Trapping Alkaline Earth Rydberg Atoms Optical Tweezer Arrays

Neutral atom qubits with Rydberg-mediated interactions are a leading platform for developing large-scale coherent quantum systems. In the majority of experiments to date, the Rydberg states are not trapped by the same potential that confines ground state atoms, resulting in atom loss and constraints on the achievable interaction time. In this Letter, we demonstrate that the Rydberg states of an alkaline earth atom, ytterbium, can be stably trapped by the same red-detuned optical tweezer that also confines the ground state, by leveraging the polarizability of the ${\mathrm{Yb}}^{+}$ ion core. Using the previously unobserved ${}^3{S}_1$ series, we demonstrate trapped Rydberg atom lifetimes exceeding $100\mu \mathrm{s}$, and observe no evidence of auto- or photoionization from the trap light for these states. We measure a coherence time of $T_2=59\mu \mathrm{s}$ between two Rydberg levels, exceeding the $28\mu \mathrm{s}$ lifetime of untrapped Rydberg atoms under the same conditions. These results are promising for extending the interaction time of Rydberg atom arrays for quantum simulation and computing, and are vital to capitalize on the extended Rydberg lifetimes in circular states or cryogenic environments.

Figure 1

(a) Cartoon of the experiment, showing a six-tweezer array, the Rydberg electron wave function, and the ${\mathrm{Yb}}^{+}$ ion core. (b) Radial probability distributions of Rydberg wave functions relative to the optical tweezers (green). (c) Calculated trap depth for ${}^3{S}_1$ states, normalized to the trap depth for the ${}^1{S}_0$ ground state for the same power and beam waist (here, $0.65\mu \mathrm{m}$).

Figure 2

(a) Survival probability of the $n=75$ ${}^3{S}_1$ state with (black, $\tau =108\mu \mathrm{s}$) and without (red, $\tau =28\mu \mathrm{s}$) the traps. Inset: Trapped Rydberg lifetime $\tau$ of the ${}^3{S}_1$ Rydberg state vs trap power at $n=75$. (b) Trapped Rydberg lifetime of the ${}^3{S}_1$ state vs principal quantum number $n$, with (black) and without (red) the trap. The dashed red line shows the untrapped lifetime of a ground state atom under the same conditions. (c) Trap depth of the ${}^3{S}_1$ Rydberg state vs principal quantum number. The green line is the theoretical trap depth using the calculation from Fig. 1. (d) Relevant Yb energy levels for Rydberg excitation.

Figure 3

(a) Microwave spectrum of the $n=75$ ${}^3{S}_1$ to $n=74$ ${}^3{P}_0$ transition with (black) and without (red) the traps, demonstrating the magic trapping condition. The black data are shifted for clarity and the solid lines are Lorentzian fits. (b) Microwave spectra of the $n=75$ ${}^3{S}_1$ $M_J=-1$ to $n=74$ ${}^3{P}_2$ $M_J=-2,-1$, 0 transitions, showing the tensor light shift of different $M_J$ levels from the ponderomotive potential. For each transition, zero detuning indicates the measured transition frequency without the trap, indicated in the figure. The solid vertical lines show the predicted $M_J^2$ dependence of the tensor light shift.

Figure 4

(a) Two-photon Rabi oscillations between $n=75$ ${}^3{S}_1$ and $n=74$ ${}^3{S}_1$. The solid line is a cosine fit with exponential decay time $\tau =42\mu \mathrm{s}$. Control data without microwave pulses (black data) shows $T_1$ for comparison. (b) Ramsey measurement of $T_2^{*}$. The orange line is a simulation that takes into account dephasing from the differential light shift between the two levels (90 kHz) and a finite atomic temperature ($13\mu \mathrm{K}$), yielding a $1/e$ decay time of $22\mu \mathrm{s}$. (c) Hahn echo measurement. The black line is an exponential fit that yields $T_2=59\mu \mathrm{s}$.