pith. sign in

arxiv: 2401.00096 · v3 · submitted 2023-12-29 · ⚛️ physics.chem-ph · cond-mat.mtrl-sci

A foundation model for atomistic materials chemistry

Ilyes Batatia , Philipp Benner , Yuan Chiang , Alin M. Elena , D\'avid P. Kov\'acs , Janosh Riebesell , Xavier R. Advincula , Mark Asta
show 80 more authors
Matthew Avaylon William J. Baldwin Fabian Berger Noam Bernstein Arghya Bhowmik Filippo Bigi Samuel M. Blau Vlad C\u{a}rare Michele Ceriotti Sanggyu Chong James P. Darby Sandip De Flaviano Della Pia Volker L. Deringer Rokas Elijo\v{s}ius Zakariya El-Machachi Fabio Falcioni Edvin Fako Andrea C. Ferrari John L. A. Gardner Mikolaj J. Gawkowski Annalena Genreith-Schriever Janine George Rhys E. A. Goodall Jonas Grandel Clare P. Grey Petr Grigorev Shuang Han Will Handley Hendrik H. Heenen Kersti Hermansson Christian Holm Cheuk Hin Ho Stephan Hofmann Jad Jaafar Konstantin S. Jakob Hyunwook Jung Venkat Kapil Aaron D. Kaplan Nima Karimitari James R. Kermode Panagiotis Kourtis Namu Kroupa Jolla Kullgren Matthew C. Kuner Domantas Kuryla Guoda Liepuoniute Chen Lin Johannes T. Margraf Ioan-Bogdan Magd\u{a}u Angelos Michaelides J. Harry Moore Aakash A. Naik Samuel P. Niblett Sam Walton Norwood Niamh O'Neill Christoph Ortner Kristin A. Persson Karsten Reuter Andrew S. Rosen Louise A. M. Rosset Lars L. Schaaf Christoph Schran Benjamin X. Shi Eric Sivonxay Tam\'as K. Stenczel Viktor Svahn Christopher Sutton Thomas D. Swinburne Jules Tilly Cas van der Oord Santiago Vargas Eszter Varga-Umbrich Tejs Vegge Martin Vondr\'ak Yangshuai Wang William C. Witt Thomas Wolf Fabian Zills G\'abor Cs\'anyi
This is my paper

Pith reviewed 2026-05-18 10:11 UTC · model grok-4.3

classification ⚛️ physics.chem-ph cond-mat.mtrl-sci
keywords machine learning force fieldsfoundation modelsatomistic simulationsmolecular dynamicsmaterials chemistrytransferable potentialsab initio modeling
0
0 comments X p. Extension

The pith

A single machine-learned force field can run stable molecular dynamics across diverse molecules and materials.

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

This paper shows that a general-purpose atomistic machine learning model trained on a public dataset of moderate size can produce forces accurate enough to run stable molecular dynamics on many different chemical systems. The MACE-MP-0 model achieves qualitative and sometimes quantitative accuracy on solids, liquids, gases, chemical reactions, interfaces, and even a small protein. If correct, this removes the need to develop and validate a separate potential for each new system of interest. Sympathetic readers care because it makes high-quality atomistic simulations faster for experts and more accessible for beginners in chemistry and materials science.

Core claim

The authors establish that it is possible to create a general-purpose atomistic ML model, trained on a public dataset of moderate size, that is capable of running stable molecular dynamics for a wide range of molecules and materials. They demonstrate the MACE-MP-0 model on properties of solids, liquids, gases, chemical reactions, interfaces, and the dynamics of a small protein. The model can be applied out of the box as a starting or foundation model for any atomistic system and fine-tuned on a handful of application-specific data points to reach ab initio accuracy.

What carries the argument

The MACE-MP-0 model, a transferable machine-learned force field trained on the Materials Project dataset that serves as a foundation for simulations across chemical systems.

If this is right

  • Experienced users obtain reliable simulation results much faster without developing new potentials for each system.
  • Beginners encounter a lower barrier to entry for performing atomistic modeling.
  • The model can be fine-tuned on just a handful of application-specific data points to reach ab initio accuracy.
  • Atomistic modeling becomes feasible for almost all materials using one starting foundation model.
  • The workflow shifts from building custom models from scratch to starting from a general model and refining as needed.

Where Pith is reading between the lines

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

  • This foundation approach could speed up initial screening of candidate materials in discovery workflows before committing to more expensive calculations.
  • Similar foundation models might be trained for related domains such as catalysis or soft matter where transferability across chemistries is also valuable.
  • Active learning loops could be combined with the base model to efficiently add data only for out-of-distribution cases.
  • The model offers a practical starting point for hybrid simulations that mix machine-learned forces with experimental constraints.

Load-bearing premise

The training dataset and model architecture are representative enough that the resulting force field remains stable and accurate on chemical systems and configurations outside the training distribution without requiring extensive additional data or retraining.

What would settle it

Running molecular dynamics on a novel molecule or material configuration far from the training distribution and observing either unstable trajectories or force errors that exceed ab initio reference levels would falsify the general-purpose claim.

read the original abstract

Atomistic simulations of matter, especially those that leverage first-principles (ab initio) electronic structure theory, provide a microscopic view of the world, underpinning much of our understanding of chemistry and materials science. Over the last decade or so, machine-learned force fields have transformed atomistic modeling by enabling simulations of ab initio quality over unprecedented time and length scales. However, early ML force fields have largely been limited by: (i) the substantial computational and human effort of developing and validating potentials for each particular system of interest; and (ii) a general lack of transferability from one chemical system to the next. Here we show that it is possible to create a general-purpose atomistic ML model, trained on a public dataset of moderate size, that is capable of running stable molecular dynamics for a wide range of molecules and materials. We demonstrate the power of the MACE-MP-0 model - and its qualitative and at times quantitative accuracy - on a diverse set of problems in the physical sciences, including properties of solids, liquids, gases, chemical reactions, interfaces and even the dynamics of a small protein. The model can be applied out of the box as a starting or "foundation" model for any atomistic system of interest and, when desired, can be fine-tuned on just a handful of application-specific data points to reach ab initio accuracy. Establishing that a stable force-field model can cover almost all materials changes atomistic modeling in a fundamental way: experienced users get reliable results much faster, and beginners face a lower barrier to entry. Foundation models thus represent a step towards democratising the revolution in atomic-scale modeling that has been brought about by ML force fields.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The manuscript introduces MACE-MP-0, a machine-learned force field trained on the Materials Project dataset of periodic inorganic crystals. It claims that this single model enables stable, physically reasonable molecular dynamics trajectories for a broad range of systems—including molecules, liquids, gases, chemical reactions, interfaces, and a small protein—while delivering qualitative and sometimes quantitative accuracy relative to ab initio methods. The model is positioned as a foundation model that can be applied out-of-the-box or fine-tuned on a handful of system-specific points to reach higher accuracy, thereby reducing the need for bespoke potential development.

