Session X08: Qubit Coherence and Noise Characterization

Anomalous Charge Noise in Superconducting Qubits

We describe experiments to probe fluctuations in offset charge in weakly charge-sensitive superconducting qubits that depart from the transmon regime. We explore a range of device geometries to elucidate the surprising scale dependence of charge noise in these qubits, and we describe experiments to investigate possible spatial correlation in charge fluctuations in neighboring devices. Since quantum error correction schemes such as the two-dimensional surface code rely on the assumption of uncorrelated noise, these studies could have implications for scale-up towards fault-tolerant superconducting qubit arrays.   

Phonon traps reduce the quasiparticle density in superconducting circuits

Out of equilibrium quasiparticles (QPs) are a main source of decoherence in high impedance superconducting quantum circuits. Despite significant progress in the understanding of QP dynamics, pinpointing their origin and decreasing their density remain outstanding tasks. The cyclic process of recombination and generation of QPs implies the exchange of phonons between the superconducting film and the underlying substrate. Reducing the number of substrate phonons with frequencies above the spectral gap of the superconductor should result in a reduction of QPs [1]. Indeed, we demonstrate that surrounding high impedance resonators made of granular aluminum (grAl) with lower gapped thin film aluminum islands increases the internal quality factors of the resonators in the single photon regime, suppresses the noise, and reduces the rate of observed QP bursts [2]. The aluminum islands are positioned far enough from the resonators to be electromagnetically decoupled, thus not changing the resonator frequency, nor the loading. We therefore attribute the improvements observed in grAl resonators to phonon trapping at frequencies close to the spectral gap of aluminum, well below the grAl gap.

[1] Valenti et al., Phys. Rev. App 11, 054087 (2019)
[2] Henriques, Valenti et al., arXiv:1908.04257   

Photon-assisted charge-parity switches in superconducting qubits

Here we identify a mechanism by which stray photons, with energy large enough to break Cooper pairs, can cause decoherence in superconducting qubits. Similar to decoherence induced by steady-state nonequilibrium quasiparticles, these photon-assisted tunneling (PAT) processes are associated with a change in the charge parity of the qubit. This enables the separation of these quasiparticle-related processes from other contributions to decoherence. Our theory predicts similar rates of PAT-induced relaxation and excitation of the qubit, in agreement with recent experiments. This talk will detail the similarities and differences between PAT processes and decoherence induced by nonequilibrium quasiparticle tunneling.   

Microscopic charging and in-gap states in superconducting granular aluminum

Granular aluminum (grAl) is a viable material for high-impedance quantum circuits [1], thanks to its high kinetic inductance, low microwave frequency losses [2], and fabrication by deposition of pure aluminum in an oxygen atmosphere. The material’s microstructure – pure aluminum grains in non stoichiometric aluminum-oxide – can be modeled as a Josephson junction array [3], in which the oxide thickness is controlled by the oxygen partial pressure during deposition. Here we present scanning tunneling microscope measurements of the local electronic structure of superconducting grAl and we confirm an increased superconducting gap in the grains of films with ρ ≈ 300 μΩcm and ρ ≈ 2000 μΩcm. In the high resistivity film, we find Coulomb charging effects, a first indication for grain decoupling, and in-gap states on individual grains, which could contribute to flux noise and dielectric loss not only in devices employing grAl, but also in pure aluminum circuits, where such states could form in Josephson junctions or in oxidized aluminum surfaces.

[1] L. Grünhaupt & M. Spiecker et al., Nat. Mat. 18 (2019)
[2] L. Grünhaupt et al., PRL 121 (2018)
[3] N. Maleeva et al., Nat. Commun. 9 (2018)   

Probing non-equilibrium quasiparticle populations in superconducting quantum circuits.

Non-equilibrium quasiparticle (QP) populations in superconducting quantum circuits are a limiting source of decoherence. When properly phase biased, nanobridge Josephson junctions provide sub-gap bound states for QPs and so function as QP traps [1]. The occupation of a trap state by a QP alters the inductive contribution of the junction to its host circuit. Thus, the population of QPs trapped in the junction can be directly measured by probing the resonance of a capacitively shunted nanobridge with standard superconducting qubit hardware. By measuring QP trapping in real time and performing statistical analyses, one can infer information about the QP generation, annihilation, and trapping mechanisms. We discuss these single-shot fidelity measurements of QPs trapping in the junctions of a nanobridge SQUID embedded in a superconducting resonator. We further discuss the use of our device as a tool for measuring QP properties and for testing methods of QP mitigation. [1] E. M. Levenson-Falk, F. Kos, R. Vijay, L. Glazman, and I. Siddiqi, Phys. Rev. Lett. 112, 047002 (2014).   

