Charge sensitivity in the transmon regime

TL;DR

This paper investigates how charge noise and parity switching events affect dephasing times in transmon qubits, revealing that charge environment fluctuations can limit qubit coherence even in highly insensitive designs.

Contribution

The study introduces a protocol to detect parity switching events in transmons and demonstrates their impact on dephasing times, highlighting charge noise as a limiting factor.

Findings

Parity switches correlate with dephasing time fluctuations.

Same parity states exhibit longer $T_2$ than averaged over parities.

Charge noise can limit transmon coherence despite high $E_J/E_C$ ratios.

Abstract

Transmons are widely adopted in quantum computing architectures for their engineered insensitivity to charge noise and correspondingly long relaxation times. Despite this advantage, transmons often exhibit large fluctuations in dephasing times across different devices and also within qubits on the same device. Existing transmon qubits are assumed to be insensitive to charge noise. However, very little recent attention has been paid to the dependence of dephasing on the local charge environment. In this study, we see fluctuations in the dephasing time, $T_{\phi}$, which correlate to charge offset. While charge offset fluctuations are slow, parity switches are fast processes tied to the charge offset and can affect $T_{\phi}$ in Ramsey experiments. We implement a protocol to detect parity switching events using single-shot methods, which are interleaved within a Ramsey measurement. We find that events that remain in the same parity state have a higher $T_2$ than measurements averaged over both parities. Our results show that transmons can be limited by charge-noise, even with $E_\text{J}/E_\text{C} \approx 50$. Consequently, parity flip rates must be considered as a device characterization metric.