Significance. If the generalization and stability claims hold under systematic scrutiny, the work would represent a substantial advance in atomistic modeling by providing a reusable starting point that lowers the barrier for both experts and newcomers. The demonstration that a moderate-sized public dataset can yield a transferable force field capable of handling both inorganic solids and organic/biomolecular systems would support the emerging paradigm of foundation models in chemistry and materials science.

major comments (2)
  1. [Results / Applications sections] The central claim of out-of-the-box stability and accuracy across chemically diverse systems (molecules, reactions, interfaces, proteins) rests on selected qualitative examples rather than systematic quantitative benchmarks. No tables or figures report aggregate metrics such as mean absolute force errors on held-out organic clusters, fraction of trajectories remaining stable beyond 100 ps, or failure rates for bonding motifs and elements underrepresented in the MP training set.
  2. [Methods / Training data description] The representativeness assumption—that MP solids data suffice for extrapolation to molecular and biomolecular configurations—is load-bearing for the 'wide range' and 'foundation model' assertions, yet the manuscript provides no explicit test of this (e.g., force-error distributions on gas-phase organic molecules or protein fragments absent from the training distribution).
minor comments (2)
  1. [Methods] Notation for the model architecture and message-passing layers could be clarified with an explicit equation or diagram in the methods section to aid reproducibility.
  2. [Figures] Figure captions for MD trajectory visualizations should state the simulation length, timestep, and thermostat used so readers can assess stability claims directly.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive review and positive assessment of the potential significance of MACE-MP-0 as a foundation model. We have revised the manuscript to incorporate additional systematic quantitative benchmarks and explicit tests of extrapolation from the Materials Project training distribution, while preserving the focus on broad applicability demonstrated through diverse case studies.

read point-by-point responses

  1. Referee: The central claim of out-of-the-box stability and accuracy across chemically diverse systems (molecules, reactions, interfaces, proteins) rests on selected qualitative examples rather than systematic quantitative benchmarks. No tables or figures report aggregate metrics such as mean absolute force errors on held-out organic clusters, fraction of trajectories remaining stable beyond 100 ps, or failure rates for bonding motifs and elements underrepresented in the MP training set.

    Authors: We agree that aggregate quantitative metrics strengthen the central claims. The original manuscript prioritized illustrative demonstrations of versatility across domains to establish the foundation-model concept. In the revised manuscript we have added a dedicated subsection with new tables and figures reporting: (i) mean absolute force errors on a held-out set of organic clusters and gas-phase molecules, (ii) the fraction of MD trajectories that remain stable beyond 100 ps for a statistically meaningful collection of systems spanning molecules, liquids, interfaces and the protein example, and (iii) performance breakdowns highlighting errors on bonding motifs and elements that are underrepresented in the MP training set. These additions are presented alongside the existing qualitative examples. revision: partial

  2. Referee: The representativeness assumption—that MP solids data suffice for extrapolation to molecular and biomolecular configurations—is load-bearing for the 'wide range' and 'foundation model' assertions, yet the manuscript provides no explicit test of this (e.g., force-error distributions on gas-phase organic molecules or protein fragments absent from the training distribution).

    Authors: The MP dataset consists of periodic inorganic crystals, yet the MACE architecture learns local many-body environments that prove transferable. To make this explicit, the revised manuscript now includes force-error distributions and direct comparisons on gas-phase organic molecules and protein fragments that lie outside the training distribution. These tests show that errors are larger than on in-distribution solids but remain low enough to support stable dynamics in the reported cases; we also discuss the limits of this extrapolation and the value of subsequent fine-tuning. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical ML training and evaluation on external datasets

full rationale

The paper presents an empirical machine learning force field (MACE-MP-0) trained on the external Materials Project dataset and evaluated on separate molecular, liquid, gas, reaction, interface, and protein systems. No equations, derivations, or first-principles results are shown that reduce claimed performance or stability to quantities defined by the same fitted parameters or by self-citation chains. The central claims rest on standard supervised training followed by out-of-sample testing rather than any self-definitional or fitted-input-called-prediction structure. Self-citations, if present for architecture details, are not load-bearing for the generalization assertions, which are supported by direct simulation results on held-out problems.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the empirical generalization of a neural-network force field trained on a finite public dataset; no new physical axioms are introduced, but the model implicitly assumes that the chosen architecture and data coverage suffice for stability across unseen chemistries.

free parameters (1)
  • neural network weights
    All model parameters are fitted to the training dataset of atomic energies and forces.
axioms (1)
  • domain assumption The chosen message-passing architecture can represent interatomic interactions sufficiently accurately when trained on the given dataset.
    Invoked implicitly when asserting that the trained model produces stable dynamics on diverse systems.

pith-pipeline@v0.9.0 · 6305 in / 1313 out tokens · 32396 ms · 2026-05-18T10:11:30.929809+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • Foundation.DimensionForcing alexander_duality_circle_linking unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    MACE-MP-0 uses the MACE architecture which unified the atomic cluster expansion (ACE) and equivariant graph neural networks.

