Simons Collaboration: Cracking the glass problem: PI Accomplishments
A major contribution this past year was our large collaborative project headed by Manning with Liu and Berthier (but including many more Collaboration Affiliates) to quantify correlations between structural defects and deformation in computer glasses. Understanding and predicting the dynamics of structural defects that give rise to deformation is the key to developing predictive theoretical models for flow and failure in glasses. They worked collaboratively to compare directly all of these proposed indicators on the same systems using the Swap Monte Carlo method (developed within this collaboration) to compare defects in computer glasses across the brittle to ductile transition. One major conclusion is that there are a subset of indicators that do robustly predict deformation across different materials preparation, while others perform surprisingly poorly. They further demonstrated that the density of defects changes as a function of material preparation and strain in a manner that is highly correlated with the macroscopic material response. Manning and collaborators also describe a new, highly effective method for identifying defects and is applicable to a broad range of (non-spherical) computer glasses.
Biroli and collaboration affiliates searched for the Gardner Transition in glycerol, a prototypical molecular glass by measuring the third harmonics susceptibility χ₃ down to 10 K. They found that χ₃ decreases by several orders of magnitude and becomes roughly constant suggesting that the physics is described by weakly-interacting localized excitations rationalizing the absence of the spin-glass Gardner phase. This hints that a Gardner phase may be suppressed in standard molecular glasses and suggests ways to favor its existence by changing the preparation protocol.
Biroli and other collaboration members studied dynamical instantons and activated processes in mean-field glass models by focusing on the energy landscape of a simple mean-field model of glasses. They showed that the dynamical mean-field equations admit two solutions: one corresponds to returning to the original minimum, and the other to reaching a new minimum past the barrier and characterize the properties of such minima reached by activated barrier-crossing.
Berthier and Biroli studied the role of fluctuations in the yielding transition of two-dimensional glasses. They provided strong evidence that stable glasses yield via a nonequilibrium discontinuous transition in the thermodynamic limit. A critical point separates this brittle yielding from the ductile one observed in less stable glasses.
Biroli, Charbonneau, Corwin, and Zamponi and collaborators reconciled percolation and glassiness in the random Lorentz gas (RLG), (a minimal model of transport in heterogeneous media). They found that mean field solutions fall in the same universality class as glasses, and that instantonic corrections destroy the glass transition in finite dimensions. This advance suggests that the RLG can be used as a toy model to develop a first-principles description of hopping in structural glasses. Charbonneau went onto study the high-dimensional void percolation criticality in random Lorentz gas and saw hints of mean-field caging.. In high-dimensional systems, he observed that the standard percolation physics is complemented by a dynamical slowdown of the tracer dynamics reminiscent of mean-field caging. Biroli, Charbonneau, Zamponi and affiliates of the Collaboration also studied the caging in a RLG and suggested that as d increases the behavior of the RLG converges to the glassy description, and that percolation physics is recovered thanks to finite-dimensional corrections. Comparing the dynamical solution with the mode- coupling theory results reveals that it does properly capture the discontinuous nature of the d => ∞ RLG.
Berthier and collaborators studied the random field Ising model criticality in a glass-forming liquid. They demonstrated a random-field Ising model critical point and a first-order transition line, in agreement with recent field-theory approaches. This study demonstrates random-field Ising criticality in the thermodynamic limit for a three-dimensional supercooled liquids at equilibrium.
Berthier and collaborators developed a strategy and numerical models for multi-component metallic glasses for which the swap Monte Carlo algorithm produces stable configurations equivalent to experimental systems cooled more than 107 times slower than in conventional simulations. This allows a deeper understanding of the properties of metallic glasses.
Nagel and his student have studyied multiple memory formation in glassy landscapes. As shown by Sastry, Liu and Nagel, cyclically sheared jammed packings form memories of the amplitude of training by falling into periodic orbits. Simple models that treat clusters of rearranging particles as isolated two-state systems fail to account for the long training times and multi-period orbits observed. Adding interactions between rearranging clusters overcomes these deficiencies and allow simultaneous encoding of multiple memories. These memories are different from those found in other systems and contain information about the strength of the interactions.
