Finite-dimensional realization of overlap-phase coexistence, the Maxwell construction, and metastable-decay dynamics in the Franz-Parisi effective potential of glassy systems

The 1998 Franz-Parisi paper 'Effective potential in glassy systems: theory and simulations' introduced the replica-coupled overlap effective potential: coupling two real replicas of a glass former with an attractive field conjugate to their overlap $q$, it predicted at mean-field level (spherical $p$-spin) a first-order transition line between a confined (high-$q$) and a deconfined (low-$q$) phase in the $(\epsilon, T)$ plane, terminating in a glassy critical point, and it supported this picture by Monte Carlo simulations of binary soft-sphere mixtures. The paper left open what the picture means in finite dimensions: what spatial structures realize phase coexistence, whether the Maxwell construction used to infer the transition-line topology is microscopically justified in short-range systems, what the nature of the glassy critical point is, by what mechanism the dominant phase replaces the metastable one (its simulations showed bulk-like rather than surface-driven invasion), and how the transition line can be located precisely. Later work resolved part of this: field-theoretical and numerical studies establish random-field Ising universality for the glassy critical point in $d=3$ (and absence of a transition in $d=2$), and the $(\epsilon, T)$ phase diagram has been mapped numerically in several models. The surviving open core is the quantitative microscopic theory of overlap coexistence in finite dimensions - inhomogeneous overlap-field solutions, interface properties, the local effective potential, and the strength of the self-induced random field that would ground the Maxwell construction - together with the dynamical mechanisms of metastable-phase decay and ergodicity restoration, whose thermodynamic-limit interpretation is still contested in current simulations.

Background

When a liquid is supercooled below its melting point it can remain fluid for a very long time, but its structural relaxation time grows by many orders of magnitude until, at the laboratory glass transition temperature $T_g$, it no longer flows on experimental timescales. A central theoretical question is whether this slowdown is driven by an underlying thermodynamic transition. In the random first-order transition (RFOT) scenario, descending from mean-field models such as the spherical $p$-spin spin glass (a solvable model of continuously valued spins interacting in groups of $p$, the archetypal RFOT system), the number of metastable amorphous free-energy minima grows exponentially with the number of particles $N$, and their entropy density - the configurational entropy or complexity $\Sigma(T)$ - is expected to vanish at a Kauzmann temperature $T_K$, where an ideal glass transition would occur; above it, mode-coupling theory (MCT), an exact description of the dynamics in these infinite-range models, locates a purely dynamical transition at $T_c > T_K$.

Because individual amorphous states cannot be told apart by any ordinary order parameter, Franz and Parisi proposed to use the similarity between two copies ('real replicas') of the same system. The overlap $q(x, y)$ between a configuration $x$ and a reference configuration $y$ is the fraction of particles of $x$ lying within a short distance (of order a third of the interparticle spacing) of some particle of $y$. Adding to one replica's Hamiltonian an attractive coupling $-\epsilon q(x, y)$ (the 'quenched' construction, in which the reference is a thermalized configuration at a possibly different temperature $T'$; the 'annealed' construction treats the two copies on the same footing) defines an effective potential $V(q)$, the minimal work needed to hold one replica at fixed overlap with the other. In the spherical $p$-spin model, $V(q)$ develops two minima, and the extended phase diagram in the $(\epsilon, T)$ plane contains a first-order transition line between a 'confined' phase (high $q$: the constrained replica stays near the reference) and a 'deconfined' phase (low $q$: the constrained replica is statistically independent of the reference), ending at a critical point; the latent heat of the transition vanishes at the critical point and at $T = 0$, and at small coupling the line starts as $T(\epsilon) = T_c + c\,\epsilon$ in the quenched construction, while the annealed construction gives $T(\epsilon) = T_c + c\,\epsilon^{1/2}$. Since direct observation of first-order transitions is hampered by exponentially long metastable lifetimes, the 1998 paper used the Maxwell construction - the standard replacement of the non-convex part of a mean-field free energy by the straight tie line joining the coexisting phases - as a heuristic to infer the topology of the transition line in short-range systems.