  • Foundation.PhiForcing phi_forcing unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    We demonstrate the power of the MACE-MP-0 model on a diverse set of problems including properties of solids, liquids, gases, chemical reactions, interfaces and even the dynamics of a small protein.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 18 Pith papers

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

  1. Lang2MLIP: End-to-End Language-to-Machine Learning Interatomic Potential Development with Autonomous Agentic Workflows

    cs.LG 2026-05 unverdicted novelty 7.0

    Lang2MLIP is an LLM multi-agent framework that automates end-to-end development of machine learning interatomic potentials from natural language input for heterogeneous materials systems.

  2. Force-Aware Neural Tangent Kernels for Scalable and Robust Active Learning of MLIPs

    cs.LG 2026-05 unverdicted novelty 7.0

    Force-aware NTKs and chunked acquisition enable scalable, robust active learning for MLIPs, achieving lowest energy and force errors on OC20 and remaining competitive on other benchmarks.

  3. Effective dynamic constants for nonequilibrium third-principles simulations

    cond-mat.mtrl-sci 2026-05 unverdicted novelty 7.0

    A protocol is introduced to derive effective inertial and viscous-damping constants for nonequilibrium polarization dynamics in soft-mode ferroelectric PbTiO3.

  4. Pretrained Model Representations as Acquisition Signals for Active Learning of MLIPs

    cs.LG 2026-05 unverdicted novelty 7.0

    Pretrained MLIP latent spaces yield NTK and activation kernels that outperform standard acquisition functions in active learning for reactive MLIP training, reducing required labels by 38% for energy and 28% for force errors.

  5. Fast and Accurate Prediction of Lattice Thermal Conductivity via Machine Learning Surrogates

    cond-mat.mtrl-sci 2026-05 conditional novelty 6.0

    Machine learning models, especially certain deep neural networks, can predict lattice thermal conductivity with useful accuracy across different generalization tests while being orders of magnitude faster than first-p...

  6. Compact SO(3) Equivariant Atomistic Foundation Models via Structural Pruning

    cs.LG 2026-05 unverdicted novelty 6.0

    Structural pruning of SO(3) equivariant atomistic models from large checkpoints yields 1.5-4x fewer parameters and 2.5-4x less pre-training compute than small models trained from scratch, while outperforming them on m...

  7. MatterSim-MT: A multi-task foundation model for in silico materials characterization

    cond-mat.mtrl-sci 2026-05 unverdicted novelty 6.0

    MatterSim-MT is a foundation model pretrained on over 35 million first-principles structures that predicts material structure, dynamics, and thermodynamics while enabling multi-task simulations of phonon splitting, fe...

  8. VibroML: an automated toolkit for high-throughput vibrational analysis and dynamic instability remediation of crystalline materials using machine-learned potentials

    cond-mat.mtrl-sci 2026-04 unverdicted novelty 6.0

    VibroML automates remediation of dynamic instabilities in crystalline materials by combining MLIPs with genetic algorithms for polymorph search, finite-temperature MD validation, and compositional alloying to yield st...

  9. Neutron and X-ray Diffraction Reveal the Limits of Long-Range Machine Learning Potentials for Medium-Range Order in Silica Glass

    cond-mat.mtrl-sci 2026-04 conditional novelty 6.0

    Long-range ML potentials improve liquid silica but remain insufficient for experimental medium-range order in the glass, as shown by diffraction data and structural analyses.

  10. LitMOF: An LLM Multi-Agent for Literature-Validated Metal-Organic Frameworks Database Correction and Expansion

    cs.DB 2025-12 conditional novelty 6.0

    LitMOF repairs 8,771 erroneous MOF structures and adds 12,646 previously missing ones from literature, yielding a database of 186,773 computation-ready entries and demonstrating that uncorrected errors distort adsorpt...

  11. Bayesian E(3)-Equivariant Interatomic Potential with Iterative Restratification of Many-body Message Passing

    cs.LG 2025-10 unverdicted novelty 6.0

    Bayesian E(3)-equivariant MLPs with joint energy-force NLL loss achieve competitive accuracy while enabling uncertainty-guided active learning, OOD detection, and calibration.

  12. Machine Learning Phonon Spectra for Fast and Accurate Optical Lineshapes of Defects

    cond-mat.mtrl-sci 2025-08 unverdicted novelty 6.0

    Machine learning interatomic potentials fine-tuned on first-principles relaxation data accurately reproduce phonon spectra and optical lineshapes for defects, matching explicit calculations and experiments.

  13. Open Materials 2024 (OMat24) Inorganic Materials Dataset and Models

    cond-mat.mtrl-sci 2024-10 accept novelty 6.0

    OMat24 provides over 110 million DFT calculations and EquiformerV2 models that reach state-of-the-art performance on material stability and formation energy prediction.

  14. MatterSim: A Deep Learning Atomistic Model Across Elements, Temperatures and Pressures

    cond-mat.mtrl-sci 2024-05 unverdicted novelty 6.0

    MatterSim delivers a single deep learning force field that simulates inorganic materials across elements, 0-5000 K, and up to 1000 GPa with near first-principles accuracy for lattice dynamics, mechanics, and Gibbs fre...

  15. Systematic Fine-Tuning of MACE Interatomic Potentials for Catalysis

    physics.chem-ph 2026-05 conditional novelty 5.0

    Fine-tuned MACE MLIPs achieve lower mean absolute errors on catalytic reaction energies and barriers than from-scratch models, with a large fine-tuned model performing best on both metallic and oxide systems including...

  16. Accurate and Efficient Interatomic Potentials for Dislocations in InP

    cond-mat.mtrl-sci 2026-04 unverdicted novelty 4.0

    New ACE and MACE potentials for InP achieve at most 4% error on partial dislocation formation energies versus DFT, outperforming literature models by factors of 4-12 while being computationally faster.

  17. Comparing fine-tuning strategies of MACE machine learning force field for modeling Li-ion diffusion in LiF for batteries

    cond-mat.mtrl-sci 2025-10 conditional novelty 4.0

    MACE-MPA-0 predicts Li diffusion Ea of 0.22 eV in LiF, fine-tuned version with 300 points gives 0.20 eV, close to DeePMD reference of 0.24 eV, using far less training data.

  18. From Knowledge to Action: Outcomes of the 2025 Large Language Model (LLM) Hackathon for Applications in Materials Science and Chemistry

    cond-mat.mtrl-sci 2026-05 unverdicted novelty 2.0

    Hackathon submissions indicate LLMs are moving from general assistants toward composable multi-agent systems for structuring scientific knowledge and automating tasks in materials science and chemistry.

