⚛️ quantum physics

Symmetries and model reduction in a $\cos 2\varphi$ transmon qubit

This paper develops a symmetry-preserving model reduction for the experimentally realized $\cos 2\varphi$ transmon qubit by deriving an effective single-band Born-Oppenheimer model that clarifies its protection mechanisms and identifies a parameter regime capable of suppressing flux-noise dephasing without compromising charge-noise protection.

Original authors: Erwan Roverc'h, Kyrylo Gerashchenko, Alvise Borgognoni, Benoit Douçot, Alexandru Petrescu, Rémi Robin, Zaki Leghtas

Published 2026-11-02

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Erwan Roverc'h, Kyrylo Gerashchenko, Alvise Borgognoni, Benoit Douçot, Alexandru Petrescu, Rémi Robin, Zaki Leghtas

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the race to build quantum computers, scientists are constantly searching for ways to protect fragile information from the noisy world around it. Imagine trying to keep a delicate house of cards standing in a hurricane; the slightest gust of wind can knock it over. In the realm of superconducting circuits, which are the leading hardware for quantum computing, this "wind" comes in the form of tiny fluctuations in electric charge and magnetic fields. These fluctuations cause errors, destroying the quantum state before it can be used. To solve this, researchers design special circuits that embed a kind of built-in shield directly into their structure. These shields work like strict rules of nature that forbid the quantum information from changing its state unless a very specific, rare event occurs. By making it impossible for common noise to disturb the system, these "protected" qubits could hold their information for much longer than current technology allows. One promising design for such a protected qubit is the cos 2φ transmon, a circuit that uses a specific arrangement of superconducting loops and junctions to create a symmetry that guards against the most common types of errors.

A team of researchers has now taken a deep dive into this specific circuit, mapping out exactly how its complex internal parts work together to create this protection. The circuit they studied is not a simple loop but a sophisticated assembly containing three different modes of oscillation, or ways the electricity can move, all coupled together. Because these modes interact in complicated ways, it is difficult to see the big picture of how the protection works or to predict how the circuit will behave under different conditions. To cut through this complexity, the researchers applied a powerful mathematical strategy known as the Born-Oppenheimer approximation. This method relies on the fact that some parts of the circuit are "heavy" and move slowly, while others are "light" and move very quickly. By treating the fast-moving parts as if they instantly adjust to the position of the slow-moving parts, the team was able to simplify the entire system into a much more manageable model. This reduction allowed them to see the underlying symmetry clearly: the circuit effectively creates a landscape where the quantum state is trapped in a valley that repeats every half-turn, making it impossible for certain types of noise to push the state out of its safe zone.

The researchers found that this simplified model, when used correctly, accurately predicts the energy levels of the full, complex circuit. They discovered that the protection relies on a specific symmetry called Cooper-pair parity, which ensures that the qubit's two main states are fundamentally different in a way that common noise cannot bridge. However, their analysis also revealed a critical weakness. While the circuit is excellent at resisting noise related to electric charge, it remains vulnerable to fluctuations in magnetic flux threading through the loop. In the standard operating regime, where the circuit behaves like a single, simple band of energy, the researchers confirmed a long-standing trade-off: improving protection against magnetic noise inevitably makes the circuit more sensitive to electric charge noise, and vice versa. This limitation had previously suggested that there was a hard ceiling on how well this type of qubit could perform.

Yet, the study did not stop at confirming this limitation. By extending their model to include higher energy bands—essentially looking at the next layer of complexity in the circuit's behavior—the team uncovered a way to break this trade-off. They identified a specific set of parameters where the circuit operates beyond the simple single-band description. In this regime, the higher energy bands actively participate in shaping the qubit's state, providing an extra degree of freedom. This allows the researchers to tune the circuit so that it suppresses magnetic noise dephasing without sacrificing its protection against charge noise. The result is a pathway to a qubit that can potentially achieve much longer coherence times than previously thought possible, all while maintaining the same fundamental protection against other sources of error like dielectric loss. The findings suggest that by carefully engineering the circuit to operate in this multi-band regime, scientists can design a quantum bit that is robust against the two most dominant sources of noise, opening a new door for building stable quantum computers.

Technical Summary: Symmetries and Model Reduction in a $\cos 2\phi$ Transmon Qubit

Problem Statement
Protected superconducting qubits, such as the $\cos 2\phi$ transmon, offer a pathway to longer coherence times by embedding symmetries directly into the Hamiltonian to forbid transitions between qubit states. However, the physical realization of these qubits often involves complex circuits with multiple nonlinearly coupled modes (e.g., a three-mode circuit based on a Kinetic Interference coTunneling Element, or KITE). This complexity obscures the relationship between circuit parameters and qubit protection, making it difficult to navigate the parameter space to maximize coherence. Specifically, there is a need for reduced low-energy models that inherit circuit symmetries to elucidate selection rules and identify optimal regimes that suppress dominant noise sources (charge noise, flux noise, and dielectric loss) without introducing new vulnerabilities.

