pith. sign in

arxiv: 2503.15389 · v1 · submitted 2025-03-19 · ❄️ cond-mat.mtrl-sci · physics.chem-ph· physics.comp-ph

Revealing the proton slingshot mechanism in solid acid electrolytes through machine learning molecular dynamics

Menghang Wang , Jingxuan Ding , Grace Xiong , Ni Zhan , Cameron J. Owen , Albert Musaelian , Yu Xie , Nicola Molinari
show 3 more authors
Ryan P. Adams Sossina Haile Boris Kozinsky
This is my paper

Pith reviewed 2026-05-22 23:34 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci physics.chem-phphysics.comp-ph
keywords proton conductionsolid acid electrolytesmolecular dynamicsmachine learning force fieldspolyanion rotationCsH2PO4CsHSO4slingshot mechanism
0
0 comments X

The pith

A proton slingshot mechanism, with polyanion rotation followed by O-H reorientation, governs fast conduction in CsH2PO4 and CsHSO4.

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

Machine learning molecular dynamics simulations run over nanosecond timescales show proton transport in these solid acids proceeds by protons being carried partway by rotating polyanions and then completing long-range jumps after O-H bond reorientation. This process allows substantial independent proton motion assisted only by limited rotations, which differs from the standard revolving paddlewheel picture. The two compounds display different transport behaviors because of their different proton concentrations, with CsH2PO4 showing two rotational rates and an O-sharing correlation not seen in CsHSO4. The work concludes that lowering proton concentration would speed up rotations and raise conductivity.

Core claim

In CsH2PO4 and CsHSO4, fast proton conduction proceeds through a proton slingshot mechanism: protons are initially carried by rotating polyanions, followed by O-H bond reorientation that enables long-range jumps. This challenges the conventional revolving paddlewheel model and reveals significant independent proton motion assisted by limited rotations. Qualitative differences in transport arise from different proton concentrations, with CsH2PO4 exhibiting two distinct rates of rotational motions and a correlation between O-sharing and proton transport, while CsHSO4 shows single-rate behavior. Higher proton concentration in CsH2PO4 produces frustrated polyanion orientations and slower overall

What carries the argument

proton slingshot mechanism in which initial polyanion rotation carries the proton before O-H reorientation completes the long-range jump

If this is right

  • The revolving paddlewheel model is incomplete because independent proton motion assisted by limited rotations accounts for a substantial fraction of transport.
  • CsH2PO4 displays two distinct rotational motion rates with different activation energies while CsHSO4 shows only single-rate behavior.
  • The higher proton concentration in CsH2PO4 produces frustrated PO4 orientations, slower rotations, and an O-sharing correlation with transport that is absent in CsHSO4.
  • Reducing proton concentration would accelerate polyanion rotations and enhance conductivity in these solid-acid compounds.
  • The interplay between proton hopping and disordered polyanion dynamics supplies a unified framework for ionic mobility across solid-acid compounds.

Where Pith is reading between the lines

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

  • The same slingshot process may operate in other solid electrolytes that contain rotating anions, which could be checked by running equivalent nanosecond simulations on those materials.
  • Tuning proton density within an existing anion framework could serve as a design lever for higher conductivity without requiring entirely new chemistries.
  • The role of frustrated orientations suggests that controlled structural disorder might be used to balance rotation assistance and independent hopping in related ionic conductors.

Load-bearing premise

The equivariant neural network force fields reproduce the true potential energy surface and dynamics of the two compounds over nanosecond timescales without creating artificial long-range jumps or rotation rates.

What would settle it

Neutron scattering or NMR measurements on CsH2PO4 that detect only a single rotational activation energy and no correlation between O-sharing and proton jumps would falsify the claimed two-rate behavior and slingshot mechanism.

Figures

Figures reproduced from arXiv: 2503.15389 by Albert Musaelian, Boris Kozinsky, Cameron J. Owen, Grace Xiong, Jingxuan Ding, Menghang Wang, Nicola Molinari, Ni Zhan, Ryan P. Adams, Sossina Haile, Yu Xie.

