NHMFL > ASC > PJL > Preprint Archive
A Historical Introduction to Superconductivity and Magnetism
The
Early Years of Superconductivity
In the early years
of superconductivity, progress to application was slow an intermittent.
On the 10th of July 1908, Heike
Kamerlingh Onnes, Professor of Experimental Physics at the University
of Leiden (Holland), was able to liquefy helium for the first time.
Not only was he able to determine a boiling point for helium, at 4.3 K,
but he was also able to further reduce the temperature to 1.7 K by reducing
the pressure of the helium bath. He soon set about measuring the electrical
resistance of metals in the new temperature regime. The resistance of
metals is strongly dependent on temperature and once the dependency has
been accurately measured, the resistance can be used a simple and convenient
tool for low temperature thermometry. The low temperature behavior of
metals was also seen as a tool to study electron theory (Einstein had
applied to be assistant of Onnes in 1901 but had been rejected). The Leiden
group initially extended measurements on platinum wires, into the liquid
helium range, and observed that their electrical resistance fell continuously
with temperature to a minimum but finite value. The minimum resistance
decreased as the impurity level of the metal decreased. Looking for a
material with a higher available purity, they showed that the resistance
of high purity gold fell to an even lower but measurable value. Seeking
an even higher purity metal, considerable effort was then expended distilling
pure mercury. When the resulting high purity mercury was tested, the electrical
resistance fell steeply but continuously, as expected. At the boiling
point of the helium (4.2 K) the resistance of the mercury wire had fallen
to 500 times less than it had been at the melting point of the mercury.
What happened next came as a complete surprise. As the mercury wires were
slowly cooled below 4.2 K, Gilles
Holst (who had been an assistant at Leiden for two years and would
later become the first director of the Philips
Research Laboratories) measured a sudden and massive drop in the electrical
resistance. As best as they could measure, in just a few hundredths of
a degree the resistance dropped to less than one millionth of the melting
point value, and eventually to a thousand millionth of it. In 1912 Onnes
termed the new electrical state that the mercury had entered below 4 K,
the superconductive state. Having worked so hard to purify the mercury
he was further surprised to find that adding gold and cadmium to the mercury
did not stop it from entering the superconducting state. He also observed
that very high currents could be passed though the mercury until a threshold
current density was reached (as high as 1000 A/mm2 at 2.45
K) at which point the mercury would return to the normal electrical state
(ref. 1). This threshold value, that we now term the critical current,
is perhaps the most important for practical application. Also significantly,
in December 1912, he discovered that others metals, indeed metals that
could be reasonably made into wires at room temperature, namely tin (3.8
K) and lead (6 K, later raised to 7.2 K) could be made superconducting
(2). The first two superconducting solenoids were quickly manufactured
by G.
J. Flim (1875-190), the Chief of the Technical Department of the (Leiden)
Cryogenic Department. At the Third International Congress of Refrigeration,
held in Chicago in September 1913, Onnes predicted that superconductivity
would enable the production of coils that could generate fields (100,000
gauss or 10 T) well in excess of that possible by conventional conductors
(3). For comparison, the flux density between the poles of a "horseshoe"
permanent magnet is 0.1 T. Less accurately, he predicted that such a development
should not be far away. He was about to discover a roadblock to high field
superconductivity. In a footnote to his Chicago address, he observed that
a 0.05 T (500 gauss) field, developed in a simple superconducting solenoid,
was sufficient to revert the superconductor to its normal state. By 1914
Leiden had produced curves of resistance as a function of applied field
and had developed an empirical fit to the temperature dependence of the
critical field (Hc): Hc(T) » Hc0(1-(T/Tc)²).
It seems surprising to us now that it was not until 1916, that the interdependence
of critical current and critical field, was shown from an analysis of
the Leiden data by F. Silsbee of the National Bureau of Standards in America
(4). For many years, it appeared that low critical current and the suppression
in critical current very with small applied fields would make superconductors
impractical for any application other than laboratory studies of solid
sate physics.