Reference graph

Works this paper leans on

299 extracted references · 299 canonical work pages · cited by 18 Pith papers · 4 internal anchors

  1. [1]

    Self-consistent equations including exchange and correlation effects,

    W. Kohn and L. J. Sham, “Self-consistent equations including exchange and correlation effects,”Phys. Rev., vol. 140, pp. A1133–A1138, Nov 1965

    work page 1965

  2. [2]

    Quantum ESPRESSO toward the exascale,

    P. Giannozzi, O. Baseggio, P. Bonfà, D. Brunato, R. Car, I. Carnimeo, C. Cavazzoni, S. de Gironcoli, P. Delugas, F. Ferrari Ruffino, A. Ferretti, N. Marzari, I. Timrov, A. Urru, and S. Baroni, “Quantum ESPRESSO toward the exascale,”The Journal of Chemical Physics, vol. 152, p. 154105, 04 2020

    work page 2020

  3. [3]

    Cp2k: An electronic structure and molecular dynamics software package-quickstep: Efficient and accurate electronic structure calculations,

    T. D. Kühne, M. Iannuzzi, M. Del Ben, V. V. Rybkin, P. Seewald, F. Stein, T. Laino, R. Z. Khaliullin, O. Schütt, F. Schiffmann, et al., “Cp2k: An electronic structure and molecular dynamics software package-quickstep: Efficient and accurate electronic structure calculations,”The Journal of Chemical Physics, vol. 152, no. 19, 2020

    work page 2020

  4. [4]

    Ab initio molecular dynamics for liquid metals,

    G. Kresse and J. Hafner, “Ab initio molecular dynamics for liquid metals,”Phys. Rev. B, vol. 47, p. 558, 1993

    work page 1993

  5. [5]

    Density functional theory in the solid state,

    P. J. Hasnip, K. Refson, M. I. J. Probert, J. R. Yates, S. J. Clark, and C. J. Pickard, “Density functional theory in the solid state,”Phil. Trans. R. Soc. A., vol. 372, p. 20130270, 2014

    work page 2014

  6. [6]

    Computational predictions of energy materials using density functional theory,

    A. Jain, Y. Shin, and K. A. Persson, “Computational predictions of energy materials using density functional theory,”Nat Rev Mater, vol. 1, p. 714, 2016

    work page 2016

  7. [7]

    Density functional theory in materials science,

    J. Neugebauer and T. Hickel, “Density functional theory in materials science,”WIREs Comput Mol Sci, vol. 3, pp. 438–448, 2013

    work page 2013

  8. [8]

    Finnis,Interatomic forces in condensed matter, vol

    M. Finnis,Interatomic forces in condensed matter, vol. 1. Oxford Series on Materials Mod, 2003

    work page 2003

  9. [9]

    Generalized neural-network representation of high-dimensional potential- energy surfaces,

    J. Behler and M. Parrinello, “Generalized neural-network representation of high-dimensional potential- energy surfaces,”Phys. Rev. Lett., vol. 98, p. 146401, Apr 2007

    work page 2007

  10. [10]

    Gaussian approximation potentials: The accuracy of quantum mechanics, without the electrons,

    A. P. Bartók, M. C. Payne, R. Kondor, and G. Csányi, “Gaussian approximation potentials: The accuracy of quantum mechanics, without the electrons,”Phys. Rev. Lett., vol. 104, p. 136403, Apr 2010

    work page 2010

  11. [11]

    Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials,

    A. Thompson, L. Swiler, C. Trott, S. Foiles, and G. Tucker, “Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials,”Journal of Computational Physics, vol. 285, pp. 316–330, 2015

    work page 2015

  12. [12]

    SchNet – A deep learning architecture for molecules and materials,

    K. T. Schütt, H. E. Sauceda, P.-J. Kindermans, A. Tkatchenko, and K.-R. Müller, “SchNet – A deep learning architecture for molecules and materials,”The Journal of Chemical Physics, vol. 148, p. 241722, 03 2018

    work page 2018

  13. [13]

    Machine Learning Interatomic Potentials as Emerging Tools for Materials Science,

    V. L. Deringer, M. A. Caro, and G. Csányi, “Machine Learning Interatomic Potentials as Emerging Tools for Materials Science,”Advanced Materials, vol. 31, no. 46, p. 1902765, 2019

    work page 2019

  14. [14]

    Atomic cluster expansion for accurate and transferable interatomic potentials,

    R. Drautz, “Atomic cluster expansion for accurate and transferable interatomic potentials,”Phys. Rev. B, vol. 99, p. 014104, Jan 2019

    work page 2019

  15. [15]

    Retrospective on a decade of machine learning for chemical discov- ery,

    O. A. von Lilienfeld and K. Burke, “Retrospective on a decade of machine learning for chemical discov- ery,”Nature Communications, vol. 11, no. 1, p. 4895, 2020. 19

    work page 2020

  16. [16]

    E (3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials,

    S. Batzner, A. Musaelian, L. Sun, M. Geiger, J. P. Mailoa, M. Kornbluth, N. Molinari, T. E. Smidt, and B. Kozinsky, “E (3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials,”Nature communications, vol. 13, no. 1, p. 2453, 2022

    work page 2022

  17. [17]

    Recent advances and outstanding challenges for machine learning interatomic potentials,

    T. W. Ko and S. P. Ong, “Recent advances and outstanding challenges for machine learning interatomic potentials,”Nature Computational Science, pp. 1–3, 2023

    work page 2023

  18. [18]

    Gaussian process regression for materials and molecules,

    V. L. Deringer, A. P. Bartók, N. Bernstein, D. M. Wilkins, M. Ceriotti, and G. Csányi, “Gaussian process regression for materials and molecules,”Chemical Reviews, vol. 121, no. 16, pp. 10073–10141,

    work page

  19. [19]

    Mace: Higher order equivariant message passing neural networks for fast and accurate force fields,

