🔬 condensed matter

Rotating two-dimensional Bose--Einstein condensates in power-law traps: collapse, Thomas--Fermi scaling, and the inverse-square endpoint

This paper investigates the scaling behavior, collapse dynamics, and Thomas-Fermi limits of rotating two-dimensional Bose-Einstein condensates in power-law traps across a range of exponents, revealing universal collapse mechanisms, rotation-induced instabilities, and specific behaviors at the inverse-square endpoint.

Original author: Dinh-Thi Nguyen, Bao-Duy Le

Published 2026-11-04

📖 5 min read🧠 Deep dive

CC BY 4.0

Original author: Dinh-Thi Nguyen, Bao-Duy Le

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

In the coldest corners of the universe, where atoms lose their individual identities and merge into a single, giant quantum wave, a strange new kind of matter emerges. This is the Bose-Einstein condensate, a state of matter first predicted nearly a century ago and created in laboratories only recently. In this state, thousands of atoms behave as if they are a single particle, moving in perfect unison. Scientists can trap these clouds of atoms using invisible walls of light, shaping them into various forms and spinning them around to study how they react. The behavior of these clouds depends heavily on how the atoms interact with one another. Sometimes they push each other away, creating a stable, expanding cloud. Other times, they pull toward one another, a tendency that can cause the entire cloud to collapse inward, shrinking until it vanishes. Understanding the precise balance between these forces, and how the shape of the trap influences this balance, is crucial for controlling these delicate quantum systems.

A team of researchers has now mapped out exactly how these clouds behave when trapped in a wide variety of shapes and spun at different speeds. They focused on a family of traps that range from simple bowls to more complex, steep-sided containers, and they looked at both the pushing and pulling forces between atoms. Their work reveals a simple, unifying rule that governs the size of the cloud, regardless of whether the atoms are repelling or attracting each other. When the atoms push apart, the cloud spreads out to a size determined by the strength of that push and the shape of the trap. When they pull together, the cloud shrinks, but only until it reaches a critical point where it collapses. The researchers found that the speed at which this collapse happens is dictated entirely by the local shape of the trap right where the cloud is sitting, while the overall size of the trap matters less.

One of the most striking discoveries concerns what happens when these clouds are spun. If the atoms are pushing each other apart, spinning the trap simply causes the cloud to develop a lattice of tiny whirlpools, known as vortices, which are a hallmark of rotating quantum fluids. However, if the atoms are pulling together, the story changes dramatically. As the rotation speed increases, the cloud does not just spin faster; it eventually becomes unstable and moves off-center. The researchers found that this happens the moment the spin rate exceeds a specific threshold related to how stiff the cloud is. In certain steep traps, this causes the collapsing cloud to settle into a ring shape, orbiting the center of the trap like a planet. This ring is not just a hollow center; it is a stable, rotating lump of matter that has broken away from the middle. The team calculated that for a specific type of steep trap, this shift occurs when the rotation speed reaches a value that is roughly the sixth power of the distance the cloud has moved, a relationship they confirmed with high precision.

The study also explored the extreme edge of these interactions, a theoretical limit where the trap becomes infinitely deep at the center. In this scenario, known as the inverse-square potential, the usual rules of balance break down completely. The researchers discovered that no matter how strongly the atoms push or pull, the cloud always falls toward the center. If the atoms are pushing, they cannot stop this fall; instead, they form a vast, thin halo of atoms that stretches out exponentially far from the center, while a tiny core remains at the very bottom. This halo is a universal feature, meaning its structure depends only on the long-range shape of the trap, not on the specific details of how the trap is smoothed out at the very center. This finding suggests that in these extreme conditions, the cloud is dominated by a tug-of-war between the pull of the trap and the spreading effect of the atoms' own repulsion, resulting in a structure that grows larger as the repulsion gets stronger, rather than shrinking.

