Pith. sign in

REVIEW 1 major objections 5 minor 79 references

Fine-tuning a universal machine-learning potential on harmonic phonon data cuts error fourfold and enables temperature-dependent phonon prediction for 4,669 compounds, showing alkali metals and perovskites dominate anharmonic shifts.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 10:40 UTC pith:PQVA6G7E

load-bearing objection A useful high-throughput APRN screening dataset with a solid 12-material validation, but the harmonic-only fine-tuning assumption on anharmonic PES is asserted, not shown, and no artifacts are released. the 1 major comments →

arxiv 2601.09123 v1 pith:PQVA6G7E submitted 2026-01-14 cond-mat.mtrl-sci

Data-Driven Exploration and Insights into Temperature-Dependent Phonons in Inorganic Materials

Huiju Lee , Zhi Li , Jiangang he , Yi Xia This is my paper
classification cond-mat.mtrl-sci
keywords temperature-dependent phononsanharmonic phonon renormalizationmachine learning interatomic potentialM3GNetself-consistent harmonic approximationperovskiteslattice thermal conductivityhigh-throughput screening
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper aims to make temperature-dependent phonon calculations—normally prohibitively expensive—feasible for thousands of materials at once. It claims that fine-tuning a universal graph-neural-network interatomic potential on harmonic phonon data reduces phonon frequency errors by a factor of four and, crucially, preserves the potential's ability to describe anharmonicity. Coupling this refined potential to a streamlined single-shot implementation of the self-consistent harmonic approximation yields renormalized phonons at 300 K for 4,669 inorganic compounds. From this dataset the authors extract systematic trends—strong anharmonic shifts in alkali metals, perovskites, and related frameworks—and train machine-learning models that point to weak bonding, large atomic radii, and specific coordination motifs as drivers. First-principles validation on a dozen strongly anharmonic materials indicates that neglecting these effects can misestimate lattice thermal conductivity by factors of two to four.

Core claim

On its own terms, the paper's central claim is that a fine-tuned machine-learning interatomic potential, trained only on small-displacement force data generated from harmonic force constants, becomes accurate enough for anharmonic lattice dynamics. The fine-tuned M3GNet potential reduces the phonon frequency mean absolute error from 1.09 THz to 0.28 THz, and a single-shot stochastic self-consistent harmonic approximation using this potential reproduces measured phonons in SrTiO3 while yielding renormalized phonons for 4,669 compounds. The dataset reveals strong anharmonic shifts in alkali metals and perovskite-related frameworks, concentrated in weakly bonded, oversized coordination sites. F

What carries the argument

The central object is the fine-tuned M3GNet potential—a graph-neural-network interatomic potential pre-trained across the periodic table—used as the potential energy surface inside a streamlined stochastic self-consistent harmonic approximation (SSCHA). SSCHA computes temperature-dependent effective harmonic force constants by averaging the Hessian of the potential over configurations sampled from a quantum covariance matrix; the authors run it in a single shot without self-consistent iteration and use space-group symmetry to average only symmetry-inequivalent atom pairs. Two metrics, R_full and R_onsite, quantify the relative change in the full and onsite force-constant matrices after renor

Load-bearing premise

The entire screening rests on the assumption that fine-tuning a neural-network potential on small-displacement harmonic data (0.01–0.05 Å) implicitly corrects its anharmonic potential-energy surface, an assumption validated directly on only a dozen compounds out of 4,669.

What would settle it

Measure 300 K phonon dispersions by inelastic neutron or X-ray scattering for a high-R_onsite compound such as Cs2RbGaF6; if the observed low-frequency shifts deviate substantially from the fine-tuned-potential SSCHA predictions while matching harmonic DFT, the screening pipeline's anharmonic accuracy on unvalidated compounds is refuted.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • A fourfold reduction in phonon error means the fine-tuned potential can replace DFT for initial phonon screening in this database, cutting the cost of temperature-dependent phonon calculations by orders of magnitude.
  • Strong anharmonic renormalization concentrates in alkali metals and perovskite-derived structures, so those families are the natural candidates for low-thermal-conductivity thermoelectrics and for materials whose stability changes with temperature.
  • Random-forest feature rankings single out weak bonding, large atomic radii, and unfilled orbital counts as practical descriptors for predicting anharmonicity without running lattice-dynamics calculations.
  • First-principles checks on twelve materials show that ignoring anharmonic renormalization can change predicted lattice thermal conductivity by factors of two to four, implying that harmonic-only database predictions for strongly anharmonic compounds are unreliable.
  • Because the streamlined SSCHA produces auxiliary phonons that still match measured trends, the same pipeline can be extended to phonon linewidths, broader temperatures, and full self-consistency when higher accuracy is needed.

