Understanding Superconductivity
The discovery in the last few years of a new class of superconductors, which carry electrical currents without any loss of energy at relatively high temperatures, has brought about a storm of interest in the potential usefulness of these materials. It has led to a frantic race to find even better materials and to a level of media attention quite unprecedented in the recent history of science and technology. Newspaper, magazine, and television reports have envisioned breakthroughs in cheaper, more efficient power transmission, magnetic lévitation, ultrafast computers, medical imaging, and more. We are told that a new era of technology is dawning.
Is all this really about to happen, or is it simply media hype—a pipe dream to be forgotten when the practical difficulties of putting these substances to work are better understood? Not a few earlier technologies have been almost as highly touted but failed in the marketplace for one reason or another, among them solar photovoltaics, thermoelectric heating and cooling, turbine automobiles, and titanium-based structural materials.
The answer to the question isn’t clear even to the experts, but a good idea of the possibilities and difficulties can be obtained by understanding how superconductivity has developed and what it has and hasn’t achieved so far. The history goes back almost to the turn of the century.
Heike Kamerlingh Onnes, a physicist at the University of Leyden in Holland, succeeded in liquefying helium in 1908. Helium boils at about 4 Kelvins—that is, 4 Celsius degrees above absolute zero—and Kamerlingh Onnes’s new ability to achieve and maintain such a low temperature enabled him to discover superconductivity in 1911. He found that the electrical resistance of mercury disappears completely at temperatures below 4.2 K. He later found that an electrical current established in a superconducting circuit would persist undiminished for hours, days, or even weeks without the aid of a battery or any other energy source, as long as the temperature remained low enough.
For three-quarters of a century since then, this strange effect has had an enduring interest for physicists trying to understand all its puzzling ramifications. Their work has won no fewer than four Nobel Prizes, the first to Kamerlingh Onnes in 1913 and the most recent to Johannes Georg Bednorz and Karl Alex Müller of IBM in 1987. Bednorz and Müller discovered the new class of oxide superconductors, which has won all the popular attention.
Scientists have from the beginning envisioned all sorts of practical applications for superconductors, most obviously lossless power lines, motors, and transformers and highstrength electromagnets that consume no power. A few practical uses have been realized, particularly in the last twenty years, but progress has been very slow because of the low temperatures necessary. Temperatures below 20 K can be attained only with the aid of complex, expensive, and cumbersome liquidhelium equipment. So until very recently superconductivity has been almost entirely a toy of research scientists. The physicists themselves were long pessimistic about the pace of development. 1 remember a lecture given about fifteen years ago by Bernd Matthias, who was one of the discoverers of the high-temperature superconductors of the time. He concluded by discussing the glacial progress by which transition temperatures had risen to about 23 K by 1973. Afterward I asked him when we might see temperatures above the boiling point of liquid nitrogen (77 K) or, better yet, room temperature (300 K).
“Ah, McKelvey,” he replied, “you can’t be serious. We’re unlikely to see forty Kelvins during our lifetimes, and that may not be far from the ultimate limit.” He was right for himself, for he died quite suddenly a few years ago, but today the record high transition temperature (above which a material loses superconductivity) is above 120 K. That is the measure of Bednorz and Müller’s breakthrough.
In 1897 J. J. Thomson, a physics professor at Cambridge University, identified the electron and showed that electrons are detached from atoms when the atoms become ionized. It soon became clear that the flow of electric current through a metallic conductor could be traced to the motion of free electrons within the metal. A conductor could therefore be regarded as a container holding a “gas” of free, negatively charged electrons and a lattice of immobile, positively charged ions. In an electrically neutral substance the sum of all these charges is zero. In an insulator the electrons are not free but are tightly bound to the atoms of the lattice. Therefore, little or no current will flow when an external voltage is applied.
