On Bose-Einstein condensates in disordered media

Marius Lemm, Simone Rademacher, Jingxuan Zhang · 2025

Read On Bose-Einstein condensates in disordered media by Marius Lemm, Simone Rademacher, Jingxuan Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-08925.txt.

Introduction

Marius Lemm, Simone Rademacher, and Jingxuan Zhang prove that disorder slows a Bose-Einstein condensate's ripples to a crawl. Their system holds interacting bosons in the mean-field regime, sitting in a potential that is either random or quasi-periodic, and starts from a condensate localized in one spot.

The fluctuations around that condensate, which carry sound and transport, propagate with only a small velocity because the disorder keeps tripping them. The result holds in any spatial dimension, giving a rare example of a disordered many-body system with rigorously proven slow transport. The technical novelty is an interaction-picture analysis built directly on the Anderson-localized one-body dynamics, the single-particle states that disorder pins in place. Such proofs matter because slow transport in interacting disordered systems is usually argued from numerics, not theorems.

Catalog record

CREATORMarius Lemm, Simone Rademacher, Jingxuan Zhang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE82 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-08925

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

On Bose-Einstein condensates in disordered media Marius Lemm, Simone Rademacher, Jingxuan Zhang Abstract . We consider the quantum dynamics of interacting bosons in the mean-field regime when they are subjected to a disordered potential, which is either random or quasi-periodic. Starting from a spatially localized Bose-Einstein condensate, we prove that fluctuations around the condensate propagate with a small velocity due to the disorder. This provides an example of a disordered many-body system with provably slow transport behavior in any spatial dimension. The main technical novelty is an interaction picture analysis relative to the Anderson-localized one-body dynamics. August 10, 2026 1. Introduction Disorder affects quantum dynamics profoundly in ways that have no classical analog. The most famous example is that of Anderson localization [ And58 ] , the complete absence of all transport for a single quantum particle, which occurs in any spatial dimension at sufficiently strong disorder. Rigorous proofs of this are landmarks works in mathematical physics: in the random case by Fröhlich-Spencer [ FS83 ] and by Aizenman-Molchanov [ AM93 ] , and in the quasi-periodic case by Bourgain-Goldstein-Schlag [ BGS02 ] and by Bourgain [ Bou07 ] . Both directions have grown into large subfields of mathematical physics and we refer to the books [ Bou05 , AW15 ] and the review [ MJ17 ] for additional background. Pertinent directions for us are that a weaker form of dynamical localization has also been proved for the nonlinear Schrödinger equation [ WZ08 , CSZ20 , CSW25 , CSW24 ] and that Anderson localization was extended to systems of multiple particles [ CS09 , FW15 ] with bounds that deteriorate as the particle number N grows large. In contrast to this large body of literature, the effect of disorder on quantum many-body systems is far less understood. By a many-body system, we mean a system with arbitrarily large particle number N , also known as positive-density systems. Understanding the quantum effects of disorder uniformly in N remains a central problem for both mathematicians and condensed-matter physicists. A particularly famous instance of this problem is to establish many-body localization (MBL) in spatial dimension d=1 when the disorder is sufficiently strong compared to the interaction strength. There is a great deal of physics literature on MBL, but from the mathematical side, there are only a few rigorous results. In particular, for strongly disordered 1D spin chains, [ Imb16 ] proved MBL under a level repulsion assumption and [ DRGHP24 ] established sub-diffusion. Other rigorous studies focused on 1D chains in spectral subspaces of low energy [ Mas17 , BW17 , EKS18 , EK24b , EK24a ] or chains with heavy-tailed disorder [ GMN22 , BEGG23 , Bal25 ] . Other recent works considered stability of the MBL phase under sparse perturbations [ TB24 , TB25 ] and MBL in hyperbolic lattices [ YNL24 ] . By contrast, in dimensions d>1 , it is now commonly believed that the MBL phase is dynamically unstable [ DRH17 , LHDR17 ] . Still, it is physically expected that the coherence-destroying effects of disorder should lead to a substantial slowdown effect on the quantum many-body dynamics also