Recovery dynamics of a gap-engineered transmon after a quasiparticle burst

Heekun Nho, Thomas Connolly, Pavel D. Kurilovich, Spencer Diamond, Charlotte G. L. Bøttcher, Leonid I. Glazman, Michel H. Devoret · 2025

Read Recovery dynamics of a gap-engineered transmon after a quasiparticle burst by Heekun Nho, Thomas Connolly, Pavel D. Kurilovich, Spencer Diamond, Charlotte G. L. Bøttcher, Leonid I. Glazman, Michel H. Devoret (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-08104.txt.

Introduction

Heekun Nho and six co-authors watch a superconducting qubit recover after a burst of stray particles knocks it off balance. Radiation hits spray quasiparticles through quantum chips, briefly wrecking coherence and threatening error correction. Their 2025 experiment uses transmon qubits whose energy gaps were engineered to shelter them, and monitors the qubits continuously to catch bursts as they happen. Gap engineering cut the burst rate fivefold. Yet that help fell far short of predictions that assumed the particles cool instantly.

The shortfall itself is the discovery: the particles stay hot and linger, so the shelter leaks. Knowing why recovery is slow guides the next chip designs. The authors tie the lingering to how the burst energy spreads. Their measurements give error-correction teams real numbers for how long a burst poisons a qubit. Radiation, from cosmic rays to materials, is a quiet enemy of quantum machines.

Catalog record

CREATORHeekun Nho, Thomas Connolly, Pavel D. Kurilovich, Spencer Diamond, Charlotte G. L. Bøttcher, Leonid I. Glazman, Michel H. Devoret
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-08104