In an ideal insulator—if one existed—no current would flow no matter how large an electric field was present. In a “perfect” conductor all the current applied flows with no loss of energy. In such a substance a battery or other energy source is needed only to accelerate the free electrons to a desired velocity. They will then circulate indefinitely. Superconducting materials are in this sense perfect conductors, as long as you maintain the circumstances needed to sustain the superconducting state: a low temperature and a low magnetic field.
In normal conductors, frequent collisions with the lattice stop some electrons, reducing the current they carry to zero. They must then be reaccelerated by the energy source to maintain the flow of current. The electron energy dissipated in the collision process eventually turns up as heat, and the effect of the collision process manifests itself as electrical resistance. Resistance can be useful, of course, if your objective is to toast bread, but if you wish to distribute and sell electric power, it is detrimental, and if you can get rid of it altogether, you will be able to sell all the power you generate, not just the fraction you can persuade to reach your customers.
One way to lower the resistance of normal conductors is by lowering the frequency at which electron-lattice collisions occur. This can be done by lowering the temperature. Normally, as long as you reduce temperature, resistance will decline, but it will never go away entirely. One reason Kamerlingh Onnes was studying the resistance of mercury filaments at liquid-helium temperatures in 1911 was to learn about how resistance varies with temperature. Another reason was to investigate some of the ideas of quantum theory. Quantum theory had been introduced during the preceding decade, and though it had begun a revolution in physics, it was by no means universally accepted.
Kamerlingh Onnes’s initial observations, in April 1911, indicated that the electrical resistance of mercury—a solid at liquid-helium temperature—essentially vanishes over a very narrow range of temperatures between 4.22 and 4.19 K. A year later he and his colleagues showed that tin and lead also become superconducting below temperatures of 3.7 and 7.2 K respectively. By 1914, when World War I interrupted their work, the group had found that supercurrents would persist undiminished for hours in a superconducting lead coil. More recently it has been found that such currents will circulate for years with no external input of energy.
The group also found that the presence of a magnetic field lowers the transition temperature—the temperature at which the abrupt shift to superconductivity occurs. And because currents in a superconductor always generate magnetic fields, as you increase the current in any superconducting coil or circuit you inevitably lower the transition temperature. There is thus a maximum critical current that a superconducting circuit can carry at any given temperature and still remain superconducting. The critical current decreases as the temperature rises, and it becomes zero at the nominal transition temperature. It depends on the size and shape of the conductor and the geometry of the circuit as well as on the material. The critical fields and currents for superconductors like mercury, tin, and lead are too small to allow very strong solenoid magnets to be made from these materials—a fact that dashed Kamerlingh Onnes’s initial hopes for practical application.
After World War I, superconductivity research picked up where it had left off in 1914. At first Kamerlingh Onnes’s group undisputedly led the field, but after his retirement, in 1924, others entered the fray. Many new elements —among them zinc, cadmium, thallium, indium, gallium, and aluminum—were shown to be superconductors, but all their transition temperatures were low, from 0.85 K for zinc to 3.4 K for indium. Also, it became apparent that many elements, including sodium, potassium, copper, silver, and gold, never exhibit superconductivity. In 1928 Walther Meissner, in Berlin, found a new group of superconducting materials, including tantalum, thorium, and vanadium, and in 1930 he found that niobium had the highest transition temperature thus far—9.2 K.
Much time and effort was devoted to theoretical questions, such as what causes superconductivity, why some but not all metals are superconductors, and what determines the transition temperature. The behavior of the electron gas in a superconductor at the transition temperature appeared to be similar in many ways to that of a liquid at the boiling point, suggesting that the gas undergoes some change into a more highly ordered, condensed phase below the transition temperature, leaving the electrons ordered, coupled, or correlated in such a way that their motion is unaffected by the collisions that normally occur. This view has turned out to be correct, but the exact nature of what happens remained a mystery for a long time.
