betaFeedback

WorksMathematics

An entropic barrier to equilibration in the discrete nonlinear Schrödinger equation: Rigorous bounds for the microcanonical Brownian motion

Abstract

The relaxation of the discrete nonlinear Schrödinger equation is known to slow down dramatically when the energy density ε\varepsilon exceeds a critical value εc\varepsilon_c, where the system enters a condensed phase. Replacing the Hamiltonian dynamics by Brownian motion on the microcanonical surface of NN sites at fixed mass and energy, we prove rigorous upper and lower bounds on the equilibration time τeqτ_{\mathrm{eq}}, which is governed by the first non-zero eigenvalue λ1λ_1 of the Laplacian on this surface. In the condensed phase the equilibration time is stretched exponential, not exponential, in the system size. The upper bounds on λ1λ_1 come from trial functions: of one occupation for ε<εc\varepsilon<\varepsilon_c, where the problem reduces to a Sturm-Liouville equation, and of two occupations for ε>εc\varepsilon>\varepsilon_c, where they cross the bottleneck through which the condensate moves between sites. The lower bounds on λ1λ_1 come from a Bochner-Reilly argument, with a vector field that compensates the concavity of the boundary. As a result, τeq=O(1)τ_{\mathrm{eq}}=O(1) for ε<εc\varepsilon<\varepsilon_c, while for ε>εc\varepsilon>\varepsilon_c, asymptotically in NN, (2−1)ε−εc≤ln⁡τeq/N≤ε−εc(\sqrt{2}-1)\sqrt{\varepsilon-\varepsilon_c}\leq\lnτ_{\mathrm{eq}}/\sqrt{N}\leq\sqrt{\varepsilon-\varepsilon_c}. We conjecture that the left-hand bound is sharp and the right-hand one is not.

Introduction

The success of thermodynamics in describing the macroscopic world is based on the fact that most laboratory systems achieve equilibrium states in a reasonably short time. When the microscopic laws of motion are known and trusted, time averages of observables in or about equilibrium states are described by the appropriate distributions dictated by the statistical mechanics of the corresponding microscopic problem. Such substitution is based on the ergodic hypothesis, which allows one to replace time averages of observables (to wash out microscopic fluctuations) to ensemble averages. This intuition, due to Boltzmann [1] for classical dynamics, and ported by Von Neumann for quantum dynamics [2], served so well to understand the physical world, that it became every physicist’s intuition when approaching the study of a non-exactly solvable (non-integrable) model made of many particles. It is then surprising when the ergodic hypothesis fails and, as everything which is counterintuitive, it is worth a better look because it could imply the necessity for a paradigm shift in the theory [3].

The failure modes of the ergodic hypothesis almost always involve a timescale which diverge very fast with the system size. Therefore, one can build a macroscopic timescale from a microscopic timescale and the inversion of the infinite-time and the infinite-volume limits (also known as thermodynamic limit) implicitly assumed in the ergodic hypothesis, is impossible.

Sometimes the divergence is understood as spontaneous symmetry breaking: a ferromagnet magnetises in one direction although its Hamiltonian is rotation invariant. Sometimes the broken symmetry is not apparent in the microscopic Hamiltonian, as in mean-field spin glasses [4, 5] or in Anderson [6, 7] and many-body localization [8, 10] [9], where it is revealed only after a rewriting of the quenched average of the partition function [11, 12].

Ergodicity breaking therefore, for the purpose of our discussion, arises from the birth of a timescale for equilibration, τeq\tau_{\mathrm{eq}}, which diverges fast, say faster than any polynomial, in the thermodynamic limit. Typically, a polynomial divergence is confined to scale-invariant critical points (critical slowing down) or is due to hydrodynamic modes (Thouless time) . Ergodicity breaking timescales are, instead, fatser than those. However, they are difficult to control quantitatively and/or analytically. For example, a divergence of the form τeq∼ecN\tau_{\mathrm{eq}} \sim e^{cN} can be easily mistaken, in numerical experiments, for a large polynomial τeq∼Na\tau_{\mathrm{eq}} \sim N^{a}, with a∼10a \sim10 or for a stretched exponential τeq∼eNδ\tau_{\mathrm{eq}} \sim e^{N^{\delta}} (with δ<1\delta< 1), although these laws might have different physical mechanisms at their origins. A time NaN^{a} with a≫1a \gg1 is usually due, for example, to subdiffusion [13] and we do not call a subdiffusive system non-ergodic.

