pith. sign in

arxiv: 2511.19746 · v2 · submitted 2025-11-24 · ❄️ cond-mat.mtrl-sci

An activation-relaxation technique study of two-level system impact on internal dissipation using DFT-based moment tensor potential

Renaude Girard , Carl L\'evesque , Normand Mousseau , Fran\c{c}ois Schiettekatte This is my paper

Pith reviewed 2026-05-17 05:18 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci
keywords amorphous silicontwo-level systemsinternal frictionmoment tensor potentialactivation-relaxation techniquemachine-learned potentialbond exchangedissipation
0
0 comments X

The pith

DFT-trained moment tensor potential finds twice as many complex two-level systems in amorphous silicon.

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

This paper combines a machine-learned moment tensor potential trained on density functional theory data with the activation-relaxation technique to locate and classify two-level systems in amorphous silicon. It shows that overall material properties such as radial distribution functions, defect levels, and internal friction match both experimental measurements and results from a modified Stillinger-Weber potential. The atomistic details differ, however, with complex two-level systems based on Wooten-Winer-Weaire bond exchange appearing roughly twice as often. These systems also tend to remain isolated and oscillate independently rather than interacting. Understanding these differences matters because two-level systems control low-temperature dissipation in amorphous materials used in sensors and quantum technologies.

Core claim

The central claim is that MTP-based models of amorphous silicon recover experimental structural and dissipative properties yet produce a higher density of complex two-level systems, specifically those involving Wooten-Winer-Weaire bond exchanges, at about twice the rate found with a modified Stillinger-Weber potential, while the density of simpler bond-hopping two-level systems remains comparable and the identified systems mostly act as isolated, independent oscillators.

What carries the argument

Moment Tensor Potential trained on DFT data, paired with the Activation-Relaxation Technique nouveau to sample and classify two-level systems according to their atomic rearrangement mechanisms.

If this is right

  • Average structural and dissipative properties including radial distribution functions and internal friction remain consistent with experimental data.
  • The density of simple bond-hopping two-level systems stays similar across the two potentials.
  • Two-level systems appear mostly isolated and oscillate independently in the MTP models.
  • More complex two-level systems involving bond exchange are roughly twice as common as in the Stillinger-Weber models.

Where Pith is reading between the lines

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

  • The method could be applied to other amorphous solids to predict how their low-temperature dissipation scales with preparation history.
  • Verification against direct quantum calculations on the discovered complex TLS configurations would test whether the potential captures tunneling rates correctly.
  • Isolated TLS behavior may simplify models of decoherence in devices that use amorphous silicon layers at millikelvin temperatures.

Load-bearing premise

The moment tensor potential accurately reproduces the energy barriers and atomic configurations of rare, low-energy two-level system events that were never included in its training data.

What would settle it

Direct density functional theory calculations of energy barriers and relaxed configurations for the specific two-level system geometries identified by the MTP-ARTn search, to test whether they agree with the potential's predictions.

Figures

Figures reproduced from arXiv: 2511.19746 by Carl L\'evesque, Fran\c{c}ois Schiettekatte, Normand Mousseau, Renaude Girard.

Figure 1
Figure 1. Figure 1: FIG. 1. Representation of the configurational energy of a TLS [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Attempt frequency, or prefactor, given by the har [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Top panel: The mean radial distribution function [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Energy asymmetry as a function of the forward energy [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Histogram of the density of TLS events per 1000 [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8. Density of TLS species per 1000 atoms in term of [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9. Loss angle contribution for MTP, mSW (recomputed [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
read the original abstract

We use a recently-developed machine-learned Moment Tensor Potential (MTP) trained on data generated with the density functional theory (DFT) and tailored to amorphous silicon coupled with the Activation-Relaxation Technique nouveau (ARTn) to identify and classify two-level systems (TLS). The samples generated using MTP recover experimental results and provide average structural and dissipative properties similar to those obtained with a modified Stillinger-Weber potential, including radial distribution function, defect concentration and internal friction. Atomistic details, however, are significantly different, including the density and type of TLS. In particular, we find that while the density of TLS involving a bond-hopping mechanism is similar for the two potentials, more complex TLSs, such as those involving a Wooten-Winer-Weaire bond exchange, are about twice as common. Analysis also shows that TLSs, for MTP-based models, are mostly isolated and oscillate independently from each other.

