2025-11-10T03:03:50.456760

Long persistence of localization in a disordered anharmonic chain beyond the atomic limit

De Roeck, Huveneers, Prośniak
We establish rigorous bounds on the decorrelation time and thermal transport in the disordered Klein-Gordon chain with a quartic on-site potential, governed by a parameter $λ$. At $λ= 0$, the chain is harmonic, and any form of transport is fully suppressed by Anderson localization. For the anharmonic system, at $λ> 0$, our results show that decorrelation and transport can occur only on time scales that grow faster than any polynomial in $1/λ$ as $λ\to 0$. From a technical perspective, the main novelty of our work is that we don't restrict ourselves to the atomic limit. Instead, we develop perturbation theory around the harmonic system with a fixed harmonic interaction between nearby oscillators. This allows us to compare our mathematical results with previous numerical work and contribute to resolving an ongoing debate, as detailed in a companion paper arXiv:2308.10572.
academic

Long persistence of localization in a disordered anharmonic chain beyond the atomic limit

Basic Information

  • Paper ID: 2308.11243
  • Title: Long persistence of localization in a disordered anharmonic chain beyond the atomic limit
  • Authors: Wojciech De Roeck (K.U.Leuven University), François Huveneers (King's College London), Oskar A. Prośniak (University of Luxembourg)
  • Classification: math-ph math.MP (Mathematical Physics)
  • Publication Date: January 3, 2025 (arXiv version 2)
  • Paper Link: https://arxiv.org/abs/2308.11243

Abstract

This paper establishes rigorous bounds on decorrelation times and heat transport in disordered Klein-Gordon chains subject to quartic on-site potentials, a system controlled by parameter λ. When λ=0, the chain is harmonic and Anderson localization completely suppresses any form of transport. For anharmonic systems (λ>0), the results demonstrate that decorrelation and transport can only occur on timescales that grow faster than any polynomial in 1/λ as λ→0. From a technical perspective, the principal novelty of this work lies in extending beyond the atomic limit by developing perturbation theory around harmonic systems with fixed harmonic interactions, enabling comparison of mathematical results with previous numerical work.

Research Background and Motivation

Research Questions

This paper investigates the fate of Anderson localization in the presence of genuine many-body interactions (anharmonicity), a problem of significant mathematical and physical importance. For classical Hamiltonian systems at positive temperature, it is widely believed that small, generic interactions destroy localization and integrability, leading to hydrodynamic behavior with diffusive or superdiffusive transport.

Problem Significance

  1. Theoretical Importance: Understanding the competition between localization and interactions is a central problem in statistical physics
  2. Numerical Controversy: Existing numerical studies show wave packet propagation following w(t) ∝ t^(1/6), which contrasts sharply with the "asymptotic localization" theory prediction of propagation slower than any polynomial in time
  3. Technical Challenges: Rigorous mathematical treatment beyond the atomic limit is extremely difficult

Limitations of Existing Approaches

  1. Atomic Limit Restriction: Previous mathematical work has focused primarily on the atomic limit where η∝γ→0, with both η and γ treated as perturbation parameters
  2. Numerical-Theoretical Gap: Significant discrepancies exist between numerical simulations and theoretical predictions
  3. Technical Difficulties: Handling the complete Anderson insulator rather than simple decoupled sites presents new technical challenges

Core Contributions

  1. Beyond the Atomic Limit: First rigorous mathematical results established with η fixed and γ as the perturbation parameter
  2. Asymptotic Localization Proof: Rigorous bounds on decorrelation rates exclude any behavior other than "asymptotic localization"
  3. Novel Perturbation Theory: Development of systematic perturbation expansion applicable to controlling local observables in extended systems
  4. Technical Innovations:
    • Control of KAM-type "denominators" involving linear combinations of eigenfrequencies of the linear system
    • Formal methods for handling localized but non-compactly-supported eigenfunctions
    • Establishment of Z-set theoretical framework

Methodological Details

Problem Definition

Investigation of long-time dynamics of disordered Klein-Gordon chains: q¨x=ωx2qx+η(Δq)xγqx3\ddot{q}_x = -\omega_x^2 q_x + \eta(\Delta q)_x - \gamma q_x^3 where ωx2\omega_x^2 are independent identically distributed random variables, η\eta is the harmonic coupling, and γ\gamma is the anharmonic coupling.

Model Architecture

1. Hamiltonian System

H(q,p)=xΛL(px22+ωx2qx22+η2(qxqx+1)2+γ4qx4)H(q,p) = \sum_{x\in\Lambda_L}\left(\frac{p_x^2}{2} + \frac{\omega_x^2 q_x^2}{2} + \frac{\eta}{2}(q_x-q_{x+1})^2 + \frac{\gamma}{4}q_x^4\right)

2. Dimensionless Formulation

Introduction of effective anharmonicity parameter: λ=β1γ\lambda = \beta^{-1}\gamma, where β is the inverse temperature.

3. Harmonic Decomposition

Hhar=12k=1ΛL([p,ψk]2+νk2[q,ψk]2)=k=1ΛLEkH_{har} = \frac{1}{2}\sum_{k=1}^{|\Lambda_L|}([p,\psi_k]^2 + \nu_k^2[q,\psi_k]^2) = \sum_{k=1}^{|\Lambda_L|}E_k where EkE_k is the energy of the k-th mode.

Technical Innovations

1. Perturbation Expansion

For observable f, establish decomposition: f={H,u}+λngf = -\{H, u\} + \lambda^n g through recursive solution of commutator equations: {Hhar,u(i)}=f(i),f(i+1)={Han,u(i)}-\{H_{har}, u^{(i)}\} = f^{(i)}, \quad f^{(i+1)} = \{H_{an}, u^{(i)}\}

2. Small Denominator Control

The key technical difficulty is controlling denominators of the form 1σ1νk1++σnνkn\frac{1}{\sigma_1\nu_{k_1} + \cdots + \sigma_n\nu_{k_n}}, where νkj2\nu_{k_j}^2 are eigenvalues of the Anderson operator.