Monte Carlo simulations of replica-coupled binary soft-sphere mixtures (66 and 512 particles) supported this picture, but the paper's conclusions explicitly acknowledged what remained open: the physical meaning of phase coexistence in finite dimensions ('we do not know what a supposed coexistence of the confined and the deconfined phase would mean'), the microscopic basis of the Maxwell construction in short-range glasses, the nature of the glassy critical point, the observed overlap dynamics - the dominant phase replaced the metastable one by a bulk-like mechanism rather than the surface-driven domain growth familiar from spinodal decomposition (the coarsening of two phases after a quench through a first-order transition) - and the need for far more extensive numerics to locate the transition line and test the small-$\epsilon$ scaling and latent-heat predictions. A key structural fact discovered later is that in a finite-dimensional system the fixed amorphous reference configuration acts on the overlap field as quenched, self-induced disorder, which connects the problem to the random-field Ising model (RFIM), the Ising magnet with quenched random local fields, whose critical behavior differs from that of the pure model (lower critical dimension 2, upper critical dimension 6, and perturbative dimensional reduction). This observation is the seed of the modern effective theory of the glass order parameter and of much of the subsequent literature reviewed here.

Problem Statement

The overarching objective, stated by the 1998 source paper and inherited by the modern research program it started, is to establish the finite-dimensional fate of the replica-coupled effective-potential picture of the glass transition in short-range glass formers: does the mean-field scenario - a first-order transition line between a confined (high-overlap) and a deconfined (low-overlap) phase in the $(\epsilon, T)$ plane, terminated by a glassy critical point, with the Maxwell construction giving the topology of the line - survive in real space, in what spatial form, and with what dynamical consequences?

Within this objective, later literature has already settled two of the issues raised in the source paper's conclusions: the glassy critical point's critical behavior lies in the random-field Ising model universality class (established by field theory and by finite-size-scaling simulations in $d=3$, with no transition in $d=2$), and the $(\epsilon, T)$ transition line has been located numerically in several three-dimensional models with quantified dependence on reference temperature and overlap tolerance. These resolved parts are recorded for context and need not be redone; the universality-class identification concerns the critical behavior of the transition as observed in finite-size simulations and is conditional on the first-order line and its critical point surviving the thermodynamic limit, whose status is itself contested and forms part of item (1) below. The surviving open core is:

(1) The microscopic theory of overlap-phase coexistence in finite dimensions. Construct and characterize the spatially inhomogeneous solutions of the overlap field theory for finite-range glassy models - droplets, interfaces, and coexisting domains of high and low overlap - and thereby determine what coexistence of the confined and deconfined phases means in real space. This requires computing the interface free energy (amorphous surface tension) and its fluctuations, the local Franz-Parisi potential and its spatial covariance, and the strength of the self-induced random field; from these, derive (or delimit) the validity of the Maxwell construction used to infer the transition-line topology, including the fate of the small-coupling predictions such as the quenched scaling $T(\epsilon) = T_c + c\,\epsilon$ and the latent-heat behavior, and the criterion under which the random field destroys the transition entirely. This includes settling the current disagreement on the thermodynamic-limit status of the first-order line and critical point (finite-size-scaling evidence for coexistence versus the finite-size-effect interpretation).

(2) The dynamical mechanisms. Identify how the dominant phase replaces the metastable high-$q$ phase under quenched and annealed protocols - in particular whether the mechanism is the bulk-like invasion observed in the 1998 simulations or surface nucleation and growth - together with the decay pathways of the metastable phase and the timescales on which ergodicity is restored when free-energy barriers are finite; and relate the static overlap correlation length to the slowdown of the unconstrained liquid.

The outcome is not assumed to be universal, so the class of glass formers (spatial dimension, interaction range, overlap tolerance) is itself part of the question.

