Pith. sign in

REVIEW 4 major objections 5 minor 59 references

Emergent rigidity percolation of five-fold aggregates enables controllable glass properties

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read The glass transition in metallic liquids is a nonequilibrium rigidity percolation of five-fold atomic clusters.

desk verdict A striking correlation between five-fold cluster percolation and Tg in Cu50Zr50, but the causal rigidity-percolation claim is not yet supported by the data. read the letter →

arxiv 2506.02588 v1 pith:TB5MDXB2 submitted 2025-06-03 cond-mat.soft cond-mat.dis-nncond-mat.mtrl-scicond-mat.stat-mechphysics.chem-ph

classification cond-mat.softcond-mat.dis-nncond-mat.mtrl-scicond-mat.stat-mechphysics.chem-ph
keywords glasstransitionrigiditypercolationfive-foldsymmetryicosahedralclustersmetallicglassesshearmodulusnonequilibriumdiverginglengthscale
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 the glass transition in supercooled metallic liquids is an underlying continuous nonequilibrium transition, not a mere dynamical crossover: stable five-fold (icosahedral-like) clusters aggregate until a system-spanning rigid network forms, and that percolation event is the glass transition. Molecular-dynamics simulations of Cu50Zr50 are used to show that the fraction of atoms in the largest five-fold cluster, the correlation length, and the storage modulus $G'$ all change sharply at the same temperature, where the average coordination number approaches the isostatic value $Z = 6$. If the scenario is correct, it provides the long-sought diverging length scale at the glass transition and explains why a finite shear modulus appears, while also making glass stiffness and toughness tunable through composition and cooling protocol.

What carries the argument

The load-bearing theoretical object is the reversible-polymerization master equation for cluster concentrations, in which every pair of clusters can merge at a size-independent rate but breakup only detaches single units; because all aggregation steps are allowed while only single-unit breakup is allowed, detailed balance is broken. Its exact solution predicts a nonequilibrium percolation (gelation) transition with cluster-mass exponent $\tau = 5/2$, fractal dimension $d_f = 2.0$, and a power-law cluster-size distribution without exponential cutoff at the transition. In the simulations, the aggregating units are atoms with local five-fold symmetry (the order parameter $f_5$ exceeding 0.6), the structural order parameter driving percolation is the average coordination number $Z$, and the mechanical readout is the storage modulus $G'$ computed from oscillatory-shear molecular dynamics; the onset at $Z \approx 6$ is interpreted as the isostatic rigidity point where affine and nonaffine elastic contributions balance.

What would settle it

Cool the same alloy at a significantly slower cooling rate and locate the diverging correlation length, the largest five-fold cluster fraction, and the $G'$ onset as functions of temperature: if the percolation threshold and the modulus jump separate, or if the correlation length saturates instead of diverging at the glass transition temperature, the causal claim that rigidity percolation drives the glass transition would be refuted.

Watch

Extended reading notes

Core claim

The central claim is that the liquid-to-glass crossover is a continuous nonequilibrium percolation transition of stable five-fold atomic clusters. In the simulations, atoms with local five-fold symmetry $f_5 > 0.6$ form clusters whose largest member grows from small isolated aggregates at high temperature into a system-spanning network at the glass transition temperature; the growth is quantified by the fraction $f_z$ of atoms in the largest cluster, by a correlation length $\xi$ that diverges as the average coordination $Z$ approaches about 6, and by the loss of the exponential cutoff in the cluster-size distribution near 800 K. The sharp rise of the storage modulus $G'$ at the same temperature locates the mechanical rigidity transition, and the paper identifies the transition point with the isostatic rigidity condition $Z = 6$ at which the nonaffine contributions to the shear modulus balance the affine ones. The aggregation and breakup rates of the clusters are asymmetric and violate detailed balance, confirming the nonequilibrium character, and the structural relaxation time tracks the lifetime of the largest cluster. The authors conclude that the glass transition coincides with, and is caused by, this rigidity percolation, and that the percolating cluster's correlation length is the long-sought diverging length scale.