Disordered solids change their response as they age. Liu, Nagel and their joint postdoc studied how aging under an applied stress affects the nonlinear response. They could modify the elastic properties and proposed two models for the evolution that capture the dramatic effects on the material’s nonlinear elastic properties. They demonstrated that aging can be used to create complex behavior in the nonlinear regime and showed that by driving the system periodically, this plasticity can be exploited to train in desired properties, both in the global moduli and in local “allosteric” interactions. Periodic driving can couple an applied “source” strain to a “target” strain over a path in the energy landscape. This coupling allows control of the system’s response even at large strains well into the nonlinear regime, where it can be difficult to achieve control simply by design.
Charbonneau and his postdoc studies memory formation in jammed hard spheres. Liquids equilibrated below an onset density share similar inherent states, while above that density their inherent states markedly differ. By reassessing the onset they found it both thermodynamically and dimensionally robust and uncover a second type of memory associated with a Gardner-like change in behavior along the jamming algorithm.
Charbonneau, Corwin, Parisi and collaborators studied finite-size effects in the critical properties of jammed configurations. Jamming criticality defines a universality class that includes glasses and constraint-satisfaction problems with the interesting feature that small interparticle forces and gaps are distributed according to non-trivial power laws. Systems at the jamming point are thus only marginally stable. They conclude that gaps are correlated over considerably longer scales than forces. These results help to delineate the domain of the jamming universality class.
Berthier, Charbonneau and collaborators studied finite-dimensional vestige of spinodal criticality above the dynamical glass transition. The dynamical transition of glass-forming liquids is a spinodal instability preempted by thermally activated processes that also limit how close the instability can be approached. They use swap Monte Carlo to efficiently thermalize configurations beyond the mode-coupling crossover, and analyze their dynamics. They find strong softening of the mean-field singularity in d=3 that is progressively restored as d increases above d=8.
Wyart and collaborators studied an experimental system that allows a test of jamming theories of frictionless particles. They investigated a model frictionless granular layer flowing down an inclined plane. They found that (i) thin frictionless granular layers are devoid of hysteresis, yet the layer stability is increased as it gets thinner and that (ii) rheological flow rules can be collapsed into a unique master curve. This collapse supports that conclusion that the isostatic length governs the effect of boundaries on flow.
Manning also studied the nonlinear response of jammed packings. Previous work had demonstrated that the linear response of jammed packings exhibits finite-size scaling highlighting that the zero-pressure, hard- sphere limit is singular. In other words, soft spheres are fundamentally different in linear response compared to hard spheres. Her group demonstrated that something similar is true for the nonlinear response that characterizes the full potential energy landscape.
Manning, Corwin and collaborators have established a direct link between sheared systems and dense active matter, using simulations and mean field theory. Specifically, they showed that shear is just a special case of a more general random forcing field that is a special limit of self-propelled active matter simulations. This more general forcing is an exciting tool to probe the rheology of glasses, and will help develop predictive theories for the emerging field of dense active matter.
Berthier, Reichman and Zamponi and collaborators studied amorphous solids exhibiting low temperature anomalies whose origin has been ascribed to localized tunneling defects. They created in silico systems spanning from hyper-quenched to ultra-stable glasses. They located tunneling defects with energy splitting smaller than the temperature below which quantum effects are relevant. They found that as the stability of a glass increases, its energy landscape as well as the manner in which it is probed tend to deplete the density of tunneling defects, as observed in recent experiments. They explored the real-space nature of tunneling defects and found that they are mostly localized to a few atoms but are occasionally dramatically delocalized.
Franz and collaborators studied the critical energy landscape of linear soft spheres and showed that when overcompressed beyond the jamming point fall in an amorphous solid phase which is critical, mechanically marginally stable and share many features with the jamming point itself. The relevant minima of the energy display an isostatic contact. Excitations around such energy minima are non-linear, system spanning, and have a set of non-trivial critical exponents that with the critical exponents appearing at jamming. Therefore, linear soft spheres appear as a novel class of finite dimensional systems that self-organize into new, critical, marginally stable, states. Franz and collaborators also studied the gradient descent dynamics in the mixed p-spin spherical model. They studied the inherent structures. In large systems, the dynamics agrees with the mean-field dynamical equations. They confirmed the existence of an onset initial temperature, within the liquid phase where the inherent structures depend on initial temperature. These results strengthen the analogy between mean-field spin glass models and supercooled liquids.
