ÐÓ°ÉÔ­´´

IN 1986, two researchers at IBM in Zurich, Switzerland, made what seemed
a momentous discovery. An unusual kind of ‘pottery’ made from oxides of
lanthanum, strontium and copper could conduct electricity without resistance
at 30 degrees above absolute zero. In other words, the ceramic material
was superconducting at 30 K. This temperature does not sound very high,
but until then, physicists had seen superconductivity at temperatures only
below 24 K. In fact, most physicists thought that superconductivity could
not exist above 35 K.

So, the two researchers, Georg Bednorz and Alex Muller, were working
in a field in which most people had given up hope of finding anything exciting.
They even had to disguise their work from their supervisor in order to be
able to do it. After Bednorz and Muller announced their results, researchers
around the world quickly confirmed the discovery. Within a few months, Paul
Chu and his associates at the universities of Texas and Alabama found a
new class of ceramics, made from oxides of yttrium, barium, copper and oxygen
oxides, that became superconducting at an even higher temperature, 93 K.
And that is when the excitement really began. It looked as though materials
that were superconducting at room temperature were just around the corner,
and the door was about to open on a golden era of physics, chemistry and
technology.

The programme for the 1987 meeting of the American Physical Society
in New York City had gone to bed in December before the discovery was widely
known, so it contained nothing about high-temperature superconductivity.
But the organisers of the meeting obligingly arranged for a special evening
session on the topic just in case anyone had anything to contribute. Four
thousand people attended. The session began at seven o’clock in the evening
and finally broke up at six o’clock the following morning. The front page
story of The New York Times called it the ‘Woodstock for physicists’. The
press heralded the high-temperature superconductors as the greatest discovery
since the invention of the transistor. Pundits postulated that the materials
would have far reaching applications in power transmission, transport, energy
storage and electronics.

The economics of high-temperature superconductivity was set to change
society. Why? Because although the lower temperature superconductors were
already being used in specialised areas of science, they required liquid
helium to keep them cool enough to remain superconducting. Cooling helium
gas to below its boiling point of 4 K was expensive. The new superconducting
oxides required only liquid nitrogen, which boils at 77 K, to keep them
working. It is much cheaper. To see the comparison: liquid helium, delivered
with the morning milk to our door at Caltech, costs $5 a litre, which is
about the cost of a litre of cheap vodka. Liquid nitrogen, on the other
hand, costs 12 cents per litre, which is less than we pay for bottled drinking
water.

Researchers also hoped that they would soon discover materials that
were superconducting at room temperature. These would change the way we
use energy and also speed up communications. A new technology would emerge
that would alter many aspects of our lives.

Virtually every major university, electronic and chemical company started
research programmes to examine the new compounds. ÐÓ°ÉÔ­´´s became so excited
that when they thought they had found a new superconducting material they
preferred The New York Times as the primary journal for superconductivity.
Most technologically advanced countries started national initiatives. No
country wished to be left behind in the race to exploit these new wonder
materials.

In August 1987, for example, the US government organised a conference
where businessmen and scientists came together to explore the commercial
possibilities of superconductivity that would ensure America’s economic
competitiveness. Foreigners were not permitted to attend. There was a flood
of extremely expensive seminars for businessmen wanting to keep up to date.
There was also a flood of expensive newsletters (again, not for scientists
but for businessmen) and scientific journals (not for scientists but for
libraries which feel obliged to buy every scientific journal that comes
out). So far as I know, these seminars, newsletters and journals have been
the only successful commercial applications of high-temperature conductivity.

So, what has happened in the past three years? Will these new superconducting
materials fulfil their early promise? Or has the euphoria of an unexpected
and remarkable discovery clouded the judgment of scientists, businessmen
and politicians alike? Well, for one thing, the highest temperature reached
so far is about 125 K, still far below room temperature (300 K). For another
thing, real scientific progress in understanding why these materials behave
as they do has been impeded by the extreme difficulty of making high-quality
samples to work on. Neither of these problems was visible early on. Nevertheless,
the real obstacles to the superconducting pottery millennium could have
been seen from the beginning, had anyone taken the trouble to analyse just
how these new materials were going to be used to change our lives.