Figure 1
Figure 1. Figure 1: Superprotonic structures. The superprotonic phase of (a) CsH2PO4 [15] and (b) CsHSO4 [21]. The oxygen sites in CsH2PO4 have a one-sixth average occupancy, where as those in CsHSO4 have a one-half average occupany. Both PO4 and SO4 polyanions are orientationally disordered. The H position in CsH2PO4 is determined according to [20], with details provided in Supporting Information. using the Vienna ab initio … view at source ↗
Figure 2
Figure 2. Figure 2: MLFF workflow, diffusion coefficients, and MSD. (a) [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Single proton trajectory and two types of polyanion rotations. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Two-rate behavior and effect of bound protons (a) [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: O –H bond reorientation (a) Three consecutive ML MD snapshots illustrating significant O – H bond reorientation occurring on the carrier P1, despite minimal rotation of the carrier itself at the moment. (b) Distributions of mean dH/dO ratios (defined in Eq. 4) for both productive and nonproductive rotations in CsH2PO4, along with outlier distributions for each. (c) Distribution of cosine angles between ins… view at source ↗
Figure 6
Figure 6. Figure 6: O-sharing (a) ML MD snapshots illustrating two O-sharing configurations during a productive carrier rotation. The configuration is boxed with shared O anion marked red. At 0 ps, the shared O2 repels the O – H bond, reorienting it away from provider P2. At 0.55 ps, carrier P1 brings the hopping proton (cyan) near O3 via rotation, creating another O-sharing configuration. At 0.75 ps, O3 offloads its previous… view at source ↗
Figure 7
Figure 7. Figure 7: Proton slingshot mechanism in CsHSO4 (a) Residence time distributions from a 4ns ML MD simulation of CsHSO4 at 450 K. Both O-pair and S-pair residence time distributions indicate a single-rate polyanion rotation. (b) Joint probability density of N¯ p(τO−O) and τO−O at 450 K with marginal probability of N¯ p (right) suggests a single peak in the population of bound protons on the neighbor polyanion of CsHSO… view at source ↗
Figure 8
Figure 8. Figure 8: Polyanion orientations in CsHSO4 and CsH2PO4 a Illustration of a normalized X –O vector in an XO4 polyanion and its spherical angles (ϕ, θ). Cumulative probability distribution of the spherical angles of a single X –O vector in one selected polyanion for b CsHSO4 and c CsH2PO4 from their respective 4 ns ML MDs at 550 K. Each red dot indicates the initial spherical angles of the tracked X – O bond. d VCF ca… view at source ↗
read the original abstract

In solid acid solid electrolytes CsH$_2$PO$_4$ and CsHSO$_4$, mechanisms of fast proton conduction have long been debated and attributed to either local proton hopping or polyanion rotation. However, the precise role of polyanion rotation and its interplay with proton hopping remained unclear. Nanosecond-scale molecular dynamics simulations, driven by equivariant neural network force fields, reveal a nuanced proton slingshot mechanism: protons are initially carried by rotating polyanions, followed by O$-$H bond reorientation, and the combined motion enables long-range jumps. This challenges the conventional revolving paddlewheel model and reveals significant independent proton motion that is assisted by limited rotations. Despite structural similarities, we identify qualitative differences in transport mechanisms between CsH$_2$PO$_4$ and CsHSO$_4$, caused by different proton concentrations. CsH$_2$PO$_4$ exhibits two distinct rates of rotational motions with different activation energies, contrasting with CsHSO$_4$'s single-rate behavior. The higher proton concentration in CsH$_2$PO$_4$ correlates with frustrated PO$_4$ polyanion orientations and slower rotations compared to SO$_4$ in CsHSO$_4$. Additionally, we reveal a correlation between O-sharing and proton transport in CsH$_2$PO$_4$, a unique feature due to extra proton per polyanion compared to CsHSO$_4$. Our findings suggest that reducing proton concentration could accelerate rotations and enhance conductivity. This work provides a unified framework for understanding and optimizing ionic mobility in solid-acid compounds, offering new insights into the interplay between proton hopping and disordered dynamics in polyanion rotation.

