How a trap may explain phase locking in a cold-atom quantum simulator

A quantum simulator is a controllable physical system built to reproduce the behavior of another, often harder-to-study system. In this case, two elongated clouds of ultracold bosonic atoms are coupled by tunneling. Their relative phase—the difference between the phase of one cloud and the other—can act like the field in the sine-Gordon model, a well-studied interacting quantum field theory.

Experiments have supported this connection in equilibrium. After a sudden change, however, the clouds have shown phase locking: their relative phase settles into a narrow range. Calculations for an ideal, uniform sine-Gordon system instead predict rephasing, in which the average relative phase approaches a nonzero value without the same locking behavior. The question for Bence Fitos and Gábor Takács was whether realistic features of the atomic setup could explain that difference while keeping the field-theory description useful.

Simulating the coupled clouds

The authors constructed an effective one-dimensional field-theory model of the two gases. It includes the relative phase, which follows sine-Gordon dynamics, and a common mode describing fluctuations shared by both clouds. Crucially, the model also includes the longitudinal harmonic trap. The trap makes the atomic density vary along the clouds, so the system is not spatially uniform.

They simulated the dynamics using the truncated Wigner approximation, a semiclassical computational method. In broad terms, it represents quantum fluctuations by sampling many starting points in a classical phase space and evolving them with classical equations of motion. The team compared versions of the model with and without three features: the nonlinear cosine interaction, the trap-induced inhomogeneity, and coupling between the relative and common modes. They also examined refinements such as temperature and uncertainty in the initial atom number.

For comparison with experiment, the authors selected parameters based on a reported experimental setup and focused on the relative phase near the center of the clouds. They tracked both the average phase and the average cosine of the phase. The latter helps distinguish genuine locking from a simple drift in the average: a value near one indicates that phase fluctuations are concentrated around alignment.

What the model finds

In the simulations, the harmonic trap’s spatial inhomogeneity can produce phase locking. With the trap included, the average cosine of the relative phase approaches a value close to one, regardless of whether the other two model features are switched on. The timescale is associated with the longitudinal trap. The coupling to the common mode does not appear to be important for this early locking behavior.

The authors also find that a flat-bottom, or box-shaped, trap supports dynamics closer to the standard sine-Gordon expectation. This suggests a practical way to test the interpretation: shape the trap to reduce the density variation and compare the resulting dynamics. It is a prediction supported by their model, not a reported experimental demonstration in this paper.

The model’s picture is therefore not that the atoms cease to simulate sine-Gordon physics. Rather, the trap changes the background on which the field evolves. Once that inhomogeneity is included, the authors argue, the observed out-of-equilibrium behavior can fit within a one-dimensional field-theory account.

Limits and open questions

The conclusion is qualified. The simulations do not reproduce the experimentally reported scaling of the relaxation time as system parameters change. The authors point to possible missing physics, including three-dimensional corrections, and note that the field-theory description becomes unreliable near the low-density edges of the trapped clouds. They also use a simplified model of the initial state; a more realistic account of how the clouds are prepared remains future work.

Long-time revival behavior is especially uncertain. The simulated phase locking occurs before signals from the edges are expected to reach the center, which is why the authors consider their explanation of early locking more robust than their predictions for later revivals. Their results also indicate that finite temperature can damp the dynamics and reduce the eventual phase-locking signal.

This is a theoretical study, not a new experiment. Its conclusions depend on the effective one-dimensional model and the truncated Wigner approximation, and the paper identifies unresolved questions about dimensionality and boundaries. The source is an arXiv preprint, which may not have undergone peer review; readers should treat the proposed explanation and box-trap prediction as results to be tested further.