It was known early on that superconducting materials have unique magnetic properties. As Maxwell’s theory of electromagnetism predicted, a perfect conductor shields its interior from externally applied magnetic fields by setting up a kind of mirror of surface currents, whose magnetic field exactly cancels the external field inside the substance and opposes it outside. Since opposite magnetic poles repel each other, a strong repulsive force arises between the perfect conductor and an external magnet. This image repulsion is what causes the now familiar phenomenon of magnetic lévitation associated with superconductors, and it has given us the image of a magnetically levitated train gliding along smoothly and noiselessly without ever touching its track. (Such a system of currents also arises in a normal conductor or a superconductor above the transition temperature, but it experiences energy losses as the result of resistance and quickly dies out, allowing the external magnetic field to penetrate the substance.)
What happens to a superconductor cooled through its transition temperature in the presence of an external magnetic field is surprising. You might expect its own internal magnetic field to remain unchanged. But in 1933 Walther Meissner and his colleague H. Ochsenfeld found that this doesn’t happen at all. Instead, as the temperature is lowered below the transition point, surface currents suddenly appear and expel any existing magnetic field from the interior, where the field immediately drops to zero. It stays there as long as the material is superconducting. This phenomenon is known as the Meissner effect. Its significance is far-reaching, for it confirms that the superconducting state is characterized not simply by the absence of resistance but also by the lack of any interior magnetic field.
Actually the external field does not vanish exactly at the surface but penetrates a small distance, known as the penetration depth. The surface currents extend that far in too. This penetration depth is temperature-dependent; it is only a few hundred atomic spacings at the lowest temperatures and becomes very large as the transition temperature is approached.
During World War II the study of uperconductivity was again set side while physicists worked on iatters like nuclear weapons and radar. Afterward scientists started hacking away at it once more, discovering new superconducting materials, many of which were alloys and intermetallic compounds rather than pure elements. The importance of coherence length—the maximum distance over which two electrons can maintain the coupling that gives rise to superconductivity—began to be appreciated.
It emerged that superconductors can be divided into two rather different types. In the earliest known superconductors the coherence length is much greater than the field-penetration depth; these are now known as Type I superconductors. In the other kind, Type II superconductors, the coherence length is much smaller than the field-penetration depth. These substances tend to have much higher transition temperatures and critical fields. Their mechanical properties and response to external magnetic fields are also somewhat different from those of Type I materials. Most of the efforts toward practical applications have been made using Type II materials.
In 1957 there finally emerged a theory that explained the electron-coupling mechanism responsible for superconductivity. It was called the BCS theory and is still the state of the art regarding our understanding of the physics of superconductivity. It was named for its creators, John Bardeen, Leon Cooper, and J. Robert Schrieffer, who received the Nobel Prize in physics in 1972 for this work.
The BCS theory demonstrates that in any superconductor there is a set of quantized energy levels for coupled superconducting electrons at lower energies than those of uncoupled normal-state electrons. These two groups of energy levels are separated by an energy gap, in which there are no allowed states. In the lower energy states, coupled electron pairs, called Cooper pairs, can move and carry current with no interference from the collisions that affect normal electrons in the highenergy, uncoupled states, and the material is superconducting. Any interaction that would promote electrons into the higher-energy normal states has to supply enough energy to break the coupling and thus “lift” them across the gap of forbidden energies. The energy gap is a function of temperature, decreasing as the temperature rises. At the transition temperature it disappears altogether, and the material reverts to its normal state. Since the size and behavior of the energy gap depends on the strength of the coupling interaction, which differs in different materials, different substances will have different transition temperatures. The BCS theory also explains the Meissner effect and the phenomena of critical field and penetration depth.
The only problem is that electrons do not by themselves attract one another and couple; in fact, their negative charges are mutually repulsive. The BCS theory dispels this apparent contradiction by showing that the superconductor’s lattice itself can set up an attractive interaction between electrons. The lattice is composed of massive, slow-moving, positively charged ions, which at equilibrium occupy a network of regularly spaced lattice sites and balance the negative charge of the electrons. Instantaneously, however, an electron exerts a strong attractive force on surrounding positive ions. The ions then move off their normal lattice sites toward the electron. This creates a region of excess positive charge in the area. But the ions are heavy and move slowly in comparison with electrons; the electron speeds on its way, and the excess positive charge left behind can attract a second electron, coupling it to the first to form a superconducting electron pair. Because of the attractive interaction, the paired electrons have lower energy than uncoupled “normal” electrons. This coupling mechanism seemed adequate to explain superconductivity in every material until Bednorz and Müller appeared on the scene in 1986.
