Pith. sign in

REVIEW 5 major objections 5 minor 49 references

Mechanistic Insights into the Oxygen Evolution Reaction on Nickel-Doped Barium Titanate via Machine Learning-Accelerated Simulations

T0 review · 5 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Nickel doping lowers the rate-limiting oxygen evolution barrier on barium titanate from 1.57 eV to 1.20 eV, and identifies oxo-oxo bond formation as the bottleneck.

desk verdict A serious MLP+MetaD study with a likely-true qualitative conclusion, but the headline barrier numbers don't match the table, the negative overpotential is an error, and the combined TS3 barrier rests on an unproven assumption about the CV projection. read the letter →

arxiv 2412.15452 v1 pith:FG4S7U2Y submitted 2024-12-19 cond-mat.mtrl-sci cond-mat.dis-nn

Kajjana Boonpalit , Nongnuch Artrith This is my paper · ORCID
classification cond-mat.mtrl-scicond-mat.dis-nn
keywords oxygenevolutionreactionbariumtitanatenickeldopingmachinelearningpotentialmetadynamicsfreeenergybarrierperovskiteoxideexplicitwaterinterface
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to establish that replacing a surface titanium atom with nickel lowers the kinetic barrier of the oxygen evolution reaction on barium titanate in contact with liquid water, and identifies which elementary step controls the rate. The authors build a machine-learning interatomic potential trained on density functional theory data, then run well-tempered metadynamics simulations of the full water-covered surface. They find the rate-limiting step is the combined oxo-oxo bond formation and deprotonation that produces an adsorbed O2 species, with free energy barriers of 1.57 eV on pristine BaTiO3 and 1.20 eV on Ni-doped BaTiO3. They also find molecular O2 desorption is easy, about 0.16 eV and 0.13 eV, so it is not rate limiting. If correct, this explains experimentally observed overpotential reduction by Ni doping and provides a mechanistic handle for designing better perovskite OER catalysts.

What carries the argument

The machinery is a machine-learned interatomic potential trained on density functional theory data and used to run well-tempered metadynamics simulations of a BaTiO3(001) slab with 128 explicit water molecules, 592 atoms in total. The reaction is biased along two collective variables: the coordination number between the adsorbed oxygen $O_s$ and hydrogen, $\mathrm{CN}(O_s{-}H)$, and the coordination number between $O_s$ and surrounding water oxygens, $\mathrm{CN}(O_s{-}O_{aw})$. These variables separate the $H_2O^*$, $OH^*$, $O^*$, and $O_2^*$ states on the free energy surface, though $OOH^*$ is not resolved as a distinct minimum; step 3 (oxo-oxo bond formation) and step 4 (proton abstraction) are therefore combined into a single TS3 barrier. A separate metadynamics run uses the distance from the O2 center of mass to the surface metal as the collective variable to measure O2 desorption.

What would settle it

Run the same metadynamics with a third collective variable that distinguishes the OOH* intermediate from the O2* product (for example, the coordination of the reacting water's hydrogen to the surface oxygen), or carry out static transition-state searches for the oxo-oxo bond formation step; if the barrier drops below 1.20 eV or the TS3 assignment changes, the central claim fails.

Watch

Extended reading notes

Core claim

The central claim is that Ni doping lowers the free energy barrier of the oxygen evolution reaction at the BaTiO3(001)/water interface, and that the rate-determining step is the combined oxo-oxo bond formation and proton-abstraction event that produces an adsorbed O2 species. On pristine BaTiO3 the computed barrier is $\Delta G^{\ddagger}_{O \to O_2} = 1.57$ eV; on Ni-doped BaTiO3 it is 1.20 eV. The corresponding theoretical overpotentials are 0.34 V and -0.03 V, and the predicted 0.34 V reduction matches the experimentally measured reduction from 0.80 V to 0.46 V on Ni-doped hollow porous spheres. The paper further claims that O2 desorption is not rate-determining, with barriers near 0.16 eV (BaTiO3) and 0.13 eV (Ni@BaTiO3), and that Ni doping also lowers the endothermicity of O2* formation from 1.37 eV to 0.97 eV.

