91̽»¨

Abstract:

Robust qubit memory is essential for quantum computing, both for near-term devices operating without error correction, and for the long-term goal of a fault-tolerant processor. We directly measure the memory error ε_m for a ⁴³Ca⁺ trapped-ion qubit in the small-error regime and find ε_m<10^−4 for storage times t ≲ 50  ms. This exceeds gate or measurement times by three orders of magnitude. Using randomized benchmarking, at t = 1  ms we measure ε_m=1.2(7)×10^−6, around ten times smaller than that extrapolated from the T^∗₂ time, and limited by instability of the atomic clock reference used to benchmark the qubit.

Abstract:

Atomic physics experiments commonly use millitesla-scale magnetic fields to provide a quantization axis. As atomic transition frequencies depend on the magnitude of this field, many experiments require a stable absolute field. Most setups use electromagnets, which require a power supply stability not usually met by commercially available units. We demonstrate the stabilization of a field of 14.6 mT to 4.3 nT rms noise (0.29 ppm), compared to noise of >100 nT without any stabilization. The rms noise is measured using a field-dependent hyperfine transition in a single 43Ca+ ion held in a Paul trap at the center of the magnetic field coils. For the 43Ca+ "atomic clock" qubit transition at 14.6 mT, which depends on the field only in second order, this would yield a projected coherence time of many hours. Our system consists of a feedback loop and a feedforward circuit that control the current through the field coils and could easily be adapted to other field amplitudes, making it suitable for other applications such as neutral atom traps.