Beyond the physics of the clouds themselves, the researchers uncovered a subtle but important flaw in the computer simulations used to study them. When scientists use digital models to find the stable state of these clouds, they often rely on a step-by-step calculation that mimics the passage of time. The team showed that near the point of collapse, this method introduces a tiny, invisible error that makes the cloud appear slightly wider than it really is. This error is so small that standard checks used to verify the accuracy of the simulation fail to detect it. The researchers demonstrated that by changing the way the calculation steps are taken, this bias can be eliminated, ensuring that future simulations of these fragile quantum states are far more accurate. This correction is vital for anyone trying to predict the behavior of these systems near the point of collapse, where even a tiny error can lead to a completely wrong conclusion.

The implications of these findings extend to real-world experiments. Scientists can now predict exactly how many atoms a trap can hold before it collapses, a number that depends only on the strength of the attraction and not on the specific shape of the trap. They also know that if they spin the trap fast enough, an attractive cloud will leave the center and orbit the axis, a behavior that could be observed directly in experiments. For repulsive clouds, the team confirmed that the cloud will empty out from the center and form a ring at a specific rotation speed, a transition that has been observed but not fully understood until now. The work provides a complete picture of how these quantum clouds respond to rotation and interaction, from the gentlest pushes to the most violent collapses, offering a clear guide for future experiments in the realm of ultra-cold atoms.

Technical Summary: Rotating Two-Dimensional Bose–Einstein Condensates in Power-Law Traps

Problem Statement
This work investigates the ground-state properties and stability of two-dimensional Bose–Einstein condensates (BECs) described by the Gross–Pitaevskii (GP) equation under power-law trapping potentials $V_s(r) = r^s/s$ , where the exponent $s$ ranges from $-2$ to $4$. The study addresses the interplay between kinetic energy, contact interactions (attractive or repulsive), and external rotation across this family of traps. A central focus is the scaling behavior of the condensate size near the collapse threshold for attractive interactions and the Thomas–Fermi (TF) limit for repulsive interactions, particularly in the presence of rotation and singular potentials (specifically the inverse-square potential at $s=-2$ ).

Methodology
The authors employ a combination of analytical scaling arguments and rigorous numerical simulations:

Scaling Analysis: The study utilizes dilation arguments to derive scaling laws for the condensate width $\ell$ and energy. It establishes that under norm-preserving dilation, kinetic and contact energies scale identically ( $\lambda^2$ ), while the trap energy scales as $\lambda^{-s}$ .
Numerical Tools:
- 2D Imaginary-Time Propagation: Normalized gradient flow using Strang splitting to compute stationary states for both attractive and repulsive regimes, with and without rotation.
- Radial Solvers: High-precision radial solutions (finite volume, bordered Newton iteration, Richardson extrapolation) serve as benchmarks for non-rotating states.
- Bogoliubov Spectra: Calculation of excitation frequencies to determine stability thresholds (specifically the dipole frequency) for centered states.
- Variational Bounds: Minimization of the energy over a gauge-boosted Townes trial function to bound ground-state energies from above.
Regularization: Singular traps ( $s \le 0$ ) are regularized with a cutoff $\epsilon$ (e.g., $V_{s,\epsilon} = (r+\epsilon)^s/s$ ) to handle numerical divergences, with specific attention paid to the limit $\epsilon \to 0$ for the inverse-square case.

Key Contributions and Results

1. Scaling Laws and Collapse Dynamics

Attraction: The condensate collapses at the Townes coupling $g^*$ regardless of the trap shape or rotation speed (provided the trap confines). The approach to collapse follows the law $\ell \propto \delta^{1/(s+2)}$ , where $\delta = 1 - |g|/g^*$ . This law holds for seven different exponents ( $s \in \{1, 2, 3, 4, -0.5, -1, -1.5\}$ ).
Regularized Cores: For singular traps with a finite cutoff $\epsilon$ , if the condensate width becomes comparable to $\epsilon$ , the scaling law transitions to a "cone law" $\ell \propto \delta^{1/3}$ , determined by the local linear shape of the regularized core rather than the far-field power law.
Repulsion: For repulsive gases, the cloud expands to a Thomas–Fermi radius $R_{TF} \propto g^{1/(s+2)}$ . The TF energy error is shown to split exactly into a density-gradient part and a flow part (mismatch between vortex flow and rigid rotation), with the flow part being comparable to or larger than the gradient part in singular traps.

