Microwave SQUID multiplexing proposed for
Researchers propose adapting microwave SQUID multiplexing to read out large arrays of massive cryogenic calorimeters, enabling arrays of thousands of
A new readout concept could enable arrays of thousands of massive cryogenic calorimeters for rare-event physics searches. The proposal, detailed in a recent arXiv submission, applies microwave SQUID multiplexing (µMUX) to large, slow calorimeters like those used in dark matter and neutrino experiments.
Currently, arrays of these highly sensitive detectors are limited to just tens of channels. Each transition-edge sensor (TES) requires its own dc-SQUID amplifier chain and several wires. This wiring complexity becomes impractical for larger systems. The new concept seeks to break this bottleneck.
The multiplexing challenge for slow signals
The detectors in question are massive calorimeters, often using bismuth germanate (BGO), sapphire, or tellurium dioxide (TeO₂) absorbers, cooled to about 20 millikelvin. They produce exceptionally slow thermal signals. These signals need only about a kilohertz of sampling bandwidth per channel.
For such systems, the limiting factor is not raw bandwidth. It is current sensitivity. The researchers' analysis shows the multiplexer's input coil must achieve a sensitivity between 0.1 and 1 microampere per flux quantum (µA/Φ₀). This requirement exceeds the capability of current µMUX devices by a factor of 10 to 30. It must be matched to the detector's intrinsic current noise, which sits between 9 and 10 picoamperes per root hertz (pA/√Hz).
Noise performance and scaling
The team developed a comprehensive noise model for the multiplexed readout chain. It accounts for contributions from the high-electron-mobility transistor (HEMT) amplifier, two-level system noise, and the SQUID itself. They then analyzed how this total noise scales with the number of channels multiplexed together, denoted as N_mux.
| Parameter | Value |
|---|---|
| Required input-coil sensitivity | 0.1 to 1 µA/Φ₀ |
| Detector current noise | 9 to 10 pA/√Hz |
| Total flux noise (readout) | ~1.2 µΦ₀/√Hz |
| Maximum N_mux with <2% resolution degradation | 1000 |
The results are promising. With a total readout flux noise of approximately 1.2 micro flux quanta per root hertz (µΦ₀/√Hz), the multiplexing system degrades the detector's baseline energy resolution by less than 2% for multiplexing factors up to 1000. This penalty is negligible compared to the resolution limit set by the TES sensor itself.
A practical array design
The theoretical framework has been translated into a concrete single-tower design. This unit would consist of 52 BGO crystals, each weighing 100 grams, read out by a single HEMT amplifier via the microwave SQUID multiplexer. Such a tower forms the building block for a larger, multi-tower array with a total mass of several kilograms.
These kilogram-scale arrays are the target. They are designed for sensitive searches for physics beyond the Standard Model. Primary applications include the detection of coherent elastic neutrino-nucleus scattering (CEνNS) and interactions from potential dark matter particles. The proposed readout architecture makes these large-scale experiments practically feasible. It moves beyond the constraints of single-channel wiring.
The work, submitted on August 27, 2026, provides a detailed pathway to scaling up cryogenic calorimeter arrays. It leverages technology developed for faster microcalorimeters and adapts it for the unique demands of massive, slow detectors. The final design hinges on achieving the specified increase in input-coil sensitivity.