Kurchan and his postdoc have focused on the subject of complex landscapes which has applications from computer science to biology. Recent interest has developed in landscapes for which the variables are complex because the “sign problem” is a major obstacle and can be attacked by deforming the sampling space into complex variables. In these cases, one needs to understand the location of saddle points, how they are connected, and typical questions of “complexity”. However, there are no studies extending the methods of the theory of complexity to complex variables. This work will widen our theoretical view of complexity in general. Kurchan and collaborator mapped out-of-equilibrium quantum models into equilibrium. They developed the duality approach for quantum systems in contact with a thermal bath. The method provides (a) a mapping of the original model to a simpler one, containing only a few particles and
(b) shows that any dynamic process of this kind with generic baths may be mapped onto one with equilibrium baths. As in the classical case, the whole construction becomes intelligible by considering the dynamical symmetries of the problem.
Wyart and collaborators also proposed a thermal origin of the quasilocalized excitations in glasses. Key aspects of glasses are controlled by the presence of excitations in which a group of particles can rearrange. Observations indicate that their density is dramatically reduced and their size decreases as the liquid temperature is lowered. They proposed an interpretation of mean-field theories of the glass transition that unifies disparate proposals in which the modes beyond the gap act as an excitation reservoir, from which a pseudogap distribution is populated with its magnitude rapidly decreasing at lower T. This picture helps unify rarefaction as well as the decreasing size of excitations upon cooling, together with a stringlike relaxation occurring near the glass transition.
Personnel working on Cracking the Glass Problem
PI: Ludovic Berthier
Postdocs: Yoshihiko Nishikawa, Anshul Parmar
Graduate Student: Benjamin Guiselin, Yann-Edwin Keta
Affiliates: Atsushi Ikeda
PI: Giulio Biroli
Postdocs: D. Facoetti, Giampaolo Folena, Chen Liu, Misaki Ozawa, Franco Pellegrini
Graduate Students: Stéphane d’Ascoli, Felix Roy
Affiliate: Chiara Cammarota
PI: Patrick Charbonneau
Postdocs: Peter Morse
Graduate Student: Yi Hu
Affiliate: Grzegorz Szamel
PI: Eric Corwin
Postdoc: Andrew Hammond
Graduate Students: Francesco Arceri, Cameron Dennis, James Sartor
PI: Silvio Franz
Graduate Students: Giampaolo Folena, Antonio Slocchi
Affiliate: Srikanth Sastry
PI: Jorge Kurchan
Postdoc: Jaron Kent-Dobias
Affiliate: Gilles Tarjus
PI: Andrea Liu
Postdocs: Miguel Ruiz-Garcia, Ge Zhang
Graduate Students: Ian Graham, Cathy Li, Sean Ridout, Jason Rocks
PI: Giulio Biroli
Postdoc: David Richard, Sudeshna Roy
Graduate Student: Ethan Stanifer
Affiliate: Edan Lerner
PI: Sidney Nagel
Graduate Student: Chloe Lindeman, Nidhi Pashine
Administrative assistant: Zena Anderson
PI: Giorgio Parisi
Postdocs: Thibauld Lesieur, Gabriele Sicuro, Georgios Tsekenis
Affiliate: Tommaso Rizzo
PI: David Reichman
Postdocs: Chitrak Bhadra, D. Facoetti, Dmytro Khomenko
Affiliate: Kunimasa Miyazaki
PI: Matthieu Wyart
Post-docs: Francesco Cagnetta, Marko Popovic, Stefano Spigler
Graduate Students: Mario Geiger Wencheng Jie, Jonas Paccolat, Leonardo Petrini, Riccardo Ravasio
Affiliates: Carolina Brito, Markus Müller
PI: Francesco Zamponi
Postdocs: Simone Ciarella, Harukuni Ikeda, Dmytro Khomenko, Alessandro Manacorda, Felix- Cosmin Mocanu, Anna Paola Muntoni
Affiliate: Hajime Yoshino
Visitors: Esssen Tjhung
University of Chicago Account:
Shared Postdocs: Horst-Holger Boltz (joint Kurchan/Liu), Rahul Chacko (Liu), Davide Facoetti (Biroli/Kurchan/Zamponi), Varda Faghir Hagh (joint Corwin/Manning/Nagel)
Events sponsored by the Simons Collaboration: Cracking the Glass Problem
Due to the pandemic, the workshops and conferences that were scheduled by the collaboration were cancelled starting in March 2020. There was one conference at Les Houches on “Recent progress in glassy systems”, which took place in late February, that was held before all travel was restricted. The details of that conference is given below.
