Engineering Anderson Localization in Arbitrary Dimensions with Interacting Quasiperiodic Kicked Bosons

This paper demonstrates that combining interparticle interactions with quasiperiodic driving in a kicked boson system creates synthetic dimensions, enabling the emulation of Anderson localization and its critical behavior in arbitrary dimensions up to four.

The Gist

Imagine you are walking through a dense, foggy forest. If the trees are scattered randomly, you might get lost and end up walking in circles, never making it to the other side. In physics, this is called Anderson Localization: a wave (like a particle of light or an electron) gets "stuck" in a disordered environment because the waves bouncing off obstacles cancel each other out, trapping the energy in one spot.

For a long time, scientists knew this happened easily in 1D (a line) and 2D (a flat sheet). But in 3D (our real world), things get tricky. Usually, if the disorder isn't strong enough, the wave can find a path through and keep moving (diffusing). To see the "stuck" behavior in 3D or higher, you need a very specific, messy environment.

The Big Idea of This Paper
The researchers in this paper asked a clever question: Can we build a "forest" with as many dimensions as we want, just by playing with how we push particles and how they bump into each other?

They used a system of ultracold atoms (bosons) that are being kicked repeatedly, like a drum being hit over and over. Here is how they engineered different "worlds" (dimensions) using two main tools:

1. The Tools of the Trade

Tool A: The "Bump" (Interactions)
Imagine two people walking in a hallway. If they are ghosts, they walk right through each other. But if they are solid, they bump into each other. In physics, when two particles interact (bump), they create a complex relationship.

• The Analogy: Think of two dancers. If they don't touch, they just spin in their own spots (1D). But if they hold hands and dance together, their movements become linked. This "link" effectively adds a new direction to their dance floor.
• The Result: Just by having two interacting particles, the system naturally behaves like a 2D world.

Tool B: The "Rhythm" (Quasiperiodic Kicks)
Now, imagine the person kicking the drum doesn't do it at a steady beat. Instead, they kick it to a rhythm that mixes two different, non-repeating tempos (like a jazz drummer playing two different time signatures at once).

• The Analogy: If you walk in a straight line, that's 1D. If you walk in a straight line but also sway side-to-side to a second rhythm, you are now tracing a 2D path. If you add a third rhythm (spinning while walking and swaying), you are moving in 3D.
• The Result: By adding these extra, non-repeating rhythms to the kicks, the researchers can "stretch" the system into 3D, 4D, or even higher dimensions.

2. The Experiment: Building Dimensions

The team combined these two tools. They took two interacting atoms and kicked them with different rhythms.

• Scenario 1 (No extra rhythm): Two interacting atoms = 2D world.
  • Outcome: The atoms get stuck (localized). No "phase transition" happens here because 2D is too small to have a true "stuck vs. moving" switch.
• Scenario 2 (One extra rhythm): Two interacting atoms + 1 rhythm = 3D world.
  • Outcome: Magic happens! They found a critical point. If the kicks are weak, the atoms get stuck (Localized). If the kicks are strong, they start moving freely (Diffusive). Right in the middle, they behave in a weird, critical way. This is the famous Anderson Transition.
• Scenario 3 (Two extra rhythms): Two interacting atoms + 2 rhythms = 4D world.
  • Outcome: The same transition happens, but now in a 4-dimensional space.

3. Why This Matters

Usually, studying 4D physics is impossible because we live in 3D. You can't build a 4D box in a lab.

But this paper shows that you don't need a 4D box. You just need a 1D line of atoms, but you "trick" the math by making them interact and kick them with complex rhythms. The atoms act as if they are living in 4D.

The "Aha!" Moment:
The researchers proved that these two tricks (interactions and rhythms) simply add up.

• 2 particles + 0 rhythms = 2 Dimensions.
• 2 particles + 1 rhythm = 3 Dimensions.
• 2 particles + 2 rhythms = 4 Dimensions.

They measured exactly how the atoms behaved at the "tipping point" (the transition from stuck to moving) and found that the math matched the predictions for 3D and 4D physics perfectly.

Summary

Think of this paper as a master class in dimensional engineering.

• The Problem: We want to study how waves get stuck in high-dimensional spaces, but we can't build those spaces.
• The Solution: Use a simple line of atoms, make them bump into each other, and kick them with a complex, non-repeating rhythm.
• The Result: The atoms pretend to live in 3D or 4D, allowing scientists to study complex quantum phenomena that were previously impossible to test.

It's like taking a flat piece of paper (1D) and, by folding it in a specific, rhythmic way, making it behave exactly like a complex, multi-layered sculpture (4D), all without ever leaving the table.

Technical Summary

1. Problem Statement

Anderson localization, the suppression of wave propagation due to disorder, typically exhibits a metal-insulator transition (Anderson transition) only in dimensions $d \geq 3$. In lower dimensions ($d=1, 2$), all states are localized in the absence of interactions. While the quantum kicked rotor (QKR) serves as a paradigmatic model for studying dynamical localization (the momentum-space analog of Anderson localization), standard 1D QKR systems are limited to $d=1$.

The central challenge addressed in this work is the engineering of synthetic dimensions to realize Anderson localization and its associated phase transitions in arbitrary effective dimensions ($d > 2$) within a controllable quantum system. Specifically, the authors investigate how two distinct mechanisms—interparticle interactions and quasiperiodic driving—can be combined to generate these synthetic dimensions in a system of interacting bosons.

2. Methodology

Physical Model

The authors study a system of $N$ identical bosons confined to a ring of length $L$, described by the Lieb-Liniger model with repulsive point-like interactions (strength $c$). The system is subjected to a sequence of delta-kicks (periodic driving) with a time-dependent amplitude $\kappa(t)$.