Any application of these materials has to make use of three special
properties of superconductors. The first is that the electrical resistance
is zero, and I mean zero, not just very small. Induce a current in superconducting
lead, for example, and the current will flow forever without a noticeable
decrease. The second property depends on the subtle relationship between
superconductivity and magnetic fields. Magnetism destroys superconductivity.
If you apply an outside magnetic field to a superconductor, it protects
itself by expelling the applied field. This is called the Meissner effect.
If you make the external magnetic field stronger, the superconductor finds
it harder to expel the field. Eventually, if the field grows strong enough,
the superconductor is unable to expel it and the material ceases to be superconducting.
In the early 1960s, somebody discovered that certain superconductors could
support extremely large magnetic fields before they were destroyed. If you
have a superconductor that can exist in a high field, it can also be used
to create a high field. You make a superconductor into a wire and coil it,
then you run a very large current through the coil. This creates a very
high magnetic field inside the coil.

The third property of superconductivity is known as ‘tunnelling’. If
you put two pieces of superconductor very close together, then some of the
superconducting current can leak across from one piece to the other. This
is called the ‘Josephson effect’. There is no voltage across this socalled
tunnel, or Josephson junction. Unlike the case of the single piece of superconductor,
however, when the current is flowing through the tunnel junction, the relation
between current and voltage becomes exquisitely sensitive to tiny outside
influences, such as very small magnetic fields and electromagnetic radiation.

The problems of putting superconductors to work

Technologists have come up with a host of exciting applications using
these properties, but they do not always point out the many obstacles. First,
a technology based on superconductivity has its own problems, and secondly,
it is sometimes difficult to see what advantages the new superconductors
have over the old ones.

To see what I mean, let us consider the first property – the conduction
of electricity with no resistance at all. The first application that comes
to mind is electrical power transmission. Suppose we could make a national
power grid out of superconducting material, then there would be no losses
due to electrical resistance. In the power grid system in the US, the line
losses due to electrical resistance amount to 10 per cent of the power generated.
In other words, if you could substitute the present grid with one made entirely
of superconducting material, you would save only 10 per cent of the power
generated. That is not entirely negligible, but the point is, the present
system is already pretty efficient.

A superconducting power system would probably work on direct current
rather than the normal alternating current. This is because superconductors
are truly ‘lossless’ only when the current drifts in one direction, not
when it switches directions. The technology does exist for a direct current
power grid, but alternating current is somewhat more convenient. It is easier
to step up the voltage for long-distance transmission and to step down the
voltage at the other end for safe domestic use.

The next point is refrigeration. The advantage of high-temperature superconductors
is that they require less refrigeration than the old-fashioned kind. In
the 1960s, engineers carried out extensive studies to examine the possibility
of a superconducting power transmission system. This was long before the
discovery of high-temperature superconductors, so the system had to be cooled
with liquid helium. One of the findings of the study was that the cost of
refrigeration was a negligible part of the cost of the system. You do not
have to cool the system down every day; once it is cold, it will stay cold
forever. To compensate for the small amount of heat that leaks into the
well-insulated system, all you do is build refrigerators that tap out a
little of the power that the system is being used to transmit. So, the savings
in refrigeration from using high-temperature superconductors represent only
a fraction of what is already a negligible cost.

On the other hand, part of the reason that this system would have been
extremely expensive to build in the 1960s was the high cost of the superconducting
materials. They are difficult to fabricate into the kinds of wires that
would be needed. Although the new superconducting materials are easy to
stir up in the laboratory, they seem to be even more difficult to make into
wires than the old materials were. So, what seems to have happened is that
we have solved the wrong problem. All in all, the new superconductors do
not seem to be a big step forward for high power transmission.