Load-bearing premise

The quantitative identification of the transition's universality class rests on the assumption that clusters break only by losing one unit at a time, but the simulations show breakup also produces multi-unit fragments, and the measured exponents ($d_f \approx 1.80$, $\sigma \approx -2$) deviate from the predicted values ($d_f = 2.0$, $\tau = 5/2$).

Editorial extensions

If this is right

  • The glass transition carries a genuine diverging length scale — the correlation length of the rigid percolating five-fold cluster — not merely a relaxation-time divergence.
  • Rigidity emerges through a continuous transition at an isostatic coordination $Z = 6$, so the shear modulus grows roughly linearly with $Z - 6$ above the threshold.
  • Structural arrest is governed by the lifetime of the percolating cluster, linking the slowdown of $\tau_\alpha$ to a structural rather than purely dynamic mechanism.
  • Alloy composition and cooling rate, which control the five-fold cluster population, become practical dials for setting modulus, damping, and toughness of metallic glasses.
  • Metallic-glass formation is placed in the same universality class as colloidal gelation and rigidity percolation in disordered media, so tools from one system transfer to the other.

Reading between the lines

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

  • A natural next step the paper does not take is to test the percolation scaling directly against system size: if $d_f$ and $\tau$ drift with box size, the transition is non-mean-field and the exact-solution exponents are only approximate.
  • The mechanism suggests an experimental probe: fluctuation electron microscopy or diffraction-derived five-fold order parameters should show the same diverging correlation length approaching $T_g$, giving a measurable structural precursor.
  • If the claim generalizes, glass formers with weak five-fold order would be predicted to show a weaker or absent percolation signature at $T_g$, which would clarify whether this is the universal mechanism or one specific to metallic systems.
  • The paper's plasticity argument implies that deliberately fragmenting the percolating network — e.g., through compositional heterogeneities — could trade a small amount of stiffness for enhanced ductility, a testable processing rule.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper uses molecular dynamics simulations of the model metallic glass Cu50Zr50 to argue that the glass transition coincides with, and is caused by, a nonequilibrium rigidity percolation transition of five-fold symmetric atomic clusters. The authors report a growing largest cluster, a diverging correlation length, a cluster-size distribution approaching a power law, and a sharp rise in the storage modulus upon cooling, all near the glass transition temperature Tg. They identify the percolation threshold with the Maxwell isostatic point Z=6 and connect the observed exponents to an exact solution of a reversible polymerization master equation with single-unit breakup. The paper concludes that this mechanism provides the long-sought diverging length scale at the glass transition and offers a route to engineer metallic glass properties.

Significance. If the central claim were established, the paper would be highly significant: it would unify the glass transition with rigidity percolation, identify a diverging correlation length, and connect local five-fold order to macroscopic mechanical response in a concrete model system. The paper contains a substantial set of molecular dynamics analyses, including cluster-size distributions, fractal dimensions, cluster lifetimes, and oscillatory shear response, and it explicitly verifies asymmetry between aggregation and breakup rates. However, the quantitative and causal claims rest on assumptions that the authors themselves admit are not satisfied, and the identification of the critical point with Z=6 is not independently established. As it stands, the paper offers a suggestive phenomenological correlation rather than a demonstrated mechanism.