Load-bearing premise

The load-bearing premise is that the two reaction coordinates chosen for the metadynamics—how many hydrogens are bound to the surface oxygen and how many water oxygens are bound to it—fully capture the path of the rate-limiting step, so that the computed 1.20 eV and 1.57 eV barriers are real and not an artifact of how the simulation was biased.

Editorial extensions

If this is right

  • Ni doping reduces the rate-limiting OER free energy barrier from 1.57 eV to 1.20 eV, a drop of 0.37 eV that matches the experimentally observed overpotential reduction of 0.34 V.
  • O2 desorption is not rate-determining on either surface, with desorption barriers of about 0.16 eV (BaTiO3) and 0.13 eV (Ni@BaTiO3), so product release will not block the active site.
  • The rate-determining step is the formation of the oxo-oxo bond (step 3), not the later proton abstraction, which is barrierless on the computed free energy surface.
  • Ni doping lowers the endothermicity of O2* formation from 1.37 eV to 0.97 eV, making the overall OER both kinetically and thermodynamically more favorable.
  • The machine-learning potential trained for this five-element water/slab system provides a basis for future simulations of the lattice-oxygen-mediated mechanism, which the authors explicitly leave for later work.

Reading between the lines

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

  • Extension beyond the paper: if the same barrier ordering persists at operating electrode potentials, Ni-doped BaTiO3 could close the activity gap with iridium and ruthenium oxides, since its theoretical rate-limiting barrier of 1.20 eV already sits in the range of good OER catalysts.
  • Extension beyond the paper: the indistinct OOH* minimum suggests the conventional four-step OER picture may overstate the role of OOH* as a stable intermediate on this surface; descriptors built from OOH* adsorption energies may therefore be the wrong reactivity measure here.
  • Extension beyond the paper: the data-generation strategy of seeding with a broad pretrained potential and refining with active learning on unstable structures is directly transferable to other doped perovskites, allowing kinetic barriers to be screened without expensive ab initio molecular dynamics.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The paper develops a machine-learned interatomic potential (ANN-based, RPBE+D3 reference) for BaTiO3 and Ni-doped BaTiO3 with explicit water, and uses well-tempered metadynamics to map the free energy surface of the oxygen evolution reaction (OER) through the conventional four-step adsorbate mechanism and O2 desorption. The authors report that the rate-determining step is oxo-oxo bond formation (TS3), with a combined free energy barrier of 1.57 eV for BTO and 1.20 eV for Ni@BTO, and a corresponding theoretical overpotential reduction from 0.34 V to -0.03 V. They conclude that Ni doping lowers both kinetic and thermodynamic barriers, consistent with prior DFT and experimental comparisons. The methodological contribution includes an active-learning dataset of 16,162 configurations and MLP-MetaD simulations at explicit water interfaces.

Significance. If the quantitative barrier reduction is correct, this is a valuable demonstration of MLP-accelerated metadynamics for OER at explicit oxide/water interfaces, going beyond thermodynamic scaling relations to include kinetic barriers and solvent effects. The qualitative conclusion that Ni doping enhances OER activity on BaTiO3 is consistent with earlier DFT and experimental work, and the MLP pipeline with active learning is a reproducible strength. However, the central quantitative claim depends on the validity of a two-collective-variable projection of a proton-coupled reaction, and several internal inconsistencies in the reported barrier values and their statistical treatment currently prevent the quantitative result from being accepted as established.

