Benjamin J. Brown

The high-fidelity storage of quantum information is crucial for quantum computation and communication. Many experimental platforms for these applications exhibit highly biased noise, with good resilience to spin depolarisation undermined by high dephasing rates. In this work, we demonstrate that the memory performance of a noise-biased trapped-ion qubit memory can be greatly improved by incorporating error correction of dephasing errors through teleportation of the information between two repetition codes written on a pair of qubit registers in the same trap. While the technical requirements of error correction are often considerable, we show that our protocol can be achieved with a single global entangling phase gate of remarkably low fidelity, leveraging the fact that the gate errors are also dominated by dephasing-type processes. By rebalancing the logical spin-flip and dephasing error rates, we show that for realistic parameters our memory can exhibit error rates up to two orders of magnitude lower than the unprotected physical qubits, thus providing a useful means of improving memory performance in trapped ion systems where field-insensitive qubits are not available.

Quantum error-correcting codes are many-body entangled states that are prepared and measured using complex sequences of entangling operations. Each element of such an entangling sequence introduces noise to delicate quantum information during the encoding or reading out of the code. It is important therefore to find efficient entangling protocols to avoid the loss of information. Here we propose an experiment that uses only global entangling operations to encode an arbitrary logical qubit to either the five-qubit repetition code or the five-qubit code, with a six-ion Coulomb crystal architecture in a Penning trap. We show that the use of global operations enables us to prepare and read out these codes using only six and ten global entangling pulses, respectively. The proposed experiment also allows the acquisition of syndrome information during readout. We provide a noise analysis for the presented protocols, estimating that we can achieve a six-fold improvement in coherence time with noise as high as $\sim 1\%$ on each entangling operation.