major comments (4)
  1. [Sec. 2.1 and Sec. 2.6] The theoretical anchor of the paper is the exact solution of the master equation (1) with breakup limited to single-unit detachment (Eq. (2)), which gives the Fisher exponent tau=5/2, fractal dimension df=2.0, and exponent sigma=-5/2. In Sec. 2.6, however, the authors state that "the exact assumptions of Eq. (2) are not strictly satisfied here, since breakup events in our simulations are not exclusively limited to single-particle detachment." The measured values (df about 1.80, sigma about -2) deviate from the predicted ones. The paper rationalizes this as non-mean-field or system-specific effects, but if the key assumption of the exact solution is violated, the predicted exponents do not apply to the simulated system at all. The universality-class claim therefore has no quantitative theoretical grounding; the observed scaling is only a qualitative coincidence. This is load-bearing because the conclusion that the transition is a nonequilibrium percolation transition with a specific universality class rests on these exponents.
  2. [Sec. 2.3 and Figure S7] The critical coordination number Zc is determined as the maximum average coordination reached upon cooling close to 0 K (Figure S7) and is then used in the scaling relation fz~(Zc-Z)^-sigma and identified with the Maxwell isostatic point Z=6. This is a circular procedure: the percolation threshold should be estimated independently, for example from the peak of the susceptibility or from spanning probabilities, rather than from the saturation value of Z(T). Using the saturation value as the critical point makes the 'coincidence' with Z=6 partly tautological. Moreover, the predicted mechanical signature G'~(Z-6) from nonaffine response theory is never tested: Figure 11 plots G' versus T/Tg, not versus Z. Without a direct test of G' versus Z and an independent estimate of Zc, the central identification of the transition with the Maxwell isostatic point is unsubstantiated.
  3. [Conclusions and Sec. 2.3] The abstract and conclusions state that the glass transition 'coincides with, and is caused by' a nonequilibrium rigidity percolation transition. The simulations demonstrate a correlation between the growth of five-fold clusters, the emergence of a percolating cluster, and the rise of G' near Tg, but they do not establish causation. For example, the authors do not test whether suppressing or enhancing five-fold cluster aggregation moves Tg in the predicted way, nor do they compare with a system in which five-fold clusters are destabilized. The causal language therefore exceeds what the data can support; at most the data show a coincidence between percolation of the chosen structural motif and the kinetic glass transition.
  4. [Sec. 2.5 and Figure 7(b)] The evolution of the power-law exponent sigma is presented as evidence for the percolation transition, but the analysis is limited to fits of the cluster-size distribution without reported error bars or a clear criterion for the onset of the power-law regime. The statement that sigma 'stabilizes around -2' after the transition is based on a few temperature points, and the inset showing sigma1 is not accompanied by a quantitative comparison with the theoretical prediction. Given that the measured exponents deviate from the mean-field values, the paper should report the uncertainty in these fits and discuss whether the apparent stabilization is significant.
minor comments (5)
  1. [Sec. 2.1] The phrase 'This break detailed balance' should read 'This breaks detailed balance.'
  2. [Sec. 2.1] Reference [17] is cited for the statement that the model breaks detailed balance, but [17] is an IEEE instrumentation paper; the intended reference is likely [27] (Krapivsky et al.) or a related statistical physics source.
  3. [Sec. 2.2 and Sec. 4] The cutoff distance for cluster definition is described in Sec. 2.2 as 'the first intersection point of the potential' and in Sec. 4 as 'the inflection point (~2.80 Å)'; these descriptions should be reconciled, and the stated sensitivity tests in the Supplementary should be summarized in the main text.
  4. [Figure 7(b)] The text says sigma stabilizes around -2, 'somewhat larger' than the mean-field prediction of -5/2; this wording is confusing because -2 is less negative, not larger in magnitude. Please rephrase for clarity.
  5. [Sec. 2.7] The definition of cluster lifetime (duration until the cluster size reduces by more than 10%) is arbitrary; the sensitivity of the comparison in Figure 9(b) to this threshold should be discussed or a more standard definition (e.g., based on a correlation function) should be used.

Circularity Check

1 steps flagged · score 6.0 of 10

The Maxwell-isostatic identification rests on defining Zc as the low-temperature saturation value and then using that same Zc in the percolation scalings.

  1. fitted input called prediction [Sec. 2.3, paragraph following Fig. 3; also the paragraph defining Zc and the isostatic condition]
    "We observe a sharp increase in fz as the system approaches the critical coordination number Zc ≈ 6, indicating rapid growth of the largest five-fold aggregate. Notably, we identify a clear crossover in the scaling exponent σ, defined by the relation fz~(Zc − Z)^−σ, occurring near this critical value Zc, which corresponds precisely to the glass transition temperature Tg. The critical value (Zc ≈ 6) was also rigorously determined by continuously cooling close to 0 K to find its maximum value (see Figure S7, Supporting Information), essentially matching our expectation of 6."