major comments (5)
  1. [Section 2.2 and Section 4.4] The central claim that TS3 is the rate-determining step with barriers of 1.57 and 1.20 eV relies on the assertion that the OOH* intermediate is indistinct and that the transition states of oxo-oxo bond formation and proton abstraction 'do not interfere' on the free energy surface. The two collective variables, CN(Os-H) and CN(Os-Oaw), do not include an explicit coordinate for the proton transfer from OOH* to a solvent water molecule. A saddle point in this two-dimensional projection is not necessarily the true transition state of the combined reaction, and sequential barriers can merge into a single apparent maximum. The manuscript states this as an assumption, not as a tested property. A concrete test would be to repeat the MetaD with an additional CV for the O_w-H coordinate, or to perform committor analysis at the putative TS3, before the reported barrier heights can be considered mechanistically quantitative.
  2. [Table C5 and Section 2.2] The headline barriers are internally inconsistent. The text reports ΔG‡O→O2 = 1.57 eV for BTO and 1.20 eV for Ni@BTO, but Table C5 lists three individual FES values per system whose means are 1.48 ± 0.07 eV and 1.26 eV, respectively. The reported mean for Ni@BTO is 1.26 eV, not 1.20 eV, and no individual Ni@BTO run in the table gives 1.20 eV (the values are 1.31, 1.20, 1.26; the mean of 1.20, 1.26, 1.31 is not 1.26). The paper must either correct the tabulated statistics or state explicitly which runs and averaging procedure produce the claimed 1.57/1.20 values. As written, the quantitative comparison between theory and experiment is not reproducible from the data presented.
  3. [Section 2.3] The conversion of the activation barrier to a 'theoretical overpotential' by subtracting 1.23 eV yields -0.03 V for Ni@BTO. A negative theoretical overpotential is unphysical for a barrier-based estimate; the overpotential should be derived from the potential-dependent free energy change or from the largest free energy step at U=0, not from subtracting 1.23 eV from an activation free energy at zero applied potential. This step needs to be justified or removed, because the claim of 'close agreement' with the experimental overpotential reduction (0.34 V) currently rests on an undefined quantity.
  4. [Section 4.4 and Figure C3/C4] The free energy surfaces are 'calculated only until the O2 formation event occurs', as stated in Section 4.4. Because well-tempered metadynamics estimates free energies from accumulated bias, truncating the simulation at the first O2 formation event means the reported TS3 barrier depends on when deposition is stopped and on the time needed to cross the barrier, rather than on a converged free energy surface. The paper does not provide a convergence check (e.g., hill-to-hill variation or block analysis) for the TS3 region. Given that the central result is a 0.3-0.4 eV barrier difference, the stopping criterion is a load-bearing choice that must be shown not to bias the relative barriers.
  5. [Section 4.2, Appendix B, and Figure B1] The MLP force error (MAE 132-143 meV/Å on the validation set) is large relative to the reported free energy barriers (0.06-0.24 eV for early steps) and to the TS3 difference between BTO and Ni@BTO. The RDF validation in Figure B1 is performed on only 15 ps of AIMD and the caption itself notes the simulations 'may not achieve equilibrium'. The manuscript does not quantify how MLP force and energy errors propagate into the metadynamics free energies. At minimum, the authors should report the MLP ensemble spread for the MetaD barriers (Figure C6 shows an ensemble comparison only for relative energies, not for barriers) and discuss the error budget for the 0.37 eV doping-induced reduction.
minor comments (5)
  1. [Section 2.2] The sentence 'There is no any potential impact on its interpretation of oxo-oxo bond formation energy barrier' is grammatically unclear and should be rewritten to state precisely what was tested.
  2. [Section 2.2 / Figure 5 caption] The phrase 'averaged over the FESs three MetaD runs' is missing a preposition; it should read 'averaged over the FESs from three MetaD runs'.
  3. [Appendix B, Table B4] The text contains 'as as shown' in the sentence about MACE computational time; this should be corrected.
  4. [Data availability] The repository URL is a placeholder ('https://github.com/atomisticnet/XXXXX'); the actual DOI or repository link should be provided for reproducibility.
  5. [Table C5] The standard deviation row for Ni@BTO reports ±0.00 for several quantities despite visible spread among the individual FES values; this is either a formatting error or a statistical error and should be corrected.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the central OER barriers are emergent MLP-MetaD results; only a minor non-load-bearing self-citation exists.