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 reports nanosecond-scale molecular dynamics simulations of CsH₂PO₄ and CsHSO₄ driven by equivariant neural network force fields. It identifies a 'proton slingshot' sequence (polyanion rotation carrying the proton, followed by O-H reorientation) that enables long-range jumps, challenges the revolving paddlewheel model, notes significant independent proton motion assisted by limited rotations, and highlights qualitative differences between the two compounds arising from proton concentration (two distinct rotational rates in CsH₂PO₄ versus one in CsHSO₄; O-sharing correlation unique to CsH₂PO₄).

Significance. If the reported trajectories faithfully reflect the underlying potential energy surface, the work supplies a concrete, atomistic picture of the coupling between polyanion dynamics and proton transport that could guide electrolyte optimization. The use of equivariant NN potentials to reach nanosecond timescales is a clear technical strength, enabling direct observation of rare long-range jumps that shorter ab initio runs cannot access.

major comments (2)
  1. [Methods] Methods (force-field training and validation paragraph): no composition of the DFT training set, no force/energy MAE on held-out configurations, and no direct comparison of simulated polyanion rotational correlation times or proton jump frequencies to experiment or ab initio benchmarks are supplied. Because the slingshot mechanism and the claimed distinction from the paddlewheel model are extracted entirely from the NN-driven trajectories, the absence of these checks is load-bearing for the central claim.
  2. [Results] Results (trajectory analysis and rate extraction): the reported two-rate rotational behavior in CsH₂PO₄ and the attribution of independent proton motion lack error bars or convergence tests on the jump statistics and correlation functions; without these, the statistical robustness of the qualitative differences versus CsHSO₄ cannot be assessed.
minor comments (2)
  1. [Abstract] Abstract: the phrase 'proton slingshot mechanism' is introduced without a concise operational definition or reference to a supporting figure; a one-sentence definition would improve immediate clarity.
  2. Figure captions (throughout): several panels lack explicit labels for the time windows or temperature ranges used to extract the quoted rates, making it harder to connect visuals to the numerical claims.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their positive evaluation of the work's significance and for highlighting the technical value of the equivariant NN potentials. We address each major comment below and will revise the manuscript to incorporate the requested details and statistical checks.

read point-by-point responses

  1. Referee: [Methods] Methods (force-field training and validation paragraph): no composition of the DFT training set, no force/energy MAE on held-out configurations, and no direct comparison of simulated polyanion rotational correlation times or proton jump frequencies to experiment or ab initio benchmarks are supplied. Because the slingshot mechanism and the claimed distinction from the paddlewheel model are extracted entirely from the NN-driven trajectories, the absence of these checks is load-bearing for the central claim.

    Authors: We agree that these validation details are essential for establishing the reliability of the trajectories. In the revised manuscript we will expand the Methods section to report the exact composition of the DFT training set (number of configurations per compound and temperature), the force and energy MAEs evaluated on held-out test configurations, and direct comparisons of the computed polyanion rotational correlation times against available experimental and prior ab initio data. For proton jump frequencies we will add a brief discussion of consistency with existing literature benchmarks, while noting the scarcity of direct experimental measurements at the relevant timescales. revision: yes

  2. Referee: [Results] Results (trajectory analysis and rate extraction): the reported two-rate rotational behavior in CsH₂PO₄ and the attribution of independent proton motion lack error bars or convergence tests on the jump statistics and correlation functions; without these, the statistical robustness of the qualitative differences versus CsHSO₄ cannot be assessed.

    Authors: We acknowledge that error bars and convergence diagnostics are required to substantiate the reported differences. In the revised version we will recompute all rotational correlation functions and jump-rate statistics from multiple independent trajectories, include standard-error estimates on the extracted rates and activation energies, and add convergence tests (e.g., block-averaging and subsampling analyses) to demonstrate that the two-rate behavior in CsH₂PO₄ and the single-rate behavior in CsHSO₄ are statistically robust. revision: yes

Circularity Check

0 steps flagged

No circularity: mechanism extracted from direct MD trajectories

full rationale

The paper runs nanosecond MD with equivariant NN force fields on CsH2PO4 and CsHSO4, then reports observed sequences (polyanion rotation followed by O-H reorientation enabling jumps) directly from the trajectories. No equations, fitted parameters, or self-citations are invoked to define or force the 'slingshot' description; it is presented as an empirical finding from the simulation output. The central claim therefore does not reduce to its inputs by construction. Absence of reported validation metrics on rotation rates or jump frequencies is a correctness concern, not a circularity issue.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 1 invented entities