    The percolation threshold Zc is not obtained from a percolation criterion (spanning probability, susceptibility peak, or a free critical-point fit). Instead it is defined as the maximum or saturation value of the average coordination Z reached upon cooling to 0 K. All the scaling evidence (fz~(Zc−Z)^−σ and ξ~(Zc−Z)) is then presented using this same Zc, so the apparent critical growth is anchored at a point chosen by definition. Identifying that value with the Maxwell isostatic condition (G'~Z−6) converts the definition into the paper's central conclusion: the transition 'coincides with' isostaticity because Zc was chosen as the terminal coordination plateau, not because a percolation threshold was measured independently.

full rationale

The paper contains substantial independent simulation content: the growth of five-fold clusters, emergence of a system-spanning cluster near Tg, the asymmetry between aggregation and breakup rates, and the sharp rise in G' are all real observations. The underlying master-equation model and nonaffine response theory are also external to this work. However, the paper's central quantitative claim—that the glass transition coincides with the Maxwell isostatic point—partly reduces to a definitional choice. Zc is set equal to the low-temperature saturation value of Z, and the same Zc is inserted into the scaling laws used to demonstrate critical divergence and to identify the isostatic point. The paper also concedes that the exact breakup assumption behind the theoretical exponents is not satisfied in the simulations, and the measured exponents (df ≈ 1.80, σ ≈ −2) deviate from the externally derived mean-field values (df = 2.0, σ = −5/2), so the universality-class claim is not confirmed by that theory. Because the coincidence with isostaticity is constructed through the definition of Zc rather than through an independent percolation-threshold measurement, the central claim is partially circular, but it is not entirely forced because the qualitative percolation phenomenon is independently visible in the data.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central claim rests on several choices and assumptions: a geometric definition of five-fold clusters (f5>0.6, cutoff 2.80 Å), an extracted critical coordination Zc about 6 from the same cooling runs, and a mean-field polymerization model whose breakup rule is not satisfied by the simulated dynamics. The measured exponents deviate from the model, so the model functions more as a narrative frame than as a falsifiable derivation.

free parameters (5)
  • LFFS threshold f5 > 0.6 = 0.6
    Defines which atoms count as five-fold and enter cluster analysis; taken from prior work [35,36] by the authors' group; no sensitivity scan in main text.
  • Cluster cutoff distance = 2.80 Å
    First inflection of the EAM potential, corresponding to first RDF peak; determines bonds and therefore all cluster sizes, percolation statistics, and aggregation/breakup rates.
  • Critical coordination Zc = 6
    Obtained by cooling the same simulation close to 0 K and taking the maximum average coordination; used as the percolation threshold in fz~(Zc-Z)^-sigma and equated with the Maxwell isostatic value; not an independent benchmark.
  • Largest-cluster lifetime reduction threshold = 10%
    Cluster lifetime is defined as the time until the largest cluster shrinks by more than 10%; a chosen criterion that affects the correlation with structural relaxation time.
  • Power-law exponents sigma1 and sigma2 = about -2 (stabilized)
    Fitted to the cluster-mass distributions at two size ranges; presented as evidence for percolation, but the fitted values deviate from the model prediction of -5/2.
assumptions (6)
  • ad hoc to paper Master equation (1) with rates (2) has exact solution giving c_k ~ k^{-5/2} at the gel point.
    Used to predict Fisher exponent 5/2 and df=2.0; the breakup-only-single-unit condition is violated by the simulated dynamics (admitted in Sec 2.6), so this axiom is not satisfied by the system it is applied to.
  • standard math Hyperscaling relation tau = d/d_f + 1.
    Used to convert tau=5/2 into fractal dimension df=2.0 in three dimensions (Sec 2.1).
  • domain assumption Maxwell isostatic condition Z=6 and G' ~ (Z-6) from nonaffine response theory of spherical particles with central forces.
    Applied to a metallic glass with many-body EAM potential (Sec 2.3, 3); transferability from central-force sphere networks to metallic bonding is assumed, not demonstrated.
  • domain assumption Global average coordination Z is the correct control parameter for the rigidity percolation of the five-fold subnetwork.
    Z is averaged over all atoms, while the percolating network is restricted to atoms with f5>0.6; the paper never measures the internal coordination of the percolating cluster, so the isostatic argument operates on a different observable.
  • domain assumption Atoms with f5>0.6 form icosahedral-like structures and are the unique structural motifs driving the transition.
    Statement in Sec 2.2 based on previous work; four-fold and six-fold symmetries are argued to be unimportant from their distributions, but their percolation behavior is not tested.
  • domain assumption The glass transition temperature Tg determined by the E-3kBT crossover at a single cooling rate is the same transition probed by the percolation analysis.
    Tg depends on cooling rate; the coincidence between percolation and Tg is established in one cooling protocol and could shift if the protocol changed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Emergent rigidity percolation of five-fold aggregates enables controllable glass properties." pith.science (2026). https://pith.science/paper/TB5MDXB2