Theorem 5: There exist constants CmC_m such that for any ϵ>0\epsilon > 0: P(Qϵ)CmImϵ12m(m+1)P(Q \leq \epsilon) \leq C_m|I|^m\epsilon^{\frac{1}{2m(m+1)}}

3. Z-Set Theory

Introduction of Z-set concept to handle properties of localized functions, satisfying:

  • Probabilistic Bounds: P(ZI(x)M)CMμP(Z_I(x) \geq M) \leq \frac{C}{M^\mu}
  • Locality: ZI(x)ZI(x,)(x)Cec|Z_I(x) - Z_{I(x,\ell)}(x)| \leq Ce^{-c\ell}

Experimental Setup

Theoretical Framework

This is pure mathematical theoretical work without numerical experiments, establishing rigorous mathematical bounds.

Main Theorem Statements

Theorem 1 (Decorrelation Bounds): lim supL1ΛLk=1ΛLCk(t)Cn(λcn+(λnt)2)\limsup_{L\to\infty}\frac{1}{|\Lambda_L|}\sum_{k=1}^{|\Lambda_L|}C_k(t) \leq C_n(\lambda^{c_n} + (\lambda^n t)^2)

Theorem 2 (Energy Transport Bounds): lim supL(J0(t))2Vn(1+(λnt)2)\limsup_{L\to\infty}\langle(J_0(t))^2\rangle \leq V_n(1 + (\lambda^n t)^2)

Theorem 3 (Thermal Conductivity Bounds): lim supL(J(λnτ))2Cnλn\limsup_{L\to\infty}\langle(\mathcal{J}(\lambda^{-n'}τ))^2\rangle \leq C_{n'}\lambda^n

Experimental Results

Main Findings

1. Asymptotic Localization Proof

  • For most modes k, decorrelation must proceed extremely slowly
  • Transport can only occur on timescales growing faster than any polynomial in 1/λ1/λ

2. Energy Transport Suppression

  • Integrated heat flux J0(t)J_0(t) remains bounded over long times
  • All derivatives of thermal conductivity κ(λ)κ(λ) vanish at λ=0: nκλn(0)=0\frac{\partial^n κ}{\partial λ^n}(0) = 0

3. Non-equilibrium Generalization

Theorem 4: Similar bounds hold for non-equilibrium initial states satisfying appropriate conditions.

Technical Achievements

1. Denominator Estimates

Successful control of small denominator problems involving linear combinations of multiple eigenfrequencies, the key technical barrier to transcending the atomic limit.

2. Locality Control

Through Z-set theory, establishment of systematic methods for handling localized but non-compactly-supported functions in Anderson insulators.

Atomic Limit Studies

  • Bourgain & Wang, De Roeck & Huveneers and others established asymptotic localization for η∝γ→0
  • Numerical studies Pikovsky & Shepelyansky, Flach et al. support this picture

Numerical Propagation Experiments

  • Multiple studies show propagation law w(t)t1/6w(t) ∝ t^{1/6}
  • Contrasts with theoretical prediction of super-polynomial slow propagation

Technical Methods

  • Application of KAM theory to localized systems
  • Minami estimates and their generalizations
  • Rigorous results on Anderson localization

Conclusions and Discussion

Main Conclusions

  1. Universality of Asymptotic Localization: Asymptotic localization persists even beyond the atomic limit
  2. Extremely Slow Transport: All forms of transport are suppressed to super-polynomial timescales
  3. Numerical-Theoretical Reconciliation: Provides mathematical foundation for resolving discrepancies between numerical and theoretical predictions

Limitations

  1. Tightness of Bounds: Established results are upper bounds; actual behavior may be slower
  2. Finite-Size Effects: Rigorous results hold only in the thermodynamic limit
  3. Model-Specific: Results specific to Klein-Gordon chains; generalizability remains to be verified

Future Directions

  1. Numerical Verification: Longer numerical simulations needed to verify theoretical predictions
  2. Other Models: Extension to nonlinear Schrödinger equations and other systems
  3. Experimental Realization: Search for possible experimental verification pathways

In-Depth Evaluation

Strengths

  1. Theoretical Breakthrough: First rigorous asymptotic localization results beyond the atomic limit
  2. Technical Innovation: Development of new methods for handling complex denominator problems and localized functions
  3. Mathematical Rigor: All results supported by complete mathematical proofs
  4. Problem Importance: Addresses fundamental questions in statistical physics

Weaknesses

  1. Potentially Non-Tight Bounds: Upper bounds may be overly conservative
  2. Technical Complexity: Proofs are extremely technical and difficult to generalize
  3. Remaining Numerical Discrepancy: Still requires explanation for different numerical behavior
  4. Limited Practical Utility: Pure theoretical results with limited practical applications

Impact

  1. Theoretical Contribution: Provides important support for many-body localization theory
  2. Methodological Value: Technical methods potentially applicable to other problems
  3. Controversy Resolution: Helps resolve important controversies in the field

Applicable Scenarios

  • Localization problem research in theoretical physics
  • Long-time dynamics analysis of disordered systems
  • Classical analogues of many-body quantum systems

References

This paper cites 49 important references covering classical and recent research on Anderson localization, KAM theory, numerical simulations, and related topics.


Summary: This paper makes important contributions to mathematical physics by rigorously proving the universality of asymptotic localization phenomena through mathematical analysis, providing profound insights into understanding the competition between interactions and localization in disordered systems. While highly technical, it provides a solid theoretical foundation for addressing fundamental problems in this field.