Designing an Asymmetric Zero-Pi Qubit for Insensitivity to Charge Noise and Quasiparticle Tunneling

Abstract

Superconducting qubits are a promising platform for quantum computers, but they are vulnerable to decoherence from energy relaxation (\(T_1\)) and dephasing (\(T_2\)). The \(0-\pi\) qubit is a superconducting qubit that offers intrinsic protection against decoherence. To obtain \(T_1\) protection, the \(0-\pi\) has disjoint support: deep wells in its potential that exponentially suppress tunneling between its logical states. To obtain \(T_2\) protection, the \(0-\pi\) parameters can be tuned such that the logical states are first-order insensitive to flux noise and exponentially insensitive to charge noise. However, driving between the logical states of the \(0-\pi\) requires passing through an ancillary level that is sensitive to charge noise and quasiparticle tunneling events. This thesis introduces unequal Josephson junctions (\(\Delta E_J\ne0\)) in the \(0-\pi\) circuit to create an "asymmetric \(0-\pi\)" qubit. Using Python scripts, the asymmetric \(0-\pi\) parameter space was searched for regimes that reduce the ancillary level's sensitivities to charge noise and quasiparticles, but also retain the symmetric \(0-\pi\)'s \(T_1\) and \(T_2\) protection. These simulations revealed three qualitatively different regimes, corresponding to small, mid, and large \(\Delta E_J\) values. While the large \(\Delta E_J\) regime has clearly reduced sensitivity to both charge noise and quasiparticle tunneling, it loses the disjoint support necessary for \(T_1\) protection. Meanwhile, the small \(\Delta E_J\) regime displays partial protection against charge noise and quasiparticle tunneling, and it also successfully replicates the strengths of the symmetric \(0-\pi\), making it the most promising regime in the long term. Unfortunately, realizing a device in this regime is difficult given current fabrication error margins. This observation motivates further investigation of the mid \(\Delta E_J\) regime, which in general shows neither protection against quasiparticles nor disjoint support. However, its reliably large charge dispersions allow greater room for fabrication error and thus make it ideal for a proof-of-concept device to check that the asymmetric \(0-\pi\) theory can be reproduced in experiment. An initial device in the mid \(\Delta E_J\) regime has been fabricated, and the measurement of this device is currently in preparation.