Direct Observation of the Three-Dimensional Anderson Transition with Ultracold Atoms in a Disordered Potential

By implementing a novel energy-resolved scheme to prepare narrow atomic matter waves, the researchers achieved the first direct and model-independent observation of the three-dimensional Anderson transition in a disordered potential, resolving long-standing discrepancies between previous experiments and theory.

The Gist

Imagine you are standing in a massive, dark forest. You want to walk from one side to the other.

In a normal forest, you can see the paths, and even if there are some bushes, you can weave through them and keep moving. This is like "diffusion"—the steady, predictable movement of particles.

Now, imagine a "Magic Ghost Forest." In this forest, the trees aren't just physical obstacles; they are made of a strange energy that creates "interference." As you try to walk, the trees create overlapping shadows and echoes of your own footsteps. If the forest is thick enough, these echoes collide with you in such a perfect, chaotic way that they actually cancel out your forward motion. You aren't stuck because a tree is physically blocking you; you are stuck because the "math" of your movement has folded in on itself. You are trapped by nothing but interference.

This is Anderson Localization, and for decades, scientists have been trying to prove exactly where the "line in the sand" is—the exact moment a forest goes from being "walkable" to "impassable."

The Problem: The "Blurry Lens"

Until now, trying to find this line was like trying to find the exact moment a light turns from dim to bright while wearing blurry glasses. Previous experiments used clouds of atoms that were "messy"—they had a wide range of energies all mixed together. Because the atoms were all moving at different speeds, scientists couldn't tell if a single atom was stuck or if the whole group was just a blurry mess. They had to guess where the transition happened using math models.

The Breakthrough: The "Precision Sniper"

The team at the Institut d’Optique did something revolutionary. Instead of throwing a messy cloud of atoms into the forest, they used a technique (using radio-frequency pulses) that acts like a precision sniper.

They can pick a specific "energy" for the atoms—essentially choosing exactly how fast they are "walking"—and then drop them into the disordered laser forest.

1. The Slow Walkers (Localized): They sent in slow atoms. As predicted, the atoms hit the "interference" and stopped dead. They didn't spread out; they just sat there, trapped by the quantum echoes.
2. The Fast Runners (Diffusive): They sent in fast atoms. These atoms had enough "oomph" to punch through the interference, spreading out across the forest like a spilled liquid.
3. The "Goldilocks" Zone (The Mobility Edge): This is the holy grail. They found the exact speed where the atoms do something weird: they don't stop, but they don't run either. They move in a strange, stuttering way called "subdiffusion."

Why does this matter?

By finding this "Mobility Edge" (the exact speed where walking becomes impossible) without having to rely on guesses or models, the researchers have provided a "gold standard" for quantum physics.

It’s like finally finding the exact temperature at which water turns to ice. Once you know that exact point, you can start studying much more complex things: How do these particles behave if they "talk" to each other (interactions)? How does it work in different dimensions?

This experiment isn't just about trapped atoms; it's about mastering the ability to control the fundamental "chaos" of the quantum world.

Technical Summary

Technical Summary: Direct Observation of the Three-Dimensional Anderson Transition with Ultracold Atoms in a Disordered Potential

1. Problem Statement

Anderson localization is a quantum interference phenomenon where wave transport is completely halted by disorder. In three-dimensional (3D) systems, scaling theory predicts a quantum phase transition at a critical energy known as the mobility edge ($E_c$), which separates localized (insulating) states from diffusive (conducting) states.

While Anderson localization has been studied in electronic and classical wave systems, direct experimental observation of the 3D mobility edge in matter waves has been historically elusive. Previous attempts using ultracold atoms were hindered by strong energy broadening in the atomic clouds. Because the atoms were not prepared with a narrow energy distribution, researchers could only provide indirect, model-dependent estimates of the mobility edge, leading to significant discrepancies between experimental results and theoretical numerical predictions.

2. Methodology

The researchers implemented a novel, energy-resolved experimental scheme to overcome the limitations of energy broadening.

• State-Dependent Disorder: The team used a bichromatic laser speckle field consisting of a "principal" laser and a "compensating" laser. By finely tuning their wavelengths and intensities, they created a potential that is essentially disorder-free for atoms in state $|1\rangle$ but highly disordered for atoms in state $|2\rangle$.
• Energy-Resolved Loading: They utilized a two-photon radio-frequency (rf) pulse to transfer a small fraction of a $^{87}\text{Rb}$ Bose-Einstein Condensate (BEC) from state $|1\rangle$ to state $|2\rangle$. By adjusting the rf frequency, they could target a specific energy $E_f$ with high precision. The energy width $\Delta E$ was Fourier-limited by the pulse duration ($t_{rf}$), achieving a resolution of $\Delta E/h \approx 14\text{ Hz}$—orders of magnitude narrower than previous experiments.
• Clock States & Levitation: To ensure long observation times (up to 5 seconds), they used "clock states" ($|F=2, m_F=1\rangle$ and $|F=1, m_F=-1\rangle$) at a "magic" magnetic field ($B^* = 3.23\text{ G}$). This configuration allowed for magnetic levitation to counteract gravity while remaining immune to magnetic noise and minimizing spin-exchange collisions.
• Dynamics Tracking: After the transfer, the optical trap was removed, and the expansion of the atomic cloud was monitored via fluorescence imaging. They analyzed the cloud size $\sigma(t)$ and the 1D integrated density profiles $n_{1d}(z, t)$.

3. Key Contributions

• Direct Observation: Unlike previous studies that inferred the transition, this work directly observes the change in transport dynamics (from exponential density tails to Gaussian profiles) as a function of energy.
• Model-Independent Determination: By analyzing the scaling exponent $\kappa$ of the cloud's squared size ($\sigma^2 \propto t^\kappa$), they extracted the mobility edge purely from experimental data without relying on underlying theoretical models.
• Resolution of Discrepancies: The study provides a high-precision measurement that resolves long-standing conflicts between prior cold-atom experiments and state-of-the-art numerical simulations.

4. Results

• Identification of the Transition: The experiment revealed three distinct dynamical regimes:
  1. Localized ($E_f < E_c$): The cloud size saturates ($\kappa \approx 0$) and density profiles show exponential decay.
  2. Critical ($E_f = E_c$): The cloud exhibits anomalous subdiffusive expansion ($\kappa = 2/3$).
  3. Diffusive ($E_f > E_c$): The cloud expands linearly in time ($\kappa = 1$) with Gaussian density profiles.
• Precise Mobility Edge: For a disorder amplitude of $V_R/h = 416\text{ Hz}$, the experimental mobility edge was determined to be $E_c^{exp}/h = 237(12)\text{ Hz}$. This is in excellent agreement with the numerical prediction of $E_c^{num}/h = 240(6)\text{ Hz}$.
• Universal Scaling: The researchers demonstrated that the mobility edge follows the predicted scaling behavior relative to the disorder strength $\eta$, matching numerical curves across a wide range of disorder amplitudes.

5. Significance

This work represents a landmark achievement in quantum simulation. By successfully implementing an energy-resolved probe, the authors have turned ultracold atoms into a high-precision tool for studying quantum criticality.

The methodology opens new avenues for investigating:

• The role of dimensionality and symmetry classes in localization.
• Many-body localization (MBL) and how particle-particle interactions modify the mobility edge.
• Wave-function multifractality and other complex phenomena at the critical point of phase transitions.