Skip to content

Предельные параметры СИС-переходов в теории и технологические возможности их достижения

MetadataDetails
Publication Date2023-01-01
JournalФизика твердого тела
AuthorsM. A. Tarasov, A. A. Lomov, А.М. Чекушкин, А.А. Гунбина, М.Ю. Фоминский
AnalysisFull AI Review Included

This research investigates the critical limiting parameters of Superconductor-Insulator-Superconductor (SIS) Josephson junctions and proposes epitaxial growth strategies to bridge the gap between theoretical performance (Vc up to 2 mV) and current practical limits (Vc typically 200 µV).

  • Core Problem: Practical SIS junctions suffer from high specific capacitance, hysteresis (high McCamber parameter βc), and leakage currents due to granular film morphology and rough, non-uniform tunnel barriers.
  • Solution Focus: Achieving atomically smooth, epitaxial interfaces by matching crystal structure (preferably cubic) and lattice constants across the substrate, superconductor, and barrier layers.
  • NbN Achievement: Quasi-epitaxial NbN/AlN/NbN junctions grown on MgO (100) at low temperatures (less than 100 °C) demonstrated high quality, achieving IcRn = 3.5 mV and a critical current density (Jc) of 250 µA/µm2.
  • Al Film Optimization: A two-stage magnetron sputtering process for Al on Si(111) using a high-temperature (400 °C) seed layer was developed, resulting in smoother films (roughness 5 nm for 150 nm thickness) and significantly higher film hardness (10000-15000 MPa).
  • Hysteresis Control: Epitaxial growth is projected to eliminate the need for resistive shunting by achieving a McCamber parameter βc less than 1, significantly improving the limiting characteristics of superconducting devices.
  • Material Selection: The ideal structure requires cubic crystal symmetry for all layers (substrate, superconductor, and barrier), which is challenging as common barrier materials like Al2O3 and AlN are naturally hexagonal.

Theoretical and achieved parameters for SIS junctions, focusing on NbN and Al systems.

ParameterValueUnitContext
Theoretical Vc (NbN/I/NbN)1400µV1 nm oxide thickness
Theoretical Jc (NbN/I/NbN)250µA/µm21 nm oxide thickness
Theoretical RnS (NbN/I/NbN)5.6Ω·µm21 nm oxide thickness
Theoretical Specific Capacitance (C)24fF/µm2NbN/I/NbN, 1 nm oxide
Theoretical McCamber Parameter (βc)0.2-NbN/I/NbN, 1 nm oxide (Hysteresis-free)
Achieved IcRn (Epitaxial NbN)3.5mVLow-temp growth on MgO (100)
Achieved Jc (Epitaxial NbN)250µA/µm2Low-temp growth on MgO (100)
Standard Nb/AlOx/Nb Jc200µA/µm2HYPRES technology [8]
High Jc (Nb/AlOx/Nb)1.4mA/µm2Achieved with amorphous Si barrier
NbN Lattice Constant (Cubic)0.445nmEpitaxial film [4]
AlN Deposition Rate (Epitaxial)0.05nm/sCritical for high-quality barrier
Al Film Hardness (Hot Seed Layer)10000-15000MPaIndicates low porosity/high coherence length
Al Film Roughness (150 nm, Magnetron)5nmOn Si substrate
Nb Work Function (100)3.55eVLowest value, preferred orientation
Nb Work Function (110)4.49eVHighest value

The study employed advanced deposition and characterization techniques to control film morphology and crystal structure.

