Abstract
The dynamics of a metastable attractive Bose-Einstein condensate (BEC) trapped by laser beams is analyzed under small intensity fluctuations. We demonstrate that such fluctuations inevitably trigger a collapse. The expected collapse time is found to be inversely proportional to the integrated covariance of the laser intensity's autocorrelation and decays logarithmically with the number of atoms. Numerical simulations of the stochastic 3D Gross-Pitaevskii equation validate these analytical predictions for small to moderate mean-field interactions.
Introduction: The Stability of Attractive Condensates
In ultracold atomic physics, species like $^7$Li or $^{85}$Rb exhibit attractive interatomic interactions. Without external intervention, these atoms would naturally clump together and collapse. To study these systems, researchers employ magnetic or optical traps to create a metastable state. However, real-world optical traps are generated by lasers with inherent intensity fluctuations. This paper investigates whether this "noise" merely causes the condensate to oscillate or if it fundamentally limits its lifetime through a process of stochastic parametric resonance.
Theoretical Framework: The Variational Approach
To model the system, the authors simplify the 3D Gross-Pitaevskii equation using a Gaussian ansatz for the condensate's wavefunction. This reduction yields an ordinary differential equation (ODE) for the condensate width $a(t)$, which can be interpreted as the motion of a particle in an effective potential $U(a)$.
As shown in Figure 1, the system resides in a local minimum at $a_0$. For the condensate to collapse, it must overcome the energy barrier at $a_1$. Laser intensity fluctuations act as a stochastic force that can eventually push the system over this barrier, leading to $a \to 0$.
Stochastic Analysis and Mean Hitting Time
By transforming the width equations into action-angle variables, the authors treat the evolution of the system's "action" as a diffusion Markov process. Solving the associated Fokker-Planck equation allows for the derivation of the "mean hitting time"—the average time required for noise to induce a collapse.
Key insights from this analysis include:
- Inevitability: Collapse is certain given enough time; noise acts cumulatively.
- Logarithmic Sensitivity: The collapse time $T_c$ decays logarithmically with the number of atoms $N$. As $N$ approaches the trap's maximum capacity, the system becomes exponentially more sensitive to noise.
- Kramers' Analogy: The process is analogous to a particle escaping a potential well due to thermal fluctuations, though here the "temperature" is replaced by the power spectrum of the laser noise.
Numerical Validation and Results
The theoretical predictions were tested against both simplified ODE simulations and full 3D stochastic Gross-Pitaevskii simulations.
As summarized in the table within Figure 4, the analytical model is remarkably accurate for small nonlinearities ($P \le 0.2$), with errors often below 1%. The histograms illustrate the distribution of collapse times, showing a characteristic "long tail" typical of stochastic exit problems. For typical experimental parameters, the predicted lifetime of 1 to 10 seconds matches observed laboratory data.
Limitations and Future Directions
While the Gaussian variational approach is highly effective for low atom counts, it begins to fail near the critical limit ($P_c$). In this regime, full 3D simulations show faster collapse than predicted, likely due to "radiation effects" where the condensate sheds atoms or deforms in ways the Gaussian model cannot capture. Furthermore, real-world effects like three-body recombination—a significant source of atom loss—are not yet included in this model. Future work will aim to incorporate these complex dynamics to better describe behavior near the absolute threshold of stability.
Summary updated