The central claim depends on the trained ML force field accurately capturing real dynamics and on post-hoc classification of trajectories as slingshot events rather than variants of known processes.

free parameters (1)
  • equivariant neural network force field parameters
    The force field is trained on quantum data, introducing many fitted parameters that determine all observed forces, rotations, and jumps.
axioms (1)
  • domain assumption The machine-learned force field provides an accurate representation of the atomic interactions in CsH2PO4 and CsHSO4 over nanosecond timescales.
    This assumption is required for the trajectories to be interpreted as revealing the true physical mechanism rather than model artifacts.
invented entities (1)
  • proton slingshot mechanism no independent evidence
    purpose: Interpretive label for the combination of limited polyanion rotation and O-H reorientation observed in trajectories.
    This is a descriptive construct for simulation results with no independent falsifiable evidence supplied outside the trajectories themselves.

pith-pipeline@v0.9.0 · 5889 in / 1659 out tokens · 54832 ms · 2026-05-22T23:34:31.811838+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

66 extracted references · 66 canonical work pages · 1 internal anchor

  1. [1]

    P-pair switches

    to describe atomic environments and uses sparse Gaussian processes (SGP) [25] to construct a Bayesian force field (BFF). The BFF predicts atomic forces and uncertainties on local atomic environments during an active learning MD simulation. When uncertainties exceed a set threshold, FLARE invokes DFT calculations to calculate interatomic forces, energy, an...

    work page

  2. [2]

    proton swings,

    with parameters from T1 data of 1H NMR and conductivity measurements, we determine the correlation time of protonic motion, τNMR, to be 330 ps at 525 K. This timescale falls between our computed slow ( λ− 1 2 ) and productive (λ− 1 3 ) rotations. The productive rotation timescale also matches our observed 500 ps equilibration period for MD to enter the di...

    work page arXiv

  3. [3]

    Baranov, A. I. Crystallography Reports 2003, 48, 1012– 1037

    work page 2003

  4. [4]

    M.; Boysen, D

    Haile, S. M.; Boysen, D. A.; Chisholm, C. R. I.; Merle, R. B. Nature 2001, 410, 910–913

    work page 2001

  5. [5]

    A.; Uda, T.; Chisholm, C

    Boysen, D. A.; Uda, T.; Chisholm, C. R. I.; Haile, S. M. Science 2004, 303, 68–70

    work page 2004

  6. [6]

    B.; Chisholm, C

    Merle, R. B.; Chisholm, C. R. I.; Boysen, D. A.; Haile, S. M. Energy and Fuels 2002, 17, 210–215

    work page 2002

  7. [7]

    S.; Jia, H.; Castro, E

    Gittleman, C. S.; Jia, H.; Castro, E. S. D.; Chisholm, C. R.; Kim, Y. S. Joule 2021, 5, 1660–1677

    work page 2021

  8. [8]

    B.; Wilson, D

    Papandrew, A. B.; Wilson, D. L.; M. Cantillo, N.; Hawks, S.; Atkinson, R. W.; Goenaga, G. A.; Zawodzinski, T. A. Journal of The Electrochemical Society 2014, 161, F679– F685

    work page 2014

  9. [9]

    B.; Paik, H.; Qian, X.; Zecevic, S.; Chisholm, C

    Lim, D.-K.; Plymill, A. B.; Paik, H.; Qian, X.; Zecevic, S.; Chisholm, C. R.; Haile, S. M. Joule 2020, 4, 2338– 2347

    work page 2020

  10. [10]

    A. I. Baranov, N. M. S., L. A. Shuvalov JETP Letter 1982, 36, 459–462

    work page 1982

  11. [11]

    I.; Khiznichenko, V

    Baranov, A. I.; Khiznichenko, V. P.; Sandler, V. A.; Shuvalov, L. A. Ferroelectrics 1988, 81, 183–186

    work page 1988

  12. [12]

    Chemistry of Materials 1996, 8, 610–641

    Kreuer, K.-D. Chemistry of Materials 1996, 8, 610–641

    work page 1996

  13. [13]

