Advanced microwave SQUID multiplexer model incorporating readout power effects and Josephson junction inhomogeneities

This paper introduces an advanced model for microwave SQUID multiplexers that accurately predicts resonance characteristics across the full range of screening parameters and accounts for Josephson junction inhomogeneities, significantly improving agreement with experimental data and enabling optimization beyond previously accessible operating regimes.

The Gist

Imagine you are trying to listen to a thousand tiny, whispering radios (superconducting sensors) all at once. These radios are so sensitive they can detect the faintest energy from a single particle hitting them. However, listening to one thousand of them individually would require a thousand wires, which is messy and impossible to manage.

To solve this, scientists invented a "microwave SQUID multiplexer" (or µMUX). Think of this as a super-conductor traffic controller. Instead of a separate wire for every radio, they all share a single highway (a microwave cable). Each radio is assigned a unique "radio station" (a specific frequency). The traffic controller listens to the whole highway and figures out which station is talking by looking at how the signal changes.

The Problem: The Old Map Was Too Simple

For a while, scientists used a mathematical "map" (a model) to predict how this traffic controller behaves. This map worked great when the radios were quiet and simple. But as they tried to make the system more powerful (by making the radios more sensitive), the old map started to fail.

It was like trying to navigate a city using a map that only works for flat, empty streets. Once you hit a steep hill or a busy intersection, the map gives you the wrong directions. Specifically, the old model broke down when the "screening parameter" (a measure of how much the radio fights back against magnetic changes) got too high. It started showing "ripples" and weird glitches that didn't exist in reality, making it impossible to design better, more sensitive detectors.

The Solution: A New, Smarter GPS

The authors of this paper built a new, advanced GPS (a numerical simulation model) to replace the old map.

1. Solving the Puzzle Step-by-Step
Instead of using a simple shortcut (a formula that approximates the answer), their new model acts like a super-smart puzzle solver. It calculates the behavior of the system by breaking it down into tiny, manageable steps and repeating them until the answer is perfect.

• The Analogy: Imagine trying to guess the exact shape of a cloud. The old model tried to guess it by drawing a few straight lines. The new model uses a 3D printer to build the cloud layer by layer, getting the shape exactly right, even if the cloud is weird or lumpy.
• The Result: This new model works perfectly even when the radios are very sensitive (up to a limit called $\beta_L < 1$), covering the entire range of designs scientists actually want to build.

2. The "Bumpy Road" Discovery
Here is the most exciting part. The old model assumed that the "tunnel" inside the radio (where the electricity jumps) was perfectly smooth and uniform, like a brand-new highway.

But in reality, these tunnels are often rough and bumpy, like a dirt road with potholes. The thickness of the barrier the electricity jumps over varies slightly from spot to spot.

• The Analogy: Think of the old model as assuming every car on the highway has the exact same engine and tires. The new model realizes that some cars have slightly worn tires or different engines (inhomogeneities).
• Why it matters: These "bumps" change how the radio behaves in a way that looks similar to making the radio more sensitive, but it's actually a different effect. If you ignore the bumps, you might think your radio is more sensitive than it really is, or you might design the system wrong.

What This Means for the Future

By including these "bumpy roads" in their new model, the authors found that their predictions matched real-world experiments much better than before.

• Better Fit: When they tested their model against real data, the "error" (the difference between the prediction and reality) dropped significantly.
• Beyond the Limit: Their model works even when the system is pushed to its limits, using high power levels that previous models couldn't handle.

The Big Picture

This paper is like upgrading from a paper map to a live, real-time GPS that accounts for traffic, road construction, and potholes.

• For Scientists: It means they can now design much larger and more sensitive arrays of detectors (for things like dark matter research or space telescopes) with confidence.
• For Everyone: It's a step toward building super-powerful sensors that can see the universe in incredible detail, all because someone figured out how to mathematically describe a "bumpy" tunnel.

In short: They built a better calculator that understands the messy reality of the microscopic world, allowing us to build better tools to explore the universe.

Technical Summary

1. Problem Statement

Microwave SQUID multiplexing ($\mu$MUX) is the leading technology for scaling superconducting detector arrays (such as TESs and MMCs) beyond $10^3$ channels. However, optimizing these systems requires accurate models that predict device behavior under varying readout conditions.