Another problem is that you have to be able to pass a very large current
through the wire without destroying the superconductivity. In fact, you
should be able to pass through about a million amps per square centimetre
of crosssectional area, which is what the old-fashioned superconductors
can do. Unfortunately, the new materials do not perform so well. They can
carry perhaps a thousandth of the current without losing their superconducting
properties. Finally, superconducting transmission systems are inherently
unstable. Just imagine that because of an accident (being struck by lightning,
for example), some small portion of the superconducting system suddenly
became a conductor. Instead of cooling down and becoming superconducting
again, the onset of electrical resistance would make the material get very
hot like an electrical heater. It would then start to heat up at the next
section which then also becomes a normal conductor, and that starts to get
very hot, heating up the next section and so on. The net result would be
a melt-down of the national power grid – a situation that most engineers
would regard as embarrassing.

The second of the properties of superconductors that we might want to
apply is the ability to make high magnetic fields. The Earth’s field is
about 0.5 of a gauss. The saturation field of an iron-core electromagnet
is 20 000 gauss. Twenty years ago, this was the strongest field that you
could get in the laboratory. Some of the high-field superconductors discovered
in the 1960s can create fields of 100 000 gauss. Today, you can walk into
your friendly neighbourhood superconducting shop and buy one off the shelf.

The new superconductors are also of the high-field type. Furthermore,
because the field that a superconducting electromagnet can generate is directly
related to temperature, the new materials should theoretically be able to
generate even larger magnetic fields. Nobody knows yet how large a field
is possible with the new superconductors, but it could well be as big as
half a million or even a million gauss.

What would we do with these high fields? My favourite idea, by far,
is the magnetically levitated train. The track on which the train rides
is just ordinary material. But inside each carriage, there are powerful
superconducting magnets. These magnets are turned on, one after the other.
These fool the track into thinking that the magnets are running backwards
along the track. The track does not like that, so it resists the motion
of the magnets, and that gives the train the impulse to move forward. The
track also repels the moving magnets, so the whole train lifts off the track
by a few centimetres and rockets forwards at a speed of 500 kilometres per
hour.

There is no technical problem with the concept. The Japanese have had
testbed track running over a few kilometres for many years. But there are
a number of problems with building it, principally the expense. The track
must be very straight and very well maintained. When you are travelling
at 50 kilometres per hour, a few centimetres off the track, you do not want
any bumps or any hairpin curves. The capital investment for this would be
enormous. In any case, we already have other ways of getting from one place
to another quickly, such as aeroplanes. There would appear to be little
economic advantage to a high-speed magnetically levitated train. There is
also some concern about having magnets of 100 000 gauss on a public conveyance.
Although there is not a shred of evidence that high magnetic fields are
dangerous to health, there is also no evidence that they are not.

Another way of using the property of generating high magnetic fields
is to store energy. The American Defense Nuclear Agency is already planning
a project on superconducting magnetic energy storage, SMES. This project
envisages being able to release 0.4 to 1 gigawatts (a billion watts of energy)
for a period of 100 seconds, or between 10 and 25 megawatts for a period
of two to three hours. You do not have to be a KGB agent to work out what
this is for; it is for star wars. Another project is to even out the peaks
and valleys in the demand on the power grid for electrical energy during
the day and night. During the night, when people are not using so much energy
you store it up, then during the day, you can release it when people most
need it. An elementary calculation tells us that if you were to build an
energy storage device out of old-fashioned superconductors and store energy
using a magnetic field of 100 000 gauss, then you need a volume of high
field of 2500 cubic metres. This would be a substantial engineering project.
The idea is to make a superconducting coil in the shape of a doughnut, 1
kilometre in diameter, and bury it underground.

The plan is to build it using conventional superconductors. Would there
be any advantage in using the new high-temperature superconductors? There
is always a small saving to be had in the cost of refrigeration but that
is not the problem. Could we take advantage of the fact that we can make
much stronger magnetic fields? It turns out that the energy stored in given
volume is proportional to the square of the magnetic field. So if you can
make a field 10 times as strong, you can store 100 times as much energy
in the same volume. Or, conversely, you can store the same amount of energy
in 1 per cent of the volume.