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 uses a DFT-trained Moment Tensor Potential (MTP) for amorphous silicon together with the Activation-Relaxation Technique nouveau (ARTn) to locate and classify two-level systems (TLS). It reports that MTP-generated samples reproduce experimental radial distribution functions and internal-friction values at a level comparable to a modified Stillinger-Weber potential, yet yield a similar density of bond-hopping TLS but roughly twice as many complex Wooten-Winer-Weaire bond-exchange TLS; the MTP TLS are also described as predominantly isolated and independently oscillating.

Significance. If the central claims are substantiated, the work would demonstrate that machine-learned potentials can resolve a richer set of TLS mechanisms than empirical potentials while still matching key experimental observables, thereby strengthening atomistic interpretations of dissipation in amorphous silicon and related glasses.

major comments (2)
  1. [MTP Training subsection] MTP Training subsection: the potential is stated to be trained on DFT data tailored to amorphous silicon, yet no explicit validation or augmentation with low-energy TLS configurations, near-degenerate minima, or saddle-point structures is described. Because the ARTn search and subsequent TLS counting rest on accurate relative energies of these rare events, even modest barrier errors (~0.05–0.1 eV) could change the sampled ensemble and directly affect the reported factor-of-two difference in complex TLS density.
  2. [TLS Statistics and Classification section] TLS Statistics and Classification section: TLS densities, type fractions, and the claim that MTP TLS are “mostly isolated” are presented without error bars, without the number of independent samples or total TLS events counted, and without convergence tests versus system size or ARTn search parameters. These omissions undermine quantitative comparison of the bond-hopping versus Wooten-Winer-Weaire populations.
minor comments (2)
  1. [Figures 1–3] Figure captions and axis labels should explicitly state whether error bands or standard deviations from multiple samples are shown; currently the RDF and internal-friction plots appear to lack them.
  2. [Methods] The precise geometric criteria used to classify a TLS as “bond-hopping” versus “Wooten-Winer-Weaire bond exchange” should be stated in a dedicated methods paragraph or table to allow reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We are grateful to the referee for their detailed and constructive comments, which have helped us improve the clarity and rigor of our manuscript. Below, we respond to each major comment in turn. We have made revisions to the manuscript to incorporate additional information and clarifications as detailed in our responses.

read point-by-point responses

  1. Referee: MTP Training subsection: the potential is stated to be trained on DFT data tailored to amorphous silicon, yet no explicit validation or augmentation with low-energy TLS configurations, near-degenerate minima, or saddle-point structures is described. Because the ARTn search and subsequent TLS counting rest on accurate relative energies of these rare events, even modest barrier errors (~0.05–0.1 eV) could change the sampled ensemble and directly affect the reported factor-of-two difference in complex TLS density.

    Authors: We acknowledge the importance of validating the MTP on configurations relevant to TLS, such as low-energy barriers and saddle points. Our training dataset includes a wide range of amorphous silicon structures generated via DFT, encompassing various local environments and energies that are representative of the material. While specific augmentation with TLS minima was not performed, the potential reproduces key experimental observables like the radial distribution function and internal friction values. To address this comment, we will revise the MTP Training subsection to include a discussion of the training data's coverage of low-energy configurations and any relevant validation metrics. We will also note that small barrier errors, if present, are unlikely to alter the qualitative finding of increased complex TLS density, as this difference is observed consistently. revision: yes

  2. Referee: TLS Statistics and Classification section: TLS densities, type fractions, and the claim that MTP TLS are “mostly isolated” are presented without error bars, without the number of independent samples or total TLS events counted, and without convergence tests versus system size or ARTn search parameters. These omissions undermine quantitative comparison of the bond-hopping versus Wooten-Winer-Weaire populations.

