Abstract

Over the past two decades significant advances have been made in the research of superconducting quantum computing and quantum simulation, in particular of the device design and fabrication that leads to ever-increasing superconducting qubit coherence times and scales. With Google’s announcement of the realization of “quantum supremacy”, superconducting quantum computing has attracted even more attention. Superconducting qubits are macroscopic objects with quantum properties such as quantized energy levels and quantum-state superposition and entanglement. Their quantum states can be precisely manipulated by tuning the magnetic flux, charge, and phase difference of the Josephson junctions with nonlinear inductance through electromagnetic pulse signals, thereby implementing the quantum information processing. They have advantages in many aspects and are expected to become the central part of universal quantum computing. Superconducting qubits and auxiliary devices prepared with niobium or other hard metals like tantalum as bottom layers of large-area components have unique properties and potentials for further development. In this paper the research work in this area is briefly reviewed, starting from the design and working principle of a variety of superconducting qubits, to the detailed procedures of substrate selection and pretreatment, film growth, pattern transfer, etching, and Josephson junction fabrication, and finally the practical superconducting qubit and their auxiliary device fabrications with niobium base layers are also presented. We aim to provide a clear overview for the fabrication process of these superconducting devices as well as an outlook for further device improvement and optimization in order to help establish a perspective for future progress.

Highlights

  • Comparison of the density of niobium film prepared by Hipims process with that by conventional magnetron sputtering[72]

  • Two designs of Josephson junctions: (a), (c) Design drawings; (b), (d) electron micrographs of Josephson junctions prepared by corresponding designs [112,113]

  • Over the past two decades significant advances have been made in the research of superconducting quantum computing and quantum simulation, in particular of the device design and fabrication that leads to ever-increasing superconducting qubit coherence times and scales

Read more

Summary

Quantum bus

图 2 Xmon 超导量子比特 [47] (a) 器件照片; (b) 约瑟夫森结区放大图; (c) 电路示意图 Fig. 2. Superconducting Xmon qubit [47]: (a) Optical micrograph; (b) magnified view of Josephson-junction area; (c) circuit diagram. 图 3 Fluxonium 量子比特 [48] (a) 器件照片; (b) 天线; (c) 3D 腔; (d) 电路示意; (e) 势能和能级示意图 Fig. 3. Superconducting fluxonium qubit [48]: (a) Optical micrograph; (b) antenna; (c) 3D resonator; (d) circuit diagram; (e) potential energy and energy levels. 这类器件结构将原来与约瑟夫森结组合 成非线性谐振电路电容的一端变为地端, 而电容另 一端设计为十字叉形, SQUID 位于在叉子形的下 端. 这种设计使得利用直流偏置 (Z 线) 调制 SQUID 变得方便、直接, 而十字叉的另外三端可以分工明 确, 分别与读出谐振腔、微波控制线 (XY 线)、以及 其他量子比特相耦合. 这种设计使比特数目的扩展 非常方便, 目前已广泛地应用于各种量子计算和量 子模拟的研究中, 例如采用这种设计 Google 成功 实现了量子优势和对时间晶体的观察. Xmon 量子 比特的退相干时间最初达到 40 μs 左右, 虽然较前 面提到的传输子量子比特略显逊色, 但在后续的制 备工艺研究中发现, 其量子退相干时间仍有较大的 提升空间. 除上述超导量子比特设计之外, 图 3 所示的 fluxonium 量子比特在磁通量子比特设计中将用一 串约瑟夫森结组合而实现自感较大的电感并将该 大电感并联在小面积约瑟夫森结旁, 由于引入了较 大电容, 该设计对电荷噪声不敏感, 再加之可以抑 制准粒子激发的特点, 使该设计的样品退相干时间 达到了毫秒量级 [55]. 类似如图 4 所示, 并联电容磁通量子比特 (Cshunt flux qubit) 也是以磁通量子比特为基础蓝 本, 将一个较大的电容并联在小面积约瑟夫森结 旁, 再通过电容耦合谐振腔实现操纵、读出 [49]. 类似如图 4 所示, 并联电容磁通量子比特 (Cshunt flux qubit) 也是以磁通量子比特为基础蓝 本, 将一个较大的电容并联在小面积约瑟夫森结 旁, 再通过电容耦合谐振腔实现操纵、读出 [49]. 该 设计使得体系势能函数将电荷量子比特和磁通量 子比特的特性结合起来, 势能曲线最终呈现四次函 数的形式 [49], 而该能级结构使得电容分流磁通量 子比特拥有较传输子量子比特更大的非线性, 该器 件的退相干时间可达 40 μs 左右

Qubit B
Deposition Sputtered HiPIMS opt HiPIMS norm
HiPIMS normal
Dolan bridge Trench Side trench
Common mode bias
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call