full rationale

The central quantitative claim—the RDS barrier lowering from ΔG‡O→O2 = 1.57 eV (BTO) to 1.20 eV (Ni@BTO)—is produced by well-tempered metadynamics on a machine-learned potential, not by fitting to the experimental overpotential. The MLP is validated against independent RPBE+D3 energies/forces (energy MAE 7–9 meV/atom, force MAE ≈132 meV/Å) and against AIMD RDFs, so the barrier is not an input parameter. Active learning did add transition-state configurations from earlier MetaD trajectories to the training set (Appendix A), but those configurations were labeled with DFT and only improve the model's accuracy in the TS region; the free-energy barrier itself is not a fitted quantity and no equation in the paper reduces it to a training label. The slab models are taken from the authors' prior work [13], but that self-citation supplies initial structures and a qualitative DFT comparison, not the kinetic barrier. Section 2.2's statement that OOH* is indistinguishable and that TS3 merges Steps 3–4 is an interpretive assumption about the reaction coordinate; it creates a correctness risk for the physical attribution of the barrier, but it is not a circular definition because the barrier value is computed from the simulated FES rather than assumed. One internal consistency note, not a circularity: the text quotes 1.57/1.20 eV while the averaged values in Table C5 are 1.48/1.26 eV; this affects reproducibility of the exact number but does not make the derivation circular. Overall, the derivation chain is self-contained against DFT-labeled data and external experimental observations, so no significant circularity is found.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The listed assumptions are background premises the barrier calculations depend on; none is validated inside the paper beyond MLP error metrics and a short AIMD comparison. The most consequential is the CV sufficiency assumption, since the authors themselves note the OOH* and O2* states are not resolved.

free parameters (3)
  • MLP loss weighting α = 0.5
    Weight between energy and force errors in Eq. 5; chosen by hand, affects MLP accuracy and the resulting barriers.
  • CN exponents m,n and r0 = m=6, n=12, r0=1.2 Å (CV1), 1.5 Å (CV2)
    Hand-chosen to define coordination-number collective variables; directly shapes the FES.
  • MetaD Gaussian height, width, deposition rate = height 0.05-0.1 eV, width 0.1, every 62.5 fs (OER); height 0.01 eV, width 0.05, every 6.25 fs (desorption)
    Well-tempered MetaD parameters varied across runs; influence barrier estimates and convergence.
assumptions (4)
  • domain assumption RPBE+D3 DFT provides a sufficiently accurate reference for OER on BTO/water interfaces.
    All 16,162 MLP labels are RPBE+D3 single-point energies; functional errors propagate directly into the free energy barriers.
  • domain assumption The conventional adsorbate evolution mechanism (Eqs. 1-4) is the operative OER pathway.
    Conclusion states LOM is not included; if lattice oxygen participates, the reported RDS may not be the active route.
  • ad hoc to paper The CVs CN(Os-H) and CN(Os-Oaw) are sufficient reaction coordinates and TS3 is unaffected by the unresolved OOH*/O2* distinction.
    Section 2.2: 'we are unable to bias and monitor the CN(O w-H)... challenging to differentiate the OOH* and O2* states' yet assumes 'there is no any potential impact.'
  • ad hoc to paper The largest free energy barrier can be converted to a theoretical overpotential by subtracting 1.23 eV.
    Used in Section 2.3 to compare with experiment; yields -0.03 V for Ni@BTO, which is unphysical.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mechanistic Insights into the Oxygen Evolution Reaction on Nickel-Doped Barium Titanate via Machine Learning-Accelerated Simulations." pith.science (2026). https://pith.science/paper/FG4S7U2Y

@misc{pith2026241215452,
  author       = {Pith},
  title        = {Pith review of: Mechanistic Insights into the Oxygen Evolution Reaction on Nickel-Doped Barium Titanate via Machine Learning-Accelerated Simulations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FG4S7U2Y}},
  note         = {Machine review of arXiv:2412.15452}
}
abstract

Electrocatalytic water splitting, which produces hydrogen and oxygen through water electrolysis, is a promising method for generating renewable, carbon-free alternative fuels. However, its widespread adoption is hindered by the high costs of Pt cathodes and IrO$_{x}$/RuO$_{x}$ anode catalysts. In the search for cost-effective alternatives, barium titanate (BaTiO$_{3}$) has emerged as a compelling candidate. This inexpensive, non-toxic perovskite oxide can be synthesized from earth-abundant precursors and has shown potential for catalyzing the oxygen evolution reaction (OER) in recent studies. In this work, we explore the OER activity of pristine and Ni-doped BaTiO$_{3}$ at explicit water interfaces using metadynamics (MetaD) simulations. To enable efficient and practical MetaD for OER, we developed a machine learning interatomic potential based on artificial neural networks (ANN), achieving large-scale and long-time simulations with near-DFT accuracy. Our simulations reveal that Ni-doping enhances the catalytic activity of BaTiO$_{3}$ for OER, consistent with experimental observations, while providing mechanistic insights into this enhancement.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 41 canonical work pages

