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

TL;DR

This work demonstrates real-time, parity-sensitive detection of correlated quasiparticle bursts in two co-located, charge-sensitive transmons coupled to a common waveguide. By driving both devices at parity-dependent frequencies and analyzing dual-tone microwave scattering, the authors resolve individual parity-switching events with 100 μs resolution and reveal bursts that enhance QP tunneling rates by roughly three orders of magnitude for ~7 ms. A cross-device analysis shows that bursts are temporally correlated across devices (zero-delay peak in $R_{12}(τ)$), with a substantial fraction of bursts coincident between detectors, and a rare subset associated with discrete offset-charge shifts consistent with ionizing events. The study identifies three regimes—uncorrelated baseline switching, correlated bursts, and offset-charge–shifting bursts—and provides a practical framework for diagnosing and mitigating correlated errors in superconducting qubit architectures, with implications for gap engineering and phonon trapping strategies.

Abstract

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.

Figures (6)

• Figure 1: Device schematic and spectroscopy. (a) Schematic of two transmon qubits (each approximated as a two-level system) coupled to a common coplanar waveguide with coupling rates $\Gamma_1$ and $\Gamma_2$. A coherent field $b_{\rm in}$ is sent through the waveguide, and the transmitted field $b_{\rm out}$, is measured at drive frequency $\omega_d$, defining the transmission coefficient $t = \frac{b_{\rm out}}{b_{\rm in}}$. (b) False-colored optical micrograph of the device, where the charge-sensitive transmons are highlighted in blue and green, and the coplanar waveguide is shown in light orange. (c) Representative magnitude of the transmission coefficient for the two detectors being in the even ($+$) or odd ($-$) charge-parity state. The corresponding transition frequencies $\omega_{1}^{\pm}$ and $\omega_{2}^{\pm}$ are indicated by dashed lines. Solid lines show fits to the theoretical transmission response obtained from master-equation simulations combined with input–output theory.
• Figure 2: Representative real-time trajectories and burst classification. Segmented time traces of the demodulated transmission signal for QP detector 1 (QPD1; top traces, green) and QP detector 2 (QPD2; bottom traces, blue), acquired during simultaneous continuous driving at the two parity-dependent transition frequencies $\omega_{01}^{\pm}$ for each device. Individual QP tunneling events are resolved as abrupt transitions between two discrete signal levels (low $\leftrightarrow$ high). In addition to the baseline switching dynamics, we observe intermittent intervals of strongly enhanced activity (“bursts”). The three panels illustrate the burst categories used throughout this work: an uncorrelated burst (I), where elevated switching occurs predominantly in one detector; a correlated burst (II), where both detectors exhibit increased switching over the same time window; and an offset-charge–shifting event (III), where a sudden change in the effective offset charge shifts $\omega_{01}^{\pm}$ and changes the parity-state discrimination. Top insets show expanded views of the time windows indicated by the dashed vertical markers.
• Figure 3: Switching-rate extraction and burst-duration statistics. (a) Histograms show the distribution of inter-event waiting times $t_{wait}$ between consecutive parity jumps detected by QPD1 (top, blue) and QPD2 (bottom, green), evaluated separately for burst intervals (left column, ms scale) and for the baseline switching dynamics (right column, s scale). Black dashed curves are single-exponential fits from which the extracted tunneling rates are acquired. (b) Burst-duration distribution. Histogram of burst durations $\tau_{burst}$ for QPD1 (blue) and QPD2 (green).
• Figure 4: Cross-correlation of burst activity between QP detectors. Intensity cross-correlation $R_{12}(\tau)$ between QPD1 and QPD2 computed from the burst-count time series, where the intensity is defined as the number of identified burst events per $\Delta t=1~\mathrm{ms}$ bin. The solid trace (“All data”) shows a pronounced peak at $\tau=0$, $R_{12}(0)>1$, indicating temporally correlated burst activity across the two devices. At large $|\tau|$ the correlation approaches the uncorrelated baseline $R_{12}(\tau)\simeq 1$. The dashed trace (“Bursts removed”), obtained after excising burst intervals from both time series, yields $R_{12}(\tau)\approx 1$ for all $\tau$, consistent with independent Poissonian dynamics.
• Figure 5: Burst-associated offset-charge shifts in two QP detectors. (a) Two-dimensional normalized histograms of the demodulated IQ response during continuous parity monitoring, showing the two well-resolved parity manifolds for each device. For offset-charge--shifting burst events, the cluster centroids exhibit a sudden displacement (red arrows), consistent with detuning of the monitoring tones due to a change in the effective offset charge. (b) Time evolution of the effective offset charge $n_g$ inferred from the parity splitting extracted from single-tone spectroscopy for QPD1 (top, blue trace) and QPD2 (bottom, green trace). Triangles indicate offset-charge--shifting burst episodes identified in the subsequent continuous QP-monitoring record (QPD1: orange; QPD2: purple); the leftmost markers correspond to a coincident offset-charge--shifting burst in both devices which results in the IQ-cluster displacements shown in panel (a). These events coincide with step-like changes in $n_g$, identifying a rare subclass of bursts associated with discrete offset-charge rearrangements.