Non-equilibrium dynamics in tunnel-coupled Bose gases

This thesis presents four theoretical studies on pairs of tunnel-coupled one-dimensional Bose gases. It first recapitulates how at low energies, the gases’ symmetric and antisymmetric combinations of density and phase are governed by two Tomonaga-Luttinger Liquids (TLL). Tunneling between the gases is believed to yield a quantum sine-Gordon model for the antisymmetric sector. The theoretical background and various applications of this model are summarized, before presenting the first new theoretical results. These show in detail how matter-wave interference measurements can access eigenvalues of the relative phase operator from the TLL, allowing to study the operator’s distribution functions and multi-point correlators. The derivation clarifies why this construction is limited to short expansion times and weak interactions and what modifications occur away from this limit. This leads to a new formula predicting longitudinal “density ripples”. The second half of the thesis was stimulated by recent and unexplained experimental results in the tunnel-coupled case, where density-phase oscillations were seen to rapidly damp out. The work first studies whether this can be explained within a translationally invariant sine-Gordon model. Treating this model in a self-consistent harmonic approximation leads to a negative conclusion. Second, the next leading perturbation due to the tunnel-coupling is investigated in a box geometry. Although this yields a non-negligible coupling between the (anti)symmetric sectors, the effects are not strong enough to explain the damping. Finally, a new low-energy theory is developed, which does not rely on the TLL and which allows to study the roles of both higher excited levels of the transverse potential and of a realistic longitudinal potential. Strong damping is observed as a result of the longitudinal potential, with a dependence on the particle number that is compatible with experimental results. This indicates that performing the experiments in a hard-wall box potential might eliminate the damping effects.