Where Pith is reading between the lines

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

  • The paper leaves untested whether fine-tuning on harmonic data alone repairs anharmonicity in general, since only twelve compounds receive direct DFT/SCPH validation. A decisive extension would run the same single-shot SSCHA with other universal potentials and compare the distribution of R_onsite across all 4,669 compounds.
  • The R_onsite heatmap could serve as an 'anharmonicity map' of the periodic table for materials design, but the paper stops at qualitative trends; converting it into a quantitative predictive descriptor would require propagating uncertainties from the potential into the force-constant ratios.
  • Because R_onsite measures force-constant changes rather than observable frequency shifts, some flagged materials may show weak phonon renormalization; pairing the metric with mode-resolved frequency shifts would sharpen screening.
  • The Cs2RbGaF6 case shows that four-phonon scattering can offset the three-phonon enhancement of thermal conductivity, implying that screening on APRN shifts alone may mispredict transport; coupling screening with higher-order scattering estimates would make thermal-conductivity predictions more reliable.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

1 major / 5 minor

Summary. The manuscript describes a high-throughput workflow for temperature-dependent phonon renormalization (APRN) in inorganic materials. The authors fine-tune the M3GNet machine-learning interatomic potential on displacement–force data generated from harmonic force constants in the MDR phonon database, then use a streamlined, single-shot stochastic self-consistent harmonic approximation (SSCHA) to compute 300 K renormalized force constants for 4,669 compounds. From this dataset they extract elemental and structural trends, train random-forest models to predict APRN from composition and local-structure features, and perform first-principles SCPH validation on 12 strongly anharmonic materials, including lattice thermal conductivity calculations. The paper claims a fourfold improvement in harmonic phonon prediction accuracy and a new capability to screen temperature-dependent phonon behavior across thousands of materials.

Significance. If the central transferability assumption is validated, the work would provide a valuable screening resource: a large APRN dataset, element/space-group trends, and ML-interpretability-derived descriptors for strongly anharmonic materials. The authors are commendably explicit about the single-shot nature of their SSCHA implementation and about the auxiliary-phonon interpretation. The 12-material SCPH validation and the SrTiO3 test are useful proof-of-concept exercises, and the commitment to release codes, data, and models is a strength. However, the screening claims and downstream trends rest on an anharmonic-transferability premise that is asserted rather than demonstrated, and the current validation strategy does not directly test the ML model's APRN predictions on a representative sample. The manuscript is therefore not yet at the level of a standard accept; the main claims are defensible but need additional load-bearing evidence.

major comments (1)
  1. [§II.A and §IV.A] The fine-tuning and the screening population both come from the same MDR database, making the harmonic baseline partly self-referential. The reported fourfold MAE reduction and the 0.28 THz accuracy are computed on a test set derived from the same harmonic force-constant database used for training, but with no demonstration that the improvement transfers to independent DFT phonon calculations for a broad set of materials. Please clarify whether the train/test split is by compound or by configuration, and provide an out-of-distribution validation on a small independent DFT-phonon set not used in fine-tuning.
minor comments (5)
  1. [§IV.B] Please specify whether the random train/test split is at the compound level or the configuration level. If configurations of the same compound appear in both training and test sets, the reported force MAE and phonon MAE may be optimistic.
  2. [Eq. (4)] R_onsite is defined with an absolute value, so the sign of the APRN shift (hardening vs. softening) is not retained. The heatmap in Fig. 3A therefore mixes positive and negative shifts; the authors should state whether the sign is analyzed anywhere, and whether the interpretation changes if signed ratios are used.
  3. [Fig. 3B / §II.C] The text first lists space groups 128, 107, 221, 87, and 225 as strongly anharmonic, then later refers to space groups 221 and 255 for perovskite structures. Please reconcile whether 255 or 225 is intended, and ensure the space-group numbering is consistent throughout.
  4. [§II.A] There is a typo: 'Perdew-Burke-Ernzerhof exchange-correlation functiional' should read 'functional.' Also, in the Fig. 4F caption, 'Mott-loffe-Regel' should be 'Mott-Ioffe-Regel.'
  5. [§IV.A] The statement that 'codes, data sets, and machine learning models will be made available via public repository upon acceptance' is appreciated, but for a data-driven manuscript of this type, providing a reviewer-accessible repository or at least a detailed data/feature table for the RF models would aid reproducibility.