    Solid State Ionics 1999, 125, 1–11

    Norby, T. Solid State Ionics 1999, 125, 1–11

    work page 1999

  14. [14]

    Physical Chemistry Chemical Physics 2020, 22, 10738–10752

    Dreßler, C.; Sebastiani, D. Physical Chemistry Chemical Physics 2020, 22, 10738–10752

    work page 2020

  15. [15]

    Physical Chemistry Chemical Physics 2022, 24, 15522–15531

    Jinnouchi, R. Physical Chemistry Chemical Physics 2022, 24, 15522–15531

    work page 2022

  16. [16]

    Solid State Ionics 2008, 179, 2345–2349

    Ishikawa, A.; Maekawa, H.; Yamamura, T.; Kawakita, Y.; Shibata, K.; Kawai, M. Solid State Ionics 2008, 179, 2345–2349

    work page 2008

  17. [17]

    Solid State Ionics 2004, 175, 557–562

    Yamada, K.; Sagara, T.; Yamane, Y.; Ohki, H.; Okuda, T. Solid State Ionics 2004, 175, 557–562

    work page 2004

  18. [18]

    Solid State Ionics 1995, 77, 97–100

    Blinc, R. Solid State Ionics 1995, 77, 97–100

    work page 1995

  19. [19]

    V.; Carlile, C

    Belushkin, A. V.; Carlile, C. J.; Shuvalov, L. A. Journal of Physics: Condensed Matter 1992, 4, 389–398

    work page 1992

  20. [20]

    Solid State Ionics 1989, 35, 143–149

    Badot, J.; Colomban, P. Solid State Ionics 1989, 35, 143–149

    work page 1989

  21. [21]

    C.; Marzari, N

    Wood, B. C.; Marzari, N. Physical Review B 2007, 76, 134301

    work page 2007

  22. [22]

    Lee, H.-S.; Tuckerman, M. E. The Journal of Physical Chemistry C 2008, 112, 9917–9930

    work page 2008

  23. [23]

    M.; Beskrovnyi, A.; Gordelii, V.; Datt, I.; Shuvalov, L

    Jirak, Z.; Dlouha, M.; Vratislav, S.; Balagurov, A. M.; Beskrovnyi, A.; Gordelii, V.; Datt, I.; Shuvalov, L. Physica Status Solidi A, Applied Research 1987, 100, 117–122

    work page 1987

  24. [24]

    S.; Patel, S

    Wang, L. S.; Patel, S. V.; Sanghvi, S. S.; Hu, Y.-Y.; Haile, S. M. Journal of the American Chemical Society 2020, 142, 19992–20001

    work page 2020

  25. [25]

    Dreßler, C.; H¨ anseroth, J.; Sebastiani, D.The Journal of Physical Chemistry Letters 2023, 14, 7249–7255

    work page 2023

  26. [26]

    Physical Review Letters 2007, 98, 146401

    Behler, J.; Parrinello, M. Physical Review Letters 2007, 98, 146401

    work page 2007

  27. [27]

    S.; Owen, C

    Vandermause, J.; Xie, Y.; Lim, J. S.; Owen, C. J.; Kozinsky, B. Nature Communications 2022, 13

    work page 2022

  28. [28]

    H.; Johansson, A.; Kozinsky, B.npj Computational Materials 2023, 9

    Xie, Y.; Vandermause, J.; Ramakers, S.; Protik, N. H.; Johansson, A.; Kozinsky, B.npj Computational Materials 2023, 9

    work page 2023

  29. [29]

    P.; Kornbluth, M.; Molinari, N.; Smidt, T

    Batzner, S.; Musaelian, A.; Sun, L.; Geiger, M.; Mailoa, J. P.; Kornbluth, M.; Molinari, N.; Smidt, T. E.; Kozinsky, B. Nature Communications 2022, 13

    work page 2022

  30. [30]

    J.; Kornbluth, M.; Kozinsky, B

    Musaelian, A.; Batzner, S.; Johansson, A.; Sun, L.; Owen, C. J.; Kornbluth, M.; Kozinsky, B. Nature Communications 2023, 14

    work page 2023

  31. [31]

