How they did it: a simple cryostat that keeps optics warm

Previous attempts to extend neutral-atom trapping lifetimes often required placing entire optical systems inside cryogenic enclosures — a nightmare of thermal contraction, vacuum compatibility, and limited optical access. The MPQ team took a different approach.

Cryogenic vacuum enhancement. The team built a cryostat whose cold tip reaches 4 K and is positioned about 30 cm from the tweezer array. Activating the cryostat dramatically reduces the partial pressure of residual hydrogen — the dominant background gas in typical vacuum chambers — even though the cold tip is not in direct contact with the trapping region .

Material choice and baking. The vacuum chamber is made predominantly of grade 2 titanium, which has superior outgassing properties compared to stainless steel, especially for hydrogen. The chamber underwent a 3-week high-temperature bakeout at 320 °C, followed by a 2-week low-temperature bakeout at about 150 °C .

The key design insight: all imaging and tweezer optics are placed outside the glass cell, operating at room temperature in air. This avoids the need for custom vacuum-compatible, cryogenic-compatible objectives that must tolerate thermal contraction and UHV conditions. The glass cell sits in the room-temperature part of the chamber, while only the cold tip deep inside the vacuum envelope is actively cooled .

The trade-off is that atoms are not shielded from room-temperature blackbody radiation, but the authors note this is compatible with future room-temperature microwave shielding for Rydberg-state experiments .

Why two hours matters for scalability

In neutral-atom quantum computing, atoms are loaded randomly into tweezers, then sorted into defect-free arrays by moving atoms around. The sorting time grows linearly with the number of atoms. When the trap lifetime is only a few minutes, the chance of losing an atom during sorting — and thus the probability of ending up with a perfect array — drops exponentially as the array size increases.

With a two-hour lifetime, that bottleneck largely disappears. Combined with full optical access enabling large-field-of-view objectives and high-power lasers for large-volume optical lattices, the MPQ team sees a clear path to arrays with tens of thousands of atoms, potentially exceeding 100,000 .

A separate breakthrough at Humboldt-Universität zu Berlin: trapping atoms with surface forces

While the MPQ result tackles the vacuum-lifetime challenge for large-scale tweezers, a separate group has solved a different problem: how to trap atoms near nanophotonic waveguides without needing multiple laser beams.

Researchers at Humboldt-Universität zu Berlin (the group of Prof. Arno Rauschenbeutel) demonstrated a hybrid nanophotonic trap for cold cesium atoms, reported in a preprint (arXiv:2509.17767, September 2025) and later published in Nature Photonics .

The trap uses surface forces — a combination of the Casimir–Polder interaction plus electrostatic charges on the nanofiber surface — to attract atoms, while a single blue-detuned evanescent laser field provides repulsion . This replaces the standard two-color (red + blue) dipole trap typically needed for nanofiber-based atom trapping. The trap minimum sits about 650 nm from the nanofiber surface and has a depth of about 1 μK .

The approach achieved an atomic storage time of 140(9) ms — longer than in the standard two-color dipole trap, despite the hybrid trap being significantly shallower. More importantly, the Ramsey coherence time reached 16.8(2) ms — over an order of magnitude longer than previous nanophotonic waveguide traps . That's because atoms spend most of their time in low-light regions, reducing differential light-shift decoherence.

The work shows that surface forces, normally considered a nuisance that must be minimized, can be harnessed as a useful trapping tool. This simplifies the optical requirements (fewer laser beams needed) and dramatically improves coherence, making the approach promising for quantum memories, quantum network nodes, and other nanophotonic quantum devices .

How they did it: a simple cryostat that keeps optics warm

Previous attempts to extend neutral-atom trapping lifetimes often required placing entire optical systems inside cryogenic enclosures — a nightmare of thermal contraction, vacuum compatibility, and limited optical access. The MPQ team took a different approach.

Cryogenic vacuum enhancement. The team built a cryostat whose cold tip reaches 4 K and is positioned about 30 cm from the tweezer array. Activating the cryostat dramatically reduces the partial pressure of residual hydrogen — the dominant background gas in typical vacuum chambers — even though the cold tip is not in direct contact with the trapping region .

Material choice and baking. The vacuum chamber is made predominantly of grade 2 titanium, which has superior outgassing properties compared to stainless steel, especially for hydrogen. The chamber underwent a 3-week high-temperature bakeout at 320 °C, followed by a 2-week low-temperature bakeout at about 150 °C .

The key design insight: all imaging and tweezer optics are placed outside the glass cell, operating at room temperature in air. This avoids the need for custom vacuum-compatible, cryogenic-compatible objectives that must tolerate thermal contraction and UHV conditions. The glass cell sits in the room-temperature part of the chamber, while only the cold tip deep inside the vacuum envelope is actively cooled .

The trade-off is that atoms are not shielded from room-temperature blackbody radiation, but the authors note this is compatible with future room-temperature microwave shielding for Rydberg-state experiments .

Why two hours matters for scalability

In neutral-atom quantum computing, atoms are loaded randomly into tweezers, then sorted into defect-free arrays by moving atoms around. The sorting time grows linearly with the number of atoms. When the trap lifetime is only a few minutes, the chance of losing an atom during sorting — and thus the probability of ending up with a perfect array — drops exponentially as the array size increases.

With a two-hour lifetime, that bottleneck largely disappears. Combined with full optical access enabling large-field-of-view objectives and high-power lasers for large-volume optical lattices, the MPQ team sees a clear path to arrays with tens of thousands of atoms, potentially exceeding 100,000 .

A separate breakthrough at Humboldt-Universität zu Berlin: trapping atoms with surface forces

While the MPQ result tackles the vacuum-lifetime challenge for large-scale tweezers, a separate group has solved a different problem: how to trap atoms near nanophotonic waveguides without needing multiple laser beams.

Researchers at Humboldt-Universität zu Berlin (the group of Prof. Arno Rauschenbeutel) demonstrated a hybrid nanophotonic trap for cold cesium atoms, reported in a preprint (arXiv:2509.17767, September 2025) and later published in Nature Photonics .

The trap uses surface forces — a combination of the Casimir–Polder interaction plus electrostatic charges on the nanofiber surface — to attract atoms, while a single blue-detuned evanescent laser field provides repulsion . This replaces the standard two-color (red + blue) dipole trap typically needed for nanofiber-based atom trapping. The trap minimum sits about 650 nm from the nanofiber surface and has a depth of about 1 μK .

The approach achieved an atomic storage time of 140(9) ms — longer than in the standard two-color dipole trap, despite the hybrid trap being significantly shallower. More importantly, the Ramsey coherence time reached 16.8(2) ms — over an order of magnitude longer than previous nanophotonic waveguide traps . That's because atoms spend most of their time in low-light regions, reducing differential light-shift decoherence.

The work shows that surface forces, normally considered a nuisance that must be minimized, can be harnessed as a useful trapping tool. This simplifies the optical requirements (fewer laser beams needed) and dramatically improves coherence, making the approach promising for quantum memories, quantum network nodes, and other nanophotonic quantum devices .