    I. Batatia, D. P. Kovacs, G. Simm, C. Ortner, and G. Csányi, “Mace: Higher order equivariant message passing neural networks for fast and accurate force fields,”Advances in Neural Information Processing Systems, vol. 35, pp. 11423–11436, 2022

    work page 2022

  20. [20]

    Atomic cluster expansion: Completeness, efficiency and stability,

    G. Dusson, M. Bachmayr, G. Csányi, R. Drautz, S. Etter, C. van der Oord, and C. Ortner, “Atomic cluster expansion: Completeness, efficiency and stability,”Journal of Computational Physics, vol. 454, p. 110946, 2022

    work page 2022

  21. [21]

    Performant implementation of the atomic cluster expansion (pace) and application to copper and silicon,

    Y. Lysogorskiy, C. v. d. Oord, A. Bochkarev, S. Menon, M. Rinaldi, T. Hammerschmidt, M. Mrovec, A. Thompson, G. Csányi, C. Ortner,et al., “Performant implementation of the atomic cluster expansion (pace) and application to copper and silicon,”npj computational materials, vol. 7, no. 1, p. 97, 2021

    work page 2021

  22. [22]

    Acepotentials. jl: A julia implementation of the atomic cluster expansion,

    W. C. Witt, C. van der Oord, E. Gelžinyt˙ e, T. Järvinen, A. Ross, J. P. Darby, C. H. Ho, W. J. Baldwin, M. Sachs, J. Kermode,et al., “Acepotentials. jl: A julia implementation of the atomic cluster expansion,”The Journal of Chemical Physics, vol. 159, no. 16, 2023

    work page 2023

  23. [23]

    The design space of E(3)-equivariant atom-centred interatomic potentials,

    I. Batatia, S. Batzner, D. P. Kovács, A. Musaelian, G. N. C. Simm, R. Drautz, C. Ortner, B. Kozinsky, and G. Csányi, “The design space of E(3)-equivariant atom-centred interatomic potentials,”Nature Machine Intelligence, vol. 7, pp. 56–67, Jan. 2025

    work page 2025

  24. [24]

    Tensor- reduced atomic density representations,

    J. P. Darby, D. P. Kovács, I. Batatia, M. A. Caro, G. L. W. Hart, C. Ortner, and G. Csányi, “Tensor- reduced atomic density representations,”Phys. Rev. Lett., vol. 131, p. 028001, Jul 2023

    work page 2023

  25. [25]

    Three Iterations of (d - 1)- WL Test Distinguish Non Isometric Clouds of d-dimensional Points,

    V. Delle Rose, A. Kozachinskiy, C. Rojas, M. Petrache, and P. Barceló, “Three Iterations of (d - 1)- WL Test Distinguish Non Isometric Clouds of d-dimensional Points,”Advances in Neural Information Processing Systems, vol. 36, pp. 9556–9573, Dec. 2023

    work page 2023

  26. [26]

    Bench- mark oxygen-oxygen pair-distribution function of ambient water from x-ray diffraction measurements with a wide Q-range,

    L. B. Skinner, C. Huang, D. Schlesinger, L. G. M. Pettersson, A. Nilsson, and C. J. Benmore, “Bench- mark oxygen-oxygen pair-distribution function of ambient water from x-ray diffraction measurements with a wide Q-range,”J. Chem. Phys., vol. 138, p. 074506, 02 2013

    work page 2013

  27. [27]

    J. E. Bertie and Z. Lan, “Infrared intensities of liquids xx: The intensity of the OH stretching band of liquid water revisited, and the best current values of the optical constants of H2O(l) at 25◦c between 15,000 and 1cm−1,”Appl. Spectrosc., vol. 50, no. 8, pp. 1047–1057, 1996

    work page 1996

  28. [28]

    Molecular Origin of the Vibrational Structure of Ice Ih,

    D. R. Moberg, S. C. Straight, C. Knight, and F. Paesani, “Molecular Origin of the Vibrational Structure of Ice Ih,”J. Phys. Chem. Lett., vol. 8, no. 12, pp. 2579–2583, 2017. PMID: 28541703

    work page 2017

  29. [29]

    A consistent and accurate ab initio parametrization of density functional dispersion correction (dft-d) for the 94 elements h-pu,

    S. Grimme, J. Antony, S. Ehrlich, and H. Krieg, “A consistent and accurate ab initio parametrization of density functional dispersion correction (dft-d) for the 94 elements h-pu,”The Journal of chemical physics, vol. 132, no. 15, 2010

    work page 2010

  30. [30]

    Crumblingcrystals: Onthedissolutionmechanism of nacl in water,

    N.O’Neill, C.Schran, S.J.Cox, andA.Michaelides, “Crumblingcrystals: Onthedissolutionmechanism of nacl in water,” 2022

    work page 2022

  31. [31]

    Perspective: How good is dft for water?,

    M. J. Gillan, D. Alfe, and A. Michaelides, “Perspective: How good is dft for water?,”The Journal of chemical physics, vol. 144, no. 13, 2016. 20

    work page 2016

  32. [32]

    The nature of the hydrated excess proton in water,

    D. Marx, M. E. Tuckerman, J. Hutter, and M. Parrinello, “The nature of the hydrated excess proton in water,”Nature, vol. 397, pp. 601–604, feb 1999

    work page 1999

  33. [33]

    The nature and transport mechanism of hydrated hydroxide ions in aqueous solution,

    M. E. Tuckerman, D. Marx, and M. Parrinello, “The nature and transport mechanism of hydrated hydroxide ions in aqueous solution,”Nature, vol. 417, pp. 925–929, jun 2002

    work page 2002

  34. [34]

    Protons and Hydroxide Ions in Aqueous Systems,

    N. Agmon, H. J. Bakker, R. K. Campen, R. H. Henchman, P. Pohl, S. Roke, M. Thämer, and A. Has- sanali, “Protons and Hydroxide Ions in Aqueous Systems,”Chem. Rev., vol. 116, pp. 7642–7672, jul 2016

    work page 2016

  35. [35]

    DMC-ICE13: Ambientandhighpressurepolymorphs of ice from diffusion Monte Carlo and density functional theory,

