Classical field simulation of vortex lattice melting in a two-dimensional fast rotating Bose gas

This paper presents a classical field simulation study using the stochastic projected Gross-Pitaevskii equation to investigate the thermal melting of a two-dimensional vortex lattice in a fast-rotating Bose gas, revealing clear signatures of the Kosterlitz-Thouless-Halperin-Nelson-Young melting scenario and demonstrating the crucial role of finite-size effects on defect proliferation and melting temperatures.

The Gist

Imagine a giant, invisible dance floor made of super-cooled atoms. This isn't just any dance floor; it's spinning incredibly fast, like a record player on maximum speed. Because the atoms are so cold and moving so fast, they don't just swirl randomly. Instead, they organize themselves into a perfect, rigid grid of tiny whirlpools, called vortices. Think of it like a massive, microscopic honeycomb made of spinning water, where every single cell is perfectly aligned with its neighbors.

This paper is about what happens when you start to warm up this frozen dance floor.

The Big Question: How Does a Perfect Grid Break?

In our everyday world, if you heat up a solid (like ice), it melts into a liquid all at once. But in this tiny, 2D world of spinning atoms, physics behaves differently. The scientists wanted to know: Does this perfect grid melt instantly, or does it go through a weird middle stage?

According to a famous theory called KTHNY (named after four physicists), the melting happens in two distinct steps, like a three-act play:

1. Act 1: The Solid (The Crystal): Everything is perfect. Every atom has exactly six neighbors, forming a beautiful hexagonal pattern.
2. Act 2: The Hexatic (The "Almost" Solid): As it gets warmer, the grid starts to get messy. The atoms lose their strict positions (they can wiggle around), but they still try to face the same direction. It's like a crowd of people who have stopped holding hands in a perfect circle, but they are still all facing the center. The structure is "loose" but still has a sense of order.
3. Act 3: The Liquid (The Chaos): Get it even warmer, and the directionality breaks too. Now, everyone is spinning and moving randomly. The honeycomb is gone; it's just a soup of swirling atoms.

What Did the Scientists Do?

The researchers couldn't just watch this happen easily in a real lab because the temperatures are so low and the systems are so small that it's hard to see the "middle act" clearly. So, they built a virtual reality simulator (a computer model) to watch the melting process in slow motion.

They used a mathematical tool called the SPGPE (which is a fancy way of saying "a computer program that simulates how atoms behave when they are hot and spinning"). They created a digital version of the spinning gas and slowly turned up the heat.

The Key Findings

1. The Two-Step Melting Confirmed: The simulation showed that the theory was right! The grid didn't just melt instantly. It went through that weird "Hexatic" middle stage where the atoms were loose but still somewhat organized.
2. The "Small Room" Problem: The scientists noticed something interesting about the size of their simulation. Because their digital dance floor was relatively small (only about 10,000 atoms), the edges of the room messed things up.
   • Analogy: Imagine trying to tile a floor. If you have a huge floor, you can make a perfect pattern in the middle. But if you have a tiny room, the walls force you to cut tiles in weird shapes, creating "defects" (gaps or misaligned tiles) right near the edge.
   • In their simulation, these "edge defects" made it harder to see the melting happen at the exact temperature the big theories predicted. The smaller the system, the more the melting temperature seemed to drop.
3. The Mystery of the "Too High" Prediction: Previous theories predicted that the melting temperature would be much higher than what was actually seen in real experiments. The scientists found that their simulation matched the real experiments (which were lower than the theory), but they still couldn't fully explain why the old theory was so wrong. They suspect it's because the old theory assumed the grid was perfectly rigid and infinite, which isn't true for these tiny, wobbly systems.

Why Does This Matter?

This isn't just about spinning atoms. It helps us understand how order turns into chaos in many different systems, from how magnets lose their magnetism to how materials behave at the quantum level.

In a nutshell: The scientists used a super-computer to watch a spinning cloud of atoms melt. They proved that melting happens in two steps, not one, and showed that the size of the system plays a huge role in how and when it melts. It's like discovering that a small ice cube melts differently than a giant iceberg, even if they are made of the same water.

Technical Summary

1. Problem Statement

The paper addresses the thermal melting of a two-dimensional (2D) vortex lattice in a fast-rotating Bose-Einstein condensate (BEC). While the melting of 2D crystals is theoretically described by the Kosterlitz–Thouless–Halperin–Nelson–Young (KTHNY) scenario—a two-step process involving a transition from a solid (crystalline) phase to a hexatic phase, and finally to a liquid phase—experimental observations in fast-rotating superfluids (specifically Ref. [21]) have shown a melting temperature significantly lower than the theoretical upper bounds predicted by KTHNY theory.

The primary objectives of this study are:

• To numerically simulate the thermal equilibrium state of a vortex lattice in a quasi-2D rotating Bose gas.
• To investigate the melting scenario and identify the signatures of the two-step KTHNY transition.
• To clarify the role of finite-size effects in the discrepancy between experimental melting temperatures and theoretical upper bounds.

2. Methodology

The authors employ a Classical Field Simulation approach using the Stochastic Projected Gross-Pitaevskii Equation (SPGPE).