  1. [1]

    Advanced Materials 33(31), 2007100 (2021)

    Yu, Z.-Y., Duan, Y., Feng, X.-Y., Yu, X., Gao, M.-R., Yu, S.-H.: Clean and affordable hydrogen fuel from alkaline water splitting: past, recent progress, and future prospects. Advanced Materials 33(31), 2007100 (2021)

    work page 2021

  2. [2]

    Nature Energy 4(6), 430–433 (2019)

    Kibsgaard, J., Chorkendorff, I.: Considerations for the scaling-up of water splitting catalysts. Nature Energy 4(6), 430–433 (2019)

    work page 2019

  3. [3]

    Advanced materials 28(2), 215–230 (2016)

    Wang, J., Cui, W., Liu, Q., Xing, Z., Asiri, A.M., Sun, X.: Recent progress in cobalt-based heterogeneous catalysts for electrochemical water splitting. Advanced materials 28(2), 215–230 (2016)

    work page 2016

  4. [4]

    Coordination Chemistry Reviews 474, 214864 (2023)

    Al-Naggar, A.H., Shinde, N.M., Kim, J.-S., Mane, R.S.: Water splitting per- formance of metal and non-metal-doped transition metal oxide electrocatalysts. Coordination Chemistry Reviews 474, 214864 (2023)

    work page 2023

  5. [5]

    ChemSusChem 14(24), 5359–5383 (2021)

    Liu, Y., Zhou, D., Deng, T., He, G., Chen, A., Sun, X., Yang, Y., Miao, P.: Research progress of oxygen evolution reaction catalysts for electrochemical water splitting. ChemSusChem 14(24), 5359–5383 (2021)

    work page 2021

  6. [6]

    a comparative review

    ˇDuroviˇ c, M., Hn´ at, J., Bouzek, K.: Electrocatalysts for the hydrogen evolution reaction in alkaline and neutral media. a comparative review. Journal of Power Sources 493, 229708 (2021)

    work page 2021

  7. [7]

    ChemCatChem 3(7), 1159–1165 (2011)

    Man, I.C., Su, H.-Y., Calle-Vallejo, F., Hansen, H.A., Mart ´ ınez, J.I., Inoglu, N.G., 32 Kitchin, J., Jaramillo, T.F., Nørskov, J.K., Rossmeisl, J.: Universality in oxy- gen evolution electrocatalysis on oxide surfaces. ChemCatChem 3(7), 1159–1165 (2011)

    work page 2011

  8. [8]

    Advanced Energy Materials 11(12), 2003755 (2021)

    Sun, Y., Li, R., Chen, X., Wu, J., Xie, Y., Wang, X., Ma, K., Wang, L., Zhang, Z., Liao, Q., et al.: A-site management prompts the dynamic reconstructed active phase of perovskite oxide oer catalysts. Advanced Energy Materials 11(12), 2003755 (2021)

    work page 2021

Show all 49 references
  1. [9]

    Small 17(43), 2101605 (2021)

    Liu, D., Zhou, P., Bai, H., Ai, H., Du, X., Chen, M., Liu, D., Ip, W.F., Lo, K.H., Kwok, C.T., et al.: Development of perovskite oxide-based electrocatalysts for oxygen evolution reaction. Small 17(43), 2101605 (2021)

    2021

  2. [10]

    Nature communications 7(1), 11053 (2016)

    Mefford, J.T., Rong, X., Abakumov, A.M., Hardin, W.G., Dai, S., Kolpak, A.M., Johnston, K.P., Stevenson, K.J.: Water electrolysis on la1- x sr x coo3- δ perovskite electrocatalysts. Nature communications 7(1), 11053 (2016)

    2016

  3. [11]

    ACS catalysis 6(2), 1153–1158 (2016)

    Rong, X., Parolin, J., Kolpak, A.M.: A fundamental relationship between reac- tion mechanism and stability in metal oxide catalysts for oxygen evolution. ACS catalysis 6(2), 1153–1158 (2016)

    2016

  4. [12]