Circularity Check

0 steps flagged

No significant circularity: APRN values are model outputs with independent DFT/experimental benchmarks, not fits to the reported trends.

full rationale

The derivation chain is not circular. Harmonic fine-tuning targets are displacement-force data generated from MDR harmonic IFCs (Methods IV.A), but the APRN values are not fit to those targets; they are obtained by evaluating the fine-tuned potential at SSCHA-displaced configurations (Eqs. 1-2) and comparing force constants before and after renormalization (Eqs. 3-4). The claimed accuracy improvement is measured on a held-out 5% split (Methods IV.A), so it is a genuine predictive benchmark rather than a fit renamed as prediction. The SrTiO3 comparison uses experimental data, and the 12-material lattice thermal conductivity study is carried out with independent DFT/SCPH calculations (Methods IV.C), providing external benchmarks. The random-forest feature-importance results are learned from the screening model's own R_full/R_onsite outputs; this makes them descriptive of the model/data pipeline rather than independent physical measurements, but it is not circular because the target variables are not composed of the input features. The statement that fine-tuning 'implicitly adjusts anharmonic components' is explicitly flagged as a posit (Section II.A) and is a validation-coverage limitation, not a circular reduction. Self-citations [21,79] are code/method citations for SCPH and 3/4-phonon calculations, not load-bearing uniqueness or ansatz assumptions.

Axiom & Free-Parameter Ledger

5 free parameters · 7 axioms · 0 invented entities

The screening pipeline rests on: (i) a fine-tuned M3GNet PES fitted to MDR harmonic forces, (ii) a non-self-consistent SSCHA approximation, and (iii) a hand-chosen R_onsite threshold. No new physical entities are introduced. The most fragile inputs are the MLIP's anharmonic transferability and the arbitrary filtering thresholds in the random forest models.

free parameters (5)
  • Fine-tuned M3GNet model weights = force MAE 0.0269 eV/Å
    Model parameters trained on MDR harmonic displacement-force data; the entire APRN screening depends on this fitted PES.
  • R_onsite screening threshold = 0.15
    Chosen by hand to define 'strong APRN'; determines the ~1,200 candidate materials.
  • Random forest outlier filters = 20% phonon error; 40% APRN frequency change
    Exclusion thresholds chosen by visual inspection, applied before training ML models; removes the strongest APRN cases from feature-importance analysis.
  • Fine-tuning displacement amplitude = 0.01–0.05 Å
    Range of random displacements used to generate training forces from harmonic IFCs; sets the scale of the anharmonic region sampled.
  • Box-Cox lambda for R_onsite = 0.074
    Fitted transformation parameter for target variable in local RF model.
axioms (7)
  • standard math Born-Oppenheimer potential energy surface and phonon quasiparticle picture
    Invoked in Introduction and throughout; the Taylor expansion and quasiparticle picture underpin all phonon calculations.
  • standard math Gibbs-Bogoliubov variational principle underlying SSCHA
    Basis of Eq. (1) and the SSCHA ensemble average; assumed valid.
  • domain assumption MDR phonon database is an accurate ground truth for harmonic IFCs
    Used as the fine-tuning target and the screening population; if MDR errors are large, the model inherits them.
  • domain assumption Pre-trained M3GNet provides a reasonable universal PES
    Starting point for fine-tuning; the original M3GNet's known phonon underestimation motivates the work.
  • ad hoc to paper Single-shot (non-self-consistent) SSCHA is sufficient for semiquantitative APRN
    Section II.B: the 'G0W0-like' one-shot approximation is stated to capture first-order correction; no systematic convergence test against full SSCHA is provided.
  • ad hoc to paper Fine-tuning on harmonic-only data implicitly corrects anharmonic components of the PES
    Section II.A: 'We posit that this fine-tuning not only refines harmonic interactions but also implicitly adjusts anharmonic components...' This is a load-bearing posit.
  • domain assumption Auxiliary SSCHA phonons correspond to experimental phonon dispersions
    Validated for SrTiO3; assumed to hold across the diverse MDR population.

pith-pipeline@v1.3.0-alltime-deepseek · 16348 in / 13318 out tokens · 122856 ms · 2026-08-03T10:40:16.385976+00:00 · methodology