It is then of utmost importance to have firm, analytical results on these timescales. In this paper, we plan to do exactly that, leveraging on a series of results on the breakdown of ergodicity occurring in the discrete, non-linear Schrödinger equation (DNLSE) [14, 15, 16] and on a series of results of the eigenvalues of the Laplacian operator on large-dimensional manifolds.

The paper is organized as follows: We first present the model, the approximations and the theorems characterizing the geometry of the microcanonical surface. These are known from previous works [14, 15, 17], and are collected in Section III. The transition at ε=εc\varepsilon= \varepsilon_{\mathrm{c}} is a dynamical transition [14, 18], where the manifold gets separated into NN distinct symmetry sectors, and the junctions between them have small volume and sectional area. Traversing from one sector, where a certain charge qiq_i is large, to another, takes time which one can link to the inverse of the smallest, non-zero eigenvalue of the Laplacian on the manifold, call it λ1\lambda_1. λ0=0\lambda_0 = 0 is the “trivial” equilibrium state eigenvalue.

This leads to the second part of the paper, where we estimate λ1\lambda_1 from above and below. To do this we need two different techniques. Upper bounds are simpler, since they are variational in nature. We can find a set of functions on the manifold on which to calculate λ1\lambda_1 and this gives a proper upper bound. The optimization in a class of problems is connected to the solution of a problem of supersymmetric quantum mechanics a la Witten [19]. We do this in Section IV.

Lower bounds are more complicated, since one has to find an inequality valid for all functions on the manifold. This is obtained by a completely different method which takes the name of Bochner [20] and Reilly [21], and is explained in detail in Sections V–VII.

The model and the results

The DNLSE is a classical problem, where the Hamiltonian of a complex field is taken as

H=−κ∑⟨i,j⟩∈G(ψi∗ψj+ψj∗ψi)+∑i=1N∣ψi∣4(1)H=-\kappa\sum_{\langle i,j\rangle\in G}\left(\psi_i^*\psi_j+\psi_j^*\psi_i\right)+\sum_{i=1}^{N}|\psi_i|^4 \tag*{(1)}

where ⟨i,j⟩∈G\langle i,j\rangle\in G is a graph over NN sites, and the canonical Poisson relations {ψi∗,ψj}=iδij\{\psi_i^*,\psi_j\}=i\delta_{ij} dictate the dynamics. Equivalently, separating ψj=(xj+ipj)/2\psi_j=(x_j+ip_j)/\sqrt{2} one could write the Hamiltonian in position-momentum variables (x,p)(x,p), while writing in the “charge-phase/action-angle” variables ψj=qjeiϕj\psi_j=\sqrt{q_j}e^{i\phi_j} we connect to the problem of Josephson junction arrays.

The dynamics conserves both the energy and the total charge Q=∑i∣ψi∣2=∑iqiQ=\sum_i|\psi_i|^2=\sum_iq_i. Localization in the DNLSE consists in the following fact: For particular initial conditions, when the energy density E/N=εE/N=\varepsilon is higher than a critical threshold εc\varepsilon_c, the energy tends to concentrate in a single oscillator for a very long time [22, 23]. Numerical calculations often fail to see the decay of such soliton [18, 24], the question arises then, what dictates the timescale of such decay.

If the kinetic energy term gets substituted by a random walk on the resulting microcanonical surface MN,εM_{N,\varepsilon} of constant Q=∑iqiQ=\sum_iq_i and E=εN=∑i∣ψi∣4=∑iqi2E=\varepsilon N=\sum_i|\psi_i|^4=\sum_iq_i^2 and qi≥0q_i\geq0 (with reflecting boundary conditions and natural measure μ\mu) most of the statistical properties of the original NLSE are conserved [25, 26]. In particular, the Brownian motion on the microcanonical surface shows the phenomenon of localization present in the original model [15] in all its qualitative and quantitative aspects, as explained in statistical mechanics terms [14, 16, 27]. Notice that, in this Brownian motion dynamics, the graph GG disappears from the problem.

Two comments on the choice of the Brownian dynamics, which we will expand a bit in the Conclusion. The metric on MN,εM_{N,\varepsilon} is a modelling choice: other reversible dynamics with the same equilibrium measure share all the static properties discussed below, but not necessarily the subleading terms of the time scales (see the Conclusions). All the results of this paper concern this Brownian motion. Their transfer to the Hamiltonian dynamics of the DNLSE rests on this substitution, which is yet to be proven.