    Authors: We agree that providing statistical details is essential for the credibility of our quantitative claims. In the revised version, we will add the number of independent amorphous samples generated and analyzed, the total count of TLS identified across all samples, and error bars derived from the standard deviation across samples for the TLS densities and type fractions. Additionally, we will include a short paragraph on convergence tests with respect to system size and ARTn search parameters, confirming that the reported trends are stable. These additions will support the comparison between MTP and Stillinger-Weber potentials and the assertion that MTP TLS are predominantly isolated. revision: yes

Circularity Check

0 steps flagged

No significant circularity; central claims are direct simulation outputs

full rationale

The paper generates amorphous silicon samples with an externally trained MTP (on DFT data) and a modified SW potential, then applies the ARTn method to locate and classify TLS events. Reported quantities such as TLS densities for bond-hopping versus Wooten-Winer-Weaire mechanisms and the conclusion that MTP TLSs are mostly isolated are obtained as direct counts and statistics from these independent runs. No step defines a TLS density or isolation metric in terms of the training parameters or prior results by construction, and the comparative claims do not reduce to a self-citation chain or fitted-input renaming. The methodology is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work relies on the assumption that the MTP potential energy surface is sufficiently accurate for TLS identification even though training data details are not supplied; no new physical entities are postulated.

axioms (1)
  • domain assumption The MTP reproduces DFT energies and forces with sufficient accuracy for locating low-energy TLS configurations.
    Invoked when claiming that MTP-based TLS statistics are physically meaningful.

pith-pipeline@v0.9.0 · 5485 in / 1263 out tokens · 46452 ms · 2026-05-17T05:18:48.030780+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

40 extracted references · 40 canonical work pages · 1 internal anchor

  1. [1]

    GWTC-4.0: An introduction to version 4.0 of the gravitational-wave transient catalog,

    The LIGO Scientific Collaboration and the Virgo Col- laboration and the KAGRA Collaboration, A. G. Abac, et al., “GWTC-4.0: An introduction to version 4.0 of the gravitational-wave transient catalog,” 2025

    work page 2025

  2. [2]

    The clustering of mas- sive Primordial Black Holes as Dark Matter: Measuring their mass distribution with advanced LIGO,

    S. Clesse and J. Garc´ ıa-Bellido, “The clustering of mas- sive Primordial Black Holes as Dark Matter: Measuring their mass distribution with advanced LIGO,”Physics of the Dark Universe, vol. 15, pp. 142–147, Mar. 2017

    work page 2017

  3. [3]

    GW150914: The Advanced LIGO Detectors in the Era of First Discover- ies,

    B. Abbott, R. Abbott, T. Abbott, M. Abernathy, F. Ac- ernese, K. Ackley, C. Adams, T. Adams, P. Addesso, R. Adhikari, V. Adya, and Affeldt, “GW150914: The Advanced LIGO Detectors in the Era of First Discover- ies,”Physical Review Letters, vol. 116, p. 131103, Mar. 2016

    work page 2016

  4. [4]

    Challenges in thermal noise for 3rd generation of gravitational wave detectors,

    R. Nawrodt, S. Rowan, J. Hough, M. Punturo, F. Ricci, and J.-Y. Vinet, “Challenges in thermal noise for 3rd generation of gravitational wave detectors,”General Rel- ativity and Gravitation, vol. 43, pp. 593–622, Feb. 2011

    work page 2011

  5. [5]

    The Advanced LIGO Gravitational Wave Detector

    S. J. Waldman, “The Advanced LIGO Gravitational Wave Detector,” Mar. 2011. arXiv:1103.2728 [gr-qc]

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  6. [6]

    The fluctuation-dissipation theorem,

    R. Kubo, “The fluctuation-dissipation theorem,”Reports on Progress in Physics, vol. 29, p. 255, jan 1966

    work page 1966

  7. [7]

    Two-level states in glasses,

    W. A. Phillips, “Two-level states in glasses,”Reports on Progress in Physics, vol. 50, pp. 1657–1708, Dec. 1987

    work page 1987

  8. [8]

    Molecular dynamics modeling of mechanical loss in amorphous tantala and titania-doped tantala,

    J. P. Trinastic, R. Hamdan, C. Billman, and H.-P. Cheng, “Molecular dynamics modeling of mechanical loss in amorphous tantala and titania-doped tantala,”Phys. Rev. B, vol. 93, p. 014105, Jan 2016

    work page 2016

  9. [9]