    Journal of Materials Chemistry A 2023, 11, 16104–16114

    Minami, S.; Jinnouchi, R. Journal of Materials Chemistry A 2023, 11, 16104–16114

    work page 2023

  32. [32]

    B.; Batzner, S.; Xie, Y.; Sun, L.; Kolpak, A

    Vandermause, J.; Torrisi, S. B.; Batzner, S.; Xie, Y.; Sun, L.; Kolpak, A. M.; Kozinsky, B. npj Computational Materials 2020, 6

    work page 2020

  33. [33]

    M.; Chisholm, C

    Haile, S. M.; Chisholm, C. R. I.; Sasaki, K.; Boysen, D. A.; Uda, T. Faraday Discuss. 2007, 134, 17–39

    work page 2007

  34. [34]

    Shigenobu Hayashi, M. M. Solid State Ionics 2004, 171, 289–293

    work page 2004

  35. [35]

    P.; Burke, K.; Ernzerhof, M

    Perdew, J. P.; Burke, K.; Ernzerhof, M. Physical Review Letters 1996, 77, 3865–3868

    work page 1996

  36. [36]

    V.; van Setten, M

    Troeye, B. V.; van Setten, M. J.; Giantomassi, M.; Torrent, M.; Rignanese, G.-M.; Gonze, X. Physical Review B 2017, 95, 024112

    work page 2017

  37. [37]

    Physical Chemistry Chemical Physics 2015, 17, 31287– 31296

    Krzystyniak, M.; Dru˙ zbicki, K.; Fernandez-Alonso, F. Physical Chemistry Chemical Physics 2015, 17, 31287– 31296

    work page 2015

  38. [38]

    F.; Cavaye, H.; Callear, S

    Parker, S. F.; Cavaye, H.; Callear, S. K. Molecules 2020, 25, 1271

    work page 2020

  39. [39]

    Electronic Structure 2021, 3, 027001

    Saßnick, H.-D.; Cocchi, C. Electronic Structure 2021, 3, 027001

    work page 2021

  40. [40]

    Computational Materials Science 1996, 6, 15–50

    Kresse, G.; Furthm¨ uller, J. Computational Materials Science 1996, 6, 15–50

    work page 1996

  41. [41]

    Kresse, G.; Furthm¨ uller, J.Physical Review B 1996, 54, 11169–11186

    work page 1996

  42. [42]

    Physical Review B 1999, 59, 1758–1775

    Kresse, G.; Joubert, D. Physical Review B 1999, 59, 1758–1775

    work page 1999

  43. [43]

    Physical Review B 2019, 99, 014104

    Drautz, R. Physical Review B 2019, 99, 014104

    work page 2019

  44. [44]

    Transactions on Machine Learning Research, 2023 2022

    Fu, X.; Wu, Z.; Wang, W.; Xie, T.; Keten, S.; Gomez- Bombarelli, R.; Jaakkola, T. Transactions on Machine Learning Research, 2023 2022

    work page 2023

  45. [45]

    P.; Ba, J

    Kingma, D. P.; Ba, J. ArXiv e-prints 2014

    work page 2014

  46. [46]

    ArXiv e-prints 2019

    Paszke, A.; et al. ArXiv e-prints 2019

    work page 2019

  47. [47]

    P.; Aktulga, H

    Thompson, A. P.; Aktulga, H. M.; Berger, R.; Bolintineanu, D. S.; Brown, W. M.; Crozier, P. S.; in ’t Veld, P. J.; Kohlmeyer, A.; Moore, S. G.; Nguyen, T. D.; Shan, R.; Stevens, M. J.; Tranchida, J.; Trott, C.; Plimpton, S. J. Computer Physics Communications 2022, 271, 108171

    work page 2022

  48. [48]

    Solid State Ionics 2017, 306, 13–19

    Aili, D.; Gao, Y.; Han, J.; Li, Q. Solid State Ionics 2017, 306, 13–19

    work page 2017

  49. [49]

    M.; Blanc, F.; Haile, S

    Kim, G.; Griffin, J. M.; Blanc, F.; Haile, S. M.; Grey, C. P. Journal of the American Chemical Society 2015, 137, 3867–3876

    work page 2015

  50. [50]