    Journal of Materials Chemistry A 9(36), 20131–20163 (2021) 33

    Ede, S.R., Luo, Z.: Tuning the intrinsic catalytic activities of oxygen-evolution catalysts by doping: a comprehensive review. Journal of Materials Chemistry A 9(36), 20131–20163 (2021) 33

    2021

  5. [13]

    Physical Chemistry Chemical Physics 18(42), 29561–29570 (2016)

    Artrith, N., Sailuam, W., Limpijumnong, S., Kolpak, A.M.: Reduced overpoten- tials for electrocatalytic water splitting over fe-and ni-modified batio 3. Physical Chemistry Chemical Physics 18(42), 29561–29570 (2016)

    2016

  6. [14]

    ACS Catalysis 11(9), 5601–5613 (2021)

    Peng, Y., Hajiyani, H., Pentcheva, R.: Influence of fe and ni doping on the oer performance at the co3o4 (001) surface: insights from dft+ u calculations. ACS Catalysis 11(9), 5601–5613 (2021)

    2021

  7. [15]

    Powder technology 217, 629–633 (2012)

    Zhang, Y., Wang, L., Xue, D.: Molten salt route of well dispersive barium titanate nanoparticles. Powder technology 217, 629–633 (2012)

    2012

  8. [16]

    Current Journal of Applied Science and Technology 43(5), 1–10 (2024)

    Onorato, G.d.C., Amaral, D.L.A.S.d., Oliveira, L.F.C.d., Brandao, H.d.M., Munk, M.: Barium titanate nanoparticles exhibit cytocompatibility in cultured bovine fibroblasts: A model for dermal exposure. Current Journal of Applied Science and Technology 43(5), 1–10 (2024)

    2024

  9. [17]

    Journal of The Electrochemical Society 123(11), 1683 (1976)

    Kennedy, J.H., Frese, K.W.: Photo-oxidation of water at barium titanate elec- trodes. Journal of The Electrochemical Society 123(11), 1683 (1976)

    1976

  10. [18]

    Materials Research Bulletin 11(8), 985–992 (1976)

    Nasby, R., Quinn, R.K.: Photoassisted electrolysis of water using a batio3 electrode. Materials Research Bulletin 11(8), 985–992 (1976)

    1976

  11. [19]

    Applied Surface Science 618, 156599 (2023) 34

    El Emam, H.K., Abdelwahab, A., El-Dek, S., El Rouby, W.M.: Performance of ni-doped batio3 hollow porous spheres supported reduced graphene oxide as an efficient bifunctional electrocatalyst for oxygen evolution reaction and oxygen reduction reaction. Applied Surface Science 61...

    2023

  12. [20]

    Chemistry–A European Journal 27(68), 17024–17037 (2021)

    Creazzo, F., Luber, S.: Water-assisted chemical route towards the oxygen evo- lution reaction at the hydrated (110) ruthenium oxide surface: Heterogeneous catalysis via dft-md and metadynamics simulations. Chemistry–A European Journal 27(68), 17024–17037 (2021)

    2021

  13. [21]

    Nature Communications 14(1), 6131 (2023)

    Zeng, Z., Wodaczek, F., Liu, K., Stein, F., Hutter, J., Chen, J., Cheng, B.: Mechanistic insight on water dissociation on pristine low-index tio2 surfaces from machine learning molecular dynamics simulations. Nature Communications 14(1), 6131 (2023)

    2023

  14. [22]

    ChemElectroChem 11(10), 202300681 (2024) https://doi.org/10.1002/celc.202300681 https://chemistry- europe.onlinelibrary.wiley.com/doi/pdf/10.1002/celc.202300681

    Boonpalit, K., Kinchagawat, J., Namuangruk, S.: Expanding the applicability domain of machine learning model for advancements in electrochemical material discovery. ChemElectroChem 11(10), 202300681 (2024) https://doi.org/10.1002/celc.202300681 https://chemistry- europe.online...

