Improving the Time Stability of Superconducting Planar Resonators

TL;DR

This study demonstrates that cleaning treatments of aluminum superconducting resonators significantly reduce flicker noise, enhancing their stability for quantum computing applications.

Contribution

It introduces a cleaning process that reduces flicker noise in superconducting resonators, improving their performance for quantum computing.

Findings

Tenfold reduction in flicker noise after cleaning.

Cleaning improves resonator frequency stability.

Results align with the tunneling model for two-level defects.

Abstract

Quantum computers are close to become a practical technology. Solid-state implementations based, for example, on superconducting devices strongly rely on the quality of the constituent materials. In this work, we fabricate and characterize superconducting planar resonators in the microwave range, made from aluminum films on silicon substrates. We study two samples, one of which is unprocessed and the other cleaned with a hydrofluoric acid bath and by heating at $880^{\circ}$C in high vacuum. We verify the efficacy of the cleaning treatment by means of scanning transmission electron microscope imaging of samples' cross sections. From 3 h-long resonator measurements at $\approx 10$ mK and with $\approx 10$ photonic excitations, we estimate the frequency flicker noise level using the Allan deviation and find an approximately tenfold noise reduction between the two samples; the cleaned sample shows a flicker noise power coefficient for the fractional frequency of $\approx 0.23 \times 10^{-15}$. Our preliminary results follow the generalized tunneling model for two-level state defects in amorphous dielectric materials and show that suitable cleaning treatments can help the operation of superconducting quantum computers.