The activities of the Collaboration were therefore transferred to a virtual format. Three notable events have taken place.
We were able to hold our annual Simons Collaboration Meeting in a three-day format on December 2 – 4, 2020. The meeting allowed for a series of talks and related breakout discussions on each topic. There was also a very lively Poster Session in which the students could present their most recent work. The talks were delivered via Zoom. What we found worked out exceedingly well was the use of the platform “gather.town” for the Discussion sessions and for the presentation of posters. This technology, new to most of us, allowed spontaneous, one-on-one discussions between groups of individuals in virtual corridors. This had much the feel of a real in-person conference.
Collaboration members also organized the CECAM (Centre Européen de Calcul Atomique et Moléculaire) workshop on January 6-8, 2021. The subject of this workshop was “Recent Advances on the Glass Problem”. This workshop also made use of the technology that we found worked so well for our Annual Meeting.
An ongoing event that was started in the midst of the pandemic was a webinar series organized by our Simons Collaboration. This started out as an ordinary seminar series in Paris that, because of social- distancing rules, was delivered over the internet. However, the graduate students and postdocs realized that this was a great opportunity to involve the entire collaboration. The talks are now given every week on Wednesdays (mornings in the US and afternoons in Europe). Every other week, the PIs of the collaboration give an internal seminar on what they are doing and what they are interested in pursuing within the Collaboration. There are not open to the general public so that there can be a free exchange of ideas. On the alternate weeks, the seminars are public and Collaborators, Affiliates and external speakers give talks related to different aspects of the glass problem. These have been widely attended by investigators from three continents – something that would not have been possible if the seminars had been held in person.
February 16 – 21, 2020 Les Houches School of Physics
Recent progress in glassy systems
Marginal Phases in Structural Glasses: P. Urbani
Excitations in Low-Temperature Glasses: E. Lerner
Quantum Disordered Systems: A. Rosso
Mathematical Methods for Diluted Mean-Field Models: A. Coja-Oglan
Theory of Machine Learning: L. Zdeborova
- E. Agoritsas,
- J. Barbier,
- P. Charbonneau,
- E. De Giuli,
- H. Ikeda,
- M. Muller,
- S. Pascazio,
- A. Sclocchi,
- G. Semerjian,
- B. Seoane,
- G. Sicuro,
- S. Spigler,
- M. Tarzia
December 2 – 4, 2020 Annual Collaboration Meeting
Organizers: L. Manning, P. Charbonneau, F. Zamponi, S. Franz, D. Reichman
Day 1: Welcome: Lisa Manning
Designing disordered systems with specific properties
- Introduction: Andrea Liu
- Carolina Brito
- Daniel Hexner
- Nidhi Pashine
- Applications to Biological Systems: Beatriz Seoane, Ojan Damavandi
- Relevance to Machine Learning: Chiara Cammarota, Marco Baity-Jesi
- Implications for glassy/jammed/rough-landscape systems: Varda Hagh, Atsushi Ikeda
- New Systems and Functionalities: Nidhi Pashine, Sri Sastry
Day 2: Dynamics and the future of MFT
- Introduction: Giulio Biroli
- Grzegorz Szamel
- Chen Liu
- Yi Hu
- Out-of-equilibrium dynamics: Elisabeth Agoritsas and Grzegorz Szamel
- DMFT + MCT: David Reichman and Simone Ciarella
- Instantons: Valentina Ros and Giampaolo Folena
- Onset: Silvio Franz and Peter Morse
Day 3: Deforming and exciting glasses at low temperature: rheology and role of two-level systems
- Introduction: Ludovic Berthier
- Elisabeth Agoritsas
- David Richard
- Francesco Arceri
December 6 – 8, 2021 CECAM Flagship Workshop: Recent advances on the glass problem
- Elisabeth Agoritsas (EPFL)
- Ludovic Berthier (CNRS)
- Patrick Charbonneau (Duke University)
- Francesco Zamponi (ENS Paris)
Day 1: Activated dynamics.
