TLS and Quasiparticle Loss in Thin-Film Aluminum CPW Resonators: A Modified Model and Design Implications

This paper presents measurements of thin-film aluminum CPW resonators used for KID development, reporting ultra-high quality factors and proposing a modified model to explain deviations from standard two-level system loss behavior at low powers and temperatures below 60 mK.

The Gist

Imagine you are trying to listen to a very faint whisper from a distant star. To do this, you need a super-sensitive microphone. In the world of astronomy, scientists use Superconducting Resonators as these microphones. They are tiny circuits made of metal that vibrate at specific frequencies, much like a guitar string. When a photon (a particle of light) hits the circuit, it changes the vibration slightly, telling the astronomer, "A star just blinked!"

However, there's a problem: these circuits are often "noisy." Just like a guitar string that is slightly frayed or sitting on a dusty table, these circuits lose energy. This loss is called noise, and it drowns out the faint whispers from space.

This paper is about a team of scientists at NASA who built a new, ultra-clean "guitar string" (a resonator) made of thin aluminum and figured out exactly why it was noisy, and how to make it quieter.

Here is the breakdown of their discovery in everyday terms:

1. The Goal: The Perfect Silence

The scientists wanted to build a detector so sensitive it could hear the faintest signals from the early universe. To do this, they needed the circuit to vibrate for a long time without losing energy. They measured how "pure" the vibration was using a score called the Quality Factor (Q). A higher score means less noise and a better detector. They achieved incredibly high scores, meaning their circuits were very quiet.

2. The Two Main "Noise Makers"

The team discovered that the noise in their circuits came from two main culprits, which they named TLS and Quasiparticles.

• TLS (Two-Level Systems): The "Jittery Neighbors"
  Imagine your circuit is a quiet library. TLS are like tiny, jittery atoms stuck in the walls or on the floor of the library. They are constantly flipping back and forth between two states (like a light switch flickering on and off). Every time they flip, they steal a tiny bit of energy from the vibrating circuit, creating noise.

  • The Problem: Usually, if you shout louder (increase the power), these jittery atoms get "tired" and stop flipping, silencing the noise. This is called "saturation."
  • The Surprise: The scientists found that at very low temperatures, these atoms didn't stop flipping as expected. They kept jittering even when the library was supposed to be quiet. The standard rulebook (the "Standard Tunneling Model") said they should stop, but they didn't.

• Quasiparticles: The "Broken Pairs"
  In a superconductor, electrons usually dance in perfect pairs (Cooper pairs). When energy hits them, they break apart into solo dancers called Quasiparticles. These solo dancers are messy and cause friction (noise) in the circuit.

  • The Fix: The team found that by using a specific design, they could keep these solo dancers under control, especially when the circuit was cold.

3. The New Rulebook (The Modified Model)

Because the "jittery neighbors" (TLS) weren't following the old rulebook, the scientists wrote a new rulebook (a modified model).

They realized that at extremely low temperatures, the "jittery atoms" interact with each other. It's like a crowd of people in a room: if everyone is moving randomly, it's chaotic. But if the room gets very cold and quiet, the people start holding hands and moving in sync. This changes how they react to the circuit's vibration.

• The Analogy: Think of the old model as assuming everyone in a crowd acts alone. The new model realizes that at very low temperatures, the crowd starts acting like a synchronized dance troupe, which changes how much noise they make. This new math allowed them to predict the noise perfectly, even at the coldest temperatures.

4. The Secret Weapon: A Wider Highway

How did they get such good results? They changed the shape of the circuit.

• The Old Way: Most circuits are like narrow, winding alleyways. The electric field (the vibration) is squeezed tight against the walls, where all the "dirt" and "jittery atoms" live. This causes a lot of noise.
• The New Way: The scientists built a wide, open highway (a wide Coplanar Waveguide).
  • Why it works: Because the highway is wide, the vibration spreads out over a larger area. It's like spreading butter on a huge slice of bread instead of a tiny cracker. The "butter" (the vibration) is less concentrated on the "crumbs" (the dirty spots on the surface).
  • The Result: This design allowed them to push the circuit harder (more power) without breaking it. This extra power was enough to silence the "jittery neighbors" completely, revealing a state where the only noise left was the "intrinsic" noise—the unavoidable background hum of the universe itself.

5. Why This Matters

This discovery is a big deal for two reasons:

1. Better Astronomy: We can now build detectors that are quieter and more sensitive, allowing us to see deeper into space and study the birth of stars and galaxies with unprecedented clarity.
2. Better Quantum Computers: These same circuits are used in quantum computers. By understanding how to silence the "jittery neighbors" and manage the "broken pairs," we can build quantum computers that make fewer mistakes and hold their data longer.

In a nutshell: The scientists built a wider, cleaner circuit, realized the old rules for noise didn't work at super-cold temperatures, wrote new rules to explain it, and found a way to silence the noise so effectively that they are now hearing the "true" sound of the universe.

Technical Summary

1. Problem Statement

Superconducting Kinetic Inductance Detectors (KIDs) are critical for sensitive sub-millimeter and far-infrared astronomy, as well as quantum computing. However, their performance is limited by signal loss mechanisms that reduce the internal quality factor ($Q_i$) and optical responsivity.

