Real-time detection of correlated quasiparticle tunneling events in a multi-qubit superconducting device

Quasiparticle tunneling events are a source of decoherence and correlated errors in superconducting circuits. Understanding and ultimately mitigating these errors calls for real-time detection of quasiparticle tunneling events on individual devices. In this work, we simultaneously detect quasiparticle tunneling events in two co-housed, charge-sensitive transmons coupled to a common waveguide. We measure background quasiparticle tunneling rates at the single-hertz level, with temporal resolution of tens of microseconds. Using time-tagged coincidence analysis, we show that individual events are uncorrelated across devices, whereas burst episodes occur about once per minute and are largely correlated. These bursts have a characteristic lifetime of 7 ms and induce a thousand-fold increase in the quasiparticle tunneling rate across both devices. In addition, we identify a rarer subset of bursts which are accompanied by a shift in the offset charge, at approximately one event per hour. Our results establish a practical and extensible method to identify quasiparticle bursts in superconducting circuits, as well as their correlations and spatial structure, advancing routes to suppress correlated errors in superconducting quantum processors.

💡 Research Summary

This paper presents a comprehensive experimental platform for real‑time detection of quasiparticle (QP) tunneling events and their correlations in a multi‑qubit superconducting device. Two charge‑sensitive transmon qubits (QPD1 and QPD2) are coupled to a common coplanar waveguide. By operating each transmon at the charge‑parity‑dependent transition frequencies (ω⁺ and ω⁻) and continuously probing the transmitted microwave field, the authors achieve single‑event resolution of QP tunneling with a temporal granularity of ~100 µs. The transmons are deliberately designed with EJ/EC≈15, giving a charge‑parity splitting of ~10 MHz—well above the resonator linewidth (Γ≈4 MHz)—so that each tunneling event produces a clear, abrupt shift between two discrete voltage levels in the demodulated signal.

The measurement chain includes a traveling‑wave parametric amplifier (TWPA) to boost signal‑to‑noise ratio, and a dual‑tone drive that simultaneously excites both parity branches of each qubit. Data are collected in 60‑second traces, with a 65 % duty cycle dedicated to QP monitoring and the remainder used for periodic spectroscopy recalibration. After discarding intervals where the parity splitting becomes comparable to the linewidth, 37 % of the data are retained for high‑fidelity analysis.

Statistical analysis separates two regimes: a baseline regime where QP tunneling follows Poisson statistics with rates Γ₁≈5.3 s⁻¹ (QPD1) and Γ₂≈3.6 s⁻¹ (QPD2), and a burst regime where the rates surge to ≈3.8 × 10³ s⁻¹, a three‑order‑of‑magnitude increase. Bursts have an average duration of 7 ms and occur at rates of 1.6 min⁻¹ (QPD1) and 1.0 min⁻¹ (QPD2). By constructing a 1 ms‑binned burst count time series for each detector and computing the cross‑correlation R₁₂(τ), the authors find a pronounced peak at zero delay (R₁₂(0)>1) that is symmetric and centered, indicating that bursts are temporally correlated across the two qubits without a measurable causal ordering within the experimental time resolution. Outside burst intervals, R₁₂(τ)≈1, confirming that background tunneling events are independent.

A third, rarer class of events—occurring at roughly 1–2 per hour—exhibits a simultaneous shift of the IQ‑plane clusters corresponding to the two parity states. This “offset‑charge‑shifting” burst reflects an abrupt change in the effective offset charge on the islands, moving the parity‑splitting frequencies and degrading the discrimination of parity states. Such charge jumps are consistent with ionizing radiation (cosmic rays, radioactive contaminants) that breaks Cooper pairs and modifies the local electrostatic environment.

The device is heavily shielded: an indium‑sealed copper enclosure, high‑energy radiation‑drain filters on all lines, and extensive low‑pass filtering. Nevertheless, bursts persist, suggesting that they originate from high‑energy photons or phonons that penetrate the shielding and generate quasiparticle avalanches. The authors discuss mitigation strategies such as gap‑engineering of the Josephson junctions, improved phonon thermalization, and possibly reducing charge sensitivity to lower the impact of bursts on qubit performance.

In summary, the work demonstrates (i) a practical, scalable method for real‑time QP detection using charge‑sensitive transmons coupled directly to a waveguide, (ii) quantitative evidence that QP bursts are temporally correlated across spatially separated qubits, and (iii) identification of a subset of bursts linked to offset‑charge jumps, likely caused by ionizing events. These findings provide a valuable diagnostic tool for large‑scale superconducting quantum processors, enabling the detection and potentially the real‑time mitigation of correlated error sources that threaten the assumptions underlying quantum error‑correction codes.