⚛️ quantum physics

Optimizing LZSM protocol for high-fidelity gates in open-system fluxonium

This paper proposes and analyzes a fast, high-fidelity quantum gate protocol for fluxonium qubits based on one-period Landau-Zener-Stückelberg-Majorana (LZSM) driving, providing analytical tools to optimize parameters, mitigate leakage, and evaluate performance in open-system regimes with strong dissipation.

Original authors: Santiago Ferreyra, Valentın Reparaz, Maria Jose Sanchez, Leandro Tosi, Daniel Dominguez

Published 2026-07-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Santiago Ferreyra, Valentın Reparaz, Maria Jose Sanchez, Leandro Tosi, Daniel Dominguez

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

Imagine you're trying to spin a coin on a table to make it land on heads. Usually, you'd give it a gentle, rhythmic nudge (like a standard radio signal) to get it spinning. But if the coin is heavy and the table is wobbly, that gentle nudge is too slow, and if you nudge it too hard, it flies off the table entirely. This is the problem scientists face with certain types of quantum "coins" (qubits) that spin very slowly. They are too slow to be useful, and trying to speed them up with standard nudges causes them to leak energy into the wrong places.

Enter the Fluxonium, a special kind of quantum circuit that acts like a coin with a very stiff, bouncy edge. Because of this stiffness, the researchers in this paper found a way to spin it incredibly fast using a different trick: a Landau-Zener-Stückelberg-Majorana (LZSM) protocol. Think of this not as a gentle nudge, but as a single, massive, perfectly timed "whack" with a hammer.

The "Whack" vs. The "Nudge"

In the old way of doing things (resonant driving), you have to keep pushing the coin for a long time to get it to flip. The faster you want it to go, the harder you push, but pushing too hard makes the coin wobble and fly off the table (this is called "leakage" into higher energy states).

The Fluxonium circuit is special because it has a "sweet spot" where it's very stable. The authors simulated a new method where they apply a single, strong pulse of magnetic force for just one cycle. It's like hitting a golf ball: instead of tapping it gently for ten seconds, you swing the club once with perfect timing. This "one-period" swing is fast and, if done right, lands the ball exactly where you want it.

The "Leakage" Trap

Here's the catch: even with this powerful swing, there's a risk. If you swing at the wrong speed, the ball might not just go to the hole; it might bounce off a tree and land in a bush (leaking into a third energy level). The paper shows that this leakage is the main enemy.

However, the researchers discovered a secret code. They found that if you tune the speed of your swing to specific "privileged" frequencies, the ball magically avoids the trees. Specifically, if the speed of your swing matches the gap between the first and second excited states divided by a whole number (like 2, 3, 4, etc.), the leakage vanishes. It's like finding a rhythm where the wind stops blowing the ball off course.

By finding these "leakage valleys," they simulated that they could perform a quantum gate (a logic operation) in just 1.92 nanoseconds (that's 1.92 billionths of a second!) with an error rate as low as 10⁻⁵. That's incredibly fast and accurate.

The "Heavy" Trade-off

The paper also looked at what happens if you make the Fluxonium circuit even "heavier" (by changing its internal electrical properties). In this "heavy-fluxonium" regime, the leakage becomes almost non-existent, and the error rate could theoretically drop to 10⁻¹⁰. That's practically perfect!

But there's a price to pay. In this heavy mode, the coin spins so slowly that the "whack" takes longer to complete because the coin itself is sluggish. The total time for the gate goes up, limited by how long you have to wait for the coin to settle. It's like trying to spin a giant, heavy steel wheel; you can hit it perfectly, but it takes longer to get it moving and stopping.

The Heat Problem

Finally, the authors simulated what happens when you add real-world messiness: heat. Even at super-cold temperatures (15 millikelvin), if the coin spins too slowly (below 200 MHz), the heat from the environment starts to jostle it, causing errors.

They found a "Goldilocks zone" for the circuit's design. If you tune the circuit so that its inductive energy is about 0.1 times its Josephson energy, and its charging energy is about 0.5 times the Josephson energy, you get the best of both worlds. In this sweet spot, the simulations suggest you can keep the gate time under 1.3 nanoseconds while keeping the total error (including heat and leakage) below 10⁻⁵.

What This Means (and What It Doesn't)

The paper doesn't claim to have built this perfect gate in a lab yet. These results come from simulations and mathematical models. They have derived the formulas and run the numbers to show that this should work.

They explicitly argue against the idea that standard, slow, gentle nudges are the best way to control these slow-spinning qubits. Instead, they suggest that this "one-period whack" (LZSM) is the superior path. They also note that while they focused on single-qubit gates, the math suggests this could eventually be extended to two-qubit gates, but that's a future step, not a current result.

In short: The paper suggests that by hitting a quantum coin with a perfectly timed, single, strong swing—and tuning the swing speed to avoid "leakage" trees—you can spin quantum logic gates faster and more accurately than ever before, provided you build the coin just right.

