It turns out that two of the coolest phenomena in physics—superconductivity and quantum tunneling—can occur together. Their "child"—the Josephson effect—earned its originator the Nobel Prize in Physics. When magnetic fields are added to the mix, a Josephson junction is the electrical analogue of single-slit light interference, which allows it to serve as a building block for some of today's quantum computers.
Advertisement
Two purely quantum phenomena that cannot be explained by classical physics are: the existence of superconductors—materials in which electric current flows without resistance thanks to the formation of electron pairs [1]; and quantum tunnelling, in which particles—in our case electrons—pass through metaphorical walls, regions where they are classically forbidden to exist [2].
One of the most remarkable (and useful) properties of superconductors is that they conduct electric current without resistance. In 1962 Brian Josephson, then a doctoral student, asked, "Is superconducting current tunnelling possible?" Can the wondrous, resistance-free current flowing through a superconductor continue to flow without resistance through a material that is not superconducting? At first glance the answer is yes, because tunnelling is a general quantum-mechanical phenomenon. The deeper question is whether this effect is measurable. Can a supercurrent be detected even when Cooper pairs have to tunnel, or will a layer only a few atoms thick reduce the tunnelling probability to a practically negligible value?
To set electric current flowing in a superconductor one needs only a slight "push" on the electron pairs—Cooper pairs—and they will keep moving forever as long as they have a path forward. In ordinary conductors, electric current loses energy as heat, so additional energy (for example, an applied voltage) must continually be supplied to keep electrons moving. It is like sliding a hockey puck on an air-hockey table: when the air is on, the puck glides with almost no friction and a single tap is enough; when the air is off, one tap is insufficient, and continuous pushing is required to keep it moving.
In a paper of fewer than three pages, Josephson proposed a theory of Cooper-pair tunnelling through an insulating layer [3]. The device he conceived—a sandwich of two superconductors separated by a non-superconducting material—is named after him: the "Josephson junction". If Cooper pairs can tunnel through the insulating layer in which they are classically forbidden to exist, then, just as in a superconductor, an electric current could flow through the junction with no applied voltage. This would mean that despite the presence of a non-superconducting layer in the sandwich, it would be effectively undetectable.
Josephson’s calculation showed not only that tunnelling is possible, but that it is measurable. As in bulk superconductors, a Josephson junction can support a current with no applied voltage. This current can increase up to a maximum called the “critical current.” The critical current of a Josephson junction is lower than that of the superconducting electrodes and depends, among other factors, on the thickness and cross-sectional area of the insulating layer. This is the “direct Josephson effect” (DC).
There is also an alternating-current effect: applying a constant voltage to a Josephson junction generates an AC current whose frequency depends on the voltage across the junction, the electron charge, and Planck’s constant. The AC Josephson effect—the ability to measure the current’s oscillation rate with great precision, combined with the fact that the electron charge and Planck’s constant are fundamental constants—has led to today’s definition of the standard unit of electric potential, the volt, which is based on voltage and current measurements in Josephson junctions. Josephson’s paper focuses on a junction with an insulating barrier but notes that similar results can be obtained with a "normal" conducting layer.
The simplest way to create a Josephson junction is by oxidizing the superconducting metal itself. An oxide layer forms naturally when the metal is exposed to air, and another metallic layer is deposited on top. The oxide thickness depends on the metal and growth conditions, but typically does not exceed a few nanometers—thin enough for Cooper pairs to tunnel through. Before Josephson’s paper, it was commonly thought that supercurrent through oxide layers arose from "nano short circuits", points where the two superconductors touch. Josephson’s work predicted a unique relationship between current and voltage in the junction, giving experimentalists a way to distinguish the two scenarios.
In 1963, Sidney Shapiro measured the AC effect in what would later be called "Shapiro steps". A plot of current versus voltage across the junction appears as steps: the current changes in multiples of the constants mentioned above, exactly as Josephson had predicted.
Josephson’s calculations even surprised him, because until then the prevailing conception was that tunnelling of Cooper pairs—the carriers of supercurrent—was almost impossible and therefore unmeasurable. One skeptic was John Bardeen, who would later receive his second Nobel Prize for his role in developing the theory of superconductivity. It turns out that even the world expert on tunnelling and superconductivity can be wrong.
In the next part we will add magnetic fields, more sandwiches, and Angela Merkel, and see how to build a quantum bit from superconductors.
English editing: Elee Shimshoni
To the next post in the series >>>
References:
Further reading:
