Thermodynamic surface reconstruction governs catalytic behavior in high-entropy alloys

Geun Ho Gu; Sathya Sheela Subramanian; Taegyeong Kim; Youngtak Kim

arxiv: 2604.25454 · v1 · submitted 2026-04-28 · ❄️ cond-mat.mtrl-sci

Thermodynamic surface reconstruction governs catalytic behavior in high-entropy alloys

Taegyeong Kim , Youngtak Kim , Sathya Sheela Subramanian , Geun Ho Gu This is my paper

Pith reviewed 2026-05-07 15:52 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci

keywords high-entropy alloyssurface reconstructioncatalytic activitythermodynamic simulationsegregation energeticsadsorption energymulticomponent alloys

0 comments

The pith

Surface ordering from thermodynamic reconstruction is required to predict catalytic activity in high-entropy alloys

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

High-entropy alloys are typically modeled with random surface mixing for catalysis predictions, but this paper shows that assumption does not hold up against experiments. Thermodynamic simulations that allow surface element segregation and ordering successfully reproduce the observed compositional activity trends, while random models do not and sometimes perform at or below random guessing. If true, this means that predictive design of catalysts from these alloys must account for the actual surface state selected by thermodynamics rather than assuming uniformity. The work identifies a validity boundary for simple models based on the degree of short-range order.

Core claim

The paper establishes that the thermodynamically selected surface state, achieved through annealing and segregation, collapses the adsorption-energy spectrum of random alloys into narrower favorable distributions, enabling accurate recovery of active compositions where homogeneous models fail or perform at random baseline. Segregation energetics reveal strong surface enrichment of preferred elements, producing chemically selective interfaces. By linking predictive error to short-range order, the work positions the reconstructed surface as the governing parameter for catalysis in multicomponent alloys.

What carries the argument

Thermodynamically annealed surfaces that incorporate segregation energetics to form chemically selective interfaces

Load-bearing premise

The thermodynamic simulations of surface ordering accurately reproduce experimental catalytic measurements without post-hoc adjustments or fitting of segregation energies to the activity data being predicted.

What would settle it

A high-entropy alloy system where a homogeneous random-mixing surface model predicts experimental activity trends more accurately than a thermodynamically annealed surface model would falsify the necessity of surface reconstruction.

read the original abstract

High-entropy alloys are widely modeled as homogeneously mixed surfaces, yet the validity of this assumption for catalytic prediction remains unclear. Here, we reproduce high-throughput experimental measurements using thermodynamic simulations and show that surface ordering is essential for accurately capturing the compositional activity landscape. Homogeneous surface models fail to reproduce experimentally observed trends and, in some regimes, perform at or below the random-selection baseline. In contrast, thermodynamically annealed surfaces restore meaningful agreement with the experimental activity landscape and substantially improve the recovery of active compositions. Segregation energetics reveal strong surface enrichment of preferred elements, producing chemically selective interfaces that collapse the broad adsorption-energy spectrum of random alloys into a narrower distribution of catalytically favorable sites. By linking predictive error to the degree of short-range order, we identify a validity boundary for homogeneous models and establish the thermodynamically selected surface state as a governing parameter for predictive catalysis in multicomponent alloys.

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.

Desk Editor's Note private letter to a colleague

Surface ordering via thermodynamic annealing matches experimental catalysis trends in HEAs better than random homogeneous models, but the independence of segregation energetics from the activity data is still unverified.

read the letter

The main thing to know is that this paper argues homogeneous random-mixing assumptions on surfaces fail to capture catalytic activity trends in high-entropy alloys, while letting the surface reach thermodynamic equilibrium through annealing restores agreement with high-throughput experiments by driving element enrichment and narrowing the distribution of active sites. They link the size of the predictive error directly to the degree of short-range order and claim this sets a validity boundary for simple models. That framing is the clearest new angle here, and it does a decent job of highlighting a practical modeling shortcut that many screening studies still rely on. The work is relevant because HEAs are being explored for real catalytic applications, so showing that surface state matters for compositional predictions is worth attention. The soft spots are exactly where the stress test points: the abstract asserts reproduction of experimental trends but supplies no quantitative metrics, error bars, or description of the fit quality, so the strength of the improvement is hard to judge. More importantly, it is not clear whether the segregation energetics driving the annealing come from independent DFT or literature values, or whether they were adjusted against the same activity landscape being predicted. If the latter, the apparent superiority over homogeneous baselines becomes circular. The paper needs to demonstrate parameter independence explicitly in the methods. This is for readers working on computational or high-throughput catalysis in multicomponent alloys who want to move past bulk-random assumptions. It deserves peer review because the question is practically important and the approach is straightforward to test, even if the current version will need revisions for transparency on data sources and quantitative validation.

