Probing qubit memory errors at the 10⁻⁵ level

Trapped atomic ions are a promising candidate system for developing a quantum computer. All elementary operations for quantum information processing have al- ready been achieved in trapped-ion systems with errors below the “fault-tolerant threshold”, meaning that these errors are sufficiently small...

Full description

Bibliographic Details
Main Author: Tarlton, James Edward
Other Authors: Lucas, David ; Thompson, Richard ; Segal, Danny
Published: Imperial College London 2018
Subjects:
530
Online Access:https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.745333
Description
Summary:Trapped atomic ions are a promising candidate system for developing a quantum computer. All elementary operations for quantum information processing have al- ready been achieved in trapped-ion systems with errors below the “fault-tolerant threshold”, meaning that these errors are sufficiently small to be removed with quan- tum error correction. However, the preservation of quantum information over the short timescales that are relevant for quantum computing has not, to our knowledge, been measured before in trapped-ion qubits. Previous investigations have measured decoherence over long times and extrapolated an exponential decay model to short times. We directly measure the qubit memory error rate over short timescales, which is made possible by the high-fidelity single-qubit gates and state preparation and mea- surement in our system. Our qubit is a hyperfine “atomic clock” transition in 43Ca+, and gates are applied with near-field microwaves. We find that the assumption of exponential decay of qubit state coherence does not apply in our system. The data is fit well by a model of 1/f frequency noise with a low-frequency cutoff. The small- est memory error that we measure is 3(1)×10−6 over 2ms. The time at which our qubit memory error is 1×10−3, an important threshold for viable quantum error correction, is ≈400ms. We also present a design study into a new ion trap electrode geometry for applying near-field microwave two-qubit gates. This design features an ‘S’-shaped meander electrode to passively null the microwave field at the ions’ locations. It has improved isolation between the meander and other electrodes compared to previous work on this, which should reduce the sensitivity of the microwave field distribution to the boundary conditions of these electrodes. We show that the trap would allow for ion chains to be trapped, transported and split with feasible DC and RF voltages.