    2024 doi

  15. [23]

    Advances in Neural Information Processing Systems 35, 11423–11436 (2022)

    Batatia, I., Kovacs, D.P., Simm, G., Ortner, C., Cs´ anyi, G.: Mace: Higher order equivariant message passing neural networks for fast and accurate force fields. Advances in Neural Information Processing Systems 35, 11423–11436 (2022)

    2022

  16. [24]

    Batatia, I., Benner, P., Chiang, Y., Elena, A.M., Kov´ acs, D.P., Riebesell, J., Advincula, X.R., Asta, M., Baldwin, W.J., Bernstein, N., Bhowmik, A., Blau, S.M., C˘ arare, V., Darby, J.P., De, S., Pia, F.D., Deringer, V.L., Elijoˇ sius, R., El-Machachi, Z., Fako, E., Ferrari,...

    2023 arXiv

  17. [25]

    The Journal of chemical physics 110(11), 5029– 5036 (1999)

    Ernzerhof, M., Scuseria, G.E.: Assessment of the perdew–burke–ernzerhof exchange-correlation functional. The Journal of chemical physics 110(11), 5029– 5036 (1999)

    1999

  18. [26]

    arXiv preprint arXiv:2311.08177 (2023)

    Kong, L., Li, J., Sun, L., Yang, H., Hao, H., Chen, C., Artrith, N., Torres, J.A.G., Lu, Z., Zhou, Y.: Overcoming the size limit of first principles molecular dynamics simulations with an in-distribution substructure embedding active learner. arXiv preprint arXiv:2311.08177 (2023)

    2023 arXiv

  19. [27]

    ACS Publications (2023)

    Maloney, M.P., Coley, C.W., Genheden, S., Carson, N., Helquist, P., Norrby, P.- O., Wiest, O.: Negative data in data sets for machine learning training. ACS Publications (2023)

    2023

  20. [28]

    Nature Catalysis 6(2), 108–111 (2023)

    Taniike, T., Takahashi, K.: The value of negative results in data-driven catalysis research. Nature Catalysis 6(2), 108–111 (2023)

    2023

  21. [29]

    Hammer, B., Hansen, L.B., Nørskov, J.K.: Improved adsorption energetics within 36 density-functional theory using revised perdew-burke-ernzerhof functionals. Phys. Rev. B 59, 7413–7421 (1999) https://doi.org/10.1103/PhysRevB.59.7413

    1999 doi

  22. [30]

    The Journal of chemical physics 132(15) (2010)

    Grimme, S., Antony, J., Ehrlich, S., Krieg, H.: 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 132(15) (2010)

    2010

  23. [31]

    Electrochimica Acta 55(27), 7975–7981 (2010)

    Tripkovi´ c, V., Sk´ ulason, E., Siahrostami, S., Nørskov, J.K., Rossmeisl, J.: The oxygen reduction reaction mechanism on pt (1 1 1) from density functional theory calculations. Electrochimica Acta 55(27), 7975–7981 (2010)

    2010

  24. [32]

    Physical Review B—Condensed Matter and Materials Physics 74(15), 153414 (2006)

    Karlberg, G.: Adsorption trends for water, hydroxyl, oxygen, and hydrogen on transition-metal and platinum-skin surfaces. Physical Review B—Condensed Matter and Materials Physics 74(15), 153414 (2006)

    2006

  25. [33]

    ChemCatChem 9(7), 1261– 1268 (2017)

    Briquet, L.G., Sarwar, M., Mugo, J., Jones, G., Calle-Vallejo, F.: A new type of scaling relations to assess the accuracy of computational predictions of catalytic activities applied to the oxygen evolution reaction. ChemCatChem 9(7), 1261– 1268 (2017)

    2017

  26. [34]

    Current Opinion in Electrochemistry, 101382 (2023)

    Binninger, T., Kowalski, P.M., Eikerling, M.H.: Oxygen desorption–critical step for the oxygen evolution reaction. Current Opinion in Electrochemistry, 101382 (2023)

    2023

  27. [35]

    ACS Applied Energy Materials 6(23), 11963– 11972 (2023)

    Zhou, Z., Kan, E., Deng, K., Ogitsu, T., Pham, T.A., Zhan, C.: Synergic effects of surface chemistry and applied potentials on the kinetics of the electrocatalytic 37 oxygen evolution reaction in iro2. ACS Applied Energy Materials 6(23), 11963– 11972 (2023)

    2023

  28. [36]