Referee Report

3 major / 2 minor

Summary. The manuscript claims that high-entropy alloy surfaces are not well-described by homogeneous mixing assumptions for catalysis. Thermodynamic annealing simulations that incorporate surface segregation and short-range ordering reproduce high-throughput experimental activity trends, whereas homogeneous and random-surface models fail or perform at baseline levels. Segregation produces chemically selective interfaces that narrow the adsorption-energy distribution, and the degree of short-range order is linked to predictive error, establishing the thermodynamically selected surface state as a governing parameter.

Significance. If the reproduction holds with independent parameters, the result would shift modeling practice in multicomponent alloy catalysis from bulk-composition or mean-field approximations toward explicit surface thermodynamics. It supplies a concrete validity boundary for homogeneous models and a mechanism (enrichment-driven site selection) that explains why random-alloy spectra overpredict inactive compositions.

major comments (3)

Abstract and Results: the central claim that 'thermodynamically annealed surfaces restore meaningful agreement' with experiment is unsupported by any quantitative metric (R², mean absolute error, fraction of active compositions recovered, or statistical test) or description of the reproduction protocol. Without these, the asserted superiority over homogeneous baselines cannot be evaluated.
Segregation energetics (Methods/Results): the independence of the segregation energies from the high-throughput activity dataset is not demonstrated. If these energies were derived or adjusted using the same catalytic measurements being predicted, the reported improvement of ordered surfaces over random baselines is circular and does not constitute an independent test of the thermodynamic-reconstruction hypothesis.
§ on short-range order and validity boundary: the quantitative link between predictive error and degree of short-range order is asserted but not shown with a specific correlation, threshold value, or cross-validation across compositions; this relation is load-bearing for the claim that surface ordering is 'essential' rather than merely correlated.

minor comments (2)

Ensure all figures reporting activity landscapes include the corresponding experimental data points with error bars or replicate counts for direct visual comparison.
Define 'random-selection baseline' and 'homogeneous surface model' explicitly in the text and legends on first use.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments, which identify opportunities to strengthen the quantitative support for our claims. We address each major comment below and will revise the manuscript to incorporate the requested details.

read point-by-point responses

Referee: Abstract and Results: the central claim that 'thermodynamically annealed surfaces restore meaningful agreement' with experiment is unsupported by any quantitative metric (R², mean absolute error, fraction of active compositions recovered, or statistical test) or description of the reproduction protocol. Without these, the asserted superiority over homogeneous baselines cannot be evaluated.

Authors: We agree that explicit quantitative metrics and a clear protocol description were omitted from the initial submission. In the revised manuscript we will report R², mean absolute error, the fraction of active compositions recovered, and appropriate statistical comparisons between the annealed-surface predictions and both the homogeneous and random baselines. We will also add a dedicated paragraph in the Results section describing the exact reproduction protocol, including how experimental activity values were normalized and compared to the simulated landscapes. revision: yes
Referee: Segregation energetics (Methods/Results): the independence of the segregation energies from the high-throughput activity dataset is not demonstrated. If these energies were derived or adjusted using the same catalytic measurements being predicted, the reported improvement of ordered surfaces over random baselines is circular and does not constitute an independent test of the thermodynamic-reconstruction hypothesis.

Authors: The segregation energies were obtained from separate, first-principles DFT calculations performed on surface slab models prior to any comparison with the experimental catalytic dataset; no fitting or adjustment to the activity measurements was performed. We will revise the Methods section to state this independence explicitly, provide the precise computational parameters used for the segregation-energy calculations, and confirm that these values were fixed before the activity-prediction comparisons were made. revision: yes
Referee: § on short-range order and validity boundary: the quantitative link between predictive error and degree of short-range order is asserted but not shown with a specific correlation, threshold value, or cross-validation across compositions; this relation is load-bearing for the claim that surface ordering is 'essential' rather than merely correlated.