Speakers: V. Ros, G. Szamel, C. Scalliet, P. Royall; chair: E. Agoritsas.
Day 2: Exploration and characterisation of the potential energy landscape.
Speakers: V. Hagh, P. Morse, P. Sollich, Y. Nishikawa; chair: F. Zamponi.
Day 3: Machine learning in glasses.
Speakers: S. Ridout, V. Bapst, L. Filion, D. Coslovich; chair: L. Berthier.
September 16, 2020 to Present
Weekly webinar series: Alternate internal and external seminars on various topics related to the glass problem. As stated above, this webinar grew out of a traditional seminar series in Paris. It now reaches investigators on three continents.
Honors to Simon Foundation Collaborators
Awarded the 2021 Wolf Prize in Physics for “ground-breaking discoveries in disordered systems, particle physics and statistical physics”.
Named member of American Philosophical Society.
Received the L’Oréal-UNESCO Young Talent fellowship for Women in Science programme 2020.
In the coming year, the Collaboration will build on the accomplishments of the past year in several exciting new directions. It will simultaneously continue to accomplish the goals that were laid out in the original proposal: to develop a unified theory of structure and excitations in glassy matter and to develop a theory for the relaxation dynamics upon approaching the glass transition. Here we will indicate only a few of the new directions that we plan to follow.
Training for stability in a background of marginal stability: Minimization of a function in a complex landscape to find low-lying states is a daunting problem that lies at the heart of many statistical physics problems ranging from machine learning to constraint satisfaction problems. Perhaps the most familiar and readily accessible example of such a situation is in finding the particularly stable configurations in a disordered packing of spherical particles – the very problem of finding the glassy ground states that our Collaboration has been addressing over the past few years. Typically, in the thermodynamic limit, a jammed system is close to the threshold for instability; that is, there is a nearly vanishing potential energy barrier that allows the system, if perturbed, to escape into a nearby alternative ground state. Such systems are only marginally stable — it only takes an infinitesimal perturbation to rearrange the packing into a new ground state. The pioneering work on Swap Monte Carlo algorithms developed within our Collaboration has given new ability how to thermalize a system to ultra-low temperatures. We are now looking for a connection of those ideas with another field developed within the Collaboration that involves training a system by adding and then freezing degrees of freedom to the system.
Model systems: Many scientific advances can be achieved by extracting the most elemental aspects of a physics problem. Over the past year the Collaboration has made substantial inroads into the study of the Random Lorentz Gas which is a minimal model of transport in heterogeneous media. Our Collaboration found that the mean-field solutions fall in the same universality class as glasses, and that instantonic corrections destroy the glass transition in finite dimensions. Our current goal for the coming year is to continue this advance and use the Random Lorentz Gas as a starting point for a first-principles description of hopping processes in structural glasses. This is one particular area that we will pursue.
Rheology: Understanding the dynamics of structural defects that give rise to deformation is the key to developing predictive theoretical models for flow and failure in glasses. Key aspects of glasses are controlled by the presence of excitations in which a group of particles can rearrange. Observations indicate that their density is dramatically reduced and their size decreases as the liquid temperature is lowered. Our Collaboration has been making important strides in predicting where these region will occur by using Machine Learning algorithms. In the coming year, the Collaboration will continue to apply such algorithms, not only to find these entities but also to understand, once the rearranging clusters are known, how they affect the various plastic flows and excitations in the material.
