CCAT: Magnetic Sensitivity Measurements of Kinetic Inductance Detectors

This paper presents and compares magnetic sensitivity measurements of three different Kinetic Inductance Detector (KID) designs for the CCAT Observatory's Prime-Cam instrument at 100 mK to evaluate their performance under the magnetic field variations expected during telescope operations.

The Gist

Imagine you are trying to listen to a very faint whisper from a distant star. To do this, you build a giant, super-sensitive ear (a telescope) that sits high up in the Atacama Desert, where the air is thin and dry. This ear uses tiny, super-cooled sensors called Kinetic Inductance Detectors (KIDs). These sensors are like tiny tuning forks made of superconducting metal; when a photon (a particle of light) hits them, they vibrate slightly, changing their pitch. By listening to these pitch changes, astronomers can create pictures of the universe.

However, there's a problem: Magnetism.

Just like how a strong magnet can mess with a speaker or a compass, the Earth's magnetic field can mess with these delicate superconducting tuning forks. As the telescope moves to scan the sky, it cuts through the Earth's magnetic field, and the sensors inside might get "confused" by this magnetic noise.

This paper is essentially a stress test for these sensors. The researchers wanted to know: If we wave a magnet near these sensors, how much does their "voice" change? Will they still be able to hear the stars?

Here is a breakdown of what they did and what they found, using some everyday analogies:

1. The Setup: The "Magnetic Wind Tunnel"

The researchers didn't just guess; they built a controlled environment to test the sensors.

• The Sensors: They used three different types of sensor designs (made of Aluminum, Titanium-Nitride, and a special mix) that are planned for the telescope.
• The Test Chamber: They put these sensors in a "dilution refrigerator," which is like a super-freezer that cools things down to almost absolute zero (colder than outer space!).
• The Magnet: They built a giant pair of coils (like a giant donut-shaped magnet) around the freezer. By turning up the electricity, they could create a "magnetic wind" that blew over the sensors, simulating the magnetic fields the telescope would encounter in the real world.

2. The Experiment: Pushing the Buttons

They slowly increased the magnetic "wind" from zero up to a strong level, and then turned it back down. They watched two things:

1. The Pitch (Frequency): Did the tuning fork's note change?
2. The Clarity (Quality Factor): Did the note become muddy or distorted?

3. The Findings: The "Hysteresis" Hangover

The results were interesting and revealed a few key behaviors:

• The "Sticky" Effect (Hysteresis): Imagine you push a heavy swing. When you stop pushing, it doesn't immediately stop; it keeps swinging a bit. Similarly, when the researchers turned off the magnetic field, the sensors didn't immediately return to their perfect state. They "remembered" the magnetic field. This is called hysteresis. It's like the sensors got a little "hangover" from the magnetic exposure, leaving them slightly less clear than before.
• Direction Matters: The sensors were much more sensitive to magnetic fields hitting them from the top (perpendicular) than from the side (parallel). Think of it like a sail: a wind blowing directly against the sail pushes it hard, but a wind blowing parallel to the sail barely moves it.
• The "Trap": The researchers suspect that tiny bits of magnetic field get "trapped" inside the metal of the sensors, like dust getting stuck in a fan blade, causing that loss of clarity.

4. The Big Question: Will the Telescope Work?

The most important part of the paper is the conclusion. The researchers asked: Is this bad enough to ruin our telescope?

The answer is: No.

Here is why:

• The Earth is Weak: The magnetic field changes the telescope will experience while scanning the sky are very small (about 50 micro-Tesla). In their tests, even at much stronger fields, the sensors only changed their pitch by a tiny, tiny amount.
• The Shield: The telescope isn't naked; it has a special "magnetic raincoat" (a shield made of a material called Cryoperm) around the sensors. This shield is like a fortress wall that blocks 99% of the magnetic noise.
• The "Common" Problem: Even if the magnetic field does cause a tiny shift, it affects all the sensors in the camera at the same time. It's like if a whole choir suddenly sang slightly flat because of a draft in the room. The astronomers can easily tell the difference between a "flat choir" (magnetic noise) and a "solo singer" (a real star).

The Bottom Line

The researchers did a thorough check-up on the sensors for the new CCAT telescope. They found that while the sensors are sensitive to magnets and can get a little "sticky" (hysteresis) when exposed to strong fields, the actual magnetic environment of the telescope is too weak to cause any real trouble.

Thanks to the telescope's built-in magnetic shields and the fact that the Earth's field is relatively gentle, the sensors should be able to listen to the universe's whispers clearly, without getting distracted by the magnetic noise of the planet they are sitting on. The telescope is ready to go!

Technical Summary

1. Problem Statement

The Cerro Chajnantor Atacama Telescope (CCAT) is a ground-based submillimeter-to-millimeter observatory featuring the 6-meter Fred Young Submillimeter Telescope (FYST). Its primary instrument, Prime-Cam, utilizes superconducting microwave kinetic inductance detectors (KIDs) to perform wide-area surveys at scan speeds of degrees per second.