Higher critical magnetic
fields and critical temperatures
Leiden enjoyed a monopoly
on liquid helium research until after the first world war. In 1923 a helium
liquefier, based on the Leiden design, started operation at the University
of Toronto. Four years later a helium liquefier capable of 10 liters per
hour was started at the Physikalisch-Technische Reichsanstalt (PTR) near
Berlin under the direction of Walther Meissner. In successive year, from
1928-1930 the PTR identified three important new superconductors; Ta (Tc of 4.4 K), Thorium (Tc of 1.4 K) and Niobium (Tc of 9.2 K)(5). An alloy of niobium, Nb-47wt.%Ti, is now by far the most
important commercial superconductor with it's widespread use in the magnets
for magnetic resonance imaging (MRI) systems in hospitals as well many
other applications (see section: Ductile Superconductors). Nb based technology
is also the current standard for digital superconducting circuits (see
Section: Superconducting Electronics). Meissner's group would go on to
find that most of the transition elements in group IV and V were superconducting.
In figure 1 we show a listing of the elemental superconductors with their
locations in the periodic table. In the same period a further important
discovery came from the group of Wander Johannes de Haas, who became co-director
of the Leiden laboratory 1924 (with Willem Hendrik Keesom). It was found
that a solid solution of 4 % bismuth in gold was found to be superconducting
at 1.9 K (6) despite neither of the components being superconducting at
ambient pressure. A similar result was found later the same year when
copper sulfide (Tc of 1.1 K) was examined by Meissner(7).
This time an insulator (sulfur) had been combined with a very good normal
conductor (Cu), to produce a superconductor. The Meissner group went on
to find a large number of carbides and nitrides with high transition temperatures,
in particular NbC (Tc >10 K).
In
1933 an important discovery was made by Meissner and his student Robert
Ochsenfeld, using cylinders of single crystal and polycrystalline lead.
They showed that in cooling a superconductor below Tc in the presence of an applied field (lower that Hc),
the existing flux inside the superconductor is suddenly expelled. This
behavior is now known as the Meissner effect(9). Furthermore, when the
external field is removed there is no trapped flux or induced dipole.
If field is applied when the superconductor is is in the superconducting
state, currents must be produced near the surface in order to maintain
constant flux. The currents near the surface, counteracting the external
field, must then be a stable. Just two years later, brothers Fritz and
Heinz London developed a set of electrodynamic equations, now known as
the London equations, which described the Meissner effect by supplementing
Maxwell's equations.(10) A consequence of these equations in the London
penetration depth, lL,
which is the maximum depth that magnetic field can penetrate into a type
I superconductor. See article: High-Tc Superconductors, Physical
Structures, and Role Of Constituents, Section: Common Features Of High-Tc Superconductors And Magnetic Noise, And Barkhausen Effect, Section: Flux
Pinning And Losses In Superconductors.
The same year as the
London's paper was published, the group of L. V. Shubnikov at Kharkov,
showed that single crystals of PbTl2 had two distinct critical
fields (11). Up to a lower critical field (Hc1), the
flux is excluded, above that field the flux begins to penetrate and increases
in its penetration until an upper critical field (Hc2)
is reached, when the flux completely penetrates and superconductivity
is extinguished. The superconductors that show this characteristic would
come to be know as Type II superconductors. This class of superconductor
includes all the technically useful superconductors including all alloy
and compound superconductors as well as the elements niobium, vanadium
and technetium. See article: Superconductors, Type I And II and article:
Superconducting Critical Current Unfortunately the importance of the work
at Kharkov was not fully appreciated in the outside the Soviet Union as
Shubnikov's group was victimized in one of Stalin's purges (with Shubnikov
eventually dying in prison in 1945). Only with his posthumous exoneration
was it possible for his Soviet colleagues to openly acknowledge his contributions
to their work.
High-T C Superconductors, Physical Structures, And Role
Of Constituents section: Common Features Of High-Tc Superconductors.
An important new material parameter, the coherence length, x,
was defined as the distance over which the density of the superconducting
electrons decreases at a superconducting-normal interface. The Ginzburg-Landau
formulations were able to predict the conditions under which Type I and
Type II behavior would occur using the parameter k = l/x. From this
work came the now familiar flux line lattice description of Type II superconductors
by Landau's student Aleksei Abrikosov eventually published in 1957(14).