• Physical Model:

  • System: A 2D Bose gas of $^{87}\text{Rb}$ atoms in a harmonic plus quartic trap.
  • Equation: The SPGPE is used to describe the finite-temperature dynamics, extending the mean-field Gross-Pitaevskii equation to include thermal fluctuations and dissipation.
  • Potential: $V(r) = \frac{M}{2} [\omega_r^2 r^2 (1 + \frac{\kappa r^2}{a_r^2})] + \frac{M}{2} \omega_z^2 z^2$, where $\kappa$ is a small parameter for the quartic correction.
  • Frame: The simulation is performed in a rotating frame with frequency $\Omega$.

• Numerical Implementation:

  • Basis Expansion: The field $\psi_C$ is expanded onto single-particle orbitals $\phi_n$ (a mix of Laguerre-Gauss and Hermite-Gauss bases) up to an energy cutoff $E_{cut}$.
  • Dynamics: The mode amplitudes evolve according to a stochastic differential equation involving a damping term ($\gamma$) and Gaussian white noise ($\eta$) satisfying the fluctuation-dissipation theorem.
  • Parameters: Simulations use $N = 10^4$ atoms (smaller than the experimental $10^5$ to reduce computational cost while maintaining 2D regime validity). Rotation frequencies $\Omega$ range from $0.95\omega_r$ to $1.0\omega_r$. Temperatures are varied to probe the melting transition.
  • Equilibration: Systems are evolved from a vacuum state to a steady state, and 12 independent samples are extracted for statistical analysis.

• Observables and Analysis:

  • Vortex Identification: Vortices are located at density minima. A Delaunay triangulation reconstructs the lattice.
  • Correlation Functions:
    • Translational Order: Pair correlation function $g(r)$, fitted to extract the translational correlation length $\ell_P$.
    • Orientational Order: Six-fold orientational correlation function $G_6(r)$, fitted to extract the orientational correlation length $\ell_G$.
  • Defect Statistics: Counting the number of vortices with 5, 6, or 7 neighbors (disclinations and dislocations).
  • Transition Criteria: Melting temperatures ($T_{s/h}$ for solid-hexatic and $T_{h/l}$ for hexatic-liquid) are defined by the behavior of $\ell_P$, $\ell_G$, and the fraction of 7-fold coordinated sites ($f_7$).

3. Key Contributions

• First Classical Field Study of Melting: This is the first study to use a classical field model (SPGPE) to simulate the thermal melting of vortex lattices in fast-rotating superfluids.
• Validation of KTHNY Scenario: The simulations provide clear numerical evidence supporting the two-step KTHNY melting scenario in this specific quantum system.
• Finite-Size Analysis: The work explicitly demonstrates how finite-size effects (due to the limited number of vortices in the simulation) influence the proliferation of defects and the determination of transition temperatures, particularly at lower rotation frequencies.
• Phase Diagram Construction: A detailed phase diagram mapping the Crystal, Hexatic, and Liquid phases as a function of rotation frequency and temperature is presented.

4. Results

• Two-Step Melting: The simulations successfully reproduce the KTHNY sequence:
  1. Solid to Hexatic ($T_{s/h}$): Characterized by the loss of quasi-long-range translational order ($\ell_P$ decreases) while orientational order ($\ell_G$) remains algebraic. Defects (dislocation pairs) unbind.
  2. Hexatic to Liquid ($T_{h/l}$): Characterized by the loss of orientational order ($\ell_G$ drops exponentially) and the proliferation of free disclinations (isolated 5- and 7-fold defects).
• Transition Temperatures:
  • Both $T_{s/h}$ and $T_{h/l}$ decrease as the rotation frequency $\Omega$ increases.
  • The hexatic-liquid transition ($T_{h/l}$) shows a stronger dependence on $\Omega$ than the solid-hexatic transition.
  • Uncertainties in $T_{s/h}$ are larger due to the smoothing of crossovers caused by finite system size.
• Comparison with Theory:
  • The observed melting temperatures are approximately half of the theoretical upper bound derived in Ref. [41] ($k_B T_m \approx \mu / (8\sqrt{3}\tilde{g})$).
  • This discrepancy mirrors experimental findings, suggesting that the incompressible lattice assumption or the neglect of finite-size effects in the analytical bound may be the cause.
• Finite-Size Effects: At lower rotation frequencies (smaller lattices), boundary effects artificially introduce 5-fold defects, complicating the analysis. The study suggests that larger system sizes (higher atom numbers) are required to fully resolve the transition boundaries and reduce the discrepancy with theoretical bounds.

5. Significance

• Bridging Theory and Experiment: The work provides a crucial numerical link between the KTHNY theoretical framework and recent experimental observations in rotating BECs, confirming that the low melting temperatures are a real physical phenomenon rather than an experimental artifact.
• Methodological Advancement: It demonstrates the utility of SPGPE for studying finite-temperature equilibrium properties in quantum gases, even in regimes where the high-occupation number assumption ($k_B T \gg \mu$) is technically marginal.
• Future Directions: The paper highlights the need for larger-scale simulations (higher $N$) to better quantify finite-size scaling and to explore the crossover from 2D to 3D melting. It also suggests that the SPGPE framework is a powerful tool for investigating non-equilibrium dynamics, such as defect formation during temperature quenches (Kibble-Zurek mechanism).

In conclusion, this study confirms the validity of the KTHNY melting scenario in fast-rotating 2D Bose gases and attributes the gap between theoretical upper bounds and observed melting temperatures largely to finite-size effects and the specific nature of the vortex lattice shear modulus in these systems.