Table of Contents
Fetching ...

Gradiometric, Fully Tunable C-Shunted Flux Qubits

Benedikt Berlitz, Alexander Konstantin Händel, Etienne Daum, Alexey V. Ustinov, Jürgen Lisenfeld

TL;DR

The paper tackles the challenge of achieving fully tunable flux qubits without sacrificing coherence by introducing a gradiometric, capacitively shunted flux qubit with independent control of potential asymmetry and barrier height through two bias lines, Φ_T and Φ_B. The qubit uses a DC-SQUID-based α(Φ_B) to tune the barrier and a half-gradiometric geometry to minimize cross-talk, with the effective potential described by U(φ) = -2 E_J cos(φ) - α(Φ_B) E_J cos(2π Φ_T/Φ_0 - 2φ). Key results show T1 up to ≈25 μs at f_q ≈ 3.32 GHz (Q ≈ 5.3×10^5) and a tunability range from 3 MHz to 21 GHz, along with strain-tuned TLS spectroscopy spanning almost an octave, validating the device as a platform for defect spectroscopy and decoherence studies. The findings highlight the device’s suitability for quantum information tasks and quantum-material research, with potential roles as intermediaries in hybrid architectures and multi-qubit couplers, while outlining paths to extend tunability and further reduce decoherence.

Abstract

Fully tunable flux qubits offer in-situ and independent controls of their energy potential asymmetry and tunnel barrier, making them versatile tools for quantum computation and the study of decoherence sources. However, only short coherence times have been demonstrated so far with this type of qubit. Here, we present a capacitively shunted flux qubit featuring improved relaxation times up to T1 = 25 $μ$s and a frequency tunability range of $\sim$ 20 GHz at the flux-insensitive sweet spot. As a model application, we demonstrate detection of two-level tunneling defects in a frequency range spanning one octave.

Gradiometric, Fully Tunable C-Shunted Flux Qubits

TL;DR

The paper tackles the challenge of achieving fully tunable flux qubits without sacrificing coherence by introducing a gradiometric, capacitively shunted flux qubit with independent control of potential asymmetry and barrier height through two bias lines, Φ_T and Φ_B. The qubit uses a DC-SQUID-based α(Φ_B) to tune the barrier and a half-gradiometric geometry to minimize cross-talk, with the effective potential described by U(φ) = -2 E_J cos(φ) - α(Φ_B) E_J cos(2π Φ_T/Φ_0 - 2φ). Key results show T1 up to ≈25 μs at f_q ≈ 3.32 GHz (Q ≈ 5.3×10^5) and a tunability range from 3 MHz to 21 GHz, along with strain-tuned TLS spectroscopy spanning almost an octave, validating the device as a platform for defect spectroscopy and decoherence studies. The findings highlight the device’s suitability for quantum information tasks and quantum-material research, with potential roles as intermediaries in hybrid architectures and multi-qubit couplers, while outlining paths to extend tunability and further reduce decoherence.

Abstract

Fully tunable flux qubits offer in-situ and independent controls of their energy potential asymmetry and tunnel barrier, making them versatile tools for quantum computation and the study of decoherence sources. However, only short coherence times have been demonstrated so far with this type of qubit. Here, we present a capacitively shunted flux qubit featuring improved relaxation times up to T1 = 25 s and a frequency tunability range of 20 GHz at the flux-insensitive sweet spot. As a model application, we demonstrate detection of two-level tunneling defects in a frequency range spanning one octave.

Paper Structure

This paper contains 8 sections, 6 equations, 7 figures, 1 table.

Figures (7)

  • Figure 1: a) Idealized circuit diagram of a 3-Junction flux qubit, with one junction having a lower critical current by a factor of $\alpha$. b) Circuit diagram of a gradiometric fully tunable flux qubit. The currents ($I_T,I_B$) in the local bias lines induce the fluxes $\Phi_T$ and $\Phi_B$, providing control of the potential asymmetry and barrier-height, respectively, which is illustrated by the orange and blue qubit potentials. c) Micrograph of a fully tunable C-shunted flux qubit (sample A), with an inset zooming in on the junction layout under a 45° angle (corresponding area of sample B).
  • Figure 2: a) principle of a $\Phi_T$-$\Phi_B$-sweep calibration measurement. A signal (red circles) is detected, when $f_\mathrm{res}$ is shifted into resonance with $f_\mathrm{probe}$ by a qubit transition. b) Calibration measurements used to identify symmetry points in the $\Phi_T$-$\Phi_B$-landscape and the bias line cross-talk (corrected in the bottom measurement). c) Numerical calculation of the qubit resonance frequency $f_\mathrm{q}$, spanning from 3 MHz to 21 GHz. d) Numerical calculation of the qubit potential for different $\alpha$-values, including the 3 lowest qubit states. Shown measurements were performed on sample A.
  • Figure 3: a) Violin- and scatter-plots of quality factor measurements at the potential symmetry point $\Phi_T=0$ on sample A at different qubit frequencies, corresponding to different $\alpha$-values at the cross-over into the double well regime ($\alpha >0.5$). The mean quality-factors (from left to right) correspond to $T_1$ of 26.8 µ s, 10.2 µ s, 7.57 µ s and 3.9 µ s. b) $T_1$-decay traces corresponding to the left-most violin. A single example trace is shown in blue and the average of all 150 traces taken over 8 hours is shown in red, with an exponential fit to the average trace in yellow. c) Rabi oscillations measured at $f_q$=3.22 GHz.
  • Figure 4: TLS-spectra in dependence of applied mechanical strain applied via a piezo-electric element. Dark lines indicate a drop in $T_1$ caused by resonant TLS defects. Spectra were taken on sample B with similar strain ranges for different frequency intervals, using a 5 $\mu$s swap pulse.
  • Figure 5: Setup schematic of the coaxial wiring, attenuation and filtering inside the cryostat. Legend: BPF=band-pass filter, LPF=low-pass filter, IR-Filter=infrared filter, MW=microwave source, AWG=arbitrary waveguide generator, HEMT=high mobility electron transistor.
  • ...and 2 more figures