See Superconducting Critical Current. In type II superconductors, the
flux tube has a radius l with
an internal normal core of radius x.
In 1953 Bern Matthias
at the Bell laboratories raised the Tc ceiling or superconductors
to 17.86 K with NbN-NbC. That discovery was followed the same year by
John Hulm's group at the University of Chicago with another 17 K superconductor
V>3Si but of a new crystal structure, A15, that would eventually
supply a series of important superconductors. Another A15, Nb3Sn,
would be added the following year at Bell Labs, with a further increase
in Tc at 18 K.
The Beginning of
Engineering Superconductivity
It was not until 1954
that the first successful superconducting magnet was made (by G. B. Yntema
at the University of Illinois), thereby ushering in the age of engineering
superconductivity. Yntema used Nb wire, which had been shown (by D. Shoenberg
at Cambridge University) to have a markedly better critical field than
any of the other known superconductors. The resulting magnet produced
a field of 0.71 T at 4.2 K. He also discovered that increasing cold work
in the strands markedly increased the current density that they could
carry. It was beginning to be clear that critical current was, to a major
extent, a property that could be increased independently of the intrinsic
bulk properties of Hc2 and Tc. By
August 1960 Stan Autler (at MIT Lincoln Laboratory) had produced a 2.5
T field at 4.2 K. Even more significantly, had applied the persistent
current in a solenoid to provide the magnetic field for a solid state
maser, perhaps the first application of superconductivity. A flurry of
activity followed focused on the high Tc, high Hc2 A15 compound Nb3Sn. Nb3Sn, at that time however,
was difficult to fabricate into magnets because of the brittle nature
of Nb3Sn and because the simple elemental powder in tube process
required a 1000 °C heat treatment. It was soon usurped by two ductile
alloy superconductors, first Nb-Zr (Tc ~ 12 K) and then
Nb-Ti (Tc = 7-10 K). Whereas the Ginzburg-Landau
theory coupled with Abrikosov's work provided and enduring phenomenological
description of superconductivity, it did not provide microscopic description.
That was to be supplied in 1957 when John Bardeen, Leon Cooper, and John
Schrieffer of the University of Illinois at Urbana, published their Nobel
prize winning theory of superconductivity (15). The superconducting electrons
of the phenomenological description proved to be two electrons (Cooper
pairs (16)) with opposite directions for both spin and momentum. The coherence
length was the size of the Cooper pair and the order parameter was proportional
to the electron energy gap, which itself was proportional to the Tc.
The BCS (Bardeen Cooper Schreiffer) theory as it has come to be know,
provided a theory adequate for low temperature and field, three years
later, however, Lev Gorkov would provide one that would be useful at high
fields (17). The key to Gorkov description was that implied a variation
in the energy gap parameter with position.The next major theoretical
advance came in 1962, from a graduate student at Cambridge University,
Brian D. Josephson. He predicted that superconducting current would tunnel
through a thin insulating layer or weak link separating two superconducting
electrodes and that a phase difference is produced between the superconducting
electrons in the two electrodes. The phase difference generates a voltage
difference between the two electrodes. The Josephson effect, as it is
now known, is the basis for superconducting electronic devices such as
the SQUID, and very high precision voltage stands, it would also earn
Josephson a Nobel prize.
Big Magnets
Much of the theoretical
development in the 1960s and 1970s was in the area of flux-pinning. Increasing
the critical current density in superconductors reduces costs because
less superconductor is required, it also makes it possible to operate
magnets at higher magnetic fields. When electricity flows through a superconductor
it produces a Lorentz force between the current and the flux lines. If
the Lorentz force is allowed to move the flux line lattice freely within
the superconductor then power is dissipated eventually the resulting heating
drives the superconductor normal. Introducing microstructural features,
such as non-superconducting precipitates and grain boundaries that pin
the flux lines in place can, however, restrict movement of the flux line
lattice. Understanding the nature of flux pinning is key to understanding
the way critical current density can be improved in superconductors. The
mechanisms and theory of flux-pinning is comprehensively reviewed in article:
Superconductivity and Magnetism.From the late 1960s,
onwards the needs of the high energy physics community propelled considerable
advances in superconducting strand technology, initially for bubble-chamber
and then accelerator magnets. In March 1983, the first superconducting
accelerator ring was completed at Fermi National Accelerator Laboratory.
