and Thermal Transport Beyond the Quasiparticle Approximation

    cond-mat.mtrl-sci 2026-07 conditional novelty 7.0 of 10

    Self-consistent spectral dressing of the three-phonon bubble broadens HgTe phonons enough to suppress lattice thermal conductivity fivefold to experiment without explicit higher-order scattering.

  2. Lattice thermal transport from phonon spectra beyond perturbation theory

    cond-mat.mtrl-sci 2026-04 unverdicted novelty 7.0 of 10

    Classical MD phonon spectral densities, inserted into the Wigner transport formula, reproduce PbTe conductivity and improve strongly anharmonic Cs3Bi2I6Cl3 without perturbative self-energies.

Reference graph

Works this paper leans on

40 extracted references · 31 canonical work pages · cited by 2 Pith papers

  1. [1]

    A. A. Maradudin and A. E. Fein, Scattering of neutrons by an anharmonic crystal, Phys. Rev. 128, 2589 (1962)

    work page 1962

  2. [2]

    W. Li, N. Mingo, L. Lindsay, D. A. Broido, D. A. Stew- art, and N. A. Katcho, Thermal conductivity of diamond nanowires from first principles, Phys. Rev. B 85, 195436 (2012)

    work page 2012

  3. [3]

    W. Li, J. Carrete, N. A. Katcho, and N. Mingo, Sheng- bte: A solver of the boltzmann transport equation for phonons, Comput. Phys. Commun. 185, 1747–1758 (2014)

    work page 2014

  4. [4]

    Z. Han, X. Yang, W. Li, T. Feng, and X. Ruan, Four- phonon: An extension module to shengbte for computing four-phonon scattering rates and thermal conductivity, Computer Physics Communications 270, 108179 (2022)

    2022

  5. [5]

    Castellano, J

    A. Castellano, J. P. A. Batista, and M. J. Verstraete, Mode-coupling theory of lattice dynamics for classical and quantum crystals, J. Chem. Phys. 159, 0174255 (2023)

    work page 2023

  6. [6]

    Castellano, J

    A. Castellano, J. P. A. Batista, O. Hellman, and M. J. Verstraete, Mode-coupling formulation of heat transport in anharmonic materials, Phys. Rev. B 111, 094306 (2025)

    work page 2025

  7. [7]

    See Supplemental Material at URL-will-be-inserted-by- 5 publisher for the derivation of the theory and the com- putational details, which include Refs [5, 6, 28–40]

    work page

  8. [8]

    Kubo, The fluctuation-dissipation theorem, Rep

    R. Kubo, The fluctuation-dissipation theorem, Rep. Prog. in Phys. 29, 255–284 (1966)

    work page 1966

Show all 40 references
  1. [9]

    Mori, Transport, collective motion, and brownian mo- tion, Prog

    H. Mori, Transport, collective motion, and brownian mo- tion, Prog. Theor. Phys. 33, 423–455 (1965)

    1965

  2. [10]

    Zwanzig, Memory effects in irreversible thermodynam- ics, Phys

    R. Zwanzig, Memory effects in irreversible thermodynam- ics, Phys. Rev. 124, 983 (1961)

    1961

  3. [11]

    Carruthers, Resonance in phonon-phonon scattering, Phys

    P. Carruthers, Resonance in phonon-phonon scattering, Phys. Rev. 125, 123 (1962)

    1962

  4. [12]

    for- bidden

    A. J. Leggett and D. t. Haar, Finite linewidths and “for- bidden” three-phonon interactions, Phys. Rev.139, A779 (1965)

    1965

  5. [13]

    Lin, S.-H

    Y.-Q. Lin, S.-H. Cao, C.-E. Hu, H.-Y. Geng, and X.- R. Chen, Intrinsic scattering channels and selection rules in four-phonon interactions, Phys. Rev. B 110, 075414 (2024)

    2024

  6. [14]

