长相干时间超导相滑移量子比特的实验研究

Developing a small and long-coherence superconducting qubit is of great significance to improve scalability and to ultimately realize a universal superconducting quantum computer. The newly emerged phase-slip junction devices based on disordered superconductor nanowires make such a superconducting qubit feasible. This proposal is aimed at discovering the decoherence mechanism of superconducting phase-slip flux qubit by experimentally studying (i) the dissipations in disordered superconductors, as well as (ii) the relation between the enormous flux noise in phase-slip flux qubit and the spatial fluctuations of superconducting gap. A phase-slip qubit dually equivalent to Josephson transmon qubit can be developed, by harnessing the dual equivalence between phase-slip junction and Josephson junction as well as the very large kinetic inductance of disordered superconductor thin film. This new type of phase-slip qubit will have a small size of only 1/100 ~ 1/50 of a transmon, and it should have reasonably long coherence time due to its flux-noise-insensitive energy gap.

发展一种既小又有长相干时间的超导量子比特对提高可扩展性并最终实现通用超导量子计算机意义重大，近年出现的基于无序超导体纳米线的相滑移结器件为实现这样的超导量子比特提供了可能。本项目旨在通过实验研究（1）无序超导体中的耗散问题（2）相滑移磁通量子比特中极大磁通噪声与超导能隙空间涨落的关系问题，揭示相滑移磁通量子比特的退相干机制。利用相滑移结与约瑟夫森结的对偶等价性和无序超导体薄膜极大的动态电感，研制一种与约瑟夫森transmon量子比特对偶等价的相滑移量子比特。这种相滑移量子比特大小应该可以控制在transmon面积的1/100~1/50，且由于其能隙对磁通噪声不敏感，应该有较长的相干时间。

超导量子电路是目前研制量子计算机最具潜力的技术路线，也基于transmon及其变体xmon发展到了能集成数百个量子比特的含噪声中等规模量子计算阶段，但如何集成更多量子比特实现可纠错的通用量子计算机，依然是一个开放问题。本项目计划研制一种大小可以控制在transmon面积的1/100~1/50且有较长相干时间的超导相滑移量子比特，以提升单芯片上超导量子比特的集成度。经过四年的努力，我们发现超导相滑移结加工难度很大且氮化铌薄膜存在极易氧化的问题，我们采用传统的铝-氧化铝-铝约瑟夫森结成功制备了新型小尺寸超导量子比特，并研究了该种超导量子比特中准粒子产生的机制及减少准粒子产生的方案。具体内容包括：（1）研制了一种面积只有transmon的1/50左右的超导量子比特，其平均能量弛豫时间超过20μs，平均相位退相干时间超过15μs；（2）基于固定频率xmon量子比特和频率可调耦合器，实现了一种保真度超过99%的快速两量子比特CZ门；（3）在xmon量子比特中对绝热极限和绝热-非绝热边界情况下的Landau-Zener-Stückelberg干涉进行了实验研究。