• Hamiltonian: The system is defined by a free part $H_0$ (Lieb-Liniger) and a kick part $H_{kick}$.
• Quasiperiodic Modulation: To introduce synthetic dimensions, the kick strength $K(t)$ is modulated by incommensurate frequencies. The authors consider:
  • $N_\omega = 0$: Constant kick strength.
  • $N_\omega = 1$: Modulation with one incommensurate frequency ($\omega_2$).
  • $N_\omega = 2$: Modulation with two incommensurate frequencies ($\omega_2, \omega_3$).
• Effective Dimensionality: The effective dimension $d$ is determined by the sum of the particle number $N$ (for finite interactions) and the number of additional driving frequencies $N_\omega$:
  $$d = N + N_\omega$$
  The study focuses on $N=2$ bosons, exploring effective dimensions $d=2, 3,$ and $4$ by varying $N_\omega$ from 0 to 2.

Numerical Approach

• Bethe Ansatz Basis: Since the Lieb-Liniger model is exactly solvable, the authors utilize the Bethe Ansatz to construct the eigenbasis of the unperturbed Hamiltonian. The wavefunctions are characterized by sets of rapidities $\{\lambda_j\}$.
• Floquet Dynamics: The system is a Floquet system. Time evolution is computed by repeatedly applying the Floquet operator $U = e^{-iH_0/\hbar} e^{-iH_{kick}/\hbar}$.
• Matrix Representation: The evolution is simulated in the truncated Bethe eigenbasis. The matrix elements of the kick operator are calculated analytically using Bessel functions and the specific Bethe wavefunction coefficients.
• Observables: The primary observable is the total energy $\langle E(t) \rangle$ of the system. To reduce fluctuations, results are averaged over 50 random realizations of the quasi-momentum $\beta$.
• Finite-Time Scaling Analysis: To distinguish between true phase transitions and finite-size effects, the authors employ finite-time scaling. They define a dimensionless scaling variable $\Lambda(K, t) = \langle E(t) \rangle t^{-2/d}$ and analyze its behavior as a function of the stochasticity parameter $K$ and time $t$.

3. Key Contributions

1. Demonstration of Additive Synthetic Dimensions: The paper establishes that interparticle interactions and quasiperiodic driving act as additive mechanisms for generating synthetic dimensions.
   • Interactions between two bosons ($N=2$) alone generate an effective 2D system.
   • Adding one incommensurate frequency ($N_\omega=1$) extends this to 3D.
   • Adding two frequencies ($N_\omega=2$) extends this to 4D.
2. Engineering Arbitrary Dimensions: The authors demonstrate a versatile framework where the effective dimensionality can be tuned continuously by adjusting the number of particles and modulation frequencies, allowing the study of Anderson transitions in dimensions that are difficult to access experimentally (e.g., $d=4$).
3. Verification of Orthogonal Universality: The study confirms that the interacting quasiperiodic kicked rotor belongs to the orthogonal universality class of the Anderson transition, validating the mapping between this driven interacting system and standard disordered models.

4. Results

Dynamical Behavior

• $d=2$ ($N=2, N_\omega=0$): No phase transition is observed. The energy saturates for all values of the kick strength $K$, consistent with the absence of an Anderson transition in 2D.
• $d=3$ ($N=2, N_\omega=1$) and $d=4$ ($N=2, N_\omega=2$): A clear metal-insulator transition is observed.
  • Localized Phase ($K < K_c$): Energy saturates at long times.
  • Diffusive Phase ($K > K_c$): Energy grows linearly with time ($\langle E(t) \rangle \sim t$).
  • Critical Point ($K = K_c$): The system exhibits anomalous diffusion with $\langle E(t) \rangle \sim t^{2/d}$.

Critical Exponents and Scaling

Through finite-time scaling analysis, the authors extracted critical exponents $\nu$ (localization length divergence) and $s$ (diffusion constant vanishing) and verified Wegner's scaling law: $s = (d-2)\nu$.

• For $d=3$ ($N_\omega=1$):
  • $\nu \approx 1.57 \pm 0.05$
  • $s \approx 1.57 \pm 0.05$
  • These values match the state-of-the-art estimates for the 3D orthogonal Anderson transition.
• For $d=4$ ($N_\omega=2$):
  • $\nu \approx 1.17 \pm 0.04$
  • $s \approx 2.28 \pm 0.18$
  • These values satisfy the scaling law $s = (4-2)\nu \approx 2.34$, consistent with recent estimates for 4D systems.

Phase Diagrams

The authors constructed phase diagrams in the $(K, \epsilon)$ plane (where $\epsilon$ is the modulation amplitude), clearly delineating the boundaries between localized and delocalized phases for $N_\omega=1$ and $N_\omega=2$.

5. Significance

• Experimental Feasibility: The proposed system can be realized using ultracold atomic gases (specifically 1D Bose gases) in optical lattices, where interactions and quasiperiodic potentials are highly controllable. This offers a clean platform to study high-dimensional quantum criticality without the geometric constraints of physical 3D or 4D lattices.
• Theoretical Insight: The work resolves the question of how interactions and driving combine to alter dimensionality, showing they are additive. It provides a robust method to probe the limits of universality in disordered quantum systems, particularly in dimensions $d \geq 4$ where mean-field theories often fail but Anderson localization persists.
• Future Directions: The framework opens avenues for studying other symmetry classes (unitary, symplectic) via synthetic gauge fields and investigating non-ergodic dynamics and multifractality at criticality. It also poses the open question of how these mechanisms scale with larger particle numbers ($N > 2$).

In conclusion, this paper provides a comprehensive theoretical proof-of-concept for engineering Anderson localization in arbitrary dimensions using interacting kicked bosons, bridging the gap between 1D experimental setups and high-dimensional quantum critical phenomena.