Low-temperature properties: Our collaboration created in silico systems over an exceedingly wide range of stability: from hyper-quenched to ultra-stable glasses. They located tunneling defects with energy splitting in the range where quantum effects are relevant. The Collaboration will continue this exciting work that has helped to put on firm theoretical footing this fundamental phenomenon that has resisted systematic attack for nearly half a century. Work remains to show how these results are applicable to other glassy systems.
Dynamical mean-field theory and aging dynamics
A Altieri, G Biroli, C Cammarota — J. Phys. A: Math. Theor. 53, 375006 (2020)
Periodic training of creeping solids
D Hexner, AJ Liu, SR Nagel — Proc. Nat. Acad. Sciences USA 117, 31690 – 31695 (2020).
Critical energy landscape of linear soft spheres
S Franz, A Sclocchi, P Urbani — SciPost Phys. 9, 012 (2020)
Large deviations of glassy effective potentials
S Franz, J Rocchi — J. Phys. A: Math. Theor. 53, 485002 (2020)
Solvable Models of Supercooled Liquids in Three Dimensions
T Rizzo, T Voigtmann — Phys. Rev. Lett. 124, 195501 (2020)
Depletion of two-level systems in ultrastable computer-generated glasses
D Khomenko*, C Scalliet*, L Berthier, DR Reichman, F Zamponi — Phys. Rev. Lett. 124, 225901 (2020)
– Editor’s Suggestion
Rethinking Mean-Field Glassy Dynamics and Its Relation with the Energy Landscape: The Surprising Case of the Spherical Mixed p-Spin Model
G Folena, S Franz, F Ricci-Tersenghi — Phys. Rev. X 10(3), 031045 (2020)
Effective Trap-like Activated Dynamics in a Continuous Landscape
MR Carbone, V Astuti, M Baity-Jesi — Phys. Rev. E 101, 052304 (2020)
Postponing the dynamical transition density using competing interactions
P Charbonneau, J Kundu — Granul. Matter 22:55 (2020)
Direct Measurement of Force Configurational Entropy in Jamming
JD Sartor, EI Corwin — Phys. Rev. E 101, 050902(R) (2020)
Vibrational properties of hard and soft spheres are unified at jamming
F Arceri, EI Corwin — Phys. Rev. Lett. 124(23), 238002 (2020)
Low-frequency vibrations of jammed packings in large spatial dimensions
M Shimada, H Mizuno, L Berthier, A Ikeda — Phys. Rev. E 101, 052906 (2020)
A solid-to-fluid transition is predicted by cell shape and alignment in an anisotropic tissue of the developing fly embryo
X Wang*, M Merkel*, LB Sutter*, G Erdemci-Tandogan, ML Manning, KE Kasza — PNAS 117(24), 13541-13551 (2020)
Role of fluctuations in the yielding transition of two-dimensional glasses
M Ozawa, L Berthier, G Biroli, and G Tarjus — Phys. Rev. Research 2, 023203 (2020)
An exploratory study of the glassy landscape near jamming
C Artiaco, P Baldan and G Parisi — Phys. Rev. E 101, 052605 (2020)
Effect of aging on the non-linear elasticity and memory formation in materials
D Hexner, N Pashine, AJ Liu, SR Nagel — Phys. Rev. Research 2, 043231 (2020)
Finite-dimensional vestige of spinodal criticality above the dynamical glass transition
L Berthier, P Charbonneau, J Kundu — Phys. Rev. Lett. 125, 108001 (2020)
Universality of the nonphononic vibrational spectrum across different classes of computer glasses
KG Lopez, D Richard, G Kapteijns, R Pater, T Vaknin, E Bouchbinder, E Lerner — Phys. Rev. Lett. 125, 085502 (2020)
Quasi-integrable systems are slow to thermalize but may be good scramblers
T Goldfriend, J Kurchan — Phys. Rev. E 102, 022201 (2020)
Random-field Ising model criticality in a glass-forming liquid
B Guiselin, L Berthier, G Tarjus — Phys. Rev. E 102, 042129 (2020)
Glass stability changes the nature of yielding under oscillatory shear
W Yeh, M Ozawa, K Miyazaki, T Kawasaki, L Berthier — Phys. Rev. Lett. 124, 220502 (2020)
Ultrastable metallic glasses in silico
A D S Parmar, M Ozawa, L Berthier — Phys. Rev. Lett. 125, 085505 (2020)
Stable glassy configurations of the Kob-Andersen model using swap Monte Carlo
A D S Parmar, B Guiselin, L Berthier — J. Chem. Phys. 153, 134505 (2020)
How to “measure” a structural relaxation time that is too long to be measured?