0 comments
read the original abstract

Phonons, quantized vibrations of the atomic lattice, are fundamental to understanding thermal transport, structural stability, and phase behavior in crystalline solids. Despite advances in computational materials science, most predictions of vibrational properties in large materials databases rely on the harmonic approximation and overlook crucial temperature-dependent anharmonic effects. Here, we present a scalable computational framework that combines machine learning interatomic potentials, anharmonic lattice dynamics, and high-throughput calculations to investigate temperature-dependent phonons across thousands of materials. By fine-tuning the universal M3GNet interatomic potential using high-quality phonon data, we improve phonon prediction accuracy by a factor of four while preserving computational efficiency. Integrating this refined model into a high-throughput implementation of the stochastic self-consistent harmonic approximation, we compute temperature-dependent phonons for 4,669 inorganic compounds. Our analysis identifies systematic elemental and structural trends governing anharmonic phonon renormalization, with particularly strong manifestations in alkali metals, perovskite-derived frameworks, and related systems. Machine learning models trained on this dataset identify key atomic-scale features driving strong anharmonicity, including weak bonding, large atomic radii, and specific coordination motifs. First-principles validation confirms that anharmonic effects can dramatically alter lattice thermal conductivity by factors of two to four in some materials. This work establishes a robust and efficient data-driven approach for predicting finite-temperature phonon behavior, offering new pathways for the design and discovery of materials with tailored thermal and vibrational properties.

Figures

Figures reproduced from arXiv: 2601.09123 by Huiju Lee, Jiangang He, Yi Xia, Zhi Li.

Figure 1
Figure 1. Figure 1: FIG. 1. (A) Illustration of the M3GNet model, based on a graph neural network (GNN) and fine-tuned with the MDR [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Overview of anharmonic phonon renormalization (APRN) calculation and validation. (A) Schematic illustrating [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. (A) Heatmap of the APRN-induced percentage change of onsite force constants ( [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. (A) Schematic workflow for screening materials with strong temperature-dependent phonons (TDPH). First, for com [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗

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

79 extracted references · 1 canonical work pages

  1. [1]

    N. W. Ashcroft and N. D. Mermin, Solid state, Physics (New York: Holt, Rinehart and Winston) Appendix C1 (1976)

    1976

  2. [2]

    Wallace,Thermodynamics of Crystals, Dover Books on Physics (Dover Publications, 1998)

    D. Wallace,Thermodynamics of Crystals, Dover Books on Physics (Dover Publications, 1998)

    1998

  3. [3]

    Parlinski, Z

    K. Parlinski, Z. Li, and Y. Kawazoe, First-principles de- termination of the soft mode in cubic zro 2, Physical Review Letters78, 4063 (1997)

    1997

  4. [4]

    Grimvall, B

    G. Grimvall, B. Magyari-K¨ ope, V. Ozoli¸ nˇ s, and K. A. Persson, Lattice instabilities in metallic elements, Re- views of Modern Physics84, 945 (2012)

    2012

  5. [5]

    K. W. B¨ oer and U. W. Pohl, Elasticity and phonons, in Semiconductor physics(Springer, 2023) pp. 113–155

    2023

  6. [6]

    P. B. Allen, Zero-point and isotope shifts: relation to thermal shifts, Philosophical Magazine B70, 527 (1994)

    1994

  7. [7]

    Cowley, The lattice dynamics of an anharmonic crys- tal, Advances in Physics12, 421 (1963)

    R. Cowley, The lattice dynamics of an anharmonic crys- tal, Advances in Physics12, 421 (1963)

    1963

  8. [8]

    Monacelli, R

    L. Monacelli, R. Bianco, M. Cherubini, M. Calandra, I. Errea, and F. Mauri, The stochastic self-consistent har- monic approximation: calculating vibrational properties of materials with full quantum and anharmonic effects, Journal of Physics: Condensed Matter33, 363001 (2021)

    2021

  9. [9]

    Hellman, I

    O. Hellman, I. A. Abrikosov, and S. I. Simak, Lattice dynamics of anharmonic solids from first principles, Phys. Rev. B84, 180301 (2011)

    2011

  10. [10]

    Tadano and S

    T. Tadano and S. Tsuneyuki, Self-consistent phonon cal- culations of lattice dynamical properties in cubic srtio 3 with first-principles anharmonic force constants, Phys. Rev. B92, 054301 (2015)

    2015

  11. [11]

    Xia, Revisiting lattice thermal transport in pbte: The crucial role of quartic anharmonic- ity, Applied Physics Letters113, 073901 (2018), https://doi.org/10.1063/1.5040887

    Y. Xia, Revisiting lattice thermal transport in pbte: The crucial role of quartic anharmonic- ity, Applied Physics Letters113, 073901 (2018), https://doi.org/10.1063/1.5040887

    work page doi:10.1063/1.5040887 2018

  12. [12]