2. Rotation and Symmetry Breaking

Dipole Criterion: For attractive clouds in traps with $s > 2$ $s > 2$ , the centered state becomes unstable when the rotation frequency $\Omega$ $Ω$ exceeds the dipole frequency $\omega_1$ $ω_{1}$ of the cloud.
- For $s=2$ (harmonic), $\omega_1$ is constant (Kohn mode), so rotation does not induce symmetry breaking below the centrifugal limit.
- For $s > 2$ , the trap is "soft" near the center; $\omega_1$ decreases as the cloud concentrates ( $\omega_1 \propto \ell^{(s-2)/2}$ ). Consequently, rotation can destabilize the centered state even for weak attraction.
Quartic Trap ( $s=4$ ): The threshold for symmetry breaking is found to be $\delta_c \simeq 0.324 \Omega^6$ . This result, derived from a Townes trial, is confirmed by Bogoliubov calculations within 3%.
Post-Instability State: Once unstable, the attractive cloud moves off-axis and collapses onto the minimum ring of the rotating-frame potential $V_{eff} = V - \frac{1}{2}\Omega^2 r^2$ . The width of this ring-localized lump is set by the curvature of the bare trap, not the effective potential.
Repulsive Rotation: In repulsive gases, rotation nucleates vortices. In quartic traps, the TF density empties at the center to form an annulus at a critical reduced rotation $\zeta_c = (3/\pi)^{1/6}$ , independent of the interaction strength $g$ .

3. The Inverse-Square Endpoint ( $s = -2$ )

Covariance: At $s=-2$ , the regularized energy is exactly covariant with the cutoff $\epsilon$ ( $E \propto \epsilon^{-2}$ ). There is no intrinsic length scale; the system falls to the center for both attractive and repulsive interactions.
Halo Structure: For repulsive interactions, the condensate forms a dense core surrounded by a universal halo with density profile $|\psi|^2 \propto 1/(gr^2)$ . The extent of this halo, $L_h$ , grows exponentially with the coupling: $L_h \propto \exp(g / [\pi(1+\kappa)])$ , where $\kappa$ characterizes the tail strength.
Universality: The halo profile and its exponential growth rate are universal (independent of the specific regularization form), whereas the core properties depend on the regularization details.

4. Numerical Artifacts Near Collapse

The paper identifies a first-order stationary bias in split-step imaginary-time propagation near the collapse threshold. This bias causes the computed width to be slightly larger than the true ground state.
Crucially, this error is invisible to standard residual tests (which measure the GP equation residual) because the error scales with the breathing mode. The bias grows as $1/\delta$ near the threshold. A symmetric step scheme or extrapolation to zero time-step is required to eliminate this error.

Significance and Claims
The authors claim three primary new results:

Rotation-Induced Collapse: The precise mechanism and threshold ( $\delta \propto \Omega^6$ ) by which rotation moves an attractive condensate off-axis in soft traps ( $s>2$ ), leading to ring collapse.
Inverse-Square Endpoint: The exact covariance of the energy with the cutoff and the derivation of the universal exponential halo for repulsive gases in the $s=-2$ limit.
Numerical Bias: The identification and characterization of a first-order stationary bias in imaginary-time methods near collapse, which is undetectable by standard convergence criteria but significantly affects the width.

The paper does not claim novelty for established results such as the Townes threshold, the general scaling law $\ell \propto \delta^{1/(s+2)}$ , or the center-of-mass origin of angular momentum in attractive clouds; rather, it tests these known laws quantitatively across the power-law family and integrates them into a unified framework. The work provides a comprehensive account of how trap geometry, interaction sign, and rotation jointly determine the stability and structure of 2D BECs, particularly in regimes involving singular potentials and near-critical collapse.

More like this