Short answer: Optical tweezer arrays hold individual neutral atoms in light traps, and they are now being pushed to thousands of atoms and onto photonic chips. A 2025 Nature paper reports an array of more than 6,100 neutral atoms in about 12,000 sites with a hyperfine qubit coherence time of 12.6 s, while two earlier papers place tweezer-trapped atoms next to nanophotonic structures. These are laboratory results, and the largest claim comes from the authors' own abstract.
What tweezer arrays are used for
The 2025 abstract says optical tweezer arrays have transformed atomic and molecular physics and now form the backbone of leading experiments in quantum computing, simulation and metrology. It notes that typical experiments trap tens to hundreds of atomic qubits, and that systems with around 1,000 atoms were realized recently without defining qubits or demonstrating coherent control. Scaling to thousands of qubits with long coherence times and low-loss, high-fidelity imaging is called an outstanding challenge, important for quantum error correction.
2025: more than 6,100 atoms
The authors realize an array of optical tweezers trapping more than 6,100 neutral atoms in around 12,000 sites. They report a coherence time of 12.6 s with an uncertainty of 0.1, which they call a record for hyperfine qubits in an optical tweezer array. Room-temperature trapping lifetimes are about 23 minutes, which they say enables an imaging survival of 99.98952% and an imaging fidelity above 99.99%, both described as record-high. They present a plan for zone-based quantum computing and demonstrate coherence-preserving qubit transport and pick-up and drop-off operations on large spatial scales, characterized with interleaved randomized benchmarking. They conclude that universal quantum computing and error correction with thousands to tens of thousands of physical qubits could be a near-term prospect. That conclusion is the authors' outlook, and the abstract does not demonstrate a computation at that scale. The abstract does not name the atomic species, so this post does not either.
2019: atoms on a photonic chip
A 2019 Nature Communications paper implements a configurable optical tweezer array over a planar photonic circuit designed for cold atom integration. The authors report trapping and high-fidelity imaging of one or more atoms directly on a photonic structure. Using an optical conveyor belt, formed by a moving optical lattice within a tweezer potential, they transport single atoms from a reservoir to close proximity to the photonic interface. They say this could allow a defect-free atom-nanophotonic hybrid lattice, and point toward on-chip many-body quantum optics.
A nanofiber array
A Japanese Journal of Applied Physics Express paper demonstrates a one-dimensional array of single atoms interfaced with an optical nanofiber, 300 nm in diameter. Single atoms are deterministically trapped in free space and along the fiber with tunable atom number and spacing. Atomic fluorescence is collected into guided modes for real-time detection, and the guided fluorescence increases with atom number and depends on spacing. The authors say the platform enables a fiber-integrated quantum interface for quantum communication and scalable networks, which is a stated aim.
Limits
The three papers address different goals, scale in the first and photonic interfaces in the other two, so none confirms the others. The scale paper reports coherence and imaging figures, not a large entangled computation. Record claims are the authors' own, and the near-term prospect is a projection.
FAQ
What is an optical tweezer array?
A set of tightly focused laser beams, each holding a single atom.
How many atoms were trapped in 2025?
More than 6,100 atoms in about 12,000 sites, per the authors.
How long did the qubits stay coherent?
12.6 s for hyperfine qubits, as reported by the authors.
Why put atoms near photonic structures?
To make tunable atom-atom interactions through photons and to interface atoms with fibres and chips.
Is a thousand-qubit computer here?
No. The abstract reports a large array and qubit transport, with computing described as a prospect.
All field notes