    Lindsay, D

    L. Lindsay, D. A. Broido, J. Carrete, N. Mingo, and T. L. Reinecke, Anomalous pressure dependence of ther- mal conductivities of large mass ratio compounds, Phys. Rev. B 91, 121202 (2015)

    2015

  7. [15]

    N. K. Ravichandran and D. Broido, Phonon-phonon in- teractions in strongly bonded solids: Selection rules and higher-order processes, Phys. Rev. X 10, 021063 (2020)

    2020

  8. [16]

    N. K. Ravichandran and D. Broido, Exposing the hidden influence of selection rules on phonon–phonon scattering by pressure and temperature tuning, Nat. Commun. 12, 3473 (2021)

    2021

  9. [17]

    R. Yang, S. Yue, Y. Quan, and B. Liao, Crystal symme- try based selection rules for anharmonic phonon-phonon scattering from a group theory formalism, Phys. Rev. B 103, 184302 (2021)

    2021

  10. [18]

    J. R. Yates, X. Wang, D. Vanderbilt, and I. Souza, Spec- tral and fermi surface properties from wannier interpola- tion, Phys. Rev. B 75, 195121 (2007)

    2007

  11. [19]

    X. Gu, Z. Fan, H. Bao, and C. Y. Zhao, Revisiting phonon-phonon scattering in single-layer graphene, Phys. Rev. B 100, 064306 (2019)

    2019

  12. [20]

    J. E. Turney, E. S. Landry, A. J. H. McGaughey, and C. H. Amon, Predicting phonon properties and thermal conductivity from anharmonic lattice dynamics calcula- tions and molecular dynamics simulations, Phys. Rev. B 79, 064301 (2009)

    2009

  13. [21]

    J. S. Kang, M. Li, H. Wu, H. Nguyen, and Y. Hu, Ex- perimental observation of high thermal conductivity in boron arsenide, Science 361, 575–578 (2018)

    2018

  14. [22]

    S. Li, Q. Zheng, Y. Lv, X. Liu, X. Wang, P. Y. Huang, D. G. Cahill, and B. Lv, High thermal conductivity in cu- bic boron arsenide crystals, Science 361, 579–581 (2018)

    2018

  15. [23]

    F. Tian, B. Song, X. Chen, N. K. Ravichandran, Y. Lv, K. Chen, S. Sullivan, J. Kim, Y. Zhou, T.-H. Liu, M. Goni, Z. Ding, J. Sun, G. A. G. Udalamatta Gamage, H. Sun, H. Ziyaee, S. Huyan, L. Deng, J. Zhou, A. J. Schmidt, S. Chen, C.-W. Chu, P. Y. Huang, D. Broido, L. Shi, G. Ch...

    2018

  16. [24]

    Lindsay, D

    L. Lindsay, D. A. Broido, and T. L. Reinecke, First- principles determination of ultrahigh thermal conductiv- ity of boron arsenide: A competitor for diamond?, Phys. Rev. Lett. 111, 025901 (2013)

    2013

  17. [25]

    T. Feng, L. Lindsay, and X. Ruan, Four-phonon scatter- ing significantly reduces intrinsic thermal conductivity of solids, Phys. Rev. B 96, 161201 (2017)

    2017

  18. [26]

    J.-M. Lihm, S. Ponc´ e, and C.-H. Park, Self-consistent electron lifetimes for electron-phonon scattering, Phys. Rev. B 110, L121106 (2024)

    2024

  19. [27]

    Lihm and S

    J.-M. Lihm and S. Ponc´ e, Non-perturbative self- consistent electron-phonon spectral functions and trans- port, arxiv , 2501.00468 (2024)

    2024 arXiv

  20. [28]

    A. P. Thompson, H. M. Aktulga, R. Berger, D. S. Bolin- tineanu, W. M. Brown, P. S. Crozier, P. J. in ’t Veld, A. Kohlmeyer, S. G. Moore, T. D. Nguyen, R. Shan, M. J. Stevens, J. Tranchida, C. Trott, and S. J. Plimpton, Lammps - a flexible simulation tool for particle-based mat...