    Energy & Environmental Science 15(6), 2519–2528 (2022)

    Binninger, T., Doublet, M.-L.: The ir–oooo–ir transition state and the mechanism of the oxygen evolution reaction on iro 2 (110). Energy & Environmental Science 15(6), 2519–2528 (2022)

    2022

  29. [37]

    The Journal of Physical Chemistry C 121(21), 11455–11463 (2017)

    Gauthier, J.A., Dickens, C.F., Chen, L.D., Doyle, A.D., Nørskov, J.K.: Solva- tion effects for oxygen evolution reaction catalysis on iro2 (110). The Journal of Physical Chemistry C 121(21), 11455–11463 (2017)

    2017

  30. [38]

    Journal of the American Chemical Society 139(1), 149–155 (2017)

    Ping, Y., Nielsen, R.J., Goddard III, W.A.: The reaction mechanism with free energy barriers at constant potentials for the oxygen evolution reaction at the iro2 (110) surface. Journal of the American Chemical Society 139(1), 149–155 (2017)

    2017

  31. [39]

    Kresse, G., Hafner, J.: Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558–561 (1993) https://doi.org/10.1103/PhysRevB.47.558

    1993 doi

  32. [40]

    Computational Materials Science 6(1), 15–50 (1996) https://doi.org/10.1016/0927-0256(96)00008-0

    Kresse, G., Furthm¨ uller, J.: Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Computational Materials Science 6(1), 15–50 (1996) https://doi.org/10.1016/0927-0256(96)00008-0

    1996 doi

  33. [41]

    Kresse, G., Furthm¨ uller, J.: Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169–11186 (1996) https://doi.org/10.1103/PhysRevB.54.11169

    1996 doi

  34. [42]

    Physical review b 59(3), 1758 (1999)

    Kresse, G., Joubert, D.: From ultrasoft pseudopotentials to the projector 38 augmented-wave method. Physical review b 59(3), 1758 (1999)

    1999

  35. [43]

    The Journal of Chemical Physics 158(16) (2023)

    L´ opez-Zorrilla, J., Aretxabaleta, X.M., Yeu, I.W., Etxebarria, I., Manzano, H., Artrith, N.: ænet-pytorch: a gpu-supported implementation for machine learning atomic potentials training. The Journal of Chemical Physics 158(16) (2023)

    2023

  36. [44]

    Artrith, N., Urban, A., Ceder, G.: Efficient and accurate machine-learning inter- polation of atomic energies in compositions with many species. Phys. Rev. B 96, 014112 (2017) https://doi.org/10.1103/PhysRevB.96.014112

    2017 doi

  37. [45]

    Journal of Physics: Condensed Matter 29(27), 273002 (2017)

    Larsen, A.H., Mortensen, J.J., Blomqvist, J., Castelli, I.E., Christensen, R., Du lak, M., Friis, J., Groves, M.N., Hammer, B., Hargus, C., et al.: The atomic simulation environment—a python library for working with atoms. Journal of Physics: Condensed Matter 29(27), 273002 (2017)

    2017

  38. [46]

    The Journal of Chemical Physics 155(7) (2021)

    Chen, M.S., Morawietz, T., Mori, H., Markland, T.E., Artrith, N.: Aenet–lammps and aenet–tinker: Interfaces for accurate and efficient molecular dynamics simula- tions with machine learning potentials. The Journal of Chemical Physics 155(7) (2021)

    2021

  39. [47]

    The Journal of chemical physics 81(1), 511–519 (1984)

    Nos´ e, S.: A unified formulation of the constant temperature molecular dynamics methods. The Journal of chemical physics 81(1), 511–519 (1984)

    1984

  40. [48]

    Physical review letters 100(2), 020603 (2008) 39

    Barducci, A., Bussi, G., Parrinello, M.: Well-tempered metadynamics: a smoothly converging and tunable free-energy method. Physical review letters 100(2), 020603 (2008) 39

    2008

  41. [49]

    Computer physics communications 185(2), 604–613 (2014) 40

    Tribello, G.A., Bonomi, M., Branduardi, D., Camilloni, C., Bussi, G.: Plumed 2: New feathers for an old bird. Computer physics communications 185(2), 604–613 (2014) 40

    2014

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.