    F.DellaPia, A.Zen, D.Alfè, andA.Michaelides, “DMC-ICE13: Ambientandhighpressurepolymorphs of ice from diffusion Monte Carlo and density functional theory,”The Journal of Chemical Physics, vol. 157, p. 134701, 10 2022

    work page 2022

  36. [36]

    Square ice in graphene nanocapillaries,

    G. Algara-Siller, O. Lehtinen, F. C. Wang, R. R. Nair, U. Kaiser, H. A. Wu, A. K. Geim, and I. V. Grigorieva, “Square ice in graphene nanocapillaries,”Nature, vol. 519, pp. 443–445, mar 2015

    work page 2015

  37. [37]

    Anomalously low dielectric constant of confined water,

    L. Fumagalli, A. Esfandiar, R. Fabregas, S. Hu, P. Ares, A. Janardanan, Q. Yang, B. Radha, T. Taniguchi, K. Watanabe, G. Gomila, K. S. Novoselov, and A. K. Geim, “Anomalously low dielectric constant of confined water,”Science, vol. 360, pp. 1339–1342, jun 2018

    work page 2018

  38. [38]

    The first-principles phase diagram of monolayer nanoconfined water,

    V. Kapil, C. Schran, A. Zen, J. Chen, C. J. Pickard, and A. Michaelides, “The first-principles phase diagram of monolayer nanoconfined water,”Nature, vol. 609, no. 7927, pp. 512–516, 2022

    work page 2022

  39. [39]

    Surface pourbaix diagrams and oxygen reduction activity of Pt, Ag and Ni(111) surfaces studied by DFT,

    H. A. Hansen, J. Rossmeisl, and J. K. Nørskov, “Surface pourbaix diagrams and oxygen reduction activity of Pt, Ag and Ni(111) surfaces studied by DFT,”Phys. Chem. Chem. Phys., vol. 10, pp. 3722– 3730, July 2008

    work page 2008

  40. [40]

    Limits to scaling relations between adsorption energies?,

    S. Vijay, G. Kastlunger, K. Chan, and J. K. Nørskov, “Limits to scaling relations between adsorption energies?,”The Journal of Chemical Physics, vol. 156, June 2022

    work page 2022

  41. [41]

    Accurate energy barriers for catalytic reac- tion pathways: an automatic training protocol for machine learning force fields,

    L. L. Schaaf, E. Fako, S. De, A. Schäfer, and G. Csányi, “Accurate energy barriers for catalytic reac- tion pathways: an automatic training protocol for machine learning force fields,”npj Computational Materials, vol. 9, no. 1, p. 180, 2023

    work page 2023

  42. [42]

    UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction

    L. McInnes, J. Healy, and J. Melville, “Umap: Uniform manifold approximation and projection for dimension reduction,”arXiv preprint arXiv:1802.03426, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  43. [43]

    MACE-MP UMAP analysis,

    R. Elijošius, “MACE-MP UMAP analysis,” github.com, 2023. 10.5281/zenodo.10426282 - https://github.com/RokasEl/mace-mp-umap

    work page doi:10.5281/zenodo.10426282 2023

  44. [44]

    Towards the computational design of solid catalysts,

    J. K. Nørskov, T. Bligaard, J. Rossmeisl, and C. H. Christensen, “Towards the computational design of solid catalysts,”Nature Chem, vol. 1, pp. 37–46, 2009

    work page 2009

  45. [45]

    From the sabatier principle to a predictive theory of transition-metal heterogeneous catalysis,

    A. J. Medford, A. Vojvodic, J. S. Hummelshøj, J. Voss, F. Abild-pedersen, F. Studt, T. Bligaard, A. Nilsson, and J. K. Nørskov, “From the sabatier principle to a predictive theory of transition-metal heterogeneous catalysis,”Journal of Catalysis, vol. 328, pp. 36–42, 2015

    work page 2015

  46. [46]

    First-principles-based multiscale modelling of heterogeneous catalysis,

    A. Bruix, J. T. Margraf, M. Andersen, and K. Reuter, “First-principles-based multiscale modelling of heterogeneous catalysis,”Nat Catal, vol. 2, pp. 659–670, 2019

    work page 2019

  47. [47]

    Cation-coordinated inner-sphere co<sub>2</sub>electroreduction at au–water interfaces,

    X. Qin, T. Vegge, and H. A. Hansen, “Cation-coordinated inner-sphere co<sub>2</sub>electroreduction at au–water interfaces,”J. Am. Chem. Soc., vol. 145, pp. 1897–1905, 2023

    work page 1905

  48. [48]

    Universality in oxygen evolution electrocatalysis on oxide surfaces,

    I. C. Man, H. Su, F. Calle-vallejo, H. A. Hansen, J. I. Martínez, N. G. Inoglu, J. Kitchin, T. F. Jaramillo, J. K. Nørskov, and J. Rossmeisl, “Universality in oxygen evolution electrocatalysis on oxide surfaces,” ChemCatChem, vol. 3, pp. 1159–1165, 2011. 21

    work page 2011

  49. [49]

    Self-activation of copper electrodes during co electro-oxidation in alkaline electrolyte,

    A. Auer, M. Andersen, E.-M. Wernig, N. G. Hörmann, N. Buller, K. Reuter, and J. Kunze-liebhäuser, “Self-activation of copper electrodes during co electro-oxidation in alkaline electrolyte,”Nat Catal, vol. 3, pp. 797–803, 2020

    work page 2020

  50. [50]

    Ternary ruthenium complex hydrides for ammonia synthesis via the associative mechanism,

    Q. Wang, J. Pan, J. Guo, H. A. Hansen, H. Xie, L. Jiang, L. Hua, H. Li, Y. Guan, P. Wang, W. Gao, L. Liu, H. Cao, Z. Xiong, T. Vegge, and P. Chen, “Ternary ruthenium complex hydrides for ammonia synthesis via the associative mechanism,”Nat Catal, vol. 4, pp. 959–967, 2021

    work page 2021

  51. [51]

    J. K. Nørskov, F. Studt, F. Abild-Pedersen, and T. Bligaard,Fundamental Concepts in Heterogeneous Catalysis. John Wiley & Sons, Inc., 2014

    work page 2014

  52. [52]