@misc{pith2026250602588,
  author       = {Pith},
  title        = {Pith review of: Emergent rigidity percolation of five-fold aggregates enables controllable glass properties},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TB5MDXB2}},
  note         = {Machine review of arXiv:2506.02588}
}
read the original abstract

Metallic glasses possess outstanding mechanical and physical properties, making them promising candidates for advanced structural and functional applications; however, the lack of understanding and control over their glass transition and solidification processes remains a significant barrier to practical design. The glass transition from liquid to amorphous solid has remained an open problem in physics despite many theories and recent advances in computational efforts. The question of identifying a clear and well-defined diverging length scale accompanying the glass transition has remained unanswered, as has the nature of the transition and, indeed, the presence of a transition at all, as opposed to a mere dynamical crossover. Here we answer these questions using numerical results and theoretical analysis showing that, in atomic (metallic) glass formers, the glass transition coincides with, and is caused by, a continuous rigidity percolation transition from a liquid-like to a solid-like material. The transition occurs as five-fold symmetric atomic clusters progressively aggregate, forming a system-spanning rigid network that marks the onset of mechanical stability. This percolation-driven rigidity growth is accompanied by a sharp increase in the shear modulus G', indicating the emergence of macroscopic solid-like behavior. Beyond this point, which coincides with the Maxwell isostatic point of the percolating structure, dynamical arrest or "freezing-in" prevents further evolution. The long-sought diverging length scale is thus identified as the percolation-driven growth of rigid five-fold clusters, providing a direct link between local structural motifs and macroscopic mechanical properties at the glass transition. These insights offer practical routes to rationally engineer metallic glasses with targeted mechanical stiffness, hardness, and toughness.

Figures

Figures reproduced from arXiv: 2506.02588 by the authors.

Figure 1
Figure 1. Simulation details of the metallic liquid. (a) Temperature dependence of E￾3kBT for Cu50Zr50 liquid upon cooling. The crossover temperature is confirmed as the glass transition temperature Tg. (b) The potential function used in this work. The first intersection of the potential function is defined as the cutoff distance for whether different atoms are considered to be in the same cluster. To explore the onset of per… view at source ↗
Figure 2
Figure 2. The distributions of (a) four-fold, and (b) six-fold symmetries at six different temperatures. The x-axis indicates the range of local symmetry with different degrees, e.g., 0.1-0.2 in [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

59 extracted references · 58 canonical work pages

  1. [1]

    Angell, Science 1995, 267, 1924

    C.A. Angell, Science 1995, 267, 1924

  2. [2]

    Debenedetti, F.H

    P.G. Debenedetti, F.H. Stillinger, Nature 2001, 410, 259. 36

  3. [3]

    Kivelson, G

    S.A. Kivelson, G. Tarjus, Nat. Mater. 2008, 7, 831

  4. [4]

    Ediger, C.A

    M.D. Ediger, C.A. Angell, S.R. Nagel, J. Phys. Chem. 1996, 100, 13200

  5. [5]

    Dyre, Rev

    J.C. Dyre, Rev. Mod. Phys. 2008, 78, 953

  6. [6]

    Kirkpatrick, P.G

    T.R. Kirkpatrick, P.G. Wolynes, Phys. Rev. B 1987, 36, 8552

  7. [7]

    Albert, T

    S. Albert, T. Bauer, M. Michl, G. Biroli, F. Ladieu, Science 2016, 352, 1308

  8. [8]

    Angell, K.J

    C.A. Angell, K.J. Rao, J. Chem. Phys. 1972, 57, 470

Show all 59 references
  1. [9]

    Chandler, J.P

    D. Chandler, J.P. Garrahan, Annu. Rev. Phys. Chem 2010, 61, 191

  2. [10]