Authors: We acknowledge that the original manuscript presented the link between predictive error and short-range order only qualitatively. In the revision we will add a quantitative analysis that includes the Pearson correlation coefficient between error and short-range-order parameter, explicit threshold values at which homogeneous models lose predictive power, and cross-validation results obtained by holding out subsets of compositions. These additions will be placed in a new subsection or supplementary figure to substantiate the validity-boundary claim. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation chain self-contained against external benchmarks

full rationale

The abstract and provided text describe thermodynamic simulations reproducing experimental activity trends via surface ordering and segregation energetics, with homogeneous models failing. No equations, self-citations, or parameter-fitting steps are quoted that reduce predictions to the same activity dataset by construction. Segregation energetics are presented as revealing enrichment independently, and the validity boundary is linked to short-range order without evidence of post-hoc tuning to the target data. This meets the criteria for an independent, externally falsifiable claim; the reader's concern about unverified independence cannot be elevated to circularity without specific quotes showing fitted inputs renamed as predictions.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Only the abstract is available, so the ledger is necessarily incomplete and based on stated claims rather than explicit equations or methods.

free parameters (1)

segregation energetics
The abstract invokes segregation energetics to explain surface enrichment but does not state whether these values are taken from prior literature, computed from first principles, or fitted to the catalytic data.

axioms (1)

domain assumption Thermodynamic simulations can faithfully reproduce experimental surface states and catalytic activity in high-entropy alloys
The central claim rests on the assertion that annealed surfaces match experiment while homogeneous models do not.

pith-pipeline@v0.9.0 · 5462 in / 1317 out tokens · 56553 ms · 2026-05-07T15:52:55.430864+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

9 extracted references · 9 canonical work pages

[1]

Discovery of a Multinary Noble Metal –Free 18 Oxygen Reduction Catalyst

(1) Löffler, T.; Meyer, H.; Savan, A.; Wilde, P.; Garzón Manjón, A.; Chen, Y.-T.; Ventosa, E.; Scheu, C.; Ludwig, A.; Schuhmann, W. Discovery of a Multinary Noble Metal –Free 18 Oxygen Reduction Catalyst. Adv. Energy Mater. 2018, 8 (34), 1802269. https://doi.org/https://doi.org/10.1002/aenm.201802269. (2) Shaikh, J. S.; Rittiruam, M.; Saelee, T.; Márquez,...

work page doi:10.1002/aenm.201802269 2018
[2]

(3) Manjón, A

https://doi.org/10.1016/j.mser.2024.100813. (3) Manjón, A. G.; Löffler, T.; Meischein, M.; Meyer, H.; Lim, J.; Strotkötter, V.; Schuhmann, W.; Ludwig, A.; Scheu, C. Sputter Deposition of Highly Active Complex Solid Solution Electrocatalysts into an Ionic Liquid Library: Effect of Structure and Composit ion on Oxygen Reduction Activity. Nanoscale 2020, 12 ...

work page doi:10.1016/j.mser.2024.100813 2024
[3]

(12) Hutu, A

https://doi.org/10.1021/jacs.5c15697. (12) Hutu, A. I.; Pervolarakis, E.; Remediakis, I. N.; Kristoffersen, H. H.; Rossmeisl, J. Scaling Relations on High -Entropy Alloy Catalyst Surfaces. Journal of Physical Chemistry C . American Chemical Society June 27, 2024, pp 10251 –10258. https://doi.org/10.1021/acs.jpcc.4c01292. (13) Medford, A. J.; Vojvodic , A....

work page doi:10.1021/jacs.5c15697 2024
[4]

(21) Tetteh, E

https://doi.org/10.1039/d5fd00082c. (21) Tetteh, E. B.; Krysiak, O. A.; Savan, A.; Kim, M.; Zerdoumi, R.; Chung, T. D.; Ludwig, A.; Schuhmann, W. Long -Range SECCM Enables High -Throughput Electrochemical Screening of High Entropy Alloy Electrocatalysts at Up -To-Industrial Current Densities. Small Methods 2024, 8 (7). https://doi.org/10.1002/smtd.2023012...

work page doi:10.1039/d5fd00082c 2024
[5]

(24) Chen, L.; Yang, J.; Zhang, Z.; Wu, F.; Yao, Y.; Zhang, A.; Duan, X.; Shan, B

https://doi.org/10.1016/j.apsusc.2024.159297. (24) Chen, L.; Yang, J.; Zhang, Z.; Wu, F.; Yao, Y.; Zhang, A.; Duan, X.; Shan, B. High Throughput Screening of Noble Metal -Free High-Entropy Alloys Catalysts for Selective Catalytic Oxidation of NH3. Appl. Surf. Sci. 2023,

work page doi:10.1016/j.apsusc.2024.159297 2024
[6]