    Hohenberg and W

    P. Hohenberg and W. Kohn, Inhomogeneous electron gas, Physical review136, B864 (1964)

    1964

  13. [13]

    Kohn and L

    W. Kohn and L. J. Sham, Self-consistent equations in- cluding exchange and correlation effects, Physical review 140, A1133 (1965)

    1965

  14. [14]

    Delaire, Orbitally driven giant phonon anharmonicity in snse, Nature Physics11, 1063 (2015)

    O. Delaire, Orbitally driven giant phonon anharmonicity in snse, Nature Physics11, 1063 (2015)

    2015

  15. [15]

    J. D. Budai, J. Hong, M. E. Manley, E. D. Specht, C. W. Li, J. Z. Tischler, D. L. Abernathy, A. H. Said, B. M. Leu, L. A. Boatner,et al., Metallization of vanadium dioxide driven by large phonon entropy, Nature515, 535 (2014)

    2014

  16. [16]

    C. Li, A. May, H. Cao, X. Chen, A. Christianson, G. Ehlers, D. Singh, B. Sales,et al., Phonon self-energy and origin of anomalous neutron scattering spectra in snte and pbte thermoelectrics, Physical review letters 112, 175501 (2014)

    2014

  17. [17]

    Hellman, J

    O. Hellman, J. Herriman, H. Smith, J. Lin, N. Shulumba, J. Niedziela, C. Li, D. Abernathy, and B. Fultz, Nuclear quantum effect with pure anharmonicity and the anoma- lous thermal expansion of silicon, Proceedings of the Na- tional Academy of Sciences115, 1992 (2018)

    1992

  18. [18]

    Errea, F

    I. Errea, F. Belli, L. Monacelli, A. Sanna, T. Koretsune, T. Tadano, R. Bianco, M. Calandra, R. Arita, F. Mauri, et al., Quantum crystal structure in the 250-kelvin super- conducting lanthanum hydride, Nature578, 66 (2020)

    2020

  19. [19]

    K. Kim, E. Vetter, L. Yan, C. Yang, Z. Wang, R. Sun, Y. Yang, A. H. Comstock, X. Li, J. Zhou,et al., Chiral- phonon-activated spin seebeck effect, Nature Materials 22, 322 (2023)

    2023

  20. [20]

    T. Feng, L. Lindsay, and X. Ruan, Four-phonon scatter- ing significantly reduces intrinsic thermal conductivity of solids, Phys. Rev. B96, 161201 (2017)

    2017

  21. [21]

    Y. Xia, V. I. Hegde, K. Pal, X. Hua, D. Gaines, S. Pa- tel, J. He, M. Aykol, and C. Wolverton, High-throughput study of lattice thermal conductivity in binary rocksalt and zinc blende compounds including higher-order an- harmonicity, Phys. Rev. X10, 041029 (2020)

    2020

  22. [22]

    Tadano and W

    T. Tadano and W. A. Saidi, First-principles phonon quasiparticle theory applied to a strongly anharmonic halide perovskite, Phys. Rev. Lett.129, 185901 (2022)

    2022

  23. [23]

    Tadano, Y

    T. Tadano, Y. Gohda, and S. Tsuneyuki, Impact of rattlers on thermal conductivity of a thermoelectric clathrate: a first-principles study, Physical review letters 114, 095501 (2015)

    2015

  24. [24]

    Y. Xia, V. Ozoli¸ nˇ s, and C. Wolverton, Microscopic mech- anisms of glasslike lattice thermal transport in cubic cu12sb4s13 tetrahedrites, Phys. Rev. Lett.125, 085901 (2020)

    2020

  25. [25]

    Verdi, L

    C. Verdi, L. Ranalli, C. Franchini, and G. Kresse, Quan- tum paraelectricity and structural phase transitions in strontium titanate beyond density functional theory, Phys. Rev. Mater.7, L030801 (2023)

    2023

  26. [26]

    B. Peng, Y. Hu, S. Murakami, T. Zhang, and B. Mon- serrat, Topological phonons in oxide perovskites controlled by light, Science Advances6, eabd1618 (2020), https://www.science.org/doi/pdf/10.1126/sciadv.abd1618

    doi:10.1126/sciadv.abd1618 2020

  27. [27]

    L. E. Bell, Cooling, heating, generating power, and re- covering waste heat with thermoelectric systems, Science 321, 1457 (2008)

    2008

  28. [28]

    Bianco, I

    R. Bianco, I. Errea, L. Paulatto, M. Calandra, and F. Mauri, Second-order structural phase transitions, free energy curvature, and temperature-dependent anhar- monic phonons in the self-consistent harmonic approxi- mation: Theory and stochastic implementation, Physical Review B96, 014111 (2017)

