arXiv:2610.25020
Fast and High-Fidelity Motional-State-Preserving Transport in Optical Tweezers
Wen-hao Wang, Jian-gong Hu, Guang-ming Huang, Gao-xiang Li, Cheng Sheng
Abstract
Reconfigurable neutral-atom processors require rapid, high-fidelity transport of atoms between optical-tweezer sites. In finite-depth Gaussian traps, fast motion can leave residual motional excitation even when the atom reaches its destination. We develop a basis-parameterized gradient-ascent pulse-engineering (GRAPE) framework for one-dimensional transport of a single $^{87}$Rb atom. A shortcut-to-adiabaticity reference is refined using 17 trajectory parameters, with ten additional parameters providing bounded trap-depth modulation; depth affects both confinement and the inverse construction of the physical trap center. For a $12.6\,μ$m transfer completed in only $30.5\,μ$s, corresponding to approximately $2.75$ periods of the reference trap, the selected joint protocol achieves a ground-state infidelity of $2.41\times10^{-7}$. Because preparing an exactly pure motional ground state is experimentally challenging, we also optimize population-weighted transport for initial distributions with mean vibrational occupations $\nbar\simeq0.16$ and $1.87$, reaching same-state probabilities of about $0.99993$ and $0.9934$. Trap-depth fluctuations can further degrade transport fidelity. Robust optimization under a measurement-informed depth-noise model improves the mean error, worst sampled error, and fidelity dispersion for the specified noise realizations. Coordinated control of tweezer position and confinement thus offers a route to rapid transport with low residual excitation for flexible atom rearrangement and dynamic connectivity in neutral-atom processors.
AI Summary
The paper studies rapid optical-tweezer transport that returns an atom to its initial vibrational-state label despite Gaussian-trap anharmonicity. It numerically models one-dimensional transport of a single $^{87}\mathrm{Rb}$ atom, refining a harmonic shortcut-to-adiabaticity reference with basis-parameterized GRAPE. Seventeen coefficients control the trajectory; ten optional coefficients modulate trap depth within $\pm5\%$. Gradients include depth’s effects on both confinement and the inverse-constructed trap center. For a $12.6\,\mu\mathrm{m}$ transfer in $30.5\,\mu\mathrm{s}$, approximately 2.75 trap periods, selected trajectory-only and joint controls give ground-state infidelities of $2.93\times10^{-7}$ and $2.41\times10^{-7}$. An earlier trajectory-only iterate reaches $2.33\times10^{-7}$, so this comparison does not establish a better best-achieved joint optimum. Optimizing one waveform for thermal populations gives weighted same-state probabilities of $0.99993$ at $\bar n\simeq0.16$ and $0.9934$ at $\bar n\simeq1.87$. Under a specified quasistatic-plus-two-frequency depth-noise model, robust refinement reduces mean error from $8.12\times10^{-5}$ to $6.99\times10^{-5}$ and worst sampled error from $1.09\times10^{-4}$ to $8.33\times10^{-5}$ across 192 paired scenarios. Results use full Gaussian wave-function propagation, adjoint-gradient checks, and frozen-control repropagation with refined grids. They are numerical predictions, not demonstrated transport fidelities. Limitations include omitted transverse dynamics, scattering and internal-state decoherence; no hard actuator-bandwidth or slew constraints; limited basis-size and initialization testing; and noise amplitudes without absolute calibration. The noise evaluation set also participates in acceptance. The multistate objective does not preserve arbitrary coherent superpositions, and sampled scans establish neither global optima nor speed limits.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.