During the 1950s and 1960s new superconducting materials were discovered by the hundreds. Some had very high critical fields, but transition temperatures crept upward at a snail’s pace. There were, however, other important advances. Brian Josephson, in England, and Ivar Giaever, in the United States, explored the phenomenon of superconducting tunneling. Contrary to intuitive notions based on classical physics, there is always some probability that a particle can tunnel directly through a potential energy barrier even if it does not have sufficient energy to overcome it. If the barrier is relatively thick, this probability is infinitesimally small; if it is thin, there is a substantial possibility that some particles will leak through. The explanation of this is intimately associated with the role of probability in quantum theory and the fact that particles act like waves on a microscopic scale.
Tunneling is most easily observed where a thin, uniform insulating film separates two metallic electrodes, forming what is called a tunnel junction. Classically there should be no current when a DC voltage difference exists between the conductors. But if the insulating layer is thin enough—about a hundred-thousandth of a millimeter thick—currents can be observed as a result of tunneling.
In normal materials the phenomenon is not of much use; in superconducting materials, however, it has some unique and valuable features. Tunneling involving single electrons can provide extremely fast logic circuits and data storage-and-retrieval devices, since slight changes in voltage can produce discontinuous changes in current and these switching operations can be much quicker than those done by conventional means. Tunneling involving Cooper pairs depends on the differences in phase of the electron waves on either side of the junction and can therefore be used to observe quantum interference between electron waves, much as an optical interferometer detects the interference of two light beams. These effects can be used to fashion extremely sensitive magnetic field detectors, magnetometers, galvanometers, and a host of related devices, as well as to generate and detect coherent microwave and far-infrared radiation. Josephson and Giaever shared the 1973 Nobel Prize for the discovery of these tunneling phenomena.
By 1965 a new class of Type II materials had been developed, with relatively high transition temperatures and very high critical fields. These materials are intermetallic compounds such as Nb 3 Sn, Nb 3 Ga, and Nb 3 Ge. In 1973 a transition temperature of 23.2 K was reported for Nb3Ge, a record that stood until 1986, when the new oxide superconductors were discovered. With the advent of these new materials, engineers for the first time began to appreciate the scope of applications that was becoming possible.
The most obvious of these have to do with conventional electromagnetic devices, in which the role of superconductivity is simply to eliminate energy losses from resistive heating and perhaps to generate magnetic fields and forces larger than those normally obtainable. Devices like power lines, transformers, generators, and motors can be made practically lossless, and in many cases shrunk significantly, if superconductors with reasonably high transition temperatures and much higher critical fields and currents become available.
Large superconducting solenoids (magnetic coils) have been proposed as energy storage units for helping utilities distribute power more evenly over a twenty-four-hour period. Magnetic fields are measured in units of teslas (T). The earth’s local magnetic field is about 0.0001 T; the fields produced by the strongest normal permanent magnets are on the order of 1 T. The critical field of most Type I superconductors is less than 0.1 T, but the new Type II materials have critical fields ranging between 10 T and 30 T, with commensurately high critical current values. Thus, though transition temperatures hadn’t risen much by the seventies, great improvement had been obtained in regard to critical fields and currents. If you were willing to put up with liquidhelium cooling, you could hope to make superconducting solenoids much stronger than any normal electromagnet. There is enough energy stored in a cubic meter of a uniform 100 T magnetic field to power a five-kilowatt household demand—lighting fifty 100-watt bulbs—for more than two hours.