With 774 6 m long dipole magnets and 210 quadrupole magnets covering a
four mile circle it exemplifies the progress that had been made. The key
features were now in place, the superconducting strand was now in the
required form of a composite of fine (>30 µm in diameter) filaments
in a high purity, high normal conductivity Cu (or Al) matrix for stability
(see Superconductors, Cryogenic Stabilization) and the strand was twisted
in order to reduce eddy currents. The strand itself was cabled with other
superconducting strands to form a thick ribbon-like conductor. Increased
understanding of the microstructural development of the superconductor
and its role in flux-pinning would further increase the critical current
density making possible the next generation of accelerators, see Article:
Ductile Superconductors. Four years later the largest superconducting
magnet yet was fabricated for the DELPHI project at the CERN particle
accelerator laboratory. The 7.4 m long, 6.2 m diameter, 84 tonnes magnet
survived a 1600 km trip to CERN by road, ship, and barge. In addition
to magnet technology the high energy physics community also benefited
from superconducting cavity technology (see Superconducting Cavity Resonators).
When the Large Electron Positron (LEP) collider was initially run in 1989,
with 128 conventional copper accelerating cavities, they provided enough
energy to take the energy of each beam to 50 GeV. By upgrading the ring
with 272 superconducting cavities from the LEP ring was eventual able
to reach 104 GeV per beam in April 2000.In 1986, Alex Müller
and Georg Bednorz, at the IBM Research Laboratory in Rüschlikon, Switzerland,
made a ceramic perovskite of lanthanum, barium, copper, and oxygen that
superconducted at 35 K (18). In fact, small amounts of this material were
later found to be superconducting at 58 K due to lead impurities. The
impact of this discovery can be gauged by the almost immediate awarding
of the Nobel Prize to the two discoverers. The following year the research
groups of Paul Chu at the University of Houston and Maw-Kuen Wu at the
University of Alabama at Huntsville substituted yttrium for lanthanum
and produced a ceramic that superconducts at 92 K (19). Now in the space of a year the highest Tc had been raised from 23.2 K (for Nb3Ge discovered in 1973 by
John Gavaler) through 35 K to 92 K, well above the temperature of liquid
Nitrogen (77 K). A further huge jump in Tc came in 1988
from Allen Hermann and Z. Z. Sheng of the University of Arkansas with
a 120 K Tl-Ca-Ba-Cu-O superconductor (20). In 1993 A. Schilling, M. Cantoni,
J. D. Guo, and H. R. Ott from Zurich, Switzerland, measured a Tc of 133 K in HgBa2Ca2Cu3O8 (21). The partial substitution
of thallium to this high Tc mercury based oxide by P.
Dai, B. C. Chakoumakos, (O.R.N.L.) G. F. Sun, K. W. Wong, (Univ. of Kansas)
Y. Xin, and D. F Lu (Midwest Superconductivity Inc.) increased the Tc to 138 K for a nominal composition of of Hg0.8Tl0.2Ba2Ca2Cu3O8+d (22).
We
have come to expect continuous advances in the properties of both HTS
and LTS superconductors and have yet to be disappointed. In the years
since the discovery of the HTS superconductors we have seen steady improvements
in their current densities, on both the laboratory and industrial scale,
as well as in the production piece lengths. In September 2000, Jochen
Mannhart research group at the University of Augsburg, Germany showed
that preferentially overdoping the grain boundaries of YBa2Cu3O7-d with Ca, relative to the grains yields
values of critical current at 77 K that far exceeded previously published
values. Their method of grain-boundary doping now represents a new and
viable approach for producing practical, cost-effective superconducting
power cable operating at liquid-nitrogen temperature (23).The
advent of high temperature superconductors (HTS) accelerated the application
of low temperature superconductors (now known termed LTS) rather than
replaced them. Not only was there a new market for LTS based test facilities
to measure the properties of new HTS superconductors but there was increased
investment in refrigeration technology and renewed public interest. Cryocoolers
now allow for the cryogen-free use of LTS bases magnets, increasing public
acceptance of the technology for everyday use. HTS in the form of current
leads (there are now several commercial vendors) make large scale LTS
applications much more energy efficient. HTS also make new applications
possible, such as microwave filters, and make old ideas, such as power
transmission lines, more viable.