    2017

  29. [29]

    O. T. Unke, S. Chmiela, H. E. Sauceda, M. Gastegger, I. Poltavsky, K. T. Schutt, A. Tkatchenko, and K.-R. Muller, Machine learning force fields, Chemical Reviews 121, 10142 (2021). 12

    2021

  30. [30]

    Chen and S

    C. Chen and S. P. Ong, A universal graph deep learn- ing interatomic potential for the periodic table, Nature Computational Science2, 718 (2022)

    2022

  31. [31]

    Batatia, P

    I. Batatia, P. Benner, Y. Chiang, A. M. Elena, D. P. Kov´ acs, J. Riebesell, X. R. Advincula, M. Asta, M. Avaylon, W. J. Baldwin,et al., A foundation model for atomistic materials chemistry, arXiv preprint arXiv:2401.00096 (2023)

    Pith/arXiv arXiv 2023

  32. [32]

    B. Deng, P. Zhong, K. Jun, J. Riebesell, K. Han, C. J. Bartel, and G. Ceder, Chgnet as a pretrained universal neural network potential for charge-informed atomistic modelling, Nature Machine Intelligence5, 1031 (2023)

    2023

  33. [33]

    Togo, Mdr phonon calculation database (2018)

    A. Togo, Mdr phonon calculation database (2018)

    2018

  34. [34]

    B. Deng, Y. Choi, P. Zhong, J. Riebesell, S. Anand, Z. Li, K. Jun, K. A. Persson, and G. Ceder, Systematic soften- ing in universal machine learning interatomic potentials, npj Computational Materials11, 1 (2025)

    2025

  35. [35]

    J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett. 77, 3865 (1996)

    1996

  36. [36]

    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

  37. [37]

    Stirling, Neutron inelastic scattering study of the lat- tice dynamics of strontium titanate: harmonic models, Journal of Physics C: Solid State Physics5, 2711 (1972)

    W. Stirling, Neutron inelastic scattering study of the lat- tice dynamics of strontium titanate: harmonic models, Journal of Physics C: Solid State Physics5, 2711 (1972)

    1972

  38. [38]

    Cowley, W

    R. Cowley, W. Buyers, and G. Dolling, Relationship of normal modes of vibration of strontium titanate and its antiferroelectric phase transition at 110 k, Solid State Communications7, 181 (1969)

    1969

  39. [39]

    Errea, M

    I. Errea, M. Calandra, and F. Mauri, Anharmonic free energies and phonon dispersions from the stochastic self- consistent harmonic approximation: Application to plat- inum and palladium hydrides, Phys. Rev. B89, 064302 (2014)

    2014

  40. [40]

    van Roekeghem, J

    A. van Roekeghem, J. Carrete, and N. Mingo, Anomalous thermal conductivity and suppression of negative thermal expansion in scf 3, Phys. Rev. B94, 020303 (2016)

    2016

  41. [41]

    Hedin, New method for calculating the one-particle green’s function with application to the electron-gas problem, Physical Review139, A796 (1965)

    L. Hedin, New method for calculating the one-particle green’s function with application to the electron-gas problem, Physical Review139, A796 (1965)

    1965

  42. [42]

    M. S. Hybertsen and S. G. Louie, Electron correlation in semiconductors and insulators: Band gaps and quasipar- ticle energies, Physical Review B34, 5390 (1986)

    1986

  43. [43]

    Mahan, Many-body physics (2000)

    G. Mahan, Many-body physics (2000)

    2000

  44. [44]

    D. B. Straus, S. Guo, A. M. Abeykoon, and R. J. Cava, Understanding the instability of the halide perovskite cspbi3 through temperature-dependent structural anal- ysis, Advanced Materials32, 2001069 (2020)

    2020

  45. [45]

    Lav´ en, E

    R. Lav´ en, E. Fransson, P. Erhart, F. Juranyi, G. E. Granroth, and M. Karlsson, Unraveling the nature of vi- brational dynamics in cspbi3 by inelastic neutron scat- tering and molecular dynamics simulations, The Journal of Physical Chemistry Letters16, 4812 (2025)

    2025

  46. [46]

    Mukhopadhyay, D

    S. Mukhopadhyay, D. S. Parker, B. C. Sales, A. A. Puret- zky, M. A. McGuire, and L. Lindsay, Two-channel model for ultralow thermal conductivity of crystalline tl3vse4, Science360, 1455 (2018)

    2018

  47. [47]