    M¨ unch, W.; Kreuer, K.; Traub, U.; Maier, J.Solid State Ionics 1995, 77, 10–14

    work page 1995

  51. [51]

    Arcon, D.; Blinc, R.; Dolinsek, J.; Shuvalov, L. A. Physical Review B 1997, 55, 8961–8966

    work page 1997

  52. [52]

    Revealing the proton slingshot mechanism in solid acid electrolytes through machine learning molecular dynamics

    Stukowski, A. Modelling and Simulation in Materials Science and Engineering 2009, 18, 015012. VIII. ACKNOWLEDGEMENTS The authors gratefully acknowledge Kyle Bystrom for insightful discussions on DFT calculations, Anders Johansson for suggesting the residence time approach 15 and guidance on using LAMMPS, and Gordon Peiker for providing feedback on the man...

    work page internal anchor Pith review Pith/arXiv arXiv 2009

  53. [53]

    E.The Journal of Physical Chemistry C2008, 112, 9917– 9930

    Lee, H.-S.; Tuckerman, M. E.The Journal of Physical Chemistry C2008, 112, 9917– 9930

    work page

  54. [54]

    E.; Hermosillo, J

    Botez, C. E.; Hermosillo, J. D.; Zhang, J.; Qian, J.; Zhao, Y.; Majzlan, J.; Chianelli, R. R.; Pantea, C.The Journal of Chemical Physics2007, 127

    work page

  55. [55]

    Yamada, K.; Sagara, T.; Yamane, Y.; Ohki, H.; Okuda, T.Solid State Ionics 2004, 175, 557–562

    work page 2004

  56. [56]

    InEuropean Powder Diffraction 3; Trans Tech Publications Ltd; Materials Science Forum, Vol

    Preisinger, A.; Mereiter, K.; Bronowska, W. InEuropean Powder Diffraction 3; Trans Tech Publications Ltd; Materials Science Forum, Vol. 166; pp 511–516

    work page

  57. [57]

    M.; Beskrovnyi, A.; Gordelii, V.; Datt, I.; Shuvalov, L.Physica Status Solidi A, Applied Research1987, 100, 117–122

    Jirak, Z.; Dlouha, M.; Vratislav, S.; Balagurov, A. M.; Beskrovnyi, A.; Gordelii, V.; Datt, I.; Shuvalov, L.Physica Status Solidi A, Applied Research1987, 100, 117–122

    work page

  58. [58]

    R.; Haile, S

    Chisholm, C. R.; Haile, S. M.Materials Research Bulletin2000, 35, 999–1005

    work page

  59. [59]

    Itoh, K.; Ukeda, T.; Ozaki, T.; Nakamura, E.Acta Crystallographica Section C Crystal Structure Communications1990, 46, 358–361

    work page

  60. [60]

    P.; Richards, W

    Ong, S. P.; Richards, W. D.; Jain, A.; Hautier, G.; Kocher, M.; Cholia, S.; Gunter, D.; Chevrier, V. L.; Persson, K. A.; Ceder, G.Computational Materials Science2013, 68, 314–319

    work page

  61. [61]

    Drautz, R.Physical Review B2019, 99, 014104

    work page

  62. [62]

    B.; Batzner, S.; Xie, Y.; Sun, L.; Kolpak, A

    Vandermause, J.; Torrisi, S. B.; Batzner, S.; Xie, Y.; Sun, L.; Kolpak, A. M.; Kozinsky, B. npj Computational Materials2020, 6

    work page

  63. [63]

    S.; Owen, C

    Vandermause, J.; Xie, Y.; Lim, J. S.; Owen, C. J.; Kozinsky, B.Nature Communications 2022, 13

    work page 2022

  64. [64]

    H.; Johansson, A.; Kozinsky, B.npj Computational Materials2023, 9

    Xie, Y.; Vandermause, J.; Ramakers, S.; Protik, N. H.; Johansson, A.; Kozinsky, B.npj Computational Materials2023, 9

    work page

  65. [65]

    He, X.; Zhu, Y.; Epstein, A.; Mo, Y.npj Computational Materials2018, 4

    work page

  66. [66]

    Virtanen, P.; et al.Nature Methods2020, 17, 261–272. 14

    work page