Improved material loss measurement using cryogenic resonator testbed

Loss measurements are critical for the superconducting qubit community as the coherence of quantum systems are limited by loss through energy relaxation. The standard procedure for measuring loss is to design and fabricate resonators that incorporate the material-under-test and measure the temperature, power, and geometry dependence. By analyzing the transmission S₂₁ data on complex plane, it is possible to extract the internal quality factor, Q_i, i.e. the inverse of the loss tangent of the material. However, there are several variations of experimental set-up, measurement and fitting algorithms used in the field. This results in uncertainty when comparing Q_i values between research groups. In this talk, we will compare the results of different measurement configurations, especially for input and output line filtering and isolation, and quantify the accuracy of loss measurement in order to identify best practices for resonator measurement and setup procedures.
  

Detection of non-equilibrium quasiparticles in charge sensitive transmons

Non-equilibrium quasiparticles are considered possible sources causing decoherence in superconducting qubits. Thus, exploring the dynamics of quasiparticles is crucial to understand their mechanism and how to minimize their presence in superconductors. Here, we study charge parity fluctuations in 2D transmon qubits with sufficiently strong coupling and low E_J/E_C ratio. We will show measurements of density of quasiparticles on various qubits to discuss impact on coherence.
  

Experimental study of flux noise in nanowire transmons subject to an applied magnetic field

Flux noise is generally the dominant dephasing mechanism in transmon qubits using SQUID loops as a tunable inductive element. Continuing experimental research ambitions to elucidate the microscopic origin and mitigation of flux noise. In this work, we experimentally investigate flux noise using an unconventional knob for circuit QED: an applied magnetic field. We first present various changes introduced to our circuit QED system to further extend the field compatibility of resonators and nanowire transmons beyond 70 mT. Next, we use standard coherence measurements and sensitivity analysis on and off flux sweetspots to investigate the dependence of flux noise on field applied along the axis of the two constituent nanowire junctions.
  

Surface spin induced 1/f flux noise dependent on SQUID geometry (1)

Magnetic flux noise is the dominant source of dephasing in tunable superconducting qubits. While the mechanism behind this 1/f flux noise is poorly understood, it has been proposed that it originates from random fluctuations of spin impurities located on the surface of SQUIDs. A previously proposed microscopic model predicts that flux noise increases with the perimeter and decreases with the wire width of the SQUID loop. Here, we demonstrate the validity of this model by measuring the flux noise amplitudes of about 50 capacitively shunted flux qubits over a wide range of geometric SQUID parameters. Our results show good agreement with the proposed model assuming independent spins, and may therefore serve as a guide on how to improve SQUID designs to minimize 1/f flux noise.   

Surface spin induced 1/f flux noise dependent on SQUID geometry (2)

The dominant source of decoherence in frequency tunable superconducting qubits is 1/f flux noise, presumably originating from local magnetic spin impurities located at the surface of their SQUID loops. Here, we measure the flux noise amplitudes of capacitively shunted flux qubits and study their dependence on geometric parameters of their SQUID loops. Our data show good agreement with a previously presented microscopic model for independent spin impurities which has so far eluded experimental verification. We discuss limitations of the proposed model for superconducting films of finite thickness and provide numerical simulations of the current distribution in our SQUIDs, which can refine the considered model. Finally, we apply and confirm our results by observing consistently low flux noise amplitudes in samples with optimized SQUID parameters – small perimeters and large wire widths.   

Characterizing non-classical non-Gaussian noise: photon shot noise fluctuations

Understanding how to accurately describe environmental noise is of crucial importance for designing control protocols that enable high-fidelity quantum operations. While standard approaches often assume classical Gaussian noise, going beyond these assumptions is important in many settings. Here, we consider both the non-classical and non-Gaussian nature of photon number fluctuations in a driven dissipative quantum resonator; such fluctuations can be a significant limitation on the performance of superconducting qubits. We extend the Keldysh approach for characterizing this quantum noise [1,2] to calculate its frequency-resolved third cumulant, the so-called bispectrum. We discuss how the quantum noise bispectrum directly reveals features of non-classicality and non-equilibrium bath dynamics. We also show how this quantum bispectrum could be measured using standard experimental tools in circuit QED.