References

1. Silvio Franz and Giorgio Parisi, 'Effective potential in glassy systems: theory and simulations', Physica A 261 (1998) 317-339, DOI 10.1016/S0378-4371(98)00315-X, https://doi.org/10.1016/S0378-4371(98)00315-X (source paper; preprint arXiv:cond-mat/9711215, https://arxiv.org/abs/cond-mat/9711215)
2. Silvio Franz and Giorgio Parisi, 'Phase diagram of coupled glassy systems: a mean-field study', Physical Review Letters 79 (1997) 2486-2489, DOI 10.1103/PhysRevLett.79.2486, https://doi.org/10.1103/PhysRevLett.79.2486
3. Miguel Cardenas, Silvio Franz and Giorgio Parisi, 'Constrained Boltzmann-Gibbs measures and effective potential for glasses in hypernetted chain approximation and numerical simulations', The Journal of Chemical Physics 110 (1999) 1726-1734, DOI 10.1063/1.478028, https://doi.org/10.1063/1.478028
4. Silvio Franz, 'First steps of a nucleation theory in disordered systems', Journal of Statistical Mechanics (2005) P04001, DOI 10.1088/1742-5468/2005/04/P04001, https://doi.org/10.1088/1742-5468/2005/04/P04001
5. M. A. Moore and J. Yeo, 'Thermodynamic glass transition in finite dimensions', Physical Review Letters 96 (2006) 095701, DOI 10.1103/PhysRevLett.96.095701, https://doi.org/10.1103/PhysRevLett.96.095701
6. Silvio Franz, Giorgio Parisi, Federico Ricci-Tersenghi and Tommaso Rizzo, 'Field theory of fluctuations in glasses', The European Physical Journal E 34 (2011) 102, DOI 10.1140/epje/i2011-11102-0, https://doi.org/10.1140/epje/i2011-11102-0
7. Silvio Franz, Giorgio Parisi and Federico Ricci-Tersenghi, 'Glassy critical points and the random field Ising model', Journal of Statistical Mechanics (2013) L02001, DOI 10.1088/1742-5468/2013/02/L02001, https://doi.org/10.1088/1742-5468/2013/02/L02001
8. Silvio Franz and Giorgio Parisi, 'Universality classes of critical points in constrained glasses', Journal of Statistical Mechanics (2013) P11012, DOI 10.1088/1742-5468/2013/11/P11012, https://doi.org/10.1088/1742-5468/2013/11/P11012
9. Giulio Biroli, Chiara Cammarota, Gilles Tarjus and Marco Tarzia, 'Random-field-like criticality in glass-forming liquids', Physical Review Letters 112 (2014) 175701, DOI 10.1103/PhysRevLett.112.175701, https://doi.org/10.1103/PhysRevLett.112.175701
10. Giulio Biroli, Chiara Cammarota, Gilles Tarjus and Marco Tarzia, 'Random-field Ising-like effective theory of the glass transition. I. Mean-field models', Physical Review B 98 (2018) 174205, DOI 10.1103/PhysRevB.98.174205, https://doi.org/10.1103/PhysRevB.98.174205
11. Giulio Biroli, Chiara Cammarota, Gilles Tarjus and Marco Tarzia, 'Random field Ising-like effective theory of the glass transition. II. Finite-dimensional models', Physical Review B 98 (2018) 174206, DOI 10.1103/PhysRevB.98.174206, https://doi.org/10.1103/PhysRevB.98.174206
12. Ludovic Berthier, 'Overlap fluctuations in glass-forming liquids', Physical Review E 88 (2013) 022313, DOI 10.1103/PhysRevE.88.022313, https://doi.org/10.1103/PhysRevE.88.022313
13. Juan P. Garrahan, 'Transition in coupled replicas may not imply a finite-temperature ideal glass transition in glass forming systems', Physical Review E 89 (2014) 030301, DOI 10.1103/PhysRevE.89.030301, https://doi.org/10.1103/PhysRevE.89.030301
14. Ludovic Berthier and Robert L. Jack, 'Evidence for a disordered critical point in a glass-forming liquid', Physical Review Letters 114 (2015) 205701, DOI 10.1103/PhysRevLett.114.205701, https://doi.org/10.1103/PhysRevLett.114.205701
15. Robert L. Jack and Juan P. Garrahan, 'Phase transition for quenched coupled replicas in a plaquette spin model of glasses', Physical Review Letters 116 (2016) 055702, DOI 10.1103/PhysRevLett.116.055702, https://doi.org/10.1103/PhysRevLett.116.055702
16. Andrea Ninarello, Ludovic Berthier and Daniele Coslovich, 'Structure and dynamics of coupled viscous liquids', Molecular Physics 113 (2015) 2707-2715, DOI 10.1080/00268976.2015.1039089, https://doi.org/10.1080/00268976.2015.1039089