• The Challenge: Current models, specifically the Standard Tunneling Model (STM), often fail to accurately predict loss behavior at extremely low temperatures ($<60$ mK) and low photon occupation numbers.
• The Gap: There is a need to better understand and model the interplay between Two-Level Systems (TLS), quasiparticle generation, and intrinsic losses to optimize resonator designs for higher sensitivity and lower noise.

2. Methodology

The authors conducted a comprehensive experimental and theoretical study using diagnostic thin-film Aluminum (Al) Coplanar Waveguide (CPW) resonators.

• Device Fabrication:
  • Material: 23 nm thick Al film sputter-deposited on a float-zone Silicon (Si) substrate.
  • Process: Included native oxide removal via HF dip and in situ reverse-bias cleaning to minimize contaminants.
  • Geometry: A test chip containing 16 resonators (15 $\lambda/4$ and 1 $\lambda/2$) coupled to a central feedline. The design utilized wide CPW geometries (10 $\mu$m width, 5 $\mu$m gap) and varying coupling capacitor finger lengths (0, 10, 30 $\mu$m) to tune coupling quality factors ($Q_c$).
• Experimental Setup:
  • Devices were tested in a dilution refrigerator with magnetic shielding.
  • Variables: Bath temperature ($T_{bath}$) ranged from 9 mK to 400 mK; read-tone power ($P_{read}$) ranged from -137 to -67 dBm.
  • Measurement: Vector Network Analyzer (VNA) used to measure transmission ($S_{21}$) to extract resonant frequencies and quality factors.
• Theoretical Modeling:
  • Loss Decomposition: Total loss ($Q_i^{-1}$) was modeled as the sum of quasiparticle loss ($Q_{i,qp}^{-1}$), TLS loss ($Q_{i,TLS}^{-1}$), and other intrinsic losses ($Q_{i,other}^{-1}$).
  • Modified TLS Model: The authors proposed a modification to the standard STM to account for deviations observed at low temperatures. The modification introduces:
    1. Temperature-dependent relaxation ($T_1$) and dephasing ($T_2$) times, where $T_2$ scales with $T_{bath}^{-\mu}$.
    2. An empirical exponent $\beta$ to account for non-uniform saturation of TLS modes by photon occupation ($\bar{n}_{ph}^\beta$).
    • Equation: $Q_{i,TLS}^{-1} \propto \tanh(\xi) \left(1 + \frac{\tanh(\xi)\bar{n}_{ph}^\beta}{D T_{bath}^\mu}\right)^{-1/2}$.

3. Key Contributions

1. Modified TLS Loss Model: The paper presents a refined model that successfully describes loss behavior at very low temperatures ($<60$ mK) where the standard STM fails. It incorporates temperature-dependent relaxation/dephasing times and a non-linear photon saturation exponent ($\beta$).
2. Design Implications for Power Handling: The study demonstrates that wide-geometry CPW resonators can handle higher read powers before entering bifurcation (non-linear instability). This allows the device to reach a "TLS-suppressed" regime without triggering quasiparticle-dominated loss.
3. Identification of Intrinsic Loss Dominance: The authors identified a specific operating regime (low $T$, intermediate power) where temperature- and power-independent intrinsic losses ($Q_{i,other}^{-1}$) become the dominant loss mechanism, rather than TLS or quasiparticles.

4. Key Results

• Performance Metrics:
  • Achieved high internal quality factors with minimum loss values of $Q_i^{-1} \approx 3.64 \times 10^{-8}$ (for $\lambda/2$) and $8.57 \times 10^{-8}$ (for $\lambda/4$).
  • Extracted kinetic inductance fractions ($\alpha$) of $\approx 0.26$ and critical temperatures ($T_c$) of $\approx 1.37$ K.
• TLS Parameters:
  • Fitted the modified model to data, yielding a TLS loss tangent product $F\delta_{TLS}^0 \approx (3.3–3.9) \times 10^{-6}$.
  • Determined the microscopic parameter $\mu \approx 1.2–1.3$, consistent with interacting TLS populations in other superconducting circuits.
  • Found the saturation exponent $\beta \approx 0.76–0.78$, indicating that TLS saturation is non-uniform compared to the standard model assumption ($\beta=1$).
• Loss Regimes:
  • High Power/Intermediate Temp: Quasiparticle loss dominates.
  • Low Power/High Temp: TLS loss dominates.
  • Low Temp/Intermediate Power (The "Sweet Spot"): The device enters a regime where $Q_{i,other}^{-1}$ (intrinsic loss) dominates, contributing only 2–15% to total loss, while TLS and quasiparticle contributions are suppressed.

5. Significance

• For KID Development: The findings provide a roadmap for designing next-generation KIDs. By optimizing resonator geometry (wide CPW) to increase power handling, engineers can operate devices in a regime where TLS noise is naturally suppressed, leading to higher sensitivity detectors for astronomy.
• For Quantum Computing: The improved understanding of TLS behavior at ultra-low temperatures and the identification of intrinsic loss limits are crucial for improving qubit coherence times.
• Modeling Accuracy: The proposed modified TLS model offers a more accurate tool for predicting device performance in the extreme low-temperature, low-power regimes required for future space-based and quantum applications.

In summary, this work bridges the gap between theoretical loss models and experimental reality in superconducting resonators, offering both a better mathematical description of TLS behavior and a practical design strategy to minimize loss in high-sensitivity detectors.