[1] A. A. Clerk, Phys. Rev. A 84, 043824 (2011)
[2] P. P. Hofer and A. A. Clerk, Phys. Rev. Lett. 116, 013603 (2016)   

Identifying Noise Sources via Multi-level Quantum Noise Spectroscopy

Identifying dominant sources of noise and mitigating or eliminating them is crucial for engineering robust quantum devices. Although existing quantum noise spectroscopy (QNS) protocols provide insight into noise sources by measuring the noise power spectral density, they do not directly identify the physical origin of the noise process. Here, we develop and experimentally validate a multi-level QNS protocol using a superconducting transmon qubit as a spectrometer. This technique has two notable implications: First, it expands the applicability and spectral range of a qubit spectrometer beyond the two-level approximation. Second, it can be applied to higher excited levels, which enables us to extract and identify the noise contributions from multiple noise sources.   

Measuring effective temperatures of qubits using correlations

Initialization of a qubit in a pure state is a prerequisite for quantum computer operation. Qubits are commonly initialized by cooling to their ground states through passive thermalization or by using active reset protocols [1, 2]. To accurately quantify quality of the initialization and to shed light on origin of spurious excitation one requires an accurate tool to measure the excited state population [3]. We propose a new technique of finding the excited state population of a qubit using measurement correlations in time. We experimentally implement the proposed method using a circuit QED platform and compare its performance with previously developed techniques [4]. The method does not rely on having high-fidelity measurement or the higher energy levels. We measure the spurious qubit population with accuracy of up to ≈0.01% and discuss its possible sources.
[1] D. Egger et al., Phys. Rev. Applied 10, 044030 (2018).
[2] P. Magnard et al., Phys. Rev. Lett.121, 060502 (2018).
[3] X. Y. Jin et al., Phys. Rev. Lett.114, 240501 (2015).
[4] A. Kulikov et al., in preparation
  

Photon Shot Noise Limited Charge Sensitivity in a Hybrid Quantum System

Hybrid mesoscopic quantum systems serve as essential building blocks for quantum information and metrology devices. One such system that displays a rich variety of behavior in the quantum-classical regime is a Cooper pair transistor embedded in a quarter wave cavity (cCPT). As one application, this device can act as a highly sensitive charge detector, and form the basis for an optomechanical system in the single photon strong coupling regime. We present a theoretical analysis of the photon shot noise limited charge sensitivity of the device, assuming a low average photon number drive. This lower bound is difficult to achieve using conventional detection approaches, owing mainly to the low frequency noise caused by the coupling of thermally activated two-level-fluctuators to the cCPT. The effects of this interaction are observed as 1/f noise in the microwave cavity fundamental resonance frequency. Using careful calibration of a feedback mechanism called Pound-Drever-Hall (PDH) locking, we should be able to observe this noise near the low photon limit. Moreover, a modified PDH scheme may possibly suppress this noise, and achieve charge sensitivity close to our theoretical, minimum shot noise prediction.   

Characterization of noise correlations in superconducting quantum circuits

Superconducting quantum circuits have been an important platform to realize a scalable quantum computing system. Qubit decoherence, which is caused by the fluctuations of its environment, remains a main obstacle to realize fault-tolerant quantum computing. To reduce decoherence, it is important to identify the source of fluctuations by precisely characterizing its correlations. Prevalent methods only consider the second order correlation under the Gaussian approximation. However, environment with higher order correlations is not uncommon in mesoscopic quantum systems. Based on a weak measurement scheme ^[1], we present an experiment to detect high-order correlations in a superconducting circuit. We verified the scheme by reconstructing the noise power spectral density and compared the results with that obtained from dynamical decoupling. We also demonstrated the characterization of higher-order correlation of the environment.

Reference:
[1]Wang P, Chen C, Peng X, et al. Characterization of arbitrary-order correlations in quantum baths by weak measurement. arXiv preprint arXiv:1902.03606 (2019).