However, there are obstacles other than those associated with cooling. The new materials of the sixties and seventies were inherently hard, brittle, difficult to form into wires and coils, and poorly suited to withstand the high mechanical stresses that act on the windings of a high-field magnet. It took a decade of intense development work to solve these problems, but eventually high-field superconducting magnets became a reality.
These magnets have become useful in very special applications where the high cost and paraphernalia of liquid-helium refrigeration can be tolerated, such as NMR (nuclear magnetic resonance) medical imaging. NMR technology depends on a strong and extremely uniform magnetic field from a solenoid large enough to accommodate the human body. The patient enters the coil; a radio pulse is then broadcast, and the nuclei of the atoms of the body react like tiny magnets, each responding to a certain resonant frequency, which differs for atoms of each individual element. These frequencies can be detected by their minute effects on a receiving coil, and the result is used to map the body’s chemical makeup. This can be done with conventional magnets, but the cost is prohibitive; with superconductors, however, power consumption is zero once a steady current has been established. The much publicized superconducting supercollider project now under way in Texas depends on massively powerful—and expensive—superconducting magnets for guiding and focusing particle beams that will break matter into its most basic constituents.
The usefulness of devices based on superconducting tunneling has also begun to be recognized, and prototypes have been made for computer memories and logic circuits, microwave devices, magnetic-field detectors, and other measuring instruments. As always, the picture has been clouded by the need for cooling to extremely low temperatures.
In 1986 Johannes Bednorz and Karl Müller discovered the superconducting properties of barium-lanthanum-copper oxide (Ba x La 2-x CuO 4 ), which has a transition temperature of about 35 K. The discovery was based on a hunch that superconductivity might result from the interaction of electrons in certain metal oxides. The two scientists had pursued their work on a part-time basis for three or four years at IBM’s Zurich research lab. They were fortunate to be working for an industrial employer with the sense—and the financial strength—to let good people now and then pursue their own scientific interests.
Bednorz and Müller first worked with nickel oxides, without success. They then turned toward compounds containing copper and oxygen and finally found success partly by scientific intuition, partly by trial and error, and partly by pursuing hints in the literature. Ba x La 2-x CuO 4 (x is a small and variable quantity) is a black, sintered, semimetallic ceramic substance, not a metal at all in the traditional sense. Its superconductivity depends on, among other things, an annealing step that must be carried out at just the right temperature. The two men published their findings in Zeitschrift für Physik in September 1986.
This German journal is respected but not widely circulated. Nevertheless, news of this sort travels fast, and scientists were soon trying not only to confirm Bednorz and Müller’s results but also to improve on them by changing the recipe. Before long the effort developed into an all-out horse race, with superconductivity at the boiling point of nitrogen, 77 K, as the first goal. Reaching this temperature would allow liquid nitrogen rather than liquid helium to be used as a coolant, greatly cutting costs and simplifying the equipment needed. It was achieved with astonishing speed. In January 1987 Paul C. W. Chu of the University of Houston and Maw-Kuen Wu of the University of Alabama announced a new oxide superconductor with a transition temperature of 95 K. This material, an oxide of barium, copper, and the rare-earth element yttrium, has the approximate composition YBa 2 Cu 3 O 7 and has come to be known as the 123 compound for its relative atomic proportions of yttrium, barium, and copper. It, too, is a sintered ceramic material dependent on proper composition, structure, and preparation, especially annealing in oxygen at the right temperature. Like other oxide superconductors, it is a Type II material with a high critical field, and it is hard, brittle, and difficult to fashion into wires or coils.
The 123 compound isn’t difficult to make, however, if you have the ingredients and the recipe. It has even been made in high school chemistry labs. You simply mix the finely powdered oxides of copper, barium, and yttrium in the indicated proportions, bake the mixture at 950° C for six hours, cool it rapidly, grind it, cold-press it into the desired form, sinter it at 950° C for twelve hours, cool it slowly, anneal it in oxygen for several hours at 450° C, and remove it from the oven, ready to use.
