History/Background of Superconductivity
In the beginning, Onnes discovered superconductivity in 1911. He earned the discovery after also becoming the first scientist to learn how to liquify helium. Liquid helium's boiling point is 4.2K, so this opened up the possibility of studying samples (or anything) at 4K, if you were patient enough to use liquid helium (LHe) for that purpose. Not very long after this liquid helium development, Onnes cooled various metals (lead, tin, mercury) to 4K, and the discovery of superconductivity was made.
There are two pinnacle qualities of a superconductor, both in theory and in practice. The first is perfect conductivity. Normal metals are said to be excellent conductors, and that is basically right. However, when a current flows through metals, there is always some loss of energy to heat. In a superconductor, this is not the case. Currents flow without loss in a superconductor, or the superconductor has exactly zero resistance. When you cool a superconductor to very low temperatures, there is a special temperature called Tc at which the resistance suddently goes to zero. Currents in a superconductor have been observed to persist for years in circuit loops, and this "persistent current" technique forms the basis of creating strong, lasting magnetic fields at will!
The second fascinating characteristic of superconductivity is perfect diamagnetism, or the Meissner Effect, named after its principle discoverer. When you cool a superconductor starting from its normal metal (high temperature) phase through its transition temperature Tc, any magnetic fields which penetrated the sample while it was a normal metal are pushed out of the sample at the moment when it becomes superconducting! What this means is that, at some critical magnetic field Hc, the a superconducting sample will no longer be able to stay superconducting and will have to allow some magnetic flux to penetrate the sample. So now there are two parameters which can be used in order to cause a transition between non-superconducting and superconducting phases of a sample: temperature (with transition temperature Tc) and magnetic field (with transition field Hc).
Flux exclusion (perfect diamagnetism) also implies that external magnetic fields will be excluded at all costs, as long as the superconducting sample remains in the superconducting phase. So if you place a superconducting sample above a magnet (and are handy enough to balance it properly), the superconductor will levitate above the magnet! You've probably seen photos of this happening in laboratory experiments. The most famous levitation photo is probably still the Japanese sumo wrestler standing on a thick slab of superconductor (at right). Quite amazing. Levitation also occurs when placing a normal magnet which exhibits imperfect diamagnetism above another magnet, but as you might imagine, the effect is much weaker.
Well, for all its glorious features, physicists were relatively at a loss to explain conventional superconductivity on a microscopic level until the year 1957. (Prior to that date, yes, there were some very good contributions, such as the London equations, the superconducting gap, and Ginzburg-Landau theory, but this is a very brief synopsis.) In that year, Bardeen, Cooper, and Schrieffer published their theory of superconductivity (which became known as BCS theory). In it, they proposed that, in the presence of an interaction between electrons, however weak, there can be a state which has an energy even lower than that of the traditional Fermi Sea, which is the model for how electroncs behave in a Fermi Liquid, or normal metal. They proposed that a virtual process between electrons through phonons could be this weak interaction. Their predicted value for the superconducting energy gaps at T approaching zero matched up with experiment several samples well. This and a few correct other predictions led most scientists to believe that BCS theory was the theory of superconductivity.
In the same year which BCS theory emerged, Abrikosov published a paper describing what he called "Type II" superconductors, making an explicit contrast to those of "Type I". Type II superconductors are a bit more complicated than Type I for several reasons. First, when you raise a magnetic field through the critical temperature Hc in a Type I superconductor, it immediately goes back to a normal metal state. However, in Type II superconductors this is not the case. There is an intermediate regime where magnetic flux can actually penetrate the sample in certain areas, but the sample is still superconducting. Inside these "areas", the sample locally goes normal, but all around the area, the sample is superconducting. These areas were given the name "vortices": they are small places where magnetic fields go through the sample. As you continue to increase the magnetic field inside this intermediate regime, more and more vortices appear (and existing vortices become more densely packed together). Finally, at a second critical temperature, all the vortices merge together and the entire sample becomes a normal metal once again. Roughly speaking, the field at which the magnetic field just starts to penetrate the sample is called Hc1, and the field where the entire sample goes back to normal is called Hc2, with Hc2 > Hc1, of course. Abrikosov also predicted that vortices themselves would seek an ordered state, and that ordered state with minimum energy was a triangular lattice. (Actually he said it was a square lattice, but later a triangular lattice was found to have slightly less energy.)
The final major chapter in this tiny history of superconductivity that I'd like to discuss concerns Josephson supercurrent. Modern theory of superconductivity attributes an order parameter &psi for any superconducting sample. &psi is complex, so it has a magnitude and phase angle, and it is a function of position in the sample. Now, if we take two samples, each with a different &psi, and press them together, what happens? Well, theoretically, they join together, forming one superconductor. :-) Okay, okay.. So what if we put a very tiny insulator between the two samples, so that they are almost touching? (Such an insulator is called a weak link.) In that case, a funny thing happens. Because each superconductor has a phase angle of its own, the order parameters quantum mechanically "interfere" with each other, and a current known as a Josephson supercurrent (a current of Cooper pairs) flows between the two samples. Even more mysterious is that this current flows (in theory) regardless of voltage applied to the junction: It flows even at zero voltage! This strange behavior was first predicted by Josephson in 1962 while still a graduate student. As an application of this phenomenon, SQUIDs (Superconducting Quantum Interference Devices) are possible, and they are used to make very precise measurements of magnetic fields.
Does the story end here? Surprisingly, no. In 1986, Georg Bednorz and Alex Muller of IBM discovered superconductivity in a very strange place: the compound LaBaCuO, a ceramic material consisting of stacked layers of atoms. More surprisingly, however, Tc for this material was found to be 30K. Quickly, new compounds with structures similar to LBCO were found, all superconducting with a higher optimum Tc than in any conventional superconductor. Furthermore, none of these materials seemed to be explainable using traditional BCS theory. Thus, the race to find a room temperature (T=300K) superconductor was begun.
Tinkham, Michael. Introduction to Superconductivity, Second Edition. Dover Publications, Inc. Mineola, New York. 1996.Next: The High-Tc Cuprate Superconductors