    Effects of elevated-temperature deposition on the atomic structure of amorphous ta2o5 films,

    K. Prasai, K. Lee, B. Baloukas, H.-P. Cheng, M. Fazio, L. Martinu, A. Mehta, C. S. Menoni, F. Schiettekatte, R. Shink, B. Shyam, G. Vajente, M. M. Fejer, and R. Bassiri, “Effects of elevated-temperature deposition on the atomic structure of amorphous ta2o5 films,”APL Materials, vol. 11, p. 121112, 12 2023

    work page 2023

  10. [10]

    Thermal noise in interferometric gravitational wave detectors due to dielectric optical coatings,

    G. M. Harry, A. M. Gretarsson, P. R. Saulson, S. E. Kit- telberger, S. D. Penn, W. J. Startin, S. Rowan, M. M. Fe- jer, D. R. M. Crooks, G. Cagnoli, J. Hough, and N. Nak- agawa, “Thermal noise in interferometric gravitational wave detectors due to dielectric optical coatings,”Classi- cal and Quantum Gravity, vol. 19, pp. 897–917, feb 2002

    work page 2002

  11. [11]

    Hydrogen-free amorphous silicon with no tunneling states,

    X. Liu, D. R. Queen, T. H. Metcalf, J. E. Karel, and F. Hellman, “Hydrogen-free amorphous silicon with no tunneling states,”Phys. Rev. Lett., vol. 113, p. 025503, Jul 2014

    work page 2014

  12. [12]

    Thermal noise reduction and absorption optimization via multimaterial coatings,

    J. Steinlechner, I. W. Martin, J. Hough, C. Kr¨ uger, S. Rowan, and R. Schnabel, “Thermal noise reduction and absorption optimization via multimaterial coatings,” Phys. Rev. D, vol. 91, p. 042001, Feb 2015

    work page 2015

  13. [13]

    Revealing the role of hydrogen in reducing optical absorption and mechanical loss in magnetron-sputtered amorphous sili- con for gravitational-wave detectors,

    M. Molina-Ruiz, R. Zhou, A. Markosyan, R. Bassiri, M. M. Fejer, A. Ananyeva, S. C. Tait, G. Vajente, A. Davenport, C. S. Menoni, and F. Hellman, “Revealing the role of hydrogen in reducing optical absorption and mechanical loss in magnetron-sputtered amorphous sili- con for gravitational-wave detectors,”Phys. Rev. Mater., vol. 9, p. 105602, Oct 2025. 10

    work page 2025

  14. [14]

    Crystalline/amorphous silicon and silica bragg reflectors for gravitational wave detectors

    E. Lalande, J. Vanier, A. W. Lussier, L. Martinu, J. Steinlechner, A. Amnache, S. Ecoffey, and F. Schiet- tekatte, “Crystalline/amorphous silicon and silica bragg reflectors for gravitational wave detectors.”in prep

    work page

  15. [15]

    Fitting the Stillinger–Weber potential to amorphous silicon,

    R. Vink, G. Barkema, W. Van Der Weg, and N. Mousseau, “Fitting the Stillinger–Weber potential to amorphous silicon,”Journal of Non-Crystalline Solids, vol. 282, pp. 248–255, Apr. 2001

    work page 2001

  16. [16]

    Internal mechanical dissipation mecha- nisms in amorphous silicon,

    C. L´ evesque, S. Roorda, F. Schiettekatte, and N. Mousseau, “Internal mechanical dissipation mecha- nisms in amorphous silicon,”Physical Review Materials, vol. 6, p. 123604, Dec. 2022

    work page 2022

  17. [17]

    Amorphous silicon with extremely low absorp- tion: Beating thermal noise in gravitational astronomy,

    R. Birney, J. Steinlechner, Z. Tornasi, S. MacFoy, D. Vine, A. S. Bell, D. Gibson, J. Hough, S. Rowan, P. Sortais, S. Sproules, S. Tait, I. W. Martin, and S. Reid, “Amorphous silicon with extremely low absorp- tion: Beating thermal noise in gravitational astronomy,” Phys. Rev. Lett., vol. 121, p. 191101, Nov 2018

    work page 2018

  18. [18]

