Third, scientists began to realize that metals, and many other materials that conduct electricity well, have a regular structure at the molecular level. The atoms and molecules of these substances are arranged in a regular three-dimensional grid pattern, termed a lattice, and held in position by interatomic electrical forces.
Finally, in 1897, J.J. Thomson found the elusive carrier of the electrical current. He originally termed it the "corpuscle," but it soon found its present name, the electron. All electrical currents are carried by electrons.
Again, lots of history before we have the tools in hand to understand the flow of electricity through conductors—but not yet, as we shall see, to explain superconductivity.
Electricity is caused by the movement of electrons. Thus a good conductor must have plenty of electrons readily able to move, which are termed free electrons. An insulator has few or no free electrons, and the electrons in such materials are all bound to atoms.
If the atoms of a material maintain exact, regularly spaced positions, it is very easy for free electrons to move past them, and hence for current to flow. In fact, electrons are not interfered with at all if the atoms in the material stand in a perfectly regular array. However, if the atoms in the lattice can move randomly, or if there are imperfections in the lattice, the electrons are then impeded in their progress, and the resistance of the material increases.
This is exactly what happens when the temperature goes up. Recalling that heat is random motion, we expect that atoms in hot materials will jiggle about on their lattice sites with the energy provided by increased heat. The higher the temperature, the greater the movement, and the greater the obstacle to free electrons. Therefore the resistance of conducting materials increases with increasing temperature.
This was all well-known by the 1930s. Electrical conduction could be calculated very well by the new quantum theory, thanks largely to the efforts of Arnold Sommerfeld, Felix Bloch, Rudolf Peierls, and others. However, those same theories predicted a steady decline of electrical resistance as the temperature went towards absolute zero. Nothing predicted, or could explain, the precipitous drop to zero resistance that was encountered in some materials at their critical temperature. Superconductivity remained a mystery for another quarter of a century. To provide its explanation, it is necessary to delve a little further into quantum theory itself.
2.11 Superconductivity and statistics. Until late 1986, superconductivity was a phenomenon never encountered at temperatures above 23 K, and usually at just a couple of degrees Kelvin. Even 23 K is below the boiling point of everything except hydrogen (20 K) and helium (4.2 K). Most superconductors become so only at far lower temperatures (see TABLE 2.2). Working with them is thus a tiresome business, since such low temperatures are expensive to achieve and hard to maintain. Let us term superconductivity below 20 K "classical superconductivity," and for the moment confine our attention to it. TABLE 2.2 shows the temperature at which selected materials become superconducting when no magnetic field is present.
Note that all these temperatures are below the temperature of liquid hydrogen (20 K), which means that superconductivity cannot be induced by bathing the metal sample in a liquid hydrogen bath, although such an environment is today readily produced. For many years, the search was for a material that would sustain superconductivity above 20 K.
For another fifteen years after the 1911 discovery of superconductivity, there seemed little hope of explaining it. However, in the mid-1920s a new tool, quantum theory, encouraged physicists to believe that they at last had a theoretical framework that would explain all phenomena of the subatomic world. In the late 1920s and 1930s, hundreds of previously-intractable problems yielded to a quantum mechanical approach. And the importance of a new type of statistical behavior became clear.
On the atomic and nuclear scale, particles and systems of particles can be placed into two well-defined and separate groups. Electrons, protons, neutrons, positrons, muons, and neutrinos all satisfy what is known as Fermi-Dirac statistics, and they are collectively known as fermions. For our purposes, the most important point about such particles is that their behavior is subject to the Pauli Exclusion Principle, which states that no two identical particles obeying Fermi-Dirac statistics can have the same values for all physical variables (so, for example, two electrons in an atom cannot have the same spin, the same angular momentum, and the same energy level). The Pauli Exclusion Principle imposes very strong constraints on the motion and energy levels of identical fermions, within atoms and molecules, or moving in an atomic lattice.
The other kind of statistics is known as Bose-Einstein statistics, and it governs the behavior of photons, alpha particles (i.e. helium nuclei), and mesons (pions and some other subnuclear particles). These are all termed bosons. The Pauli Exclusion Principle does not apply to systems satisfying Bose-Einstein statistics, so bosons are permitted to have the same values of all physical variables; in fact, since they seek to occupy the lowest available energy level, they will group around the same energy.
In human terms, fermions are loners, each with its own unique state; bosons love a crowd, and they all tend to jam into the same state.
Single electrons are, as stated, fermions. At normal temperatures, which are all well above a few Kelvins, electrons in a metal are thus distributed over a range of energies and momenta, as required by the Pauli Exclusion Principle.
In 1950, H. Fröhlich suggested a strange possibility: that the fundamental mechanism responsible for superconductivity was somehow the interaction of free electrons with the atomic lattice. This sounds at first sight highly improbable, since it is exactly this lattice that is responsible for the resistance of metals at normal temperatures. However, Fröhlich had theoretical reasons for his suggestion, and in that same year, 1950, there was experimental evidence—unknown to Fröhlich—that also suggested the same thing: superconductivity is caused by electron-lattice interactions.