    Z. Zeng, X. Shen, R. Cheng, O. Perez, N. Ouyang, Z. Fan, P. Lemoine, B. Raveau, E. Guilmeau, and Y. Chen, Push- ing thermal conductivity to its lower limit in crystals with simple structures, Nature Communications15, 3007 (2024)

    2024

  48. [48]

    Y. Xia, K. Pal, J. He, V. Ozoli¸ nˇ s, and C. Wolverton, Par- ticlelike phonon propagation dominates ultralow lattice thermal conductivity in crystalline tl 3vse4, Phys. Rev. Lett.124, 065901 (2020)

    2020

  49. [49]

    Dutta, S

    M. Dutta, S. Matteppanavar, M. V. Prasad, J. Pandey, A. Warankar, P. Mandal, A. Soni, U. V. Waghmare, and K. Biswas, Ultralow thermal conductivity in chain-like tlse due to inherent tl+ rattling, Journal of the American Chemical Society141, 20293 (2019)

    2019

  50. [50]

    K. Pal, Y. Xia, and C. Wolverton, Microscopic mecha- nism of unusual lattice thermal transport in tlinte2, npj Computational Materials7, 5 (2021)

    2021

  51. [51]

    M. K. Jana, K. Pal, A. Warankar, P. Mandal, U. V. Waghmare, and K. Biswas, Intrinsic rattler-induced low thermal conductivity in zintl type tlinte2, Journal of the American Chemical Society139, 4350 (2017)

    2017

  52. [52]

    Errea, M

    I. Errea, M. Calandra, and F. Mauri, Anharmonic free energies and phonon dispersions from the stochastic self- consistent harmonic approximation: Application to plat- inum and palladium hydrides, Physical Review B89, 064302 (2014)

    2014

  53. [53]

    Ceriotti, W

    M. Ceriotti, W. Fang, P. G. Kusalik, R. H. McKenzie, A. Michaelides, M. A. Morales, and T. E. Markland, Nu- clear quantum effects in water and aqueous systems: Ex- periment, theory, and current challenges, Chemical re- views116, 7529 (2016)

    2016

  54. [54]

    E. L. Da Silva, J. M. Skelton, S. C. Parker, and A. Walsh, Phase stability and transformations in the halide per- ovskite cssni 3, Physical Review B91, 144107 (2015)

    2015

  55. [55]

    R. X. Yang, J. M. Skelton, E. L. da Silva, J. M. Frost, and A. Walsh, Spontaneous octahedral tilting in the cubic inorganic cesium halide perovskites cssnx3 and cspbx3 (x = f, cl, br, i), The Journal of Physical Chemistry Letters 8, 4720 (2017)

    2017

  56. [56]

    A. E. Hoffman, R. A. Saha, S. Borgmans, P. Puech, T. Braeckevelt, M. B. Roeffaers, J. A. Steele, J. Hofkens, and V. Van Speybroeck, Understanding the phase tran- sition mechanism in the lead halide perovskite cspbbr3 via theoretical and experimental giwaxs and raman spec- troscopy, Apl Materials11(2023)

    2023

  57. [57]

    Klarbring, O

    J. Klarbring, O. Hellman, I. A. Abrikosov, and S. I. Simak, Anharmonicity and ultralow thermal conductiv- ity in lead-free halide double perovskites, Physical review letters125, 045701 (2020)

    2020

  58. [58]

    L. Ward, A. Dunn, A. Faghaninia, N. E. Zimmermann, S. Bajaj, Q. Wang, J. Montoya, J. Chen, K. Bystrom, M. Dylla,et al., Matminer: An open source toolkit for materials data mining, Computational Materials Science 152, 60 (2018)

    2018

  59. [59]

    M. J. Mehl, D. Hicks, C. Toher, O. Levy, R. M. Hanson, G. Hart, and S. Curtarolo, The aflow library of crystal- lographic prototypes: part 1, Computational Materials Science136, S1 (2017)

    2017

  60. [60]

    Hicks, M

    D. Hicks, M. J. Mehl, E. Gossett, C. Toher, O. Levy, R. M. Hanson, G. Hart, and S. Curtarolo, The aflow li- brary of crystallographic prototypes: part 2, Computa- tional Materials Science161, S1 (2019)

    2019

  61. [61]

    Hicks, M

    D. Hicks, M. J. Mehl, M. Esters, C. Oses, O. Levy, G. L. Hart, C. Toher, and S. Curtarolo, The aflow library of crystallographic prototypes: part 3, Computational Ma- terials Science199, 110450 (2021)

    2021

  62. [62]