Methodology
The authors develop a systematic model reduction for the $\cos 2\phi$ transmon circuit, which consists of a superconducting island shunted by a large capacitor ( $E_{CS}$ ) and a two-Cooper-pair interferometer (KITE) composed of two arms, each containing a Josephson junction ( $E_J, E_{CJ}$ ) and an inductor ( $E_L$ ).

Born-Oppenheimer (BO) Approximation: Leveraging the separation of energy scales where the capacitive ratio $\epsilon = E_{CS}/E_{CJ} \ll 1$ , the authors treat the island phase $\phi$ as a "heavy" variable (nucleus) and the interferometer phases $\theta_{L/R}$ as "light" variables (electrons).
Single-Band Reduction: By freezing the heavy mode and diagonalizing the light-mode Hamiltonian, they derive an effective single-band Hamiltonian. Crucially, they include a geometric correction ( $U_{geom}$ ) to the potential, which is significant in their parameter regime and was previously omitted in similar analyses.
Multi-Band Extension: To capture higher-order corrections and regimes where the BO gap closes, the authors extend the model to include multiple energy bands of the light modes (up to a 3-band model), treating the system wavefunction as a spinor.
Symmetry Analysis: The authors identify symmetry operators ( $S$ and $P$ ) for both the full three-mode Hamiltonian and the reduced models. They derive exact selection rules for charge and flux branch operators based on the commutation and anti-commutation relations of these symmetries.
Noise and Coherence Analysis: They project noise operators (dielectric loss, flux noise, charge noise) onto the reduced basis to determine their impact on $T_1$ (relaxation) and $T_2$ (dephasing) times, specifically analyzing the curvature of the qubit frequency with respect to noisy parameters.

Key Contributions and Results

Effective Hamiltonian Derivation: The paper derives an effective single-band BO model that explicitly reveals the $\pi$ -periodic potential characteristic of the $\cos 2\phi$ qubit. The inclusion of the geometric correction $U_{geom}$ is shown to be essential for accurately modeling the potential landscape.
Quantitative Accuracy of Multi-Band Models: Numerical benchmarks demonstrate that while the single-band model accurately describes the ground state doublet, it deviates significantly at $\phi_{ext} = \pi$ and for higher energy levels. In contrast, the 3-band BO model quantitatively captures the full 3-mode circuit spectrum over a 0–4.5 GHz bandwidth with a maximal relative error of 3%, while being computationally more efficient than diagonalizing the full Hilbert space.
Exact Selection Rules: The analysis establishes rigorous selection rules. For a symmetric KITE at $\phi_{ext} = \pi$ , transitions within the qubit doublet are protected against charge noise ( $\hat{N}$ ), symmetric phase drops ( $\hat{\theta}_\Sigma$ ), and symmetric number operators ( $\hat{n}_\Sigma$ ). However, the doublet remains vulnerable to anti-symmetric operators ( $\hat{n}_\Delta, \hat{\theta}_\Delta$ ), which break Cooper-pair parity and correspond to single Cooper-pair tunneling.
Flux/Charge Tradeoff and its Resolution:
- In the standard single-mode regime ( $E_{CS} \ll \Delta$ ), the authors confirm the "flux/charge tradeoff": optimizing for low charge-noise dephasing requires reducing $E_L$ , which simultaneously reduces the plasmon frequency, leading to thermal activation and poor flux-noise protection.
- Beyond Single-Mode: The paper identifies a parameter regime where the BO gap closes ( $E_{CS} \gtrsim \Delta$ ), necessitating the multi-band description. In this regime, higher energy bands contribute significantly to the confinement of the qubit states. This allows $E_L$ to be reduced (suppressing flux-noise dephasing) while tuning the energy gap $\Delta$ to maintain the qubit frequency and plasmon gap. This effectively escapes the flux/charge tradeoff, enabling a design with suppressed flux noise without degrading charge-noise protection or thermal stability.

Significance
The paper provides a rigorous theoretical framework for understanding and designing protected $\cos 2\phi$ transmon qubits. By deriving reduced models that preserve the underlying symmetries, the authors clarify the mechanism of Cooper-pair parity protection and identify the specific noise channels that limit coherence. Most notably, the work challenges the previously assumed limits of coherence in interference-based qubits by demonstrating that moving beyond the single-band approximation allows for the mitigation of flux-noise dephasing without sacrificing protection against charge noise. This suggests that the multi-mode nature of the circuit, often viewed as a source of decoherence, can be leveraged as a resource to shape eigenstates into a more robust, protected regime.

More like this