We study the optimization of the transport and transfer of neutral atoms between optical tweezers, both critical steps in the implementation of quantum computers and simulators. We analyze four experimentally relevant pulse shapes (piecewise linear, piecewise quadratic, minimum jerk, and a combination of linear and minimum jerk), and we also develop a protocol using shortcuts-to-adiabaticity (STA) methods to crucially incorporate the time-dependent effects of static traps. By computing a measure of the final transport error and two measures of the heating during transport, we show that our proposed STA protocol comprehensively outperforms all the experimentally inspired pulses. After further optimizing the pulse shapes, we find a lower bound on the protocol duration, compatible with the time at which the vibrational excitations exceed half of the states hosted by the moving tweezer. This lower bound is at least eight times faster than the one reported in recent experiments, which highlights the importance of including and optimizing the transfer from and to static traps, which may be the largest bottleneck to speed. Finally, our STA results demonstrate that a modulation in the depth of the moving tweezer designed to time-dependently counteract the effect of the static traps is key to reducing errors and reducing the pulse duration. To motivate the implementation of our STA pulses in future experiments, we provide a simple analytical approximation for the moving-tweezer position and depth controls.
We present our experimental and theoretical framework, which combines a broadband superluminescent diode with fast learning algorithms to provide speed and accuracy improvements for the optimization of on-dimensional optical dipole potentials, here generated with a digital micromirror device. To characterize the setup and potential speckle patterns arising from coherence, we compare the superluminescent diode to a single-mode laser by investigating interference properties. We employ machine-learning tools to train a physics-inspired model acting as a digital twin of the optical system predicting the behavior of the optical apparatus including all its imperfections. Implementing an iterative algorithm based on iterative learning control we optimize optical potentials an order of magnitude faster than heuristic optimization methods. We compare iterative model-based “offline” optimization and experimental feedback-based “online” optimization. Our methods provide a route to fast optimization of optical potentials, which is relevant for the dynamical manipulation of ultracold gases.
We study the dilute and ultracold unitary Bose gas, characterized by a universal equation of state due to the diverging s-wave scattering length, under a transverse harmonic confinement. From the hydrodynamic equations of superfluids we derive an effective one-dimensional nonpolynomial Schrödinger equation (1D NPSE) for the axial wavefunction which, however, also takes into account the transverse wavefunction. By solving the 1D NPSE we obtain meaningful analytical formulas for the dark (gray and black) solitons of the bosonic system.