In this paper we prove that the Brownian motion equilibration timescales behave in two different ways, depending we are in the normal or condensed phase, namely ε<2\varepsilon< 2 or ε>2\varepsilon> 2. First of all, however, let us define the equilibration time.

Let Pt=etΔM/2P_t = e^{t\Delta_M/2} be the semigroup of the Brownian motion on MN,εM_{N,\varepsilon} with reflecting conditions. −12ΔM-\frac{1}{2}\Delta_M is self-adjoint and non-negative on L2(μ)L^2(\mu), with spectral gap λ1/2\lambda_1/2, where λ1\lambda_1 is the first non-zero Neumann eigenvalue of −ΔM-\Delta_M, eq. (9) below.

The relaxation (or equilibration) time of the Brownian motion on MN,εM_{N,\varepsilon} is

τeq=2λ1.(2)\tau_{\mathrm{eq}} = \frac{2}{\lambda_1}. \tag*{(2)}

The ground state is the constant function ρeq≡1\rho_{\mathrm{eq}} \equiv1 on MM, with λ0=0\lambda_0 = 0.

The relaxation time is characterised by the following, which follows from the spectral theorem by expanding in eigenfunctions of −ΔM-\Delta_M. For every f∈L2(μ)f \in L^2(\mu) and every t≥0t \ge0,

∥Ptf−μ(f)∥L2(μ)≤e−t/τeq∥f−μ(f)∥L2(μ),(3)\left\|P_t f-\mu(f)\right\|_{L^2(\mu)} \le e^{-t/\tau_{\mathrm{eq}}}\left\|f-\mu(f)\right\|_{L^2(\mu)}, \tag*{(3)}

with equality when ff is an eigenfunction associated with λ1\lambda_1. Equivalently, if the initial law has density ρ0\rho_0 with respect to μ\mu, the density ρt\rho_t at time tt satisfies

∥ρt−ρeq∥L2(μ)≤e−t/τeq∥ρ0−ρeq∥L2(μ).(4)\left\|\rho_t-\rho_{\mathrm{eq}}\right\|_{L^2(\mu)} \le e^{-t/\tau_{\mathrm{eq}}}\left\|\rho_0-\rho_{\mathrm{eq}}\right\|_{L^2(\mu)}. \tag*{(4)}

In particular, the equilibrium autocorrelation of any observable, Cf(t)=Eμ[(f(q0)−μ(f))(f(qt)−μ(f))]C_f(t) = \mathbb{E}_{\mu}\left[(f(q_0)-\mu(f))(f(q_t)-\mu(f))\right], obeys ∣Cf(t)∣≤e−t/τeqVar⁡μ(f)\left|C_f(t)\right| \le e^{-t/\tau_{\mathrm{eq}}}\operatorname{Var}_{\mu}(f). The decay rate of CfC_f equals λ1/2\lambda_1/2 only if ff has a non-zero component along the eigenspace of λ1\lambda_1, if not CfC_f decays faster.

The equilibration time τeq\tau_{\mathrm{eq}} is the slowest time observed in the dynamics. Some correlation functions or operator expectations will decay faster, if they are not coupled to the slowest sector, and this is observed in numerics. However, on timescales t≫τeqt \gg\tau_{\mathrm{eq}} any signal of the initial condition will be reduced by a factor e−t/τeqe^{-t/\tau_{\mathrm{eq}}}.

In this paper we prove that, for the Brownian motion defined before, there exists an εc\varepsilon_c (εc=2\varepsilon_c = 2 in our units) such that for ε<εc\varepsilon< \varepsilon_c the equilibration time is τeq=O(1)\tau_{\mathrm{eq}} = O(1) and for ε>εc\varepsilon> \varepsilon_c, τeq=eΘ(N)\tau_{\mathrm{eq}} = e^{\Theta(\sqrt{N})}. We prove an upper and a lower bound on λ1\lambda_1 in each phase: the upper bounds in Section IV, the lower bounds in § 0.6–VII.