    Exploringcatalyticreactionnetworkswithmachine learning,

    J.T.Margraf, H.Jung, C.Scheurer, andK.Reuter, “Exploringcatalyticreactionnetworkswithmachine learning,”Nat Catal, vol. 6, pp. 112–121, 2023

    work page 2023

  53. [53]

    Neural network potentials for accelerated meta- dynamics of oxygen reduction kinetics at au–water interfaces,

    X. Yang, A. Bhowmik, T. Vegge, and H. A. Hansen, “Neural network potentials for accelerated meta- dynamics of oxygen reduction kinetics at au–water interfaces,”Chem. Sci., vol. 14, pp. 3913–3922, 2023

    work page 2023

  54. [54]

    Active learning across intermetallics to guide discovery of electrocatalysts for co2 reduction and h2 evolution,

    K. Tran and Z. W. Ulissi, “Active learning across intermetallics to guide discovery of electrocatalysts for co2 reduction and h2 evolution,”Nat Catal, vol. 1, pp. 696–703, 2018

    work page 2018

  55. [55]

    Learning design rules for selective oxidation catalysts from high-throughput experimentation and artificial intelligence,

    L. Foppa, C. Sutton, L. M. Ghiringhelli, S. De, P. Löser, S. A. Schunk, A. Schäfer, and M. Schef- fler, “Learning design rules for selective oxidation catalysts from high-throughput experimentation and artificial intelligence,”ACS Catalysis, vol. 12, no. 4, pp. 2223–2232, 2022

    work page 2022

  56. [56]

    A data-driven high-throughput workflow applied to promoted in-oxide catalysts for co2 hydrogenation to methanol,

    M. Khatamirad, E. Fako, C. Boscagli, M. Müller, F. Ebert, R. Naumann d’Alnoncourt, A. Schaefer, S. A. Schunk, I. Jevtovikj, F. Rosowski, and S. De, “A data-driven high-throughput workflow applied to promoted in-oxide catalysts for co2 hydrogenation to methanol,”Catalysis Science & Technology, vol. 13, no. 9, p. 2656–2661, 2023

    work page 2023

  57. [57]

    Estimating free energy barriers for heterogeneous catalytic reactions with machine learning potentials and umbrella integration,

    S. Stocker, H. Jung, G. Csányi, C. F. Goldsmith, K. Reuter, and J. T. Margraf, “Estimating free energy barriers for heterogeneous catalytic reactions with machine learning potentials and umbrella integration,”J. Chem. Theory Comput., vol. 19, pp. 6796–6804, 2023

    work page 2023

  58. [58]

    The open catalyst 2022 (oc22) dataset and challenges for oxide electrocatalysts,

    R. Tran, J. Lan, M. Shuaibi, B. M. Wood, S. Goyal, A. Das, J. Heras-domingo, A. Kolluru, A. Rizvi, N. Shoghi, A. Sriram, F. Therrien, J. Abed, O. Voznyy, E. H. Sargent, Z. Ulissi, and C. L. Zitnick, “The open catalyst 2022 (oc22) dataset and challenges for oxide electrocatalysts,”ACS Catal., vol. 13, pp. 3066–3084, 2023

    work page 2022

  59. [59]

    Pourbaix, ATLAS of Electrochemical Equilibria in Aqueous Solutions

    M. Pourbaix, ATLAS of Electrochemical Equilibria in Aqueous Solutions. Pergamon Press, Oxford, 1966

    work page 1966

  60. [60]

    Pourbaix,Lectures on Electrochemical Corrosion

    M. Pourbaix,Lectures on Electrochemical Corrosion. Springer Science and Business Media, 1973

    work page 1973

  61. [61]

    Origin of the overpotential for oxygen reduction at a fuel-cell cathode,

    J. K. Nørskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J. R. Kitchin, T. Bligaard, and H. Jónsson, “Origin of the overpotential for oxygen reduction at a fuel-cell cathode,”J. Phys. Chem. B, vol. 108, pp. 17886–17892, Nov. 2004

    work page 2004

  62. [62]

    Prediction of solid-aqueous equilibria: Scheme to combine first-principles calculations of solids with experimental aqueous states,

    K. A. Persson, B. Waldwick, P. Lazic, and G. Ceder, “Prediction of solid-aqueous equilibria: Scheme to combine first-principles calculations of solids with experimental aqueous states,”Phys. Rev. B Condens. Matter Mater. Phys., vol. 85, June 2012

    work page 2012

  63. [63]

    Electrochemical stability of metastable materials,

    A. K. Singh, L. Zhou, A. Shinde, S. K. Suram, J. H. Montoya, D. Winston, J. M. Gregoire, and K. A. Persson, “Electrochemical stability of metastable materials,”Chem. Mater., vol. 29, pp. 10159–10167, Dec. 2017

    work page 2017

  64. [64]

    Scaling properties of adsorption energies for hydrogen-containing molecules on transition-metal surfaces,

    F. Abild-Pedersen, J. Greeley, F. Studt, J. Rossmeisl, T. R. Munter, P. G. Moses, E. Skúlason, T. Bli- gaard, and J. K. Nørskov, “Scaling properties of adsorption energies for hydrogen-containing molecules on transition-metal surfaces,”Physical Review Letters, vol. 99, July 2007. 22

    work page 2007

  65. [65]

    Strategies to break linear scaling relationships,

    J. Pérez-Ramírez and N. López, “Strategies to break linear scaling relationships,”Nature Catalysis, vol. 2, p. 971–976, Oct. 2019

    work page 2019

  66. [66]

    Statistical learning goes beyond the d-band model providing the thermochemistry of adsorbates on transition metals,

    R. García-Muelas and N. López, “Statistical learning goes beyond the d-band model providing the thermochemistry of adsorbates on transition metals,”Nat. Commun., vol. 10, p. 4687, Oct. 2019

    work page 2019

  67. [67]