  1. Epitaxial Substrate Selection: Used single-crystal substrates including Sapphire (Al2O3), MgO (100), and Si (111) to promote ordered film growth, focusing on lattice matching.
  2. NbN/AlN/NbN Quasi-Epitaxial Growth: Reactive magnetron sputtering was used. High-quality junctions were achieved on MgO (100) at low temperatures (less than 100 °C). The AlN barrier was deposited at an extremely low rate (0.05 nm/s) to maintain a thickness less than 2 nm and force growth in the cubic phase.
  3. Nb/AlN/Nb Epitaxial Growth: Deposited on Sapphire at 480 °C. A thin (0.5 nm) Al buffer layer was inserted between the AlN barrier and the top Nb layer to improve Nb(200) orientation and prevent Nb diffusion.
  4. Two-Stage Epitaxial Al Growth (on Si(111)):
    • Stage 1 (Seed Layer): 10-20 nm Al island layer deposited at high temperature (400 °C) with a slow rate (less than 0.2 nm/s) to form monocrystalline seeds.
    • Stage 2 (Bulk Film): Main 150 nm Al film deposited at low temperature (19 °C) to preserve the smooth structure established by the seed layer.
  5. Standard Nb/AlOx/Nb Fabrication: Three-layer structure (Nb 200 nm / Al 7 nm / Nb 80 nm) deposited in one vacuum cycle. The Al barrier was oxidized in 1 mbar pressure for 20 minutes.
  6. Structural Characterization: Extensive use of X-ray diffraction (XRD), Atomic Force Microscopy (AFM), and Scanning Transmission Electron Microscopy (STEM) to analyze elemental composition, crystal structure, surface roughness, and grain size.

The realization of high-quality, hysteresis-free SIS junctions with parameters approaching theoretical limits is essential for next-generation superconducting devices.

  • Quantum Computing: High-coherence Josephson junctions are the fundamental building blocks for superconducting qubits, requiring minimal defects and low noise characteristics.
  • THz and Sub-THz Receivers: SIS mixers and generators used in radio astronomy and remote sensing require low specific capacitance and high characteristic voltage (Vc) for optimal frequency response and efficiency.
  • Superconducting Quantum Interference Devices (SQUIDs): Improved junction quality leads to higher sensitivity and lower noise in SQUIDs used for ultra-precise magnetic field measurements in medical imaging (MEG) and geophysical surveys.
  • High-Speed Digital Electronics: Achieving hysteresis-free junctions (βc less than 1) allows for the development of high-speed superconducting logic circuits (RSFQ) operating at frequencies well above 400 GHz without the need for resistive shunting.
  • Cryogenic Detectors: Used in highly sensitive bolometers and calorimeters where maximizing the superconducting energy gap (Δ) and minimizing leakage currents are critical for detection efficiency.
View Original Abstract

Tunneling Josephson junctions of the superconductor-insulator-superconductor (SIS) type have a history of more than 50 years, and theoretical estimates of the ultimate parameters of devices for receiving and processing signals based on them look very promising. In practice, in many cases, the actually achieved parameters turn out to be much worse than the theoretical ones, so for niobium SQUIDs the characteristic voltage Vc=IcRn at best reaches 200 µV, and according to theory it should be up to 2 mV. For Terahertz SIS mixers and oscillators, the main problems are a large specific capacitance, hysteresis, and leakage currents. These problems may be related to the morphology and crystal structure of superconductor films. In practice, films are granular, tunnel barriers are nonuniform, the effective area is about 10% of geometric area, leakage currents, parasitic capacitances occur. The crystal structure determines fundamentally different properties of the same elements, for example, for carbon it is diamond, graphite, fullerenes, nanotubes. Important components of a promising superconducting technology are: the use of single-crystal substrates matched in lattice constant and orientation with the grown films, optimization of growth temperature conditions, controlled formation of an oxide or nitride tunnel barrier. One option is to use a Schottky barrier for the semiconductor interlayer instead of a dielectric or normal metal one. This review presents the results of studying films by X-ray diffraction diagnostics, atomic force microscopy, and electron microscopy, showing the main bottlenecks of the existing technology with the deposition of niobium, niobium nitride, and aluminum films on oxidized standard silicon substrates, as well as the results of quasi-epitaxial growth of films on single-crystal substrates at various temperature conditions. Reproducible manufacturing of high-quality tunnel junctions can be achieved by implementing atomically smooth surfaces of tunnel contacts, which will improve the signal and noise characteristics of superconducting devices for receiving and processing information.