    Adam, J.H

    G. Adam, J.H. Gibbs, The journal of chemical physics 1965, 43, 139

  3. [11]

    Argon, Acta metallurgica 1979, 27, 47

    A. Argon, Acta metallurgica 1979, 27, 47

  4. [12]

    Lewandowski, A

    J. Lewandowski, A. Greer, Nature materials 2006, 5, 15

  5. [13]

    Donth, Journal of Non-Crystalline Solids 1982, 53, 325

    E. Donth, Journal of Non-Crystalline Solids 1982, 53, 325

  6. [14]

    Lubchenko, P.G

    V. Lubchenko, P.G. Wolynes, Annu. Rev. Phys. Chem 2006, 58, 235

  7. [15]

    Zaccone, E.M

    A. Zaccone, E.M. Terentjev, Phys. Rev. Lett. 2013, 110, 178002

  8. [16]

    Zaccone, E

    A. Zaccone, E. Scossa-Romano, Phys. Rev. B 2011, 83, 184205

  9. [17]

    P. Shen, S. Zhang, D. Ma, B. Xing, H. Chen, M. Zhang, H. Wei, B. Li, X. Miao, M. Xie, IEEE Transactions on Instrumentation and Measurement 2025

  10. [18]

    Zhang, J

    J. Zhang, J. Li, G. Tan, R. Hu, J. Wang, C. Chang, X. Wang, ACS applied materials & interfaces 2017, 9, 42192

  11. [19]

    Lebrun, F

    N. Lebrun, F. Dupla, H. Bruhier, M. Prudent, A. Borroto, C. Der Loughian, F. Bourquard, J. Pelletier, M. Rousseau, J.-P. Colombier, Applied Surface Science 2024, 160617

  12. [20]

    Hofmann, L.M

    D.C. Hofmann, L.M. Andersen, J. Kolodziejska, S.N. Roberts, J.P. Borgonia, W.L. Johnson, K.S. Vecchio, A. Kennett, Advanced Engineering Materials 2017, 19, 1600541

  13. [21]

    W. Lu, J. Ma, C. Wang, Y. Liu, Science China Technological Sciences 2024, 67, 2505

  14. [22]

    Wang, Advanced Materials 2009, 21, 4524

    W. Wang, Advanced Materials 2009, 21, 4524

  15. [23]

    Rouwhorst, C

    J. Rouwhorst, C. Ness, S. Stoyanov, A. Zaccone, P. Schall, Nat. Commun. 2020, 11, 3558

  16. [24]

    Zhang, X

    H. Zhang, X. Wang, J. Zhang, H.B. Yu, J.F. Douglas, Eur. Phys. J. E: Soft Matter 2023, 46, 50

  17. [25]

    Y. Sun, Y. Zhang, F. Zhang, Z. Ye, Z. Ding, C.Z. Wang, K.M. Ho, J. Appl. Phys. 2016, 120, 869

  18. [26]

    G. Hägg, J. Chem. Phys. 1935, 3, 42

  19. [27]

    Krapivsky, S

    P.L. Krapivsky, S. Redner, E. Ben -Naim, 2010, A Kinetic View of Statistical Physics, Cambridge University Press

  20. [28]

    Majumdar, S

    S.N. Majumdar, S. Krishnamurthy, M. Barma, J. Stat. Phys. 2000, 99, 1

  21. [29]

    W. Chu, J. Yu, N. Ren, Z. Wang, L. Hu, Phys. Chem. Chem. Phys. 2023, 25, 4151

  22. [30]

    Plimpton, Journal of computational physics 1995, 117, 1

    S. Plimpton, Journal of computational physics 1995, 117, 1

  23. [31]

    Cheng, H.W

    Y.Q. Cheng, H.W. Sheng, E. Ma, Phy. Rev. B 2008, 78, 1436

  24. [32]

    Hoover, Phys

    W.G. Hoover, Phys. Rev. A 1985, 31, 1695

  25. [33]

    Stukowski, Alexander, Modell. Simul. Mater. Sci. Eng 2010, 18, 2154

  26. [34]