    Electrochemical oxidation of co on cu single crystals under alkaline conditions,

    A. Tiwari, H. H. Heenen, A. S. Bjørnlund, D. Hochfilzer, K. Chan, and S. Horch, “Electrochemical oxidation of co on cu single crystals under alkaline conditions,”ACS Energy Letters, vol. 5, no. 11, pp. 3437–3442, 2020

    work page 2020

  68. [68]

    Rationally designed indium oxide catalysts for CO2 hydrogenation to methanol with high activity and selectivity,

    S. Dang, B. Qin, Y. Yang, H. Wang, J. Cai, Y. Han, S. Li, P. Gao, and Y. Sun, “Rationally designed indium oxide catalysts for CO2 hydrogenation to methanol with high activity and selectivity,”Science advances, vol. 6, no. 25, p. eaaz2060, 2020

    work page 2020

  69. [69]

    Machine learning the quantum-chemical properties of metal–organic frameworks for accelerated materials discovery,

    A. S. Rosen, S. M. Iyer, D. Ray, Z. Yao, A. Aspuru-Guzik, L. Gagliardi, J. M. Notestein, and R. Q. Snurr, “Machine learning the quantum-chemical properties of metal–organic frameworks for accelerated materials discovery,”Matter, vol. 4, no. 5, pp. 1578–1597, 2021

    work page 2021

  70. [70]

    High-throughputpredictionsofmetal–organicframeworkelectronicproper- ties: theoretical challenges, graph neural networks, and data exploration,

    A. S. Rosen, V. Fung, P. Huck, C. T. O’Donnell, M. K. Horton, D. G. Truhlar, K. A. Persson, J. M. Notestein, andR.Q.Snurr, “High-throughputpredictionsofmetal–organicframeworkelectronicproper- ties: theoretical challenges, graph neural networks, and data exploration,”npj Computational Materials, vol. 8, no. 1, p. 112, 2022

    work page 2022

  71. [71]

    DeePMD-kit v2: A software package for deep potential models,

    J. Zeng, D. Zhang, D. Lu, P. Mo, Z. Li, Y. Chen, M. Rynik, L. Huang, Z. Li, S. Shi, Y. Wang, H. Ye, P. Tuo, J. Yang, Y. Ding, Y. Li, D. Tisi, Q. Zeng, H. Bao, Y. Xia, J. Huang, K. Muraoka, Y. Wang, J. Chang, F. Yuan, S. L. Bore, C. Cai, Y. Lin, B. Wang, J. Xu, J.-X. Zhu, C. Luo, Y. Zhang, R. E. A. Goodall, W. Liang, A. K. Singh, S. Yao, J. Zhang, R. Wentz...

    work page 2023

  72. [72]

    Uff, a full periodic table force field for molecular mechanics and molecular dynamics simulations,

    A. K. Rappé, C. J. Casewit, K. Colwell, W. A. Goddard III, and W. M. Skiff, “Uff, a full periodic table force field for molecular mechanics and molecular dynamics simulations,”Journal of the American chemical society, vol. 114, no. 25, pp. 10024–10035, 1992

    work page 1992

  73. [73]

    Introducing ddec6 atomic population analysis: part 1. charge partitioning theory and methodology,

    T. A. Manz and N. G. Limas, “Introducing ddec6 atomic population analysis: part 1. charge partitioning theory and methodology,”RSC advances, vol. 6, no. 53, pp. 47771–47801, 2016

    work page 2016

  74. [74]

    Vapor–liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen,

    J. J. Potoff and J. I. Siepmann, “Vapor–liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen,”AIChE journal, vol. 47, no. 7, pp. 1676–1682, 2001

    work page 2001

  75. [75]

    O. M. Yaghi, M. J. Kalmutzki, and C. S. Diercks,Introduction to reticular chemistry: metal-organic frameworks and covalent organic frameworks. John Wiley & Sons, 2019

    work page 2019

  76. [76]

    Highly efficient separation of carbon dioxide by a metal-organic framework replete with open metal sites,

    D. Britt, H. Furukawa, B. Wang, T. G. Glover, and O. M. Yaghi, “Highly efficient separation of carbon dioxide by a metal-organic framework replete with open metal sites,”Proceedings of the National Academy of Sciences, vol. 106, no. 49, pp. 20637–20640, 2009

    work page 2009

  77. [77]

    Comprehensive study of carbon dioxide adsorption in the metal– organic frameworks M 2 (dobdc)(M= Mg, Mn, Fe, Co, Ni, Cu, Zn),

    W. L. Queen, M. R. Hudson, E. D. Bloch, J. A. Mason, M. I. Gonzalez, J. S. Lee, D. Gygi, J. D. Howe, K. Lee, T. A. Darwish,et al., “Comprehensive study of carbon dioxide adsorption in the metal– organic frameworks M 2 (dobdc)(M= Mg, Mn, Fe, Co, Ni, Cu, Zn),”Chemical Science, vol. 5, no. 12, pp. 4569–4581, 2014

    work page 2014

  78. [78]

    Mof-74 type variants for co 2 capture,

    J. H. Choe, H. Kim, and C. S. Hong, “Mof-74 type variants for co 2 capture,”Materials Chemistry Frontiers, vol. 5, no. 14, pp. 5172–5185, 2021

    work page 2021

  79. [79]

    Coordinatively unsaturated metal sites (open metal sites) in metal–organic frameworks: design and applications,

    Ü. Kökçam-Demir, A. Goldman, L. Esrafili, M. Gharib, A. Morsali, O. Weingart, and C. Janiak, “Coordinatively unsaturated metal sites (open metal sites) in metal–organic frameworks: design and applications,”Chemical Society Reviews, vol. 49, no. 9, pp. 2751–2798, 2020. 23

    work page 2020

  80. [80]

    Quantum informed machine-learning potentials for molecular dynamics simulations of CO2’s chemisorption and diffusion in Mg-MOF-74,

    B. Zheng, F. L. Oliveira, R. Neumann Barros Ferreira, M. Steiner, H. Hamann, G. X. Gu, and B. Luan, “Quantum informed machine-learning potentials for molecular dynamics simulations of CO2’s chemisorption and diffusion in Mg-MOF-74,”ACS nano, vol. 17, no. 6, pp. 5579–5587, 2023

    work page 2023

Showing first 80 references.