Pith. sign in

REVIEW 2 major objections 1 minor 33 references

Intermetallic compounds like LiBePt2 can reach conductivities comparable to noble metals by shifting high-scattering d-states below the Fermi level.

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 · grok-4.3

2026-06-30 16:47 UTC pith:4OWTTWWD

load-bearing objection Large ML+DFT screen flags LiBePt2 and similar compounds as high-conductivity candidates, but the surrogate model's ranking accuracy is not shown. the 2 major comments →

arxiv 2605.22167 v2 pith:4OWTTWWD submitted 2026-05-21 cond-mat.mtrl-sci cond-mat.other

High-throughput study of electrical conductivity in ordered metals

Thalis H. B. da Silva , Hai-Chen Wang , Tiago F. T. Cerqueira , Simone Di Cataldo , Silvana Botti , Miguel A. L. Marques This is my paper
classification cond-mat.mtrl-sci cond-mat.other
keywords electrical conductivityintermetallic compoundshigh-throughput screeningmachine learningab initio calculationselectron-phonon couplingmetallic transportnoble metals
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 develops a computational framework combining machine learning with high-throughput ab initio calculations to screen over 2.8 million compounds for metallic transport. It identifies several ordered intermetallics predicted to match or approach the conductivity of aluminum. Analysis shows noble metals set a practical ceiling due to their electronic structure, while certain compounds use valence electrons from light elements to move high-scattering d-states away from the Fermi level. A reader would care because this points to concrete design rules for new high-conductivity materials. Full electron-phonon calculations on top candidates match available experiments.

Core claim

By integrating machine learning with high-throughput ab initio calculations, the study screens millions of compounds and finds that while noble metals define the conductivity ceiling, compounds like LiBePt2 can achieve comparable performance by utilizing valence electrons from light elements to shift high-scattering d-states beneath the Fermi level, with full electron-phonon calculations confirming the predictions.

What carries the argument

The mechanism by which valence electrons from light elements shift high-scattering d-states below the Fermi level in ordered intermetallics to reduce scattering.

Load-bearing premise

The machine-learning surrogate model produces sufficiently accurate conductivity rankings that the top candidates selected for full calculations are not false positives.

What would settle it

Room-temperature conductivity measurement on synthesized LiBePt2 that falls substantially below the predicted value near aluminum.

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

If this is right

  • Several intermetallic candidates are predicted to have conductivities comparable to aluminum.
  • Full electron-phonon coupling calculations for the top materials agree with available experimental data.
  • The screening approach can identify novel high-performance conductors beyond noble metals.
  • Combining statistical learning with detailed ab initio calculations demonstrates predictive power for metallic transport.

Where Pith is reading between the lines

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

  • The same light-element valence-electron strategy might extend to designing materials with improved thermal conductivity or other transport properties.
  • These compounds could provide more abundant alternatives to noble metals for applications requiring high conductivity.
  • Synthesis and direct measurement of LiBePt2 would provide an immediate experimental test of the mechanism.
  • Applying the framework to disordered systems or additional structure types could uncover further candidates.

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

2 major / 1 minor

Summary. The paper presents a computational framework combining machine learning with high-throughput ab initio calculations to screen over 2.8 million ordered metallic compounds for electrical conductivity. It identifies top candidates including LiBePt2 with predicted conductivities comparable to aluminum (36.59 × 10^6 S/m), attributes this performance to valence electrons from light elements shifting high-scattering d-states below the Fermi level, and reports that full electron-phonon calculations on the ML-selected top materials agree with available experimental data.

Significance. If the ML surrogate reliably ranks candidates, the work would provide both new high-conductivity intermetallic candidates and a mechanistic route to exceed the practical limits set by noble metals, demonstrating a scalable workflow that pairs statistical screening with detailed first-principles validation.