L Berthier, M Ediger — J. Chem. Phys. 153, 44501 (2020)
Two classes of events in sheared particulate matter
P Morse, M van Deen, S Wijtmans, M Van Hecke, ML Manning — Phys. Rev. Research 2, 023179 (2020)
Brittle yielding of amorphous solids at finite shear rates
M Singh, M Ozawa, L Berthier — Phys. Rev. Materials 4, 025603 (2020)
Attractive versus truncated repulsive supercooled liquids: The dynamics is encoded in the pair correlation function
FP Landes, G Biroli, O Dauchot, AJ Liu, DR Reichman – Phys. Rev. E 101 (1), 010602 (2020)
Linear and nonlinear mechanical responses can be quite different in models for biological tissues
P Sahu*, J Kang*, G Erdemci-Tandogan, ML Manning – Soft Matter 16, 1850-1856 (2020)
Universal relaxation dynamics of sphere packings below jamming
A Ikeda, T Kawasaki, L Berthier, K Saitoh, T Hatano – Phys. Rev. Lett. 124, 058001 (2020)
Scaling description of generalization with number of parameters in deep learning
M Geiger, A Jacot, S Spigler, F Gabriel, L Sagun, S d’Ascoli, G Biroli, C Hongler, M Wyart – Journal of Statistical Mechanics: Theory and Experiment 2020, 023401 (2020)
Experimental observation of the marginal glass phase in a colloidal glass
AP Hammond, EI Corwin – PNAS 117 (11) 5714-5718 (2020)
The Jamming Energy Landscape is Hierarchical and Ultrametric
RC Dennis, EI Corwin – Phys. Rev. Lett. 124, 078002 (2020)
Brittle yielding of amorphous solids at finite shear rates
M Singh, M Ozawa, L Berthier – Phys. Rev. Mater. 4, 025603 (2020)
Small-scale demixing in confluent biological tissues
P Sahu, DM Sussman, M Rübsam, AF Mertz, V Horsley, ER Dufresne, CM Niessen, MC Marchetti, ML Manning, JM Schwarz – Soft Matter 16, 3325-3337 (2020)
Distribution of rare saddles in the p-spin energy landscape
V Ros – J. Phys A: Math Theor. 53 (12), 125002 (2020)
Marvels and pitfalls of the Langevin algorithm in noisy high-dimensional inference
SS Mannelli, G Biroli, C Cammarota, F Krzakala, P Urbani, L Zdeborová – Phys. Rev. X 10 (1), 011057 (2020)
Random-link matching problems on random regular graphs
G Parisi, G Perrupato, G Sicuro – Journal of Statistical Mechanics 2020, 033301 (2020)
Infinitesimal asphericity changes the universality of the jamming transition
H Ikeda, C Brito, M Wyart – Journal of Statistical Mechanics: Theory and Experiment 2020(3), 033302 (2020)
Nonlinear quasilocalized excitations in glasses: True representatives of soft spots
G Kapteijns, D Richard, and E Lerner – Phys. Rev. E 101, 032130 (2020)
Machine Learning Glasses
G Biroli – Nature Physics 16(4) 373 (2020)
How to iron out rough landscapes and get optimal performances: Averaged Gradient Descent and its application to tensor PCA
G Biroli, C Cammarota, F Ricci-Tersenghi – Journal of Physics A: Mathematical and Theoretical 53 174003 (2020)
From complex to simple: hierarchical free-energy landscape renormalized in deep neural networks
H Yoshino – Sci Post Phys. Core 2, 005 (2020)
Lay Abstract of year’s achievements
Crystals and glasses lie at the opposite poles of disorder. While much of solid-state physics has been concerned with understanding the physics of crystals, the other pole – where disorder plays the essential role for organization and where far-from-equilibrium effects are paramount – is much more subtle and difficult to understand. Glasses are the quintessential far-from-equilibrium material and have become the paradigm for studying many aspects of nature that include not only disordered matter but also far-from- equilibrium behavior. As our Collaboration has emphasized from the outset, understanding glasses requires a novel understanding of systems that have an extensive number of ground. Because of its importance for understanding disordered and far-from-equilibrium matter, the understanding of glasses has ramifications throughout nearly all of science. This Collaboration has been assembled with the study of glasses and jammed materials as its focus. Over the past year, several important developments have occurred.