    Simulation of single- particle displacement damage in silicon—part III: First principle characterization of defect properties,

    A. Jay, A. Hemeryck, N. Richard, L. Martin-Samos, M. Raine, A. Le Roch, N. Mousseau, V. Goiffon, P. Pail- let, M. Gaillardin, and P. Magnan, “Simulation of single- particle displacement damage in silicon—part III: First principle characterization of defect properties,”IEEE Trans. Nucl. Sci., vol. 65, no. 2, pp. 724–731, 2018

    work page 2018

  19. [19]

    Improving accuracy of interatomic potentials: more physics or more data? a case study of silica,

    I. S. Novikov and A. V. Shapeev, “Improving accuracy of interatomic potentials: more physics or more data? a case study of silica,”Materials Today Communications, vol. 18, pp. 74–80, 2019

    work page 2019

  20. [20]

    Moment Tensor Potentials: A Class of Systematically Improvable Interatomic Potentials,

    A. V. Shapeev, “Moment Tensor Potentials: A Class of Systematically Improvable Interatomic Potentials,”Mul- tiscale Modeling & Simulation, vol. 14, pp. 1153–1173, Jan. 2016

    work page 2016

  21. [21]

    A unified moment tensor potential for silicon, oxygen, and silica,

    K. Zongo, H. Sun, C. Ouellet-Plamondon, and L. K. B´ eland, “A unified moment tensor potential for silicon, oxygen, and silica,”npj Computational Materials, vol. 10, p. 218, Sept. 2024

    work page 2024

  22. [22]

    Current status of transition-state theory,

    D. G. Truhlar, B. C. Garrett, and S. J. Klippenstein, “Current status of transition-state theory,”The Journal of Physical Chemistry, vol. 100, no. 31, pp. 12771–12800, 1996

    work page 1996

  23. [23]

    Atomistic study of two-level systems in amorphous silica,

    T. Damart and D. Rodney, “Atomistic study of two-level systems in amorphous silica,”Physical Review B, vol. 97, p. 014201, Jan. 2018

    work page 2018

  24. [24]

    Force fields for silicas and aluminophosphates based on ab initiocalculations,

    B. W. H. Van Beest, G. J. Kramer, and R. A. Van Santen, “Force fields for silicas and aluminophosphates based on ab initiocalculations,”Phys. Rev. Lett., vol. 64, no. 16, pp. 1955–1958, 1990

    work page 1955

  25. [25]

    Role of potential-energy scaling in the low-temperature relaxation behavior of amorphous mate- rials,

    F. H. Stillinger, “Role of potential-energy scaling in the low-temperature relaxation behavior of amorphous mate- rials,”Phys. Rev. B, vol. 32, no. 5, pp. 3134–3141, 1985

    work page 1985

  26. [26]

    Amorphous silicon structures gen- erated using a moment tensor potential and the ac- tivation relaxation technique nouveau,

    K. Zongo, H. Sun, C. Ouellet-Plamondon, N. Mousseau, and L. K. B´ eland, “Amorphous silicon structures gen- erated using a moment tensor potential and the ac- tivation relaxation technique nouveau,” Jan. 2025. arXiv:2501.10549 [cond-mat]

    work page arXiv 2025

  27. [27]

    Event-Based Relax- ation of Continuous Disordered Systems,

    G. T. Barkema and N. Mousseau, “Event-Based Relax- ation of Continuous Disordered Systems,”Physical Re- view Letters, vol. 77, pp. 4358–4361, Nov. 1996

    work page 1996

  28. [28]

    Dynamics of lennard-jones clusters: A characterization of the activation-relaxation technique,

    R. Malek and N. Mousseau, “Dynamics of lennard-jones clusters: A characterization of the activation-relaxation technique,”Phys. Rev. E, vol. 62, pp. 7723–7728, Dec 2000

    work page 2000

  29. [29]

