Principles of Superconducting Quantum Computers

8 Josephson Junctions

The device that is the workhorse of most superconducting computers is the Josephson Junction, which consists of two superconducting regions separated by a non-superconducting layer—usually an insulator. A typical junction is fabricated from aluminum films with thicknesses in the range of 35–85 nm. The native oxide of the aluminum is used to form the insulating layer. This structure exhibits a non-linear inductance, which results in different spacings between the excited energy levels and enables us to uniquely address only two states—a key requirement for a qubit. By biasing the Josephson junction with different currents, it can also be used to realize a time-varying inductance that is used in low noise parametric amplifiers. Finally, when two junctions are placed in parallel, the effective inductance can be varied by the application of a magnetic field. This makes it possible to realize tunable qubits.

The useful dynamics result from Cooper pairs that “tunnel” across the non-superconducting layer in the junction. This tunneling phenomenon is purely quantum mechanical and very unintuitive. For this reason we begin the chapter with a basic discussion of tunneling.

8.1 Tunneling

Suppose that we have a tennis ball in the bottom of a box. The only way to get the ball out of the box classically is to lift it up and over the side. In doing this, we are giving the ball additional gravitational potential energy so that its energy exceeds the potential energy ...