ISOCTA
Institute for Scientific Operations, Cryogenics & Technical Applications

Solid-State Qubit Architectures

Figure 1: Nb/Al-AlO\u2093/Nb flux qubit chip mounted in the sample holder of the dilution refrigerator in Building 2. The qubit loop (approximately 10 \u00b5m across) is visible at the centre of the chip, surrounded by the on-chip DC SQUID readout circuit. Bonding wires connect the chip carrier to the filtered measurement lines.

Overview

This programme investigates superconducting flux and phase qubits — micrometre-scale superconducting circuits that behave as quantum two-level systems at millikelvin temperatures. The work focuses on coherent manipulation, decoherence mechanisms, and measurement techniques rather than on large-scale integration or fault-tolerant architectures.

Technical Approach

Device Fabrication

Qubit devices are fabricated in the Building 1 cleanroom using a Nb/Al-AlOx/Nb trilayer process with shadow-evaporated aluminium Josephson junctions. Junction critical current densities are targeted at 50–200 A/cm2 with junction areas of 0.05–0.2 μm2, yielding junction Ic values in the 100–400 nA range. The qubit loop dimensions (typically 5–15 μm) set the inductive energy scale relative to the Josephson coupling energy.

On-chip readout is provided by a DC SQUID inductively coupled to the qubit loop, operated as a switching magnetometer. All measurement lines are heavily filtered — copper-powder filters at the 1 K stage and RC filters at the mixing chamber — to suppress electromagnetic noise from room-temperature electronics.

Measurement

Qubit measurements are conducted in the primary dilution refrigerator in Building 2 at a base temperature below 20 mK. Coherent control is achieved using microwave pulses at the qubit Larmor frequency (typically 4–8 GHz), delivered via a capacitively-coupled on-chip microwave line. Rabi oscillations, Ramsey fringes, and spin-echo sequences are used to characterise coherence times.

Recent results (October 2003) demonstrated coherent manipulation of a flux qubit with a gate fidelity exceeding 99%, as measured by randomised benchmarking. The dominant decoherence source at the current operating point is believed to be charge noise from microscopic two-level fluctuators in the junction tunnel barrier — this is the focus of ongoing materials optimisation work. [Park & Okonkwo, Phys. Rev. B 2003]

Figure 2: Rabi oscillation data for a flux qubit at 18 mK, showing the qubit state population as a function of microwave pulse duration. The oscillation frequency scales linearly with microwave amplitude, as expected for coherent Rabi driving. The decay envelope yields a coherence time T\u2082* of approximately 120 ns for this device.

Decoherence Studies

A significant fraction of the programme is devoted to understanding decoherence mechanisms. Measurements of T1 (energy relaxation) and T2 (dephasing) times as functions of temperature, flux bias, and junction parameters provide insight into the noise environment. At present, T2 is limited to approximately 120–180 ns at the optimal bias point, with T1 typically 2–5 μs.

Improving these figures requires both materials optimisation (reducing the density of tunnelling two-level systems in the junction oxide) and electromagnetic environment engineering (improved filtering, reduced parasitic modes in the sample enclosure). These efforts are ongoing.

Related Work

  • SQUID development: The on-chip DC SQUID readout amplifiers are being optimised for lower noise temperature in a parallel programme led by Dr. Chen. [Electromagnetic Systems]
  • Cryogenic infrastructure: The sub-20 mK measurement capability depends on the Building 2 cryogenics facility and the Institute's helium liquefier. [Facilities]

← Quantum Information Division  |  ← Research overview