(25) Pan, Y.; Shan, X.; Cai, F.; Gao, H.; Xu, J.; Zhou, M

https://doi.org/10.1016/j.apsusc.2023.157354. (25) Pan, Y.; Shan, X.; Cai, F.; Gao, H.; Xu, J.; Zhou, M. Accelerating the Discovery of Oxygen Reduction Electrocatalysts: High -Throughput Screening of Element Combinations in Pt - Based High -Entropy Alloys. Angew. Chem. Int. Ed. 2024, 63 (37). https://doi.org/10.1002/anie.202407116. (26) Polak, M.; Rubinov...

work page doi:10.1016/j.apsusc.2023.157354 2023
[7]

(31) Moniri, S.; Yang, Y.; Ding, J.; Yuan, Y.; Zhou, J.; Yang, L.; Zhu, F.; Liao, Y.; Yao, Y.; Hu, L.; Ercius, P.; Miao, J

https://doi.org/10.1016/j.commatsci.2022.111762. (31) Moniri, S.; Yang, Y.; Ding, J.; Yuan, Y.; Zhou, J.; Yang, L.; Zhu, F.; Liao, Y.; Yao, Y.; Hu, L.; Ercius, P.; Miao, J. Three-Dimensional Atomic Structure and Local Chemical Order of Medium- and High -Entropy Nanoalloys. Nature 2023, 624 (7992), 564 –569. https://doi.org/10.1038/s41586-023-06785-z. 22 (...

work page doi:10.1016/j.commatsci.2022.111762 2022
[8]

(33) Zhang, R.; Zhao, S.; Ding, J.; Chong, Y.; Jia, T.; Ophus, C.; Asta, M.; Ritchie, R

https://doi.org/10.1016/j.actamat.2021.117490. (33) Zhang, R.; Zhao, S.; Ding, J.; Chong, Y.; Jia, T.; Ophus, C.; Asta, M.; Ritchie, R. O.; Minor, A. M. Short -Range Order and Its Impact on the CrCoNi Medium -Entropy Alloy. Nature 2020, 581 (7808), 283–287. https://doi.org/10.1038/s41586-020-2275-z. (34) Ye, Q.; Cheng, Y.; Cao, L. Surface Segregation-Driv...

work page doi:10.1016/j.actamat.2021.117490 2021
[9]

(35) Xie, T.; Grossman, J

https://doi.org/10.1021/acscatal.5c06492. (35) Xie, T.; Grossman, J. C. Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties. Phys. Rev. Lett. 2018, 120 (14). https://doi.org/10.1103/PhysRevLett.120.145301. (36) Kresse, G.; Furthmüller, J. Efficient Iterative Schemes for Ab Initio Total-Energy Cal...

work page doi:10.1021/acscatal.5c06492 2018

[1] [1]

Discovery of a Multinary Noble Metal –Free 18 Oxygen Reduction Catalyst

(1) Löffler, T.; Meyer, H.; Savan, A.; Wilde, P.; Garzón Manjón, A.; Chen, Y.-T.; Ventosa, E.; Scheu, C.; Ludwig, A.; Schuhmann, W. Discovery of a Multinary Noble Metal –Free 18 Oxygen Reduction Catalyst. Adv. Energy Mater. 2018, 8 (34), 1802269. https://doi.org/https://doi.org/10.1002/aenm.201802269. (2) Shaikh, J. S.; Rittiruam, M.; Saelee, T.; Márquez,...

work page doi:10.1002/aenm.201802269 2018

[2] [2]

(3) Manjón, A

https://doi.org/10.1016/j.mser.2024.100813. (3) Manjón, A. G.; Löffler, T.; Meischein, M.; Meyer, H.; Lim, J.; Strotkötter, V.; Schuhmann, W.; Ludwig, A.; Scheu, C. Sputter Deposition of Highly Active Complex Solid Solution Electrocatalysts into an Ionic Liquid Library: Effect of Structure and Composit ion on Oxygen Reduction Activity. Nanoscale 2020, 12 ...

work page doi:10.1016/j.mser.2024.100813 2024

[3] [3]