Recommended Reading
In 1986 the Applied
Superconductivity Conference celebrated the 75th Anniversary
of the discovery of superconductivity with a symposium on the history
of superconductivity. The symposium is published in full in IEEE Trans.
Magn., 23, pp. 354-415, 1986.An important text
that is both informing and entertaining is that of Per Fridtjof Dahl:Per Fridtjof Dahl, Superconductivity:
Its historical roots and development from mercury to the ceramic oxides.Recently V. L. Ginzberg
has written a review:V L Ginzburg, "Superconductivity:
the day before yesterday - yesterday - today - tomorrow ," Physics
- Uspekhi 43 (6) 573 ± 583 (2000)http://ufn.ioc.ac.ru/ufn2000/ufn00_6/ufn006b.pdfThere is now a $7
charge for this document although the Russian language version is still
available for free at:http://data.ufn.ru//ufn2000/ufn00_6/Russian/r006b.pdfA brief introduction
to the BCS theory of superconductivity and the development of the history
behind it by David Pines,
John Bardeen's first postdoctoral researcher (research assistant professor)
at the University of Illinois during the period, 1952-1955, can be found
at: http://cnls.lanl.gov/Highlights/1997-06/html/node2.htmlThe internet continues
to grow as a remarkable repository of knowledge, often a quick internet
search will yield fascinating material. For instance Ted Geballe and John
Hulm wrote a biographical memoirs of Bernd Matthias that can be found
at:http://books.nap.edu/html/biomems/bmatthias.html:
And may I suggest:"Engineering
Superconductivity," ed. Peter J. Lee, Wiley-Interscience, New
York, 2001 
Contents
of and Contributors to "Engineering Superconductivity"
Classical and still
excellent introductions to the phenomenon of superconductivity are:A. C. Rose-Innes,
F. H. Rhoderick, Introduction to Superconductivity, Oxford, UK: Pergamon,
1969.M. Tinkham, Introduction
to Superconductivity, New York: McGraw-Hill, 1975.T. P. Orlando, K.
A. Delin, Foundations of Applied Superconductivity, Reading, MA: Addison-Wesley,
1991.A. M. Campbell, J.
E. Evetts, Adv. Phys., 21: 199-428, 1972.
Bibliography
1.
H. Kamerlingh
Onnes, Further Experiments with liquid helium. H. On the electrical resistance
of pure metals etc. VII The potential difference necessary for the electric
current through mercury below 4.19 K (continuation), Comm. Physical
Lab. Leiden, 133b, 29, 1913. [Leiden references use systematic naming from Per Fridtjof Dahl - see recommended reading]
2.
H. Kamerlingh
Onnes, Further experiments with liquid helium. H. On the electrical resistance
of pure metals etc. (continued). VIII. The sudden disappearance of the
ordinary resistance of tin, and the super-conductive state of lead, Comm.
Physical Lab. Leiden, 133d, 51, 1913.
3. H. Kamerlingh
Onnes, Report on the researches made in the Leiden cryogenics laboratory
between the second and third international congress of refrigeration:
Superconductivity, Comm. Physical Lab. Leiden Suppl., 34b:
55-70, 1913.
4. F. B.
Silsbee, A note on electrical conduction in metals at low temperatures,
Washington Academy of Sciences, Journal, 6:597-602, 1916.
5. W. Meissner
and H. Franz, Messungen mit Hilfe von flüssigen Helium. VIII. Supraleitfähigkeit
von Niobium, Physikalisch-Technische Reichsanstalt, Mitteilung:
558-559, 1930.