The crystal lattice of the 123 compound has what is called a perovskite structure. It contains parallel planes, as well as linear chains, containing copper and oxygen atoms, forming an overall architecture like the steel structure of a skyscraper. It is believed that the interaction between electrons from copper and oxygen atoms in the copper-oxygen planes is responsible for superconductivity, the rest of the atoms playing no role except to keep the copper and oxygen in the right place. The optimal configuration isn’t known yet, so there’s reason to believe there may be ample room for improvement in this class of materials.
The 123 compound and similar superconductors differ profoundly from earlier superconducting substances. The BCS theory still seems adequate to describe their properties, but the picture of electrons coupled by an interaction with lattice ions is no longer appropriate. Electronlattice coupling is simply not strong enough to explain the high transition temperatures; also, some other effects associated with lattice coupling are absent in these materials. Something seems to couple electrons together, but the exact mechanism has not yet been found. The most recent theory, proposed by William A. Goddard III of Caltech, replaces the BCS coupling scheme with one involving a magnetic interaction. It predicts that while higher transition temperatures may still be attained, room temperature superconductivity simply cannot, at least with the copper-bearing oxides now being studied. Whether this is true only time—and a lot of experimental work—will reveal.
Theory didn’t catch up with experiment in the early history of superconductivity until 1957. Then the situation reversed, the field being dominated by the BCS theory, whose predictions tended to pave the way for experimental work. Now the tables have turned once more, and the theorists are scrambling to put together the pieces thrown their way by experimentalists. Is it better to work from a wellestablished theoretical framework or not? The answer isn’t clear, but it will be interesting to see what emerges from the present chaotic state of affairs. Maybe it’s best for now not to have some theorist around saying that this or that can’t be done.
The larger picture is also rapidly changing. For a very long time superconductivity research was charted totally within the physics community. During the past decade or two, things have changed. With the focus on applications, the limelight is increasingly being shared with chemists, engineers, materials scientists, and ceramicists. The new participants have introduced an increased emphasis on experimentation, empiricism, and phenomenology; their orientation is toward results rather than fundamental understanding. Whether this is desirable in the long run is debatable, but for the present it is an undeniable fact of life.
The increasingly interdisciplinary character of the research has created serious problems of communication. Physicists often do not fully understand what chemists mean by terms like oxidation potential , and chemists often do not understand physicists’ calculations involving path integrals and thermal Green’s functions. Physicists and chemists attach different meanings to the term resonance . Electrical engineers understand logic circuits and switching devices but may know little chemistry and nothing about ceramics. Ceramicists understand materials and materials processing, but may not grasp electronic circuits or device physics. A period of adjustment will be needed for these different constituencies to learn to work together more effectively.
A race is on, and the pace is very fast. The high-temperature-superconductivity session at the March 1987 meeting of the American Physical Society in New York was jammed with a thousand or so scientists. It did not break up until almost dawn the following morning and became known as the Woodstock of physics. More information is exchanged in hallways than at formal presentations. Graduate students spend all night in the labs; their professors spend most of their time on planes to and from meetings and conferences. New data are circulated in photocopied preprints, by newspapers or television, or even by word of mouth; if you only read the scientific journals, you will be hopelessly behind the times. In 1937 there were seventyfive known superconducting materials; in 1965, about six hundred. Now there are more than six thousand. The contest is being reported—and to some extent funded—by people who are not noted for length of attention span or long-term tenacity of purpose. How will this chaotic situation sort itself out, and what are the long-term prospects for superconducting materials and devices?
There is no question that the new class of oxide superconductors represents a very important scientific breakthrough. Whether this progress will be translated into important technology is far from certain. It is clear, however, that new technology will not emerge in response to six months or a year of frantic effort, no matter how much money is thrown its way. There is at present a great deal of overlap and duplication of effort; it is inevitable, even desirable, that many now absorbed in this enterprise will soon drop out. The optimization of this new class of materials, and the exploitation of their unique properties, is a long-term proposition, if past history is any guide. It is reasonable to believe that it may take twenty years or more.