Rheology: Understanding and predicting the dynamics of structural defects that give rise to dynamical deformation is the key to developing predictive theoretical models for flow and failure in glasses. For this reason, the Collaboration has put together a large group of researchers to quantify correlations between structural defects and deformation in computer glasses created with the Swap Monte Carlo method (developed within this collaboration) to compare defects in computer glasses across the brittle to ductile transition. A subset of indicators robustly predict deformation across different materials preparation. The Collaboration further demonstrated that the density of defects changes as a function of material preparation and strain in a manner that is highly correlated with the macroscopic material response.
Memory formation: Aging can only occur in systems that are not in their ordered, equilibrium state. However disordered solids change their response as they age. They thus retain a memory of how they have been manipulated. One focus of the Collaboration has been to study what kinds of memory can be stored and read out. This can occur in a wide variety of ways which have significant consequences for the behavior and functional properties that can be designed into a material. The process of directed aging is way that materials can be aged under external strains to achieve otherwise difficult-to-obtain function.
Model Systems: Over the past year the Collaboration has made substantial inroads into the study of the Random Lorentz Gas which is a minimal model of transport in heterogeneous media. Our Collaboration found that the mean-field solutions fall in the same universality class as glasses so that it can be considered as a particularly good starting point to study the nature of the excitations in real glassy systems. The Collaboration also found that instantonic corrections destroy the glass transition in finite dimensions. This provides a solid starting point for investigating the nature of dynamics in these systems. Another important model is a particulate sample composed of soft spheres interacting with a linear potential energy. The Collaboration showed that when overcompressed beyond the jamming point, this system falls in an amorphous solid phase which is critical, mechanically marginally stable and share many features with the jamming point itself. The relevant minima of the energy display an isostatic contact network.
Low-temperature properties: Quantum-mechanical two-level excitations are believed to cause the excess excitations in glasses at low temperature. Our collaboration simulated glassy systems over a broad range of properties: from hyper-quenched to ultra-stable glasses. This was only possible because of the development of Swap Monte Carlo simulations by members of our Collaboration that achieve thermalization at exceedingly low temperatures. They located tunneling defects with energy splitting in the range where quantum effects are relevant. This allows an unprecedented systematic study of the nature of these otherwise elusive excitations in disordered matter.
Public Lecture (Wednesday, March 10)
Director of Research, French National Center for Scientific Research (CNRS)
Laboratoire Charles Coulomb
In this lecture, Ludovic Berthier will describe the world of amorphous materials and why scientists are interested in predicting and tuning their physical properties. The quest for a fundamental understanding of disordered materials has repeatedly led to the idea of an ‘ideal’ glassy state or an ‘optimal’ random structure. Modern theories of disordered systems offer refined descriptions of these elusive concepts but have also raised difficult questions about their applicability for real systems. In the last few years, novel experimental and computational techniques have been developed that bring actual materials much closer to the ideal glass state. This work reveals unexpected physical properties and sheds new light on the very concept of an ideal glass state of matter.
- Talk Slides