for d>1 . This leads to: Problem 1: Show that disorder slows down many-body dynamics in any dimension. On the one hand, Problem 1 is related to studying MBL because it is also concerned with deriving bounds on transport from disorder. On the other hand, the focus is different: First, Problem 1 seeks to treat any dimension. Second, a “slow-down” could be even a very modest decrease in propagation velocity. Methodologically, this problem is challenging and important because the existing analytical toolkit to prove any kind of dynamical slow-down is extremely limited. The reason for this is that all standard techniques for deriving ballistic upper bounds on quantum many-body transport, most notably Lieb-Robinson bounds [ LR72 ] , are completely insensitive to disorder. In other words, the robustness of Lieb-Robinson bounds to the system details turns from a blessing to a curse when one is tasked with exploiting special features such as disorder. Indeed, exploiting disorder to improve Lieb-Robinson bounds has recently become an actively researched topics in 1D spin chains [ GMN22 , BEGG23 , TB24 , Bal25 ] . The recent work [ MLL25 ] then established the first general positive answer to Problem 1 by deriving a Lieb-Robinson bound with non-perturbatively small velocity in any spatial dimension. In the present paper, we investigate Problem 1 for a paradigmatic model of quantum many-body dynamics — interacting bosons in the mean-field regime [ Hep74 , GV79 , RS09 , LNS15 ] . We work on a lattice and the novel twist is that we place this system in a disordered environment modeled by a potential that is either random or quasi-periodic. 1 1 1 We mention in passing that discrete Bose gases in such disordered environments are physically realized in modern optical lattice experiments to study MBL [ CHZ+16 , SZ18 ] . We consider an initial state that is given by a spatially localized Bose-Einstein condensate and prove that its propagation is extremely slow: after time t\gg 1 , the entire system is approximately contained within distances less than \epsilon t from its initial location, where the speed \epsilon\to 0 with increasing disorder strength. This shows strong slow-down due to disorder and thus addresses Problem 1 for mean-field bosons. We now continue the introduction with a slightly more precise, but still informal summary of the setup and main result. The mathematically precise versions are then given in Section 2 . Afterwards, in Subsection 2.4 we discuss further how the result fits into the literature. 1.1. Summary of setup and main result We consider N bosons on \mathbb{Z}^{d} in the mean-field regime. The bosonic nature is encoded in the choice of Hilbert space as the permutation-symmetric elements 2 2 2 Here and in the following, all elements of \ell^{2} -spaces are take to be complex-valued functions. of \ell^{2}((\mathbb{Z}^{d})^{N}) , i.e., \psi_{N,t}(x_{\pi(1)},\ldots,x_{\pi(N)})=\psi_{N,t}(x_{1},\ldots,x_{N}),\qquad\forall\pi\in S_{N}. The system’s energy is encoded in the Hamiltonian, a linear self-adjoint operator of the form (1.1) H_{N}:=\sum_{i=1}^{N}(-\Delta_{x_{i}}+\lambda v_{\omega}(x))+\frac{1}{N}\sum_{1\leq i,j\leq N}V(x_{i}-x_{j}). Here, -\Delta_{x_{i}} denotes the discrete Laplacian acting in the x_{i} variable, v_{\omega}(x_{i}) are multiplication operators that are, e.g., generated by a family of independent and identically distributed (i.i.d.) random variables, and \lambda\in\mathbb{R} will be a large parameter called the disorder strength. For the interaction, we consider for simplicity the “on-site” form V(x_{i}-x_{j})=U\delta_{x_{i},x_{j}} , with some \mathcal{O}(1) constant U\in\mathbb{R} , but the proof generalizes to any V:\mathbb{Z}^{d}\to\mathbb{R} of compact support. The main challenge is that all bounds should be uniform in the particle number parameter N , which in particular means that the dimension of the Hilbert space is extremely large. The disorder strength \lambda will be a large constant independent of N . We study the quantum many-body dynamics associated to H_{N} . As a self-adjoint operator, H_{N} generates the associated unitary operator e^{-itH_{N}} , which is the solution operator to the N -body Schrödinger equation (1.2) \displaystyle i\partial_{t}\psi_{N,t}=H_{N}\;\psi_{N,t}. The standard initial states for the dynamic ( 1.2 ) describe uncorrelated Bose-Einstein condensates. Mathematically, this means one considers initial data given by a product function of the form (1.3) \displaystyle\psi_{N,0}(x_{1},\ldots,x_{N})=\prod_{i=1}^{N}\varphi_{0}(x_{i}), for a fixed \varphi_{0}\in\ell^{2}(\mathbb{Z}^{d}) . Due

Read the complete text