    Gao, H.-B

    L. Gao, H.-B. Yu, T.B. Schrøder, J.C. Dyre, Nature Physics 2025, 1

  27. [35]

    N. Ren, L. Hu, B. Wang, K. Song, P. Guan, Scr. Mater. 2021, 200, 113926

  28. [36]

    Y.C. Hu, F.X. Li, M.Z. Li, H.Y. Bai, W.H. Wang, Nat. Commun. 2015, 6, 8310

  29. [37]

    Z. Wang, F. Yang, A. Bernasconi, K. Samwer, A. Meyer, Phys. Rev. B 2018, 98, 024204

  30. [38]

    J. Yu, Z. Wang, L. Hu, W. Chu, Y. Bai, Scr. Mater. 2022, 216, 114737

  31. [39]

    Hansen, I.R

    J.P. Hansen, I.R. Mcdonald, 2013, Theory of Simple Liquids. Fourth Edition, Academic Press, London

  32. [40]

    Lunkenheimer, A

    P. Lunkenheimer, A. Loidl, B. Riechers, A. Zaccone, K. Samwer, Nature Physics 2023, 19, 694. 37

  33. [41]

    Landau, E.M

    L.D. Landau, E.M. Lifshitz, 1980, Statistical Physics. third Edition., Pergamon Press, Oxford

  34. [42]

    Zaccone, 2023, Theory of Disordered Solids, Springer, Heidelberg

    A. Zaccone, 2023, Theory of Disordered Solids, Springer, Heidelberg

  35. [43]

    Stauffer, A

    D. Stauffer, A. Aharony, 1994, Introduction To Percolation Theory, CRC Press, Boca -Raton FL

  36. [44]

    Sandra, O

    J.V. Sandra, O. Antoniuk, B. Weber, M. Potenza, S. Mazzoni, P. Schall, G.H. Wegdam, Phys. Rev. Lett. 2012, 109, 248302

  37. [45]

    Shelke, V.D

    P.B. Shelke, V.D. Nguyen, A.V. Limaye, P. Schall, Adv Mater. 2013, 25, 1499

  38. [46]

    Cavagna, Physics Reports 2009, 476, 51

    A. Cavagna, Physics Reports 2009, 476, 51

  39. [47]

    Li, Journal of Materials Science & Technology 2014, 30, 551

    M. Li, Journal of Materials Science & Technology 2014, 30, 551

  40. [48]

    H. Peng, M. Li, W. Wang, Physical review letters 2011, 106, 135503

  41. [49]

    Schuh, T.C

    C.A. Schuh, T.C. Hufnagel, U. Ramamurty, Acta Materialia 2007, 55, 4067

  42. [50]

    Javerzat, Physical Review Letters 2024, 132, 018201

    N. Javerzat, Physical Review Letters 2024, 132, 018201

  43. [51]

    Javerzat, M

    N. Javerzat, M. Bouzid, Physical Review Letters 2023, 130, 268201

  44. [52]

    Plimpton, J

    S. Plimpton, J. Comput. Phys. 1995, 117, 1

  45. [53]

    Ding, Y.-Q

    J. Ding, Y.-Q. Cheng, H. Sheng, E. Ma, Phys. Rev. B 2012, 85, 060201

  46. [54]

    Z. Wu, M. Li, W. Wang, K. Liu, Nat. Commun. 2015, 6, 6035

  47. [55]

    L. Ward, D. Miracle, W. Windl, O.N. Senkov, K. Flores, Phys. Rev. B 2013, 88, 134205

  48. [56]

    Rouwhorst, C

    J. Rouwhorst, C. Ness, S. Stoyanov, A. Zaccone, P. Schall, Nature communications 2020, 11, 3558

  49. [57]

    Rouwhorst, P

    J. Rouwhorst, P. Schall, C. Ness, T. Blijdenstein, A. Zaccone, Physical Review E 2020, 102, 022602

  50. [58]

    Z. Wu, M. Li, W. Wang, K. Liu, Physical Review B —Condensed Matter and Materials Physics 2013, 88, 054202

  51. [59]

    Lee, C.-M

    M. Lee, C.-M. Lee, K.-R. Lee, E. Ma, J.-C. Lee, Acta Materialia 2011, 59, 159

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.