This does not, of course, explain superconductivity. The question is, what does the lattice do? What can it possibly do, that would give rise to superconducting materials? Somehow the lattice must affect the free electrons in a fundamental way, but in a way that is able to produce an effect only at low temperatures.
The answer was provided by John Bardeen, Leon Cooper, and Robert Schrieffer, in 1957 (they got the physics Nobel prize for this work in 1972). They showed that the atomic lattice causes free electrons to pair off. Instead of single electrons, moving independently of each other, the lattice encourages the formation of electron couplets, which can then each be treated as a unit. The coupling force is tiny, and if there is appreciable thermal energy available it is enough to break the bonds between the electron pairs. Thus any effect of the pairing should be visible only at very low temperatures. The role of the lattice in this pairing is absolutely fundamental, yet at the same time the lattice does not participate in the pairing—it is more as if the lattice is a catalyst, which permits the electron pairing to occur but is not itself affected by that pairing.
The pairing does not mean that the two electrons are close together in space. It is a pairing of angular momentum, in such a way that the total angular momentum of a pair is zero. The two partners may be widely separated in space, with many other electrons between them; but, like husbands and wives at a crowded party, paired electrons remain paired even when they are not close together.
Now for the fundamental point implied by the work of Cooper, Bardeen, and Schrieffer. Once two electrons are paired, that pair behaves like a boson, not a fermion. Any number of these electron pairs can be in the same low-energy state. More than that, when a current is flowing (so all the electron pairs are moving) it takes more energy to stop the flow than to continue it. To stop the flow, some boson (electron pair) will have to move to a different energy level; and as we already remarked, bosons like to be in the same state.
To draw the chain of reasoning again: superconductivity is a direct result of the boson nature of electron pairs; e
lectron pairs are the direct result of the mediating effect of the atomic lattice; and the energy holding the pairs together is very small, so that they exist only at very low temperatures, when no heat energy is around to break up the pairing.
2.12 High-temperature superconductors. We now have a very tidy explanation of classical superconductivity, one that suggests we will never find anything that behaves as a superconductor at more than a few degrees above absolute zero. Thus the discovery of materials that turn into superconductors at much higher temperatures is almost an embarrassment. Let's look at them and see what is going on.
The search for high-temperature superconductors began as soon as superconductivity itself was discovered. Since there was no good theory before the 1950s to explain the phenomenon, there was also no reason to assume that a material could not be found that exhibited superconductivity at room temperature, or even above it. That, however, was not the near-term goal. The main hope of researchers in the field was more modest, to find a material with superconductivity well above the temperature of liquid hydrogen. Scientists would certainly have loved to find something better yet, perhaps a material that remained superconducting above the temperature of liquid nitrogen (77 K). That would have allowed superconductors to be readily used in many applications, from electromagnets to power transmission. But as recently as December 1986, that looked like an impossible dream.
The first signs of the breakthrough had come early that year. In January 1986, Alex Müller and Georg Bednorz, at the IBM Research Division in Zurich, Switzerland, produced superconductivity in a ceramic sample containing barium, copper, oxygen, and lanthanum (one of the rare-earth elements). The temperature was 11 K, which was not earth-shaking, but much higher than anyone might have expected. Müller and Bednorz knew they were on to something good. They produced new ceramic samples, and little by little worked the temperature for the onset of superconductivity up to 30 K. The old record, established in 1973, had been 23 K. By November, Paul Chu and colleagues at the University of Houston, and Tanaka and Kitazawa at the University of Tokyo had repeated the experiments, and also found the material superconducting at 30 K.
Once those results were announced, every experimental team engaged in superconductor research jumped onto the problem. In December, Robert Cava, Bruce van Dover, and Bertram Batlogg at Bell Labs had produced superconductivity in a strontium-lanthanum-copper-oxide combination at 36 K. Also in December, 1986, Chu and colleagues had positive results over 50 K.
In January 1987, there was another astonishing breakthrough. Chu and his fellow workers substituted yttrium, a metal with many rare-earth properties, for lanthanum in the ceramic pellets they were making. The resulting samples went superconducting at 90 K. The researchers could hardly believe their result, but within a few days they had pushed up to 93 K, and had a repeatable, replicable procedure. Research groups in Tokyo and in Beijing also reported results above 90 K in February.
Recall that liquid nitrogen boils at 77 K. For the first time, superconductors had passed the "nitrogen barrier." In a bath of that liquid, a ceramic wire using yttrium, barium, copper, and oxygen was superconducting.
The end of the road has still not been reached. There have been hints of superconductive behavior at 234 K. This is only -40deg.C, just a few degrees below the temperature at which ammonia boils.
Fascinating, and the natural question is, can roomtemperature superconductors, the Holy Grail of this field, ever be produced?
Unfortunately, the question cannot be answered. There is no accepted model to explain what is going on, and it would not be unfair to say that at the moment experiment is still ahead of theory.