Without discussing the strategy, we present here the results. For ε<2\varepsilon< 2, and NN sufficiently large (i.e. N>2000/(2−ε)2N > 2000/(2-\varepsilon)^2), one has

(2−ε)2100(ε−1)<λ1≲λMF,(5)\frac{(2-\varepsilon)^2}{100(\varepsilon-1)} < \lambda_1 \lesssim\lambda_{\mathrm{MF}}, \tag*{(5)}

where λMF\lambda_{\mathrm{MF}} is the lowest non-zero eigenvalue of a second-order differential equation in one variable which can be easily tabulated to arbitrary accuracy. In the limit ε→2\varepsilon\to2, λMF(2)=1/4\lambda_{\mathrm{MF}}(2) = 1/4 while for ε→1\varepsilon\to1, λMF≃1/(ε−1)\lambda_{\mathrm{MF}} \simeq1/(\varepsilon-1). The upper bound holds asymptotically, lim sup⁡N→∞λ1≤λMF\limsup_{N\to\infty} \lambda_1 \leq\lambda_{\mathrm{MF}} (Section IV A), hence “≲\lesssim”. The numerical experiments find a decay time which is much closer to the upper bound 2/λMF2/\lambda_{\mathrm{MF}} than to the lower bound, which suggests that the lower bound is not strict. We could not, however, find a better lower bound.

For ε>2\varepsilon> 2 we have (assuming the limit exists)

(2−1)ε−2≤lim⁡N→∞log⁡τeqN≤ε−2.(6)(\sqrt{2}-1)\sqrt{\varepsilon-2} \leq\lim_{N\to\infty}\frac{\log\tau_{\mathrm{eq}}}{\sqrt{N}} \leq\sqrt{\varepsilon-2}. \tag*{(6)}

(Theorems IV.1 and V.1). The stretched-exponential time scale is a consequence of the second conservation law. In condensed zero-range processes, which conserve only the mass, the condensate moves between sites on a polynomial time scale [28]. Here it cannot move without passing through configurations in which two sites share the excess energy, whose area is smaller by a factor e−(2−1)N(ε−2)e^{-(\sqrt{2}-1)\sqrt{N(\varepsilon-2)}} (Section III C).

In the next Section, we introduce some results about the geometry of the microcanonical surface that will be used in Section IV. More specialized results to be used in § 0.6 will be introduced at the beginning of that Section.

