High-efficiency loading of 2,400 Ytterbium atoms in optical tweezer arrays

Abstract

Neutral atom arrays have emerged as a powerful platform for quantum computation, simulation, and metrology. Among them, alkaline-earth-like atoms exhibit distinct advantages, including long coherence time and high-fidelity Rydberg gates. However, their scalability has lagged behind that of the alkali atoms. Here, we report 2,400 Ytterbium-174 atoms trapped in an optical tweezer array with enhanced single-atom loading efficiency of 83.5(1)%. Notably, the loading efficiency is largely maintained for array sizes ranging from dozens to thousands, exhibiting excellent scalability. We demonstrate the broad applicability of the enhanced loading method by showing that the enhancement exists robustly across a range of interatomic potentials, suggesting its utility for other atomic species. To establish the capability of the 174Yb arrays toward universal quantum computation, we propose to encode the qubit in the ground-clock state manifold and estimate a 99.9% two-qubit gate fidelity with experimentally feasible parameters. Our work advances the prospects for realizing large-scale quantum computers using alkaline-earth-like atoms.

Figures (7)

• Figure 1: High-efficiency loading of over $2,400$ Yb atoms in optical tweezer arrays.a, Averaged fluorescence image of a 2,939-site array over 100 experimental trials. b, Single-shot image of trapped single atoms, with 2,437 atoms loaded. Two small regions are zoomed in to highlight the high loading efficiency. c, A histogram presenting the site-resolved loading efficiency, with the inset illustrating this efficiency across the arrays. The average loading efficiency is 83.5(1)%. Data are collected from 85 images. d, The loading efficiency for different array sizes. Data are collected from 20 images. Error bars are smaller than the marker. Insets show single-shot images of the corresponding array sizes (image of the largest array is in b). e, Polarizabilities for the ground state $^{1}\text{S}_0$ (blue curve) and the excited state $^{3}\text{P}_1$ (red curve) of $^{174}$Yb from numerical calculation, which are used for fluorescence imaging (inset). The optical tweezer operates at the magic wavelength of 532 nm. The enhanced loading in all plots is performed with $\pi$-polarized light, $I/I_{\text{sat}}=17$, $\Delta=2\pi \times6.4 \,\text{MHz}$ and $2.3\,\text{G}$ magnetic field.
• Figure 2: Enhanced loading under different interatomic potentials.a, The enhanced loading mechanism. The solid black curves represent the molecular eigenstates in the rotating frame of the blue-detuned laser, with the gray region highlighting the crossing point. Green and orange arrows mark two pathways for inelastic collision. b, The angular distribution of the simulated inelastic collision probability $P_\text{ic}$ as a function of $\theta$, with $\theta$ being the angle between the interatomic displacement and the quantization axis ($I/I_{\text{sat}} = 100, \Delta/f_{\text{trap}} = 1.5$, close to the optimal point in c, d). The green curve corresponds to a globally repulsive potential, while the yellow curve corresponds to a potential that is repulsive only around $\theta=90^\circ$. c, d The loading efficiency of the globally and partially repulsive potential, respectively. The loading enhancement time is 500 ms. The star labels the maximum loading efficiency [81.1(4)% in c and 74.2(7)% in d]. Two subplots share the same color bar. The black dots mark the optimal parameters of the $P_\text{ic}$ predicted by the theoretical model and black contours mark 90% and 70% of the highest $P_\text{ic}$.
• Figure 3: Loading and imaging performance of the large array.a, Initial loading efficiency distribution with (blue) and without (red) implementing the rotating MOT technique, prior to the enhanced loading step. Data are collected from 100 images. The inset figures illustrate the fixed MOT and the rotating MOT. b, Histogram of fluorescence brightness across all array sites. The two peaks of the bimodal distribution correspond to zero and one atom occupancy. The inset depicts a site-resolved imaging fidelity across the array. The imaging exposure time is 200 ms. Data are collected from 1,250 images. c, Probability density distribution of imaging lifetime across the large array. The inset plots the average atom survival over all sites with respect to the imaging time, showing a lifetime of 35.1(5) s. All data presented in this figure are collected under the maximum array scale with 2,939 optical tweezers.
• Figure 4: Schemes for quantum gates in $^{174}$Yb atoms.a, Level diagram of single- and two-qubit gates. b, Calculated Rydberg-pair interaction potential for the state $|r\rangle = |\nu=49.56, L=0, J=1, m_J=1\rangle$ at a magnetic field of 200 G with atoms aligned perpendicular to the magnetic field. The color of each curve represents its overlap with the target pair state $|r,r\rangle$. c, Laser phase of the time-optimal CZ gate, assuming an interatomic distance of 2.1 $\mu$m. d, Numerical simulation of two-qubit gate error rates.
• Figure 5: a, Schematic of the experimental apparatus. Purple beams denote the 399 nm cooling light used for the 2D MOT. A weak 556 nm beam (green) entering from the left acts as a push beam, transporting atoms into the science chamber before being blocked by an aluminum mirror. The red dashed line indicates the resulting atomic trajectory. Within the science chamber, three orthogonal 556 nm beams form a 3D MOT. An objective mounted above the chamber focuses the 532 nm laser to generate optical tweezers and simultaneously collects single-atom fluorescence for detection. b, Time Sequence for MOT loading, tweezer loading, enhanced loading and imaging. PA denotes photon association. c, The MOT positions at different moments during the loading process. The red region represents the fluorescent atoms in the MOT, and the black dotted grid indicates the relative positions of the optical tweezer arrays.