    The activation-relaxation technique: ART nouveau and kinetic ART,

    N. Mousseau, L. K. B´ eland, P. Brommer, J.-F. Joly, F. El-Mellouhi, E. Machado-Charry, M.-C. Marinica, and P. Pochet, “The activation-relaxation technique: ART nouveau and kinetic ART,”J. Atom. Molec. Opt. Phys., vol. 2012, pp. 1–14, 2012

    work page 2012

  30. [30]

    Finding reaction path- ways and transition states: r-ARTn and d-ARTn as an efficient and versatile alternative to string approaches,

    A. Jay, C. Huet, N. Salles, M. Gunde, L. Martin-Samos, N. Richard, G. Landa, V. Goiffon, S. De Gironcoli, A. H´ emeryck, and N. Mousseau, “Finding reaction path- ways and transition states: r-ARTn and d-ARTn as an efficient and versatile alternative to string approaches,” J. Chem. Theory Comput., vol. 16, no. 10, pp. 6726–6734, 2020

    work page 2020

  31. [31]

    Enthalpy-entropy compensation of atomic diffusion originates from softening of low frequency phonons,

    S. Gelin, A. Champagne-Ruel, and N. Mousseau, “Enthalpy-entropy compensation of atomic diffusion originates from softening of low frequency phonons,”Na- ture Communications, vol. 11, no. 1, p. 3977, 2020

    work page 2020

  32. [32]

    Information-driven inverse approach to disordered solids: Applications to amorphous silicon,

    D. K. Limbu, R. Atta-Fynn, D. A. Drabold, S. R. Elliott, and P. Biswas, “Information-driven inverse approach to disordered solids: Applications to amorphous silicon,” Physical Review Materials, vol. 2, p. 115602, Nov. 2018

    work page 2018

  33. [33]

    High resolution radial distribution function of pure amorphous silicon,

    K. Laaziri, S. Kycia, S. Roorda, M. Chicoine, J. L. Robertson, J. Wang, and S. C. Moss, “High resolution radial distribution function of pure amorphous silicon,” Phys. Rev. Lett., vol. 82, no. 17, pp. 3460–3463, 1999

    work page 1999

  34. [34]

    Den- sity of amorphous Si,

    J. S. Custer, M. O. Thompson, D. C. Jacobson, J. M. Poate, S. Roorda, W. C. Sinke, and F. Spaepen, “Den- sity of amorphous Si,”Applied Physics Letters, vol. 64, pp. 437–439, Jan. 1994

    work page 1994

  35. [35]

    Evo- lution of the potential-energy surface of amorphous sili- con,

    H. Kallel, N. Mousseau, and F. m. c. Schiettekatte, “Evo- lution of the potential-energy surface of amorphous sili- con,”Phys. Rev. Lett., vol. 105, p. 045503, Jul 2010

    work page 2010

  36. [36]

    Computer Gen- eration of Structural Models of Amorphous Si and Ge,

    F. Wooten, K. Winer, and D. Weaire, “Computer Gen- eration of Structural Models of Amorphous Si and Ge,” Physical Review Letters, vol. 54, pp. 1392–1395, Apr. 1985

    work page 1985

  37. [37]

    Identification of Re- laxation and Diffusion Mechanisms in Amorphous Sili- con,

    G. T. Barkema and N. Mousseau, “Identification of Re- laxation and Diffusion Mechanisms in Amorphous Sili- con,”Physical Review Letters, vol. 81, no. 9, 1998

    work page 1998

  38. [38]

    Energy landscape of relaxed amorphous silicon,

    F. Valiquette and N. Mousseau, “Energy landscape of relaxed amorphous silicon,”Phys. Rev. B, vol. 68, p. 125209, Sep 2003

    work page 2003

  39. [39]

    Connected network model for the mechanical loss of amorphous materials,

    S. Blaber, D. Bruns, and J. Rottler, “Connected network model for the mechanical loss of amorphous materials,” 2025

    work page 2025

  40. [40]

    Hydrogen-Free Amorphous Silicon with No Tunneling States,

    X. Liu, D. R. Queen, T. H. Metcalf, J. E. Karel, and F. Hellman, “Hydrogen-Free Amorphous Silicon with No Tunneling States,”Physical Review Letters, vol. 113, p. 025503, July 2014

    work page 2014