(12) Hutu, A

https://doi.org/10.1021/jacs.5c15697. (12) Hutu, A. I.; Pervolarakis, E.; Remediakis, I. N.; Kristoffersen, H. H.; Rossmeisl, J. Scaling Relations on High -Entropy Alloy Catalyst Surfaces. Journal of Physical Chemistry C . American Chemical Society June 27, 2024, pp 10251 –10258. https://doi.org/10.1021/acs.jpcc.4c01292. (13) Medford, A. J.; Vojvodic , A....

work page doi:10.1021/jacs.5c15697 2024

[4] [4]

(21) Tetteh, E

https://doi.org/10.1039/d5fd00082c. (21) Tetteh, E. B.; Krysiak, O. A.; Savan, A.; Kim, M.; Zerdoumi, R.; Chung, T. D.; Ludwig, A.; Schuhmann, W. Long -Range SECCM Enables High -Throughput Electrochemical Screening of High Entropy Alloy Electrocatalysts at Up -To-Industrial Current Densities. Small Methods 2024, 8 (7). https://doi.org/10.1002/smtd.2023012...

work page doi:10.1039/d5fd00082c 2024

[5] [5]

(24) Chen, L.; Yang, J.; Zhang, Z.; Wu, F.; Yao, Y.; Zhang, A.; Duan, X.; Shan, B

https://doi.org/10.1016/j.apsusc.2024.159297. (24) Chen, L.; Yang, J.; Zhang, Z.; Wu, F.; Yao, Y.; Zhang, A.; Duan, X.; Shan, B. High Throughput Screening of Noble Metal -Free High-Entropy Alloys Catalysts for Selective Catalytic Oxidation of NH3. Appl. Surf. Sci. 2023,

work page doi:10.1016/j.apsusc.2024.159297 2024

[6] [6]

(25) Pan, Y.; Shan, X.; Cai, F.; Gao, H.; Xu, J.; Zhou, M

https://doi.org/10.1016/j.apsusc.2023.157354. (25) Pan, Y.; Shan, X.; Cai, F.; Gao, H.; Xu, J.; Zhou, M. Accelerating the Discovery of Oxygen Reduction Electrocatalysts: High -Throughput Screening of Element Combinations in Pt - Based High -Entropy Alloys. Angew. Chem. Int. Ed. 2024, 63 (37). https://doi.org/10.1002/anie.202407116. (26) Polak, M.; Rubinov...

work page doi:10.1016/j.apsusc.2023.157354 2023

[7] [7]

(31) Moniri, S.; Yang, Y.; Ding, J.; Yuan, Y.; Zhou, J.; Yang, L.; Zhu, F.; Liao, Y.; Yao, Y.; Hu, L.; Ercius, P.; Miao, J

https://doi.org/10.1016/j.commatsci.2022.111762. (31) Moniri, S.; Yang, Y.; Ding, J.; Yuan, Y.; Zhou, J.; Yang, L.; Zhu, F.; Liao, Y.; Yao, Y.; Hu, L.; Ercius, P.; Miao, J. Three-Dimensional Atomic Structure and Local Chemical Order of Medium- and High -Entropy Nanoalloys. Nature 2023, 624 (7992), 564 –569. https://doi.org/10.1038/s41586-023-06785-z. 22 (...

work page doi:10.1016/j.commatsci.2022.111762 2022

[8] [8]

(33) Zhang, R.; Zhao, S.; Ding, J.; Chong, Y.; Jia, T.; Ophus, C.; Asta, M.; Ritchie, R

https://doi.org/10.1016/j.actamat.2021.117490. (33) Zhang, R.; Zhao, S.; Ding, J.; Chong, Y.; Jia, T.; Ophus, C.; Asta, M.; Ritchie, R. O.; Minor, A. M. Short -Range Order and Its Impact on the CrCoNi Medium -Entropy Alloy. Nature 2020, 581 (7808), 283–287. https://doi.org/10.1038/s41586-020-2275-z. (34) Ye, Q.; Cheng, Y.; Cao, L. Surface Segregation-Driv...

work page doi:10.1016/j.actamat.2021.117490 2021

[9] [9]

(35) Xie, T.; Grossman, J

https://doi.org/10.1021/acscatal.5c06492. (35) Xie, T.; Grossman, J. C. Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties. Phys. Rev. Lett. 2018, 120 (14). https://doi.org/10.1103/PhysRevLett.120.145301. (36) Kresse, G.; Furthmüller, J. Efficient Iterative Schemes for Ab Initio Total-Energy Cal...

work page doi:10.1021/acscatal.5c06492 2018