major comments (2)
  1. [Abstract] Abstract: the central claim that LiBePt2 and similar compounds achieve noble-metal-comparable conductivities rests on the ML surrogate correctly identifying the true top performers from the 2.8 million compound screen so that subsequent full electron-phonon calculations are performed on genuine high-conductors rather than false positives; however, the manuscript provides no quantitative validation (error metrics, ranking precision, or hold-out performance) of the surrogate in the relevant high-conductivity regime.
  2. [Workflow description] Workflow description: without reported details on the surrogate's training set, features, or performance metrics for conductivity ranking, it remains unclear whether the selected top materials for full calculations represent actual high-performers or potential artifacts of surrogate mis-ranking.
minor comments (1)
  1. [Abstract] The abstract states 'excellent agreement with available experimental data' but does not specify the materials compared or the quantitative measures (e.g., percentage error) used to assess agreement.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. The two major comments correctly identify that the manuscript does not provide quantitative validation or workflow details for the ML surrogate used in the 2.8 million compound screen. We address each point below and will revise the manuscript to supply the requested information.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that LiBePt2 and similar compounds achieve noble-metal-comparable conductivities rests on the ML surrogate correctly identifying the true top performers from the 2.8 million compound screen so that subsequent full electron-phonon calculations are performed on genuine high-conductors rather than false positives; however, the manuscript provides no quantitative validation (error metrics, ranking precision, or hold-out performance) of the surrogate in the relevant high-conductivity regime.

    Authors: We agree that the absence of surrogate-specific validation metrics is a limitation. The current manuscript reports only that full electron-phonon calculations on the selected top candidates agree with experiment, but does not quantify the surrogate's ranking reliability or error in the high-conductivity regime. In revision we will add a new subsection (Methods and Results) that reports the surrogate training set size and composition, the features employed, mean absolute error and ranking metrics on hold-out data, and a specific assessment of precision/recall for the top 0.1 % conductivity tail. This will allow direct evaluation of whether the selected materials are likely true high-performers. revision: yes

  2. Referee: [Workflow description] Workflow description: without reported details on the surrogate's training set, features, or performance metrics for conductivity ranking, it remains unclear whether the selected top materials for full calculations represent actual high-performers or potential artifacts of surrogate mis-ranking.

    Authors: The referee is correct that these details are missing from the workflow description. The manuscript currently states only that an ML surrogate was used to screen 2.8 million compounds before full calculations on the top candidates. We will expand the Methods section to include the exact training-set construction (number of compounds with prior full calculations), the feature vector, hyper-parameter choices, and quantitative performance metrics (MAE, Spearman rank correlation, and top-k precision). We will also add a short discussion of possible surrogate limitations in the high-conductivity regime and how the subsequent first-principles validation step addresses them. revision: yes

Circularity Check

0 steps flagged

No circularity; workflow uses external experimental benchmarks and independent ab initio calculations

full rationale

The paper's chain consists of ML screening of 2.8M compounds followed by full electron-phonon calculations on selected candidates, with direct comparison to external experimental conductivity data. No equations or steps reduce by construction to fitted parameters, self-definitions, or self-citation chains; the validation against independent measurements keeps the derivation self-contained and non-circular.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract alone supplies no explicit free parameters, axioms, or invented entities; the screening relies on standard DFT and ML approximations whose details are not visible.

pith-pipeline@v0.9.1-grok · 5709 in / 1015 out tokens · 25983 ms · 2026-06-30T16:47:45.679582+00:00 · methodology

0 comments
read the original abstract

We present a computational framework that integrates machine learning with high-throughput \textit{ab initio} calculations to screen over 2.8 million compounds for metallic transport. We identify several intermetallic candidates with predicted high conductivities comparable to that of aluminum ($36.59 \times 10^6$~S/m). We perform full electron--phonon coupling calculations for the top-performing materials, yielding results in excellent agreement with available experimental data. Our analysis reveals that while the noble metals (Ag, Au, Cu) define the practical ceiling for conductivity due to their unique electronic structure and low scattering, compounds like $\text{LiBePt}_2$ can achieve comparable performance by utilizing valence electrons from light elements to shift high-scattering $d$-states beneath the Fermi level. This study not only identifies novel high-performance conductors but also demonstrates the predictive power of combining statistical learning with detailed ab initio calculations.

Figures

Figures reproduced from arXiv: 2605.22167 by Hai-Chen Wang, Miguel A. L. Marques, Silvana Botti, Simone Di Cataldo, Thalis H. B. da Silva, Tiago F. T. Cerqueira.