17. Robert L. Jack and Ludovic Berthier, 'The melting of stable glasses is governed by nucleation-and-growth dynamics', The Journal of Chemical Physics 144 (2016) 244506, DOI 10.1063/1.4954327, https://doi.org/10.1063/1.4954327
18. Giulio Biroli, Charlotte Rulquin, Gilles Tarjus and Marco Tarzia, 'Role of fluctuations in the phase transitions of coupled plaquette spin models of glasses', SciPost Physics 1 (2016) 007, DOI 10.21468/SciPostPhys.1.1.007, https://doi.org/10.21468/SciPostPhys.1.1.007
19. Jean-Marc Bomont, Jean-Pierre Hansen and Giorgio Pastore, 'An investigation of the liquid to glass transition using integral equations for the pair structure of coupled replicae', The Journal of Chemical Physics 141 (2014) 174505, DOI 10.1063/1.4900774, https://doi.org/10.1063/1.4900774
20. Jean-Marc Bomont, Giorgio Pastore and Jean-Pierre Hansen, 'Coexistence of low and high overlap phases in a supercooled liquid: an integral equation investigation', The Journal of Chemical Physics 146 (2017) 114504, DOI 10.1063/1.4978499, https://doi.org/10.1063/1.4978499
21. Ludovic Berthier and Giulio Biroli, 'Theoretical perspective on the glass transition and amorphous materials', Reviews of Modern Physics 83 (2011) 587-645, DOI 10.1103/RevModPhys.83.587, https://doi.org/10.1103/RevModPhys.83.587
22. Benjamin Guiselin, Ludovic Berthier and Gilles Tarjus, 'Random-field Ising model criticality in a glass-forming liquid', Physical Review E 102 (2020) 042129, DOI 10.1103/PhysRevE.102.042129, https://doi.org/10.1103/PhysRevE.102.042129
23. Benjamin Guiselin, Gilles Tarjus and Ludovic Berthier, 'On the overlap between configurations in glassy liquids', The Journal of Chemical Physics 153 (2020) 224502, DOI 10.1063/5.0022614, https://doi.org/10.1063/5.0022614
24. Benjamin Guiselin, Ludovic Berthier and Gilles Tarjus, 'Statistical mechanics of coupled supercooled liquids in finite dimensions', SciPost Physics 12 (2022) 091, DOI 10.21468/SciPostPhys.12.3.091, https://doi.org/10.21468/SciPostPhys.12.3.091
25. Benjamin Guiselin, Gilles Tarjus and Ludovic Berthier, 'Static self-induced heterogeneity in glass-forming liquids: overlap as a microscope', The Journal of Chemical Physics 156 (2022) 194503, DOI 10.1063/5.0086517, https://doi.org/10.1063/5.0086517
26. Chiara Cammarota, Misaki Ozawa and Gilles Tarjus, 'The Kauzmann transition to an ideal glass phase', book chapter in 'Spin Glass Theory and Far Beyond: Replica Symmetry Breaking after 40 Years' (World Scientific), preprint arXiv:2209.11049 (2022), https://arxiv.org/abs/2209.11049
27. Giulio Biroli and Jean-Philippe Bouchaud, 'The RFOT theory of glasses: recent progress and open issues', Comptes Rendus Physique 24 (2024) 9-23, DOI 10.5802/crphys.136, https://doi.org/10.5802/crphys.136 (preprint arXiv:2208.05866, 2022, https://arxiv.org/abs/2208.05866)
28. Niklas Küchler and Jürgen Horbach, 'On the growing length scale in a replica-coupled glassforming liquid', preprint arXiv:2402.04205 (2024), https://arxiv.org/abs/2402.04205
29. Gerhard Jung, Misaki Ozawa, Giulio Biroli and Ludovic Berthier, 'Numerical investigation of the equilibrium Kauzmann transition in a two-dimensional atomistic glass', Proceedings of the National Academy of Sciences 123 (2026) e2612355123, DOI 10.1073/pnas.2612355123, https://doi.org/10.1073/pnas.2612355123 (preprint arXiv:2507.03590, 2025, https://arxiv.org/abs/2507.03590)
30. Ludovic Berthier and David R. Reichman, 'Modern computational studies of the glass transition', Nature Reviews Physics 5 (2023) 102-116, DOI 10.1038/s42254-022-00548-x, https://doi.org/10.1038/s42254-022-00548-x
31. Camille Scalliet, Benjamin Guiselin and Ludovic Berthier, 'Thirty milliseconds in the life of a supercooled liquid', Physical Review X 12 (2022) 041028, DOI 10.1103/PhysRevX.12.041028, https://doi.org/10.1103/PhysRevX.12.041028
32. Hiroshi Frusawa, 'Replica field theory for a generalized Franz-Parisi potential of inhomogeneous glassy systems: new closure and the associated self-consistent equation', Entropy 26 (2024) 241, DOI 10.3390/e26030241, https://doi.org/10.3390/e26030241