    2022

  21. [29]

    I. S. Novikov, K. Gubaev, E. V. Podryabinkin, and A. V. Shapeev, The mlip package: moment tensor potentials with mpi and active learning, Mach. Learn.: Sci. Technol. 2, 025002 (2021)

    2021

  22. [30]

    Gonze, B

    X. Gonze, B. Amadon, G. Antonius, F. Arnardi, L. Baguet, J.-M. Beuken, J. Bieder, F. Bottin, J. Bouchet, E. Bousquet, N. Brouwer, F. Bruneval, G. Brunin, T. Cavignac, J.-B. Charraud, W. Chen, M. Cˆ ot´ e, S. Cottenier, J. Denier, G. Geneste, P. Ghosez, M. Giantomassi, Y. Gille...

    2020

  23. [31]

    A. H. Romero, D. C. Allan, B. Amadon, G. Anto- nius, T. Applencourt, L. Baguet, J. Bieder, F. Bot- tin, J. Bouchet, E. Bousquet, F. Bruneval, G. Brunin, D. Caliste, M. Cˆ ot´ e, J. Denier, C. Dreyer, P. Ghosez, M. Giantomassi, Y. Gillet, O. Gingras, D. R. Hamann, G. Hautier, F...

    2020

  24. [32]

    J. P. Perdew, A. Ruzsinszky, G. I. Csonka, O. A. Vydrov, G. E. Scuseria, L. A. Constantin, X. Zhou, and K. Burke, Restoring the density-gradient expansion for exchange in solids and surfaces, Phys. Rev. Lett. 100, 136406 (2008)

    2008

  25. [33]

    van Setten, M

    M. van Setten, M. Giantomassi, E. Bousquet, M. Ver- straete, D. Hamann, X. Gonze, and G.-M. Rignanese, The pseudodojo: Training and grading a 85 element op- timized norm-conserving pseudopotential table, Comput. Phys. Commun. 226, 39–54 (2018)

    2018

  26. [34]

    Hellman, I

    O. Hellman, I. A. Abrikosov, and S. I. Simak, Lattice dynamics of anharmonic solids from first principles, Phys. Rev. B 84, 180301 (2011)

    2011

  27. [35]

    Hellman, P

    O. Hellman, P. Steneteg, I. A. Abrikosov, and S. I. Simak, Temperature dependent effective potential method for accurate free energy calculations of solids, Phys. Rev. B 87, 104111 (2013)

    2013

  28. [36]

    Hellman and I

    O. Hellman and I. A. Abrikosov, Temperature-dependent effective third-order interatomic force constants from first 6 principles, Phys. Rev. B 88, 144301 (2013)

    2013

  29. [37]

    Knoop, N

    F. Knoop, N. Shulumba, A. Castellano, J. P. A. Batista, R. Farris, M. J. Verstraete, M. Heine, D. Broido, D. S. Kim, J. Klarbring, I. A. Abrikosov, S. I. Simak, and O. Hellman, Tdep: Temperature dependent effective po- tentials, J. Open Source Softw. 9, 6150 (2024)

    2024

  30. [38]

    Tamura, Isotope scattering of dispersive phonons in ge, Phys

    S.-I. Tamura, Isotope scattering of dispersive phonons in ge, Phys. Rev. B 27, 858 (1983)

    1983

  31. [39]

    Castellano, F

    A. Castellano, F. Bottin, J. Bouchet, A. Levitt, and G. Stoltz, ab initio canonical sampling based on varia- tional inference, Phys. Rev. B 106, L161110 (2022)

    2022

  32. [40]

    known” and “random

    A. Castellano, R. B´ ejaud, P. Richard, O. Nadeau, C. Du- val, G. Geneste, G. Antonius, J. Bouchet, A. Levitt, G. Stoltz, and F. Bottin, Machine learning assisted canonical sampling (mlacs) (2024). Supplementary material: Fluctuation-dissipation and virtual processes in intera...

    2024 arXiv

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.