Figure 1
Figure 1. Figure 1: FIG. 1. Schematic overview of the computational workflow [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Electronic contribution to the electrical conductivity, [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Electronic contribution to the electrical conductiv [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Symbolic regression of the experimental conductivity [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Electronic (top) and phonon (bottom) band struc [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. Electronic (top) and phonon (bottom) band struc [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗

Review history (2 revisions) →

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

33 extracted references

  1. [1]

    First-principles dynamics of electrons and phonons*.Eur

    Marco Bernardi. First-principles dynamics of electrons and phonons*.Eur. Phys. J. B., 89(11):239, November 2016

    2016

  2. [2]

    Théo Cavignac, Jonathan Schmidt, Pierre-Paul De Breuck, Antoine Loew, Tiago F Cerqueira, Haichen Wang, Anton Bochkarev, Yury Vyatcheslavovich Lyso- gorskiy, Aldo H Romero, Ralf Drautz, Silvana Botti, and Miguel A. L. Marques. AI-Driven expansion and application of the Alexandria database.JPhys Mater., April 2026

    2026

  3. [3]

    Atomistic line graph neural network for improved materials property predictions.npj Comput

    Kamal Choudhary and Brian DeCost. Atomistic line graph neural network for improved materials property predictions.npj Comput. Mater., 7(1):185, November 2021

    2021

  4. [4]

    Hart, Michal Jahnatek, Roman V

    Stefano Curtarolo, Wahyu Setyawan, Gus L.W. Hart, Michal Jahnatek, Roman V. Chepulskii, Richard H. Tay- lor, Shidong Wang, Junkai Xue, Kesong Yang, Ohad Levy, Michael J. Mehl, Harold T. Stokes, Denis O. Dem- chenko, and Dane Morgan. Aflow: An automatic frame- work for high-throughput materials discovery.Comp. Mater. Sci., 58:218–226, June 2012

    2012

  5. [5]

    Robust model benchmarking and bias- imbalance in data-driven materials science: a case study on MODNet.J

    Pierre-Paul De Breuck, Matthew L Evans, and Gian- Marco Rignanese. Robust model benchmarking and bias- imbalance in data-driven materials science: a case study on MODNet.J. Phys.: Condens. Matter, 33(40):404002, July 2021

    2021

  6. [6]

    An- harmonic stabilization of the high-pressure simple cubic phase of calcium.Phys

    Ion Errea, Bruno Rousseau, and Aitor Bergara. An- harmonic stabilization of the high-pressure simple cubic phase of calcium.Phys. Rev. Lett., 106(16):165501, April 2011

    2011

  7. [7]

    Advanced capabilities for materials modelling with Quantum ESPRESSO.J

    P Giannozzi, O Andreussi, T Brumme, O Bunau, M Buongiorno Nardelli, M Calandra, R Car, C Cavaz- zoni, D Ceresoli, M Cococcioni, N Colonna, I Carnimeo, A Dal Corso, S de 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 Kawamura, H-Y Ko, A Kokalj, E Küçükbenli, M Lazzeri, M...

    2017

  8. [8]

    Phys.: Condens

    Paolo Giannozzi, Stefano Baroni, Nicola Bonini, Mat- teo Calandra, Roberto Car, Carlo Cavazzoni, Davide Ceresoli, Guido L Chiarotti, Matteo Cococcioni, Ismaila Dabo, Andrea Dal Corso, Stefano de Gironcoli, Ste- fano Fabris, Guido Fratesi, Ralph Gebauer, Uwe Ger- stmann, Christos Gougoussis, Anton Kokalj, Michele Lazzeri, Layla Martin-Samos, Nicola Marzari...

    2009

  9. [9]

    Cohen, and Steven G

    Feliciano Giustino, Marvin L. Cohen, and Steven G. Louie. Electron-phonon interaction using Wannier func- tions.Phys. Rev. B, 76(16):165108, October 2007

    2007

  10. [10]

    High-pressure phases of cal- cium: Prediction of phase VI and upper-pressure phases from first principles.Phys

    Takahiro Ishikawa, Hitose Nagara, Naoshi Suzuki, Taku Tsuchiya, and Jun Tsuchiya. High-pressure phases of cal- cium: Prediction of phase VI and upper-pressure phases from first principles.Phys. Rev. B, 81(9):092104, March 2010. 10

    2010

  11. [11]