The Bardeen, Cooper, and Schrieffer (BCS) theory of superconductivity leads to a very weak binding force between electron pairs. Thus according to this theory the phenomenon ought not to occur at 90 K, still less at 240 K. At the same time, the theory tells us that any superconductivity, high-temperature or otherwise, is almost certainly the result of free electrons forming into pairs, and then behaving as bosons. In classical superconductivity, at just a few degrees above absolute zero, the mediating influence that operates to form electron pairs can be shown to be the atomic lattice itself. That result, in quantitative form, comes from the Cooper, Bardeen, and Schrieffer approach. The natural question to ask is, What other factor could work to produce electron pairs? To be useful, it must produce strong bonding of electron pairs, otherwise they would be dissociated by the plentiful thermal energy at higher temperatures. And any electron pairs so produced must be free to move, in order to carry the electric current.
Again, we are asking questions that take us beyond the frontiers of today's science. Any writer has ample scope for speculation.
2.13 Making it work. Does this mean that we now have useful, workhorse superconductors above the temperature of liquid nitrogen, ready for industrial applications? It looks like it. But there are complications.
Soon after Kamerlingh Onnes discovered superconductivity, he also discovered (in 1913) that superconductivity was destroyed when he sent a large current through the material. This is a consequence of the effect that Oersted and Ampère had noticed in 1820; namely, that an electric current creates a magnetic field. The temperature at which a material goes superconducting is lowered when it is placed in a magnetic field. That is why the stipulation was made in TABLE 2.2 (p. 57) that those temperatures apply only when no magnetic field is present. A large current creates its own magnetic field, so it may itself destroy the superconducting property.
For a superconductor to be useful in power transmission, it must remain superconducting even though the current through it is large. Thus we want the critical temperature to be insensitive to the current through the sample. One concern was that the new high-temperature superconductors might perform poorly here, and the first samples made were in fact highly affected by imposed magnetic fields. However, some of the new superconductors have been found to remain superconducting at current up to 1,000 amperes per square millimeter, and this is more than adequate for power transmission.
A second concern is a practical one: Can the new materials be worked with, to make wires and coils that are not too brittle or too variable in quality? Again the answers are positive. The ceramics can be formed into loops and wires, and they are not unduly brittle or fickle in behavior.
The only thing left is to learn where the new capability of high-temperature superconductors will be most useful. Some applications are already clear.
First, we will see smaller and faster computers, where the problem of carrying off heat caused by dissipation of energy from electrical currents in small components is a big problem. This application will exist, even if the highest temperature superconductors cannot tolerate high current densities.
Second, as Faraday discovered, any tightly-wound coil of wire with a current running through it becomes an electromagnet. Superconducting coils can produce very powerful magnets of this type, ones which will keep their magnetic properties without using any energy or needing any cooling. Today's electromagnets that operate at room temperature are limited in their strength, because very large currents through the coils also produce large heating effects.
Third, superconductors have another important property that we have not so far mentioned, namely, they do not allow a magnetic field to be formed within them. In the language of electromagnetic theory, they are perfectly diamagnetic. This is known as the Meissner Effect, and it was discovered in 1933. It could have easily been found in 1913, but it was considered so unlikely a possibility that no one did the experiment to test superconductor diamagnetism for another twenty years.
As a consequence of the Meissner Effect, a superconductor that is near a magnet will form an electric current layer on its surface. That layer is such that the superconductor is then strongly repelled by the magnetic field, rather than being attracted to it. This permits a technique known as magnetic levitation to be used to lift and support heavy objects. Magnets, suspended above a l
ine of superconductors, will remain there without needing any energy to hold them up. Friction-free support systems are the result, and they should be useful in everything from transportation to factory assembly lines. For many years, people have talked of super-speed trains, suspended by magnetic fields and running at a fraction of the operating costs of today's locomotives. When superconductors could operate only when cooled to liquid hydrogen temperatures and below, such transportation ideas were hopelessly expensive. With high-temperature superconductors, they become economically attractive.
And of course, there is the transmission of electrical power. Today's transmission grids are full of transformers that boost the electrical signal to hundreds of thousands of volts for sending the power through the lines, and then bring that high-voltage signal back to a hundred volts or so for household use. However, the only reason for doing this is to minimize energy losses. Line heating is less when electrical power is transmitted at low current and high voltage, so the higher the voltage, the better. With superconductors, however, there are no heat dissipation losses at all. Today's elaborate system of transformers will be unnecessary. The implications of this are enormous: the possible replacement of the entire electrical transmission systems of the world by a less expensive alternative, both to build and to operate.
However, before anyone embarks on such an effort, they will want to be sure that the technology has gone as far as it is likely to go. It would be crazy to start building a power-line system based on the assumption that the superconductors need to be cooled to liquid nitrogen or liquid ammonia temperatures, if next year sees the discovery of a material that remains superconducting at room temperature and beyond.
Borderlands of Science Page 6