    Eckert, S

    H. Eckert, S. Divilov, M. J. Mehl, D. Hicks, A. C. Zettel, M. Esters, X. Campilongo, and S. Curtarolo, The aflow library of crystallographic prototypes: Part 4, Computa- 13 tional Materials Science240, 112988 (2024)

    2024

  63. [63]

    S.-P. Guo, Z. Ma, J.-C. Li, and H.-G. Xue, First inves- tigation of the electrochemical performance ofγ-lifeo 2 micro-cubes as promising anode material for lithium-ion batteries, Journal of Materials Science52, 1469 (2017)

    2017

  64. [64]

    A. Jain, S. P. Ong, G. Hautier, W. Chen, W. D. Richards, S. Dacek, S. Cholia, D. Gunter, D. Skinner, G. Ceder, et al., Commentary: The materials project: A materials genome approach to accelerating materials innovation, APL materials1(2013)

    2013

  65. [65]

    Jong, C.-J

    U.-G. Jong, C.-J. Yu, Y.-H. Kye, S.-H. Choe, J.-S. Kim, and Y.-G. Choe, Anharmonic phonons and phase transi- tions in the vacancy-ordered double perovskite cs 2 sni 6 from first-principles predictions, Physical Review B99, 184105 (2019)

    2019

  66. [66]

    Bhumla, M

    P. Bhumla, M. Jain, S. Sheoran, and S. Bhattacharya, Vacancy-ordered double perovskites cs2bi6 (b= pt, pd, te, sn): an emerging class of thermoelectric materials, The Journal of Physical Chemistry Letters13, 11655 (2022)

    2022

  67. [67]

    K. Ueda, R. Kaneko, A. Subedi, M. Minola, B. Kim, J. Fujioka, Y. Tokura, and B. Keimer, Phonon anoma- lies in pyrochlore iridates studied by raman spectroscopy, Physical Review B100, 115157 (2019)

    2019

  68. [68]

    Kumar, V

    H. Kumar, V. Sathe, and A. Pramanik, Spin–phonon and electron–phonon coupling in pyrochlore iridates (y1–x pr x) 2ir2o7: An investigation with raman spectroscopy, The Journal of Physical Chemistry C127, 13178 (2023)

    2023

  69. [69]

    Kresse and J

    G. Kresse and J. Hafner,Ab initiomolecular dynamics for liquid metals, Phys. Rev. B47, 558 (1993)

    1993

  70. [70]

    Kresse and J

    G. Kresse and J. Hafner,Ab initiomolecular-dynamics simulation of the liquid-metal˘amorphous-semiconductor transition in germanium, Phys. Rev. B49, 14251 (1994)

    1994

  71. [71]

    Kresse and J

    G. Kresse and J. Furthm¨ uller, Comput. Mater. Sci.6, 15 (1996)

    1996

  72. [72]

    Kresse and J

    G. Kresse and J. Furthm¨ uller, Efficient iterative schemes forab initiototal-energy calculations using a plane-wave basis set, Phys. Rev. B54, 11169 (1996)

    1996

  73. [73]

    P. E. Bl¨ ochl, Projector augmented-wave method, Phys. Rev. B50, 17953 (1994)

    1994

  74. [74]

    J. P. Perdew, K. Burke, and Y. Wang, Generalized gra- dient approximation for the exchange-correlation hole of a many-electron system, Phys. Rev. B54, 16533 (1996)

    1996

  75. [75]

    Hohenberg and W

    P. Hohenberg and W. Kohn, Inhomogeneous electron gas, Phys. Rev.136, B864 (1964)

    1964

  76. [76]

    Togo and I

    A. Togo and I. Tanaka, First principles phonon calcu- lations in materials science, Scripta Materialia108, 1 (2015)

    2015

  77. [77]

    F. Zhou, W. Nielson, Y. Xia, and V. Ozoli¸ nˇ s, Lattice an- harmonicity and thermal conductivity from compressive sensing of first-principles calculations, Phys. Rev. Lett. 113, 185501 (2014)

    2014

  78. [78]

    F. Zhou, W. Nielson, Y. Xia, and V. Ozolins, Compres- sive sensing lattice dynamics. i. general formalism, Phys. Rev. B100, 184308 (2019)

    2019

  79. [79]

    Xia, Revisiting lattice thermal transport in pbte: The crucial role of quartic anharmonicity, Applied Physics Letters113(2018)

    Y. Xia, Revisiting lattice thermal transport in pbte: The crucial role of quartic anharmonicity, Applied Physics Letters113(2018)

    2018