    Anubhav Jain, Shyue Ping Ong, Geoffroy Hautier, Wei Chen, William Davidson Richards, Stephen Dacek, Shreyas Cholia, Dan Gunter, David Skinner, Gerbrand Ceder, and Kristin A. Persson. Commentary: The mate- rials project: A materials genome approach to accelerat- ing materials innovation.APL Mater., 1(1):011002, July 2013

    2013

  12. [12]

    M. I. Katsnelson, A. V. Trefilov, M. N. Khlopkin, and K. Yu. Khromov. Peculiarities of anharmonic lattice dynamics and thermodynamics of alkaline-earth metals. Philos. Mag. B, 81(12):1893–1913, December 2001

    1913

  13. [13]

    Margine, and Feliciano Giustino

    Hyungjun Lee, Samuel Poncé, Kyle Bushick, Samad Ha- jinazar, Jon Lafuente-Bartolome, Joshua Leveillee, Chao Lian, Jae-Mo Lihm, Francesco Macheda, Hitoshi Mori, Hari Paudyal, Weng Hong Sio, Sabyasachi Tiwari, Mar- ios Zacharias, Xiao Zhang, Nicola Bonini, Emmanouil Kioupakis, Elena R. Margine, and Feliciano Giustino. Electron–phonon physics from first prin...

    2023

  14. [14]

    Lide, editor.CRC Handbook of Chemistry and Physics

    David R. Lide, editor.CRC Handbook of Chemistry and Physics. CRC Press, Boca Raton, FL, internet version 2005 edition, 2005

    2005

  15. [15]

    Madsen, Jesús Carrete, and Matthieu J

    Georg K.H. Madsen, Jesús Carrete, and Matthieu J. Ver- straete. BoltzTraP2, a program for interpolating band structures and calculating semi-classical transport coeffi- cients.Comput. Phys. Commun., 231:140–145, October 2018

    2018

  16. [16]

    Mostofi, Jonathan R

    Arash A. Mostofi, Jonathan R. Yates, Giovanni Pizzi, Young-Su Lee, Ivo Souza, David Vanderbilt, and Nicola Marzari. An updated version of wannier90: A tool for ob- taining maximally-localised Wannier functions.Comput. Phys. Commun., 185(8):2309–2310, August 2014

    2014

  17. [17]

    Orb: A fast, scalable neural network potential.arXiv, 2024

    Mark Neumann, James Gin, Benjamin Rhodes, Steven Bennett, Zhiyi Li, Hitarth Choubisa, Arthur Hussey, and Jonathan Godwin. Orb: A fast, scalable neural network potential.arXiv, 2024

    2024

  18. [18]

    Malone, Cheol-Hwan Park, Steven G

    Jesse Noffsinger, Feliciano Giustino, Brad D. Malone, Cheol-Hwan Park, Steven G. Louie, and Marvin L. Cohen. EPW: A program for calculating the elec- tron–phononcouplingusingmaximallylocalizedWannier functions.Comput. Phys. Commun., 181(12):2140–2148, December 2010

    2010

  19. [19]

    Oganov, Yanming Ma, Ying Xu, Ion Errea, Aitor Bergara, and Andriy O

    Artem R. Oganov, Yanming Ma, Ying Xu, Ion Errea, Aitor Bergara, and Andriy O. Lyakhov. Exotic behavior and crystal structures of calcium under pressure.Proc. Natl. Acad. Sci., 107(17):7646–7651, April 2010

    2010

  20. [20]

    Perdew, Adrienn Ruzsinszky, Gábor I

    John P. Perdew, Adrienn Ruzsinszky, Gábor I. Csonka, Oleg A. Vydrov, Gustavo E. Scuseria, Lucian A. Con- stantin, Xiaolan Zhou, and Kieron Burke. Restoring the density-gradientexpansionforexchangeinsolidsandsur- faces.Phys. Rev. Lett., 100(13):136406, April 2008

    2008

  21. [21]

    Pires, Thalis H.B

    Paulo R. Pires, Thalis H.B. da Silva, Kun Gao, Kaja H. Hiorth, Tiago F.T. Cerqueira, Théo Cavignac, Pierre- Paul De Breuck, Hai-Chen Wang, Ðorđe Dangić, Yue- Wen Fang, Antonio Sanna, Wenwen Cui, Ion Errea, Päivi Törmä, and Miguel A.L. Marques. Machine learning drivenexplorationofhydridesuperconductorsatambient pressure.Comput. Mat. Today, 10:100052, June 2026

