Researchers have built a cavity-array microscope with 603 individually addressable optical cavities. That’s a big step forward for scalable light–matter interfaces in neutral-atom quantum tech.
Using free-space intra-cavity optics, the team forms a two-dimensional lattice of tightly packed TEM modes. These modes have wavelength-scale waists and are meant to couple efficiently to large neutral-atom arrays.
The setup achieves high finesse and cooperativity throughout the array. Researchers also dig into the loss channels, stability issues, and tricky alignment steps that matter when scaling up to tens of thousands of cavities.
Platform architecture and performance
The platform uses a new arrangement of intra-cavity optics to build a dense lattice of optical modes. This lets many atoms interact with light at once.
It’s a scalable interface with strong photon–atom coupling and readout capabilities that work with mid-circuit operations. Here’s what stands out in terms of performance.
Key performance metrics
- Number of cavities: 603 cavities, each individually resolved in a 2D array.
- Mode structure and waist: Tightly packed TEM modes, with waists around 1.15 μm.
- Finesse: Average finesse across the array: 114(17).
- Cooperativity: Single-atom peak cooperativity averages above 10 across the array.
- Loss budget: About 5.3% average round-trip loss, factoring in coatings, scattering, and aperture clipping.
- Mode degeneracy: 537 cavities are mutually degenerate within a readout-optimized linewidth (finesse ≈ 26).
- Field of view: Radius of 140 μm, limited by the numerical aperture and lens curvature.
- Atom–surface separation: 5.7 mm—chosen to cut down surface-induced decoherence for Rydberg experiments.
Optical stability and mode characterization
Optical tests showed there’s still some non-degeneracy between modes. Nanometer-scale variations in optical path length, mostly from surface bumps and mechanical stress, cause this.
To check stability, the team scanned aspheric lens positions over an 18 μm range and ran ray-tracing simulations. They found radial shifts in stability due to field curvature and small alignment mistakes.
At the same time, they saw that 537 cavities stayed mutually degenerate within a readout-optimized linewidth. That’s a pretty good sign of robust degeneracy, even if there are imperfections per cavity.
Losses, alignment, and mitigation strategies
The loss budget points to two main culprits for finesse variation: anti-reflective coating losses and scattering from rough surfaces. Local defects and dust add more differences from cavity to cavity—cleanroom handling really matters here.
Researchers outline specific control and alignment tricks to keep the array stable and reduce aberrations. They back this up with realistic simulations.
- Dominant loss sources: AR-coating losses and surface scattering.
- Per-cavity variation: Local defects and dust bring extra finesse variation across the array.
- Mitigation: Alignment protocols and stabilization techniques help minimize aberrations and keep cooperativity high.
Scaling prospects and potential applications
Looking ahead, the researchers want to swap aspheres for high-NA, wide-field microscope objectives. They also aim to double the microlens density and push scaling toward tens of thousands of independent cavities, all while keeping strong cooperativity.
This kind of roadmap could open up wild levels of parallelism in quantum operations and readout. It might let us control large neutral-atom ensembles in ways we’ve never seen before.
- Scaling strategy: Swap out aspheric lenses for high-NA, wide-field objectives. Boost microlens density to ramp up the cavity count, but don’t lose performance.
- Applications: Run quantum operations in parallel, pull off fast non-destructive mid-circuit readout, generate remote entanglement, and dig into hybrid atom–photon Hamiltonians for neutral-atom quantum processors.
Here is the source article for this story: Hundreds Of Miniature Light Traps Built For Future Quantum Technologies