There are serious, though probably not insurmountable, problems involved in making wires, coils, and other configurations from these hard, sintered materials. Their transition temperatures and critical fields are high, but in the materials now most easily available, critical currents are too low. The materials simply will not carry enough current without going normal. This may be because they are composed, apparently, of a network of superconducting filaments embedded in a non-superconducting matrix, the low critical current being associated with the small cross section of the filaments. This problem might be solved by producing the materials in crystalline form. At present, however, this is not easy to do. Thin film samples can be made that exhibit critical current densities that are much higher than those observed in the sintered ceramics. This may be good news for those concerned with switching devices and logic circuits, but it isn’t for anyone who must design high-field magnets or power-transmission systems.
At present it looks as though the initial uses of the new materials will be in logic circuitry, switching devices, memory elements, instruments, and sensors, in which the relatively large critical-current densities of thinfilm samples can be advantageously put to work. The next step beyond that might hinge on the discovery of processes for producing strong (albeit probably brittle) solenoidal coils with improved maximum current densities. This would allow the construction of good high-field electromagnets, transformers, and similar equipment. In the more distant future, presumably, lie applications that rely on strength, resilience, and flexibility, such as long-distance power-transmission and, conceivably, energy-storage systems.
It is relevant to note that electric-lighting technology went through similar stages of evolution. At first lamp filaments were made of carbon, which was brittle and subject to premature burnout. They operated in expensive, highly evacuated glass bulbs, which darkened rapidly, and they were fed from inefficient DC power lines. It took four decades before the gas-filled, AC-powered tungsten-filament lamp emerged; its ultimate production hinged on solution of the problem of forming tungsten —a brittle, fibrous, intractable metallic substance—into strong, flexible filaments as fine as a human hair. The task of working the new superconductors into useful forms is similar; the job may be more difficult, but it is not unprecedented.
Also, the challenge of refrigeration remains. Adequate cryogenic cooling can be provided at reasonable cost for many proposed applications, but the problem is not primarily economic. The obstacles lie instead in the areas of reliability, portability, simplicity, and freedom from maintenance. Using either liquid nitrogen or liquid helium imposes a burden of complexity on systems that might otherwise be simple, compact, and elegant. There is little sense in requiring a lot of bulky, cranky, unreliable cryogenic gear to cool a computer the size of a small suitcase.
If the full potential of superconducting technology is to become a reality, materials that are superconducting at room temperature will have to be found. In fact, transition temperatures considerably above room temperature will have to be attained, for although superconductivity exists at the transition temperature, the critical field and current are zero at that point. You must operate at substantially lower temperatures if you want your superconductors to perform any useful tasks. For this reason, transition temperatures of around 400 K (260° F) will be necessary for useful “room-temperature” superconductors. This brings us back to the second part of the question I asked the late Bernd Matthias: When, if ever, will room-temperature superconductivity be observed? I am now sixty-two and in good health. My actuarial life expectancy is about twenty years—knock on wood—though I may not live to see the next sunrise. Assuming the insurance companics are right, might I see superconductivity at room temperature?
I don’t know, of course, but I would hesitate to say no. We are still in a period when new ideas are being proposed and explored rapidly and very intensely; where this surge will end nobody knows. At present, new compounds are being prepared and tested by the hundreds, and older ones are being modified in every conceivable way. Transition temperatures in the 100 K range have been reported for a complex bismuth-calcium-strontiumcopper oxide, and 125 K (still-235° F) seems to have been achieved in a thallium-calciumbarium-copper oxide material. Even if the possibilities for the copper-oxygen interaction are exhausted at this point, the events of the past two years should encourage scientists and technologists to continue in hope of finding even better materials. It is now clear that nature does not require temperatures near that of liquid helium for superconductivity to manifest itself. There may be a long quest before us, one that could extend well into the next century, but there is reason to believe that the elusive goal of room-temperature superconductivity may be within our reach.