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Recovery dynamics of a gap-engineered transmon after a quasiparticle burst Preprint: APS/123-QED Heekun Nho, Thomas Connolly, Pavel D. Kurilovich, Spencer Diamond, Charlotte G. L. Bøttcher, Leonid I. Glazman, Michel H. Devoret Abstract Ionizing radiation impacts create bursts of quasiparticle density in superconducting qubits. These bursts severely degrade qubit coherence for a prolonged period of time and can be detrimental for quantum error correction. Here, we experimentally resolve quasiparticle bursts in 3D gap-engineered transmon qubits by continuously monitoring qubit transitions. Gap engineering allowed us to reduce the burst detection rate by a factor of a few. This modest reduction falls several orders of magnitude short of the reduction expected if the quasiparticles quickly thermalize to the cryostat temperature. We associate the limited effect of gap engineering with the slow thermalization of the phonons in our chips after the burst. Introduction.— Quantum error correction protocols redundantly encode a logical qubit in the entangled state of multiple physical qubits Nielsen and Chuang 2010 . This redundancy makes the encoded information robust to errors occurring on individual physical qubits, as long as two assumptions hold. First, errors should be uncorrelated between different physical qubits. Second, errors occurring on the same qubit should not have temporal correlations. In superconducting processors, common errors – such as qubit relaxation due to material defects – generally satisfy these assumptions. It has been shown, however, that there are rare error burst events that violate both conditions. During these error bursts (hereby referred to as “bursts”), the qubit relaxation rates dramatically increase across multiple qubits simultaneously for a prolonged duration Vepsäläinen et al. 2020 ; Martinis 2021 ; Wilen et al. 2021 ; McEwen et al. 2022 ; Thorbeck et al. 2023 ; Harrington et al. 2024 ; Li et al. 2024 . Mitigating bursts is therefore essential to realize large-scale quantum error correction. The underlying mechanism of the bursts is the impact of ionizing radiation on the chip hosting the qubits 1 1 1 There is another mechanism for bursts, where a phonon shower is generated by the avalanche-like release of stress in the chip Yelton et al. 2025 . Its experimental signatures and mitigation strategies should be similar to those for high-energy impacts. . A high-energy particle striking the chip produces a cascade of high-energy phonons that rapidly propagate throughout the substrate. When the phonons collide with the superconducting films of the qubits, they break Cooper pairs, producing quasiparticle excitations (QPs). When the QPs tunnel across the Josephson junction, they can exchange energy with the qubit degree of freedom. This results in an elevated qubit relaxation rate. Only when the QPs recombine or get trapped does the error burst end and the performance of the processor recover. This prompts the search for strategies to either accelerate the phonon decay Martinis 2021 ; Iaia et al. 2022 or suppress QP tunneling. On the latter front, a promising way forward is engineering of the superconducting gap Aumentado et al. 2004 ; Ferguson et al. 2006 ; Naaman and Aumentado 2007 ; Court et al. 2007 ; Shaw et al. 2008 ; Sun et al. 2012 ; Riwar and Catelani 2019 ; Serniak 2019 ; Diamond et al. 2022 ; Pan et al. 2022 ; Marchegiani et al. 2022 ; Connolly et al. 2024 ; McEwen et al. 2024 . By creating a non-uniform gap profile – which can be achieved by varying the thickness of superconducting films – it is possible to impede the QP tunneling across the qubit junction and suppress decoherence Diamond et al. 2022 ; Marchegiani et al. 2022 ; Connolly et al. 2024 ; McEwen et al. 2024 . For this approach to be effective, the difference in gaps at the two sides of the junction should exceed the qubit energy by an amount significantly larger than the QP energy. This condition has been shown to hold for the “resident” non-equilibrium QPs typically present in the superconducting films in the steady state Connolly et al. 2024 . By interacting with phonons, the energy of these QPs equilibrates with the base plate of the refrigerator even while the QP density remains strongly out of equilibrium. Yet, an important question for the suppression of QP-induced correlated errors remains unanswered: can gap engineering prevent QP tunneling following the impact of high-energy particles, when the phonon bath itself is moved away from equilibrium? In this paper, we address this question by measuring QP bursts in 3D transmons with different pairs of junction film thicknesses and, consequently, with different gap profiles. First, we benchmark our gap engineering by measuring the gap difference, \delta\Delta , for our devices with different film thicknesses. The gap difference is obtained by monitoring the tunneling rate of resident QPs at different fridge temperatures. By varying the film thickness, we fabricate the gap difference in the range of 0 to 10 \mathrm{GHz} (in frequency units). We then evaluate the effect of gap difference on the detection rate of burst events and measure qubit relaxation rates during bursts. We find that gap engineering reduces the rate of burst detection by a factor of a few. However, the reduction falls short of the four orders of magnitude suppression observed for the tunneling of the resident QPs. We believe that this discrepancy results from the elevation of the effective chip temperature during the burst to about 90 mK. The additional thermal energy allows the QPs to overcome the gap difference and tunnel across the junction, thus facilitating qubit relaxation. We directly confirm the persistence of elevated substrate temperature by comparing the qubit excitation and relaxation rates during the burst. Our results should be contrasted with those of Ref. McEwen et al. 2024 , where the gap difference – obtained by film thicknesses similar to our gap-engineered device – resulted in essentially no bursts being detected. The difference may arise from a combination of several factors. First, unlike ours, the chip in Ref. McEwen et al. 2024 has most of its area covered by the aluminum film forming the ground plane. Second, the substrate of Ref. McEwen et al. 2024 is silicon whereas ours is sapphire. Third, the chips are thermalized differently. Our chips are anchored to the package only by two isolated clamps each, whereas the chip of Ref. McEwen et al. 2024 is presumably anchored by a large number of wirebonds. Finally, unlike in our measurement, Ref. McEwen et al. 2024 probed the qubit only sparsely, once every 100 \mathrm{\mu s} . Such a measurement may miss bursts with a recovery time of less than 100 \mathrm{\mu s} . Figure 1: (a) Parity-switching rate ( \Gamma ) as a function of the fridge temperature ( T ) for devices with different film thicknesses. The film thickness is used as a knob to control the difference of gaps \>\delta\Delta at the two sides of the qubit junction. Down (up)-pointing triangles represent rates \Gamma_{0} ( \Gamma_{1} ) conditioned on the qubit residing in the ground (excited) state. When \>\delta\Delta>hf_{q} , the gap difference is efficient in obstructing the tunneling of resident QPs. Correspondingly, parity-switching rate \Gamma_{1} follows an Arrhenius law with the activation exponent \>\delta\Delta-hf_{q} , i.e., \Gamma_{1}\propto\exp\left(-\left[\>\delta\Delta-hf_{q}\right]/\>k_{\mathrm{B}}T\right) . This regime is realized in the big- \>\delta\Delta device (left panel) and

Read the complete text