• Limitation of Existing Models: The current state-of-the-art analytical model (Wegner et al.) relies on Taylor series expansions. This restricts its validity to SQUID screening parameters $\beta_L < 0.6$.
• Consequences: For modern device designs where $\beta_L$ often exceeds 0.6 (up to the hysteretic limit of $\beta_L = 1$), the analytical model produces unphysical artifacts (ripples) and fails to accurately predict resonance characteristics.
• Missing Physics: Existing models assume ideal Josephson tunnel junctions with homogeneous barriers and purely sinusoidal current-phase relations (CPR). They do not account for barrier inhomogeneities (roughness), which are common in real-world fabrication and significantly alter device behavior.

2. Methodology

The authors developed a numerical simulation framework to overcome the limitations of analytical approximations. The core of the methodology involves:

• Numerical Solution of Implicit Equations: Instead of using Taylor expansions, the model solves the implicit equation for the SQUID supercurrent ($I_S$) numerically using a fixed-point iteration method.
  • The algorithm iteratively calculates the total flux $\phi_{tot}$ and the resulting supercurrent $I_S$ until convergence.
  • It employs acceleration techniques (Aitken's $\Delta^2$ process) to ensure rapid convergence for $\beta_L < 1$.
• Harmonic Analysis: The model computes the time derivative of the supercurrent ($dI_S/dt$) over one oscillation period. Using a Discrete Fourier Transform (DFT), it isolates the fundamental frequency component ($\omega$) to determine the induced current in the load inductor, which dictates the effective inductance and resonance shift.
• Incorporation of Barrier Inhomogeneities:
  • The framework allows for arbitrary Current-Phase Relations (CPR).
  • It integrates a mesoscopic model (based on Willsch et al.) for tunneling through inhomogeneous barriers.
  • The barrier thickness is modeled as a Gaussian distribution with mean $\bar{d}$ and standard deviation $\sigma$.
  • The CPR is expressed as a Fourier series of Josephson energies ($E_{J,k}$), where higher-order terms account for the non-sinusoidal nature of the junction due to barrier roughness.

3. Key Contributions

1. Extended Validity Range: The numerical model is valid for the entire practically relevant design space of non-hysteretic rf-SQUIDs ($\beta_L < 1$), whereas previous models were limited to $\beta_L < 0.6$.
2. Modeling Junction Inhomogeneities: This is the first $\mu$MUX model to explicitly incorporate the effects of inhomogeneous tunnel barrier thickness. It demonstrates that barrier roughness introduces higher-order harmonics in the CPR, distinct from but qualitatively similar to the effects of the screening parameter.
3. Generalized Framework: The simulation structure is flexible, allowing for the modeling of non-tunneling junctions (e.g., Dayem bridges) or specific junction geometries by simply swapping the input CPR function.

4. Results

The authors validated the model against experimental data from a $\mu$MUX channel:

• Comparison with Analytical Model:
  • For $\beta_L \approx 0.65$, the analytical model exhibited significant unphysical ripples in the resonance frequency curve.
  • The numerical model eliminated these ripples and matched the exact power-independent solution in the low-power limit.
  • The coefficient of determination ($R^2$) for the fit improved from 94.6% (analytical) to 95.5% (numerical with ideal junction).
• Impact of Inhomogeneity:
  • Including barrier inhomogeneity (modeled with $\sigma = 0.48$ nm) further improved the fit, raising $R^2$ to 97.4%.
  • The model successfully reproduced the specific asymmetry in the resonance frequency shift and the reduced shift at higher readout powers that the ideal model missed.
• High-Power Performance: The model maintained excellent agreement with experimental data even at readout powers ($\Phi_{rf} \approx 0.78 \Phi_0$) far exceeding typical operating conditions, a regime where the analytical model fails completely.
• Distinction of Effects: The study clarified that while barrier inhomogeneity and high screening parameters both skew the CPR, they produce distinct signatures in the power dependence of the resonance frequency, necessitating separate characterization.

5. Significance

• Design Optimization: This model enables the optimization of $\mu$MUX devices with higher $\beta_L$ values, which were previously inaccessible to simulation. This is crucial for maximizing multiplexing factors and minimizing noise in next-generation detector arrays.
• Accurate Characterization: By accounting for barrier inhomogeneities, the model prevents the misinterpretation of experimental data (e.g., confusing barrier roughness for a high screening parameter), leading to more accurate device characterization.
• Scalability: The framework provides a robust tool for simulating large-scale detector arrays, supporting the development of cryogenic sensors for astrophysics and particle detection where high channel counts are required.

In conclusion, this work replaces a limited analytical approximation with a robust numerical framework that captures the full non-linear dynamics of $\mu$MUX devices, including the critical real-world effects of Josephson junction imperfections.