Superconductivity: Basics and Applications to Magnets

© Springer International Publishing Switzerland 2015

R.G. SharmaSuperconductivitySpringer Series in Materials Science21410.1007/978-3-319-13713-1_1

1. Introduction

R. G. Sharma¹  

(1)

Cryogenics and Applied Superconductivity Laboratory, Inter-University Accelerator Centre, New Delhi, India

R. G. Sharma

Email: [email protected]

Abstract

The study of matter at very low temperature is fascinating because the phonon activity dies down at very low temperatures and one can look into the electronic behaviour minutely. Cryogenic baths of liquefied gases provide excellent medium to cool down samples. Liquefaction of a gas is a combination of an isothermal compression followed by an adiabatic expansion. Cascade process were adopted in liquefying oxygen by Cailletet and Pictet independently in 1877. The final cooling stage has always been a Joule-Thomsen (J-T) valve. Another important breakthrough came in 1898 when James Dewar succeeded in liquefying hydrogen making a temperature range of 20–14 K accessible. The moment of triumph came in July, 1908 when years of hard work by Kamerlingh Onnes at Leiden ultimately resulted in the liquefaction of helium. A temperature range of 4.2–0.8 K thus became accessible in the laboratory. A cascade process using Lair, LO2, LN2 and LH2 and the J-T expansion valve was employed. Within 3 years of this discovery came the defining moment of the discovery of superconductivity in April, 1911 in pure Hg at just below 4.2 K.

1.1 Why Low Temperature Is So Exciting?

Temperature is one of the most important variable parameter like pressure and magnetic field which can be manipulated to change the phase of the material and thus its mechanical, thermodynamical, chemical, electronic and phonon properties. To carry out studies at low temperatures is particularly interesting. At ambient temperatures lattice vibrations (which are quantized and are called phonons) in any material are dominant and usually mask its fine properties. As the temperature is lowered the amplitude of these vibrations is reduced and at very low temperature (close to absolute zero) die down completely. In other words the dominant phonon contribution almost disappears and one can observe the quantum behaviour of matter. Normal laws valid at ordinary temperatures may not hold any more. Some of the properties in fact violates all our norms about the matter as to how it ought to behave. Superconductivity and Superfluidity are perhaps the most striking phenomena observed at low temperature which display the quantum behaviour at a macroscopic scale. An electric current can flow persistently in a superconductor (without dissipation) without a voltage and similarly a superfluid can flow effortlessly (no viscosity) through fine pores, impervious to normal liquid, without a pressure head.

1.2 How to Conduct Experiment at Low Temperatures?

The best way to carry out studies at low temperature is to have a suitable cryogenic bath which can cover the temperature range of interest. A cryogenic bath is best provided by a liquefied gas. One can pump over the liquid surface to reduce vapour pressure and obtain still lower temperature until it freezes. Thus for example, nitrogen (N2) gas boils at 77 K (Kelvin) and can be pumped down to 65 K, hydrogen (H2) boils at 20 K and be pumped down to 14 K. Helium (He) is the only stable gas which boils at the lowest temperature, that is, 4.2 K. One can pump liquid helium and can achieve a temperature of 0.8 K. There is a rare isotope of helium with a mass number of 3 called ³He. It boils at a temperature of 3.2 K and can be pumped down to 0.3 K. Most interesting thing about these two liquids is that they refuse to freeze even at zero absolute. Both can be solidified only at very high pressure. For their unique and very peculiar properties they are referred to as Quantum Liquids.

To go to temperature lower than 0.3 K, one can use what we call as ³He/⁴He dilution refrigerator (DR) and is based upon the finite solubility of ³He (6.4 %) into ⁴He down to absolute zero. ⁴He at 1 K is used as a pre-cooling agent. We can use a DR to produce a temperature of a few milli Kelvin (mK). To go to still lower temperatures this DR is used as a pre-cooling stage to an adiabatic nuclear demagnetization system and temperature of a few micro Kelvin (μK) is achieved. By using a cascade of demagnetization stages one can go down to nano Kelvin (nK) and pico Kelvin (pK). A world record of ultra low temperature ~100 pK is held by the Low Temperature Laboratory of the Helsinki University of Technology, Finland. Whenever the scientific community entered a new regime of low temperature, some discovery or the other of very fundamental importance took place.

1.3 Gas Liquefaction

Broadly speaking, a gas liquefaction is a combination of two thermodynamic processes, an isothermal compression followed by an adiabatic expansion. In the first step a gas is compressed at high pressure at constant temperature. This is achieved by removing the heat of compression by a suitable cooling mechanism. In the second stage the compressed gas is allowed to expand under adiabatic conditions wherein heat is neither allowed to enter nor escape from the system. The temperature therefore drops. The process goes on till the gas liquefies. This is schematically shown in Fig. 1.1. The expansion could be of either isenthalpic or isentropic type.

Fig. 1.1

Principle of gas liquefaction

1.3.1 Isenthalpic Process

In isenthalpic expansion one uses an expansion valve or a Joule-Thomson valve through which the high pressure gas throttles and gets cooled. There is no change in the enthalpy (zero heat transfer and zero work transfer) in this expansion process. This process is sometimes also referred to as the internal-work method as it does not remove the energy from the gas. It only moves the molecules apart against the inter-atomic forces. This process is irreversible and therefore not an efficient thermodynamical cycle. Further, the isenthalpic expansion does not always lower the temperature. If the temperature of the gas is above the maximum inversion temperature , it will heat up the gas. The temperature of the compressed gas before expansion should therefore be below its inversion temperature. It is for this reason that gases like helium (45 K), hydrogen (205 K) and neon (250 K) cannot be liquefied using a J-T valve expansion, their inversion temperatures being below the ambient temperature (~300 K). These gases have to be pre-cooled to below their inversion temperature before they suffer J-T expansion. A J-T valve is however integral to any liquefier and always forms the last stage of cooling since the liquid formation in the expander cannot be sustained. Table 1.1 lists the maximum inversion temperatures (at P = 0) of a few permanent gases. The gases therefore have to be pre-cooled to below their respective inversion temperature before entering the J-T valve.

Table 1.1

Maximum inversion temperature of some gases

1.3.2 Isentropic Process

In the isentropic process the energy is extracted as external work and always produces cold in contrast to the isenthalpic process. This is also referred to as the external work method. Gas expands in an expander which can be of reciprocating engine type or can be a turbine. The process is reversible and thus thermodynamically more efficient. With the same initial temperature of the gas this process always leads to lower temperature than obtainable with the isenthalpic process. As stated above, the operational problem associated with the expansion of the two phase mixture (liquid and gas) in an expander makes it mandatory to use a J-T expander as the last stage of the cooling cycle.

1.3.3 The Linde-Hampson Process

Oxygen was first time liquefied by Louis Cailletet of France and Raoul Pictet of Switzerland independently within days of each other in the year 1877. Pictet used a cascade process wherein a precooling stage is cooled by another precooling stage. He used liquid SO2 and then dry ice (−80 °C, solid CO2) for