Skip to main content
European Commission logo print header

Decoherence in Superconducting Devices

Final Report Summary - DESDE (Decoherence in Superconducting Devices)

Superconducting devices incorporating Josephson elements are a promising platform for future quantum technologies, in particular for quantum computation and simulation with superconducting qubits. Qubit performances can benefit from a detailed understanding of decoherence mechanisms; such an understanding can then suggest ways for the suppression of decoherence channels. This project focused mostly on the interaction between the qubit degree of freedom and quasiparticles, the intrinsic excitations in a superconductor.
Various superconducting qubit designs are under investigation, most notably transmon, flux qubit, and fluxonium. For the single-junction transmon, quasiparticle effects include not only relaxation but also dephasing due to parity switching; the developed theory compares favorably to available experimental data. For the 4-junction flux qubit, it was shown that an improved model that includes stray capacitances it is necessary for a quantitative comparison between theory and experiment. For the many-junction fluxonium, quasiparticle interference was studied in a joint theoretical and experimental effort which has led to the verification of the phase dependence of loss in a Josephson junction, a long-unconfirmed prediction. A detailed theoretical model of the fluxonium which includes the modes of the junction array supports its viability as a quantum bit. Additionally, the effect of sparse quasiparticles on the transmission of microwaves through a junction array was also studied, in this case with possible metrological applications in mind.
The results briefly described above have led to the sought-for detailed understanding of quasiparticle-induced decoherence. For the second part of the project, the goal was to find ways to limit the unwanted quasiparticle decoherence. In a joint theory-experiment study, trapping of quasiparticles by vortices was observed, along with some improvement in the qubit lifetime. A better-controllable way to trap quasiparticle is offered by normal-metal traps – that is, normal-metal islands in tunnel contact with the superconductor. A theoretical description of trapping, which takes into account tunneling from the superconductor into the normal metal and back, and relaxation of excitations in the normal metal, was developed and successfully compared to experiments. Subsequent work used the model to devise optimal implementations of traps, both to prolong the qubit relaxation time (by suppressing the steady-state quasiparticle density) and to improve qubit stability (by suppressing density fluctuations and speeding up the trapping of excess quasiparticles). In a different approach, quasiparticle-induced decoherence was suppressed via so-called “quasiparticle pumping”: the qubit is excited and when it decays, it can increase the quasiparticle energy - this makes it more likely that the quasiparticle leaves the qubit; this stochastic process is then repeated several times. In this way, quasiparticles are pumped away, rather than being trapped. Both approaches have been shown to improve qubit performance and therefore can potentially have positive impact on the development of a superconducting quantum computer and other quantum technologies.