6. W. J.
De Haas, E. van Aubel, and J. Voogd, A superconductor consisting of two
non-superconductors, Akademie der Wetenschappen, Amsterdam, Proceedings, 32: 730, 1929.
7. W. Meissner,
Messungen mit Hilfe vo flüssigem Helium. V. Suprleitfähigkeit von Kupfersulfid,
Physikalisch-Technische Reichsanstalt, Mitteilung, 571, 1929
8. W. J.
de Haas and J. Voogd, The influence of magnetic fields on supracondcutors,
Akademie der Wetenschappen, Amsterdam, Proceedings, 33:
262-270, 1930.
9. W. Meissner
and R. Oschenfeld, Ein neuer Effect bei Eintritt der Supraleitfähigkeit, Naturwiss., 21: 787-788, 1933.
10. F. London,
H. London, The electromagnetic equations of the supraconductor, Proc.
R. Soc. London, Ser, A., 149: 71-88, 1935.
11. J. N.
Rjabinin and L. V. Schubnikov, Magnetic properties and critical currents
of supercondcuting alloys, Physikalische Zeitschrift der Sowjetunion, 6: 605-607, 1935.
12. G. Aschermann,
E. Freiderich, E. Justi and J. Kramer, Supraleitfähige Verbindenungen
mit extrem hohen Sprungtemperaturen (NbH und NbN), Physik. Zeit., 42: 349-60, 1941.
13. V. L.
Ginzburg and L. D. Landau, On the theory of superconductivity, Zhurnal
Eksperimental'noi I Teoreticheskoi Fiziki, 20: 1064-1082, 1950.
14. A. A.
Abrikosov, On the magnetic properties of superconductors of the second
group, Sov. Phys. JETP, 5: 1174-1182, 1957.
15. J. Bardeen,
L. N. Cooper and J. R. Schreiffer, "Theory of superconductivity, Phys. Rev., 108: 1175-1204, 1957.
16. L. N.
Cooper, bound electron pairs in a degenerate Fermi gas, Phys. Rev., 104: 1189-1190, 1956.
17. L. P.
Gorkov, Theory of superconducting alloys in a strong magnetic field near
the critical temperature, Soviet Physics JETP, 10: 998-1004, 1960.
18. G. Bednorz,
K. A. Müller, Possible high Tc superconductivity in
the Ba-La-Cu system, Z. Phys. B, 64: 189-197, 1986.
19. M. K.
Wu, J. R. Ashburn, C. J. Torng, P. H. Hor, R. L. Meng, L. Gao, Z. J. Huang,
Y. Q. Wang, and C. W. Chu, Superconductivity at 93 K in a new mixed-phase
Y-Ba-Cu-O compound system at ambient pressure, Phys. Rev. Lett., 58: 908-910, 1987.
20. Z. Z.
Sheng and A. M. Hermann, 90 K Tl-Ba-Cu-O and 120 K Tl-Ca-Ba-Cu-O bulk
superconductors, Proc. 1988 World Congress on Superconductivity.
World Scientific, Singapore: p.365-76, 1988.
21. M.
Cantoni, A. Schilling, H. U. Nissen, and H. R. Ott, Characterisation of
superconducting Hg-Ba-Ca-Cu-oxides. Structural and physical aspects, Physica-C, 215 (1-2):11-18, 1993.
22. P. Dai,
B. C. Chakoumakos, G. F. Sun, K. W. Wong, Y. Xin, D. F. Lu, Synthesis
and neutron powder diffraction study of the superconductor HgBa2Ca2Cu3O8+d by Tl substitution, Physica-C., 243 (3-4):201-6, 1995.
23. G. Hammerl,
A. Schmehl, R. R. Schulz, B. Goetz, H. Bielefeldt, C. W. Schneider, H.
Hilgenkamp and J. Mannhart, Enhanced Supercurrent density in polycrystalline
YBa2Cu3O7-d at 77 K from calcium doping of grain
boundaries, Nature, 407: 162-164, 2000.
Excerpted from "Engineering
Superconductivity," ed. Peter J. Lee, Wiley-Interscience, New
York, 2001
Contents
of and Contributors to "Engineering Superconductivity"
Viewable
Publications