Geometry of the microcanonical surface

The surface and its regions

References

  1. [1]L. Boltzmann, Lectures on Gas Theory (University of California Press, Berkeley, 1964) translated by S. G. Brush.
  2. [2]S. Goldstein, J. L. Lebowitz, C. Mastrodonato, R. Tumulka, and N. Zanghì, Proc. R. Soc. A 466, 3203 (2010).
  3. [3]T. S. Kuhn, The Structure of Scientific Revolutions, 2nd ed. (University of Chicago Press, Chicago, 1970).
  4. [4]S. F. Edwards and P. W. Anderson, J. Phys. F: Met. Phys. 5, 965 (1975).
  5. [5]M. Mézard, G. Parisi, and M. A. Virasoro, Spin Glass Theory and Beyond (World Scientific, Singapore, 1987).
  6. [6]P. W. Anderson, Phys. Rev. 109, 1492 (1958).
  7. [7]F. Evers and A. D. Mirlin, Rev. Mod. Phys. 80, 1355 (2008).
  8. [8]D. M. Basko, I. L. Aleiner, and B. L. Altshuler, Ann. Phys. 321, 1126 (2006). Nuclear Physics B 891, 420 (2015).
  9. [10]P. Sierant, M. Lewenstein, A. Scardicchio, L. Vidmar, and J. Zakrzewski, Rep. Prog. Phys. 88 (2025), doi:10.1088/1361-6633/ad9756.DOI
  10. [11]G. Parisi, J. Phys. A: Math. Gen. 13, 1101 (1980).
  11. [12]K. Efetov, Supersymmetry in Disorder and Chaos (Cambridge University Press, Cambridge, 1999).
  12. [13]M. Žnidarič, A. Scardicchio, and V. K. Varma, Physical review letters 117, 040601 (2016).
  13. [14]G. Gradenigo, S. Iubini, R. Livi, and S. N. Majumdar, J. Stat. Mech. 2021, 023201 (2021).
  14. [15]C. Arezzo, F. Balducci, R. Piergallini, A. Scardicchio, and C. Vanoni, J. Stat. Phys. 186, 24 (2022).
  15. [16]S. Iubini and A. Politi, Phys. Rev. Lett. 134, 097102 (2025).
  16. [17]K. Nam, Ann. Probab. 48, 2525 (2020).
  17. [18]T. Mithun, Y. Kati, C. Danieli, and S. Flach, Phys. Rev. Lett. 120, 184101 (2018), 1712.03375.
  18. [19]E. Witten, J. Differential Geom. 17, 661 (1982).
  19. [20]S. Bochner, Bull. Amer. Math. Soc. 52, 776 (1946).
  20. [21]R. C. Reilly, Indiana Univ. Math. J. 26, 459 (1977).
  21. [22]J.-P. Eckmann and C. E. Wayne, Discrete Contin. Dyn. Syst. A 38, 6091 (2018), 1710.10999.
  22. [23]N. Cuneo, J.-P. Eckmann, and C. E. Wayne, Nonlinearity 30, R81 (2017).
  23. [24]S. Iubini, L. Chirondojan, G.-L. Oppo, A. Politi, and P. Politi, Phys. Rev. Lett. 122, 084102 (2019).
  24. [25]S. Iubini, R. Franzosi, R. Livi, G.-L. Oppo, and A. Politi, New J. Phys. 15, 023032 (2013).
  25. [26]S. Iubini, A. Politi, and P. Politi, Journal of Statistical Physics 154, 1057 (2014).
  26. [27]S. Chatterjee, Comm. Pure Appl. Math. 67, 1737 (2014), 1203.4027.
  27. [28]J. Beltrán and C. Landim, Probab. Theory Related Fields 152, 781 (2012).
  28. [29]R. N. Bhattacharyya and R. R. Rao, Normal Approximation and Asymptotic Expansions (Wiley, New York, 1976) reprinted in SIAM Classics in Applied Mathematics, 2010.
  29. [30]S. Chatterjee, A note about the uniform distribution on the intersection of a simplex and a sphere (2010), arXiv:1011.4043.arxiv.org/abs/1011.4043
  30. [31]H. Eyring, J. Chem. Phys. 3, 107 (1935).
  31. [32]H. A. Kramers, Physica 7, 284 (1940).
  32. [33]P. Hänngi, P. Talkner, and M. Bor kovec, Rev. Mod. Phys. 62, 251 (1990).
  33. [34]A. Bovier, M. Eckhoff, V. Gayrard, and M. Klein, J. Eur. Math. Soc. 6, 399 (2004).
  34. [35]B. Helffer and J. Sjöstrand, Comm. Partial Differential Equations 10, 245 (1985).
  35. [36]B. Helffer, M. Klein, and F. Nier, Mat. Contemp. 26, 41 (2004).
  36. [37]A. Bovier, V. Gayrard, and M. Klein, J. Eur. Math. Soc. 7, 69 (2005).
  37. [38]N. Berglund, Markov Process. Related Fields 19, 459 (2013).
  38. [39]A. Bovier and F. den Hollander, Metastability: A Potential-Theoretic Approach, Grundlehren der mathematischen Wissenschaften, Vol. 351 (Springer, 2015).
  39. [40]M. Mitrea, M. Taylor, and A. Vasy, Comm. Partial Differential Equations 30, 1445 (2005), math/0408438.
  40. [41]S. Raulot and A. Savo, J. Geom. Anal. 21, 620 (2011), 1003.0817.
  41. [42]A. A. Balinsky, W. D. Evans, and R. T. Lewis, The Analysis and Geometry of Hardy’s Inequality, Universitext (Springer, Cham, 2015).DOI
  42. [43]E. M. Stein, Singular Integrals and Differentiability Properties of Functions, Princeton Mathematical Series, Vol. 30 (Princeton University Press, Princeton, NJ, 1970).
  43. [44]L. E. Payne and H. F. Weinberger, Arch. Rational Mech. Anal. 5, 286 (1960).
  44. [45]M. Bebendorf, Z. Anal. Anwendungen 22, 751 (2003).
  45. [46]K. S. Schweizer and E. J. Saltzman, The Journal of chemical physics 119, 1181 (2003).
  46. [47]D. Reguera, G. Schmid, P. S. Burada, J. Rubi, P. Reimann, and P. Hänggi, Physical review letters 96, 130603 (2006).
  47. [48]M. Bellitti, F. Ricci-Tersenghi, and A. Scardicchio, Physical Review Research 3, 043015 (2021).
  48. [49]A. Carati and A. M. Maiocchi, Commun. Math. Phys. 314, 129 (2012).
  49. [50]A. Giorgilli, S. Paleari, and T. Penati, Ann. Henri Poincaré 16, 897 (2015).
  50. [51]G. Basile, C. Bernardin, and S. Olla, Phys. Rev. Lett. 96, 204303 (2006).

Paper details

vlnc:W_ex6cxmhs
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Claude Opus

View statement
Looking for related papers…
Contents