Trapping of Single Atoms in Metasurface Optical Tweezer Arrays

Optical tweezer arrays have emerged as a key experimental platform for quantum computation, quantum simulation, and quantum metrology, enabling unprecedented levels of control over single atoms and molecules. However, existing tweezer platforms have fundamental limitations in array geometry, size, and scalability. Here we demonstrate the trapping of single strontium atoms in optical tweezer arrays generated via holographic metasurfaces. We realize two dimensional arrays with more than 1000 trapped atoms, arranged in arbitrary geometries with trap spacings as small as 1.5 um. The arrays have a high uniformity in terms of trap depth, trap frequency, and positional accuracy, rivaling or surpassing existing approaches. This is enabled by highly efficient holographic metasurfaces fabricated from high-refractive index materials, silicon-rich silicon nitride and titanium dioxide. Leveraging sub-micrometer pixel sizes and high pixel densities, our platform allows scaling far beyond current capabilities. As a demonstration, we realize an optical tweezer array with 360,000 traps. These advances will facilitate tweezer-array based quantum applications that require large system sizes.

💡 Research Summary

In this work the authors introduce a new platform for generating large‑scale optical tweezer arrays using holographic metasurfaces, and they demonstrate the trapping of single strontium (88Sr) atoms in such arrays. The metasurfaces are fabricated from silicon‑rich silicon nitride (SRN) and titanium dioxide (TiO₂), both high‑index dielectrics that support sub‑wavelength meta‑atoms with a unit‑cell size of 290 nm, widths of 100–190 nm, and a height of 750 nm. By assigning each meta‑atom a specific geometry from a pre‑computed library, the devices impose a spatially varying phase on an incident 520 nm Gaussian beam while preserving its amplitude, thereby acting as a phase‑only hologram that simultaneously generates and focuses a two‑dimensional array of tightly focused traps.

Design of the phase pattern is performed with a Gerchberg‑Saxton algorithm, yielding a diffraction efficiency of roughly 60 % and an effective numerical aperture (NA) exceeding 0.6. The SRN devices tolerate optical intensities of at least 25 W mm⁻², whereas TiO₂ devices survive more than 2 kW mm⁻² without active cooling, demonstrating excellent power‑handling capability. The metasurfaces are mounted on a 20 mm × 20 mm substrate; each individual element has a circular aperture of 1.2–3.5 mm diameter and can be swapped rapidly by translating the substrate.

In the experimental setup, a 520 nm laser is intensity‑modulated by an acousto‑optic modulator (AOM) before illuminating the metasurface. The generated trap plane is relayed through a high‑NA (0.6) microscope objective, a 1:1 telescope, and a second high‑NA (0.5) objective into an ultra‑high‑vacuum glass cell where the atoms are trapped. Strontium atoms are first laser‑cooled on the broad 461 nm transition, with repumpers at 679 nm and 707 nm, and further cooled on the narrow 689 nm intercombination line to reach microkelvin temperatures.

The authors demonstrate a variety of tweezer geometries: a “Statue of Liberty” pattern (183 traps, 3 µm spacing), a Penrose tiling (225 traps, 4 µm spacing), a 32 × 32 square lattice (1024 traps, 2.5 µm spacing), and a “necklace” pattern with 1.45 µm spacing. In the current configuration the total number of trapped atoms is limited by the available 1 W of tweezer‑laser power, yielding arrays of a few hundred atoms.

Single‑atom preparation is shown in a 16 × 16 array. After stochastic loading, the authors apply parity projection via photoassociation on a molecular state near the 689 nm line, which removes atom pairs and leaves either zero or one atom per site. The resulting single‑atom filling fraction is 41 % (≈106 atoms), and fluorescence imaging on the 461 nm transition, combined with simultaneous Sisyphus cooling on the 689 nm transition, provides a photon‑count histogram that clearly separates zero‑atom and one‑atom events. Imaging fidelity exceeds 95 % for the 16 × 16 array and reaches 99 % for a smaller 4 × 4 array.

Uniformity of the traps is quantified by probing each site with the trapped atoms. The standard deviation of trap depth across the 16 × 16 array is 7.5 %, the radial (in‑plane) trap frequency varies by 5 %, and the axial frequency by 8 %. Positional inaccuracies are about 1.5 % of the 4 µm lattice spacing, comparable to the size of the ground‑state wavefunction. These figures are comparable to or better than those obtained with liquid‑crystal spatial light modulators (typically ~10 % depth variation) and demonstrate that metasurfaces can deliver highly uniform large‑scale arrays without the need for active feedback.

A key advantage of metasurfaces is the sub‑wavelength pixel size. The authors derive an analytical expression for the effective NA achievable with a pixel size d at wavelength λ: NA_eff = 1/√