The Core Issue: As the telescope scans the sky, it moves through the Earth's magnetic field. Additionally, internal telescope components may generate magnetic fields. Previous studies indicate that magnetic fields can degrade the performance of superconducting materials used in KIDs. The authors needed to quantify the magnetic sensitivity of specific CCAT KID designs to determine if Earth's magnetic field variations would significantly impact observational data quality or require mitigation strategies.

2. Methodology

The researchers conducted controlled experiments to measure the magnetic sensitivity of three distinct KID designs at cryogenic temperatures (100 mK).

• Device Under Test (DUT): Three "witness chips" were fabricated using the same lithographic processes and materials as the actual Prime-Cam detectors:
  1. Aluminum (Al) 280 GHz: Lumped-element polarimeter design with a meandered absorber.
  2. Titanium-Nitride (TiN) 280 GHz: Lumped-element polarimeter design with short Al patches for responsivity tuning.
  3. EoR-Spec (210–315 GHz): Non-polarized design using a meandered Hilbert curve absorber.
• Experimental Setup:
  • The chips were mounted in non-magnetic copper test boxes and placed inside a Bluefors LD dilution refrigerator (DR).
  • Field Generation: A set of room-temperature Helmholtz coils (radius 44.6 cm, 267 turns) was mounted around the DR to generate a uniform, adjustable DC magnetic field ranging from 0 to 500 µT.
  • Orientation: Measurements were taken with the magnetic field applied both perpendicular and parallel to the detector absorber plane.
  • Shielding: A mu-metal shield was used during cooldown to prevent flux trapping during the transition through the critical temperature ($T_c$). The shield was removed only once the system reached 100 mK for testing.
• Readout & Analysis:
  • Detectors were read out via RF chains with cryogenic amplifiers and an RFSoC system.
  • Scattering parameters ($S_{21}$) were swept to extract the resonant frequency ($f_0$) and internal quality factor ($Q_i$).
  • Data was fitted to a nonlinear resonator model to account for kinetic inductance effects.

3. Key Contributions

• Comparative Sensitivity Analysis: The study provides the first direct comparison of magnetic sensitivity across three different KID material and geometry designs (Al vs. TiN, and different absorber geometries) intended for the CCAT Prime-Cam instrument.
• Orientation Dependence: The paper quantifies the difference in sensitivity between magnetic fields applied perpendicular to the detector plane versus parallel to it.
• Hysteresis Characterization: The authors document the hysteretic behavior of KID parameters (frequency shift and $Q_i$) when exposed to varying magnetic fields, noting that properties do not fully recover after field removal.
• Operational Validation: The study validates the necessity of magnetic shielding during the cooldown process through the superconducting transition temperature ($T_c$) to prevent flux trapping.

4. Key Results

• Perpendicular vs. Parallel Fields: Magnetic fields applied perpendicular to the detector plane caused significant shifts in resonant frequency ($\Delta f/f_0$) and degradation in the internal quality factor ($Q_i$). In contrast, fields applied parallel to the plane resulted in effects that were several orders of magnitude lower.
• Hysteresis and Flux Trapping:
  • A clear hysteresis loop was observed: the resonator properties changed differently when the field was increasing versus decreasing.
  • After a magnetic field sweep, the quality factor ($Q_i$) did not fully recover to its initial value, suggesting the formation of trapped magnetic flux or vortices within the superconducting films.
  • Experiments without the mu-metal shield during cooldown showed a distinct "bump" in frequency shift when sweeping through negative fields, confirming that flux trapping occurs if the shield is absent during the $T_c$ transition.
• Material Comparison: While all three designs (Al 280 GHz, TiN 280 GHz, and EoR-Spec) exhibited sensitivity, the magnitude of the shift was consistent across the different materials for a given field orientation.
• Quantitative Impact: At the maximum expected change in Earth's magnetic field during telescope operation (~50 µT), the fractional frequency shift is estimated to be less than $10^{-5}$.

5. Significance and Implications

• Negligible Impact on Operations: The authors conclude that the magnetic field variations encountered by CCAT while scanning the Earth's magnetic field will have a negligible impact on detector performance.
  • The telescope receiver includes a Cryoperm A4K shield at 4 K, estimated to provide a shielding factor >200. This reduces the external 25 µT Earth field to <0.125 µT at the detector plane.
  • Even without the shield, the calculated signal shift (approx. 0.3 pW optical equivalent) is an extreme upper bound and would manifest as a common-mode signal across the array, similar to atmospheric loading, rather than introducing noise.
• Operational Protocol: The results reinforce the critical importance of maintaining a magnetically shielded environment (mu-metal) while cooling the detectors below their critical temperature ($T_c$) to prevent permanent flux trapping and performance degradation.
• Future Design: The study suggests that future magnetic sensitivity studies should utilize cold Helmholtz coils or cryogenic magnetometers to decouple chip hysteresis from the cryostat's own magnetic hysteresis.

In summary, the paper confirms that while KIDs are sensitive to magnetic fields (particularly perpendicular ones), the combination of the telescope's magnetic shielding and the common-mode nature of the expected field variations ensures that CCAT's Prime-Cam instrument will operate effectively without magnetic interference.