There is obviously a great advantage, but there is also a serious problem.
If you have a coil carrying a current that creates a magnetic field, the
field applies a force on the coil, as if the coil contained a very high
pressure. That magnetic pressure, like the energy stored per unit volume,
is proportional to the square of the field. So the magnetic pressure would
also go up by a factor of 100. At 100 000 gauss, the magnetic pressure in
the system is about 400 atmospheres. Engineers can handle such pressure.
It is roughly the pressure you would expect to find in a supercritical steam
boiler. If we go up by a factor of 10 in pressure, we reach the kind of
pressure in the barrel of a big artillery gun when it fires. That kind of
pressure is contained for very short times by using special prestressed
steel that is always under compression on the inside but not on the outside.
The explosion just evens out the stresses, and the gun barrel does not blow
up. If you go to 100 times this pressure, then you are talking about a pressure
that has never been contained before. If you ask engineers whether it can
be done, they are reluctant to say that it is absolutely impossible, but
they will allow that it had never been done, and certainly never using pottery.

Another problem associated with magnetic fields is a phenomenon that
happens if you have ever tried to open the switch on a circuit containing
a big electromagnet. The switch sparks over because the magnet does not
like to be turned off quickly. If you do that with a very big magnetic field,
the spark can be so violent that it vaporises the switch itself. The SMES
project would produce the biggest fields ever created, so there just might
be some inconvenient switching problems.

The third property of high-temperature superconductors – tunnelling
– may in the end turn out to be most practical. For instance, orbiting satellites
rely on old-fashioned superconducting tunnel-junction detectors to detect
infrared radiation, either in devices that look up the sky for astronomical
reasons or downwards towards the Earth for purposes that are usually military
secrets. The fact that the conducting state of one of these junctions is
sensitive to small magnetic fields means that you can use small magnetic
fields to turn it on or off. A number of big electronics companies, IBM,
AT&T and Sperry, were working on these projects to build a superconducting
computer, but these projects were all abandoned in 1983. The reason was
not that superconducting research was not advancing; it was. The reason
was that devices based on semiconductors were advancing much faster. The
problem with superconductivity was not refrigeration; it was material fabrication.
And as I mentioned earlier, the fabrication problem with these new superconductors
is even worse than with the old ones.

In the long run, however, the superconducting computer will win. The
reason is simple. The whole game in computers is to make them fast, which
means they have to be small because it takes time for the signal to travel
any distance. So this means packing the elements of the computer as tightly
together as you can. When you pack a lot of semiconducting elements together
they generate a lot of heat. This provides an ultimate limitation on how
small and, therefore, how fast a computer can be. That limitation does not
apply to superconducting circuit elements because they do not dissipate
heat in the same way as semiconductor circuit elements do.

Even if superconductivity of some kind is sure to have future applications,
it is too early to tell whether the new ceramic materials will ever be anything
more than curiosities. Nevertheless, the discovery of high-temperature superconductivity
has had a number of significant effects. For one, the frenzy that occurred
after the initial discoveries established the unfortunate precedent of announcing
real or imagined scientific advances in press conferences, thus probably
paving the way for the cold fusion affair. More lastingly and more significantly
(one hopes), it reminded us that nature has real surprises in store for
us, even in fields that have been around for a very long time. It is probably
not lost on the many workers quietly toiling in nearly forgotten trenches
of science that Bednorz and Muller may have been the first ever to satisfy
the actual conditions of Alfred Nobel’s will: they were awarded the 1987
Nobel Prize for work done within the past year.

David Goodstein is vice provost and professor of physics and applied
physics at California Institute of Technology, Pasadena, California. This
article is based on excerpts from a lecture published in Engineering and
Science, which is produced by the California Institute of Technology.