    2026

  22. [22]

    EPW: Electron–phonon coupling, transport and superconduct- ing properties using maximally localized wannier func- tions.Comput

    S.Poncé, E.R.Margine, C.Verdi, andF.Giustino. EPW: Electron–phonon coupling, transport and superconduct- ing properties using maximally localized wannier func- tions.Comput. Phys. Commun., 209:116–133, December 2016

    2016

  23. [23]

    First-principles calculations of charge carrier mobility and conductivity in bulk semiconduc- tors and two-dimensional materials.Rep

    Samuel Poncé, Wenbin Li, Sven Reichardt, and Feli- ciano Giustino. First-principles calculations of charge carrier mobility and conductivity in bulk semiconduc- tors and two-dimensional materials.Rep. Prog. Phys., 83(3):036501, February 2020

    2020

  24. [24]

    Rosén and G

    J. Rosén and G. Grimvall. Anharmonic lattice vibrations in simple metals.Phys. Rev. B, 27(12):7199, June 1983

    1983

  25. [25]

    Saal, Scott Kirklin, Muratahan Aykol, Bryce Meredig, and C

    James E. Saal, Scott Kirklin, Muratahan Aykol, Bryce Meredig, and C. Wolverton. Materials design and dis- covery with high-throughput density functional theory: The Open Quantum Materials database (OQMD).JOM, 65(11):1501–1509, September 2013

    2013

  26. [26]

    Transportcoefficientsfrom first-principles calculations.Phys

    T.J.Scheidemantel, C.Ambrosch-Draxl, T.Thonhauser, J.V.Badding, andJ.O.Sofo. Transportcoefficientsfrom first-principles calculations.Phys. Rev. B, 68(12):125210, September 2003

    2003

  27. [27]

    Cerqueira, Aldo H

    Jonathan Schmidt, Tiago F.T. Cerqueira, Aldo H. Romero, Antoine Loew, Fabian Jäger, Hai-Chen Wang, Silvana Botti, and Miguel A.L. Marques. Improving machine-learning models in materials science through large datasets.Mater. Today Phys., 48:101560, Novem- ber 2024

    2024

  28. [28]

    Jonathan Schmidt, Noah Hoffmann, Hai-Chen Wang, Pedro Borlido, Pedro J. M. A. Carriço, Tiago F. T. Cerqueira, Silvana Botti, and Miguel A. L. Marques. Machine-learning-assisted determination of the global zero-temperature phase diagram of materials.Adv. Mater., 35:2210788, 2023

    2023

  29. [29]

    Jonathan Schmidt, Love Pettersson, Claudio Verdozzi, Silvana Botti, and Miguel A. L. Marques. Crystal graph attention networks for the prediction of stable materials. Sci. Adv., 7(49):abi7948, December 2021

    2021

  30. [30]

    Jonathan Schmidt, Hai-Chen Wang, Tiago F. T. Cerqueira, Silvana Botti, and Miguel A. L. Marques. A dataset of 175k stable and metastable materials calcu- lated with the pbesol and scan functionals.Sci. Data, 9(1):64, March 2022

    2022

  31. [31]

    van Setten, M

    M.J. van Setten, M. Giantomassi, E. Bousquet, M.J. Ver- straete, D.R. Hamann, X. Gonze, and G.-M. Rignanese. ThePseudoDojo: Training and grading a 85 element optimized norm-conserving pseudopotential table.Com- put. Phys. Commun., 226:39–54, May 2018

    2018

  32. [32]

    Kauwe, Ryan J

    Anthony Yu-Tung Wang, Steven K. Kauwe, Ryan J. Murdock, and Taylor D. Sparks. Compositionally re- stricted attention-based network for materials property predictions.npj Comput. Mater., 7(1):77, May 2021

    2021

  33. [33]

    Opposite impact of thermal expansion and phonon anharmonicity on the phonon-limited resistivity of elemental metals from first principles.Phys

    AoWang, JunwenYin, FélixAntoineGoudreault, Michel Côté, Olle Hellman, and Samuel Poncé. Opposite impact of thermal expansion and phonon anharmonicity on the phonon-limited resistivity of elemental metals from first principles.Phys. Rev. B, 113(6):L060302, February 2026

    2026