Technical Summary: Optimizing LZSM Protocol for High-Fidelity Gates in Open-System Fluxonium

Problem Statement
Quantum gates based on resonant Rabi oscillations face inherent limitations, particularly for low-frequency qubits and systems with low anharmonicity. These limitations include slow gate speeds (bounded by the qubit frequency) and susceptibility to errors arising from counter-rotating terms and leakage into non-computational levels. While non-resonant driving protocols, specifically Landau-Zener-Stückelberg-Majorana (LZSM) interferometry, offer a pathway to fast, high-fidelity rotations using single-period sinusoidal pulses, existing analytical tools are restricted to pure two-level systems (TLS) and unitary evolution. These tools fail to account for the multi-level structure of realistic superconducting circuits and the complex dependence of system-bath coupling on strong drive amplitudes, which is critical for LZSM protocols.

Methodology
The authors develop a theoretical framework to optimize LZSM quantum gates for the fluxonium architecture, explicitly incorporating multi-level dynamics and open-system effects.

Multi-Level Coherent Dynamics:
- The study begins with a standard two-level system (TLS) model driven by a single-period sinusoidal pulse with idle times for phase correction.
- For the fluxonium circuit, the authors move beyond a bare truncation of the Hamiltonian to the lowest two levels. Instead, they employ a Floquet-Schrieffer-Wolf (FSW) transformation. This perturbative approach derives an effective two-level Hamiltonian that renormalizes the qubit frequency ( $\tilde{\omega}_q$ ) based on the drive amplitude ( $A_e$ ) and frequency ( $\omega_d$ ), accounting for the influence of higher energy levels ( $j \geq 2$ ).
- This analytical FSW model is used to map the parameter space $(A_e, \omega_d)$ where a $\pi/2$ rotation (specifically a $Y(\pi/2)$ gate) is achieved, defined by the condition $P_{01} = 1/2$ .
Leakage Analysis:
- The authors quantify leakage errors by calculating the probability of population loss to states outside the computational subspace.
- They identify that the primary leakage channel is the drive-induced transition from the first excited state $|1\rangle$ to the second excited state $|2\rangle$ , governed by the matrix element $\phi_{12}$ .
- By analyzing the Fourier transform of the driving pulse, they identify "privileged" driving frequencies $\omega_d^n = \omega_{21}/n$ (where $n \geq 2$ ) at which the transition probability to $|2\rangle$ vanishes, thereby suppressing leakage.
Open-System Fidelity Formalism:
- To address dissipation, the authors develop a robust formalism based on the Floquet-Born-Markov master equation.
- Unlike standard models that neglect drive-amplitude dependence in system-bath coupling, this approach captures the modification of relaxation ( $T_1$ ) and dephasing ( $T_2$ ) times under strong driving.
- The model assumes dielectric loss in the shunting capacitor as the dominant noise channel, characterized by a super-ohmic noise spectrum.
- Gate fidelity is calculated by transforming the master equation solution back to the static eigenbasis and comparing the resulting process matrix to the ideal target unitary.

Key Results

Analytical Optimization: The FSW analytical curve accurately predicts the parameter regions for high-fidelity gates in fluxonium, correcting the deviations seen in bare truncation models at large amplitudes.
Leakage Suppression: The study confirms that operating at specific "privileged" frequencies ( $\omega_d = \omega_{21}/n$ ) creates deep valleys in the error landscape. For a moderate fluxonium architecture ( $E_J, E_C, E_L = \{4, 1, 1\} \times 2\pi$ GHz), optimal parameters yield coherent gate errors on the order of $10^{-5}$ with gate times as short as 1.92 ns.
Architecture Dependence:
- Reducing the inductive energy ratio ( $E_L/E_J$ ) and increasing the charging energy ratio ( $E_C/E_J$ ) improves fidelity by increasing anharmonicity and reducing leakage.
- In the "heavy-fluxonium" regime (very low $E_L/E_J$ and $E_C/E_J$ ), coherent errors can theoretically reach $10^{-10}$ . However, this regime introduces longer gate times due to reduced qubit frequencies (increased Larmor period) and significant thermal occupation errors.
Open-System Performance: When including thermal fluctuations and dissipation (with $T_b = 15$ mK and $T_1^0 = 500$ µs), the optimal operating point shifts. The authors identify a regime around $E_L/E_J \approx 0.1$ and $E_C/E_J \approx 0.5$ where the total open-system error remains below $10^{-5}$ with gate times under 1.3 ns.
Error Sources: The analysis distinguishes between regimes where coherent leakage dominates (large amplitudes) and where decoherence/thermal effects dominate (low frequencies/heavy-fluxonium regime).

Significance and Claims
The paper claims to provide a comprehensive theoretical toolkit for implementing fast, high-fidelity single-qubit gates in fluxonium qubits using LZSM protocols. By deriving analytical expressions that account for multi-level structures and developing a formalism for open-system fidelity in the strong-driving regime, the work offers a practical guide for experimental calibration.

The authors assert that for specific device parameters, LZSM-based control can outperform standard resonant Rabi driving protocols, particularly for low-frequency qubits where resonant driving is inherently slow and error-prone. The work suggests that while the heavy-fluxonium regime offers superior coherence, a balanced architecture is required to mitigate thermal errors and maintain fast gate speeds. The authors also note that while the current analysis focuses on single-qubit gates, the methodology could potentially be extended to two-qubit gates using tunable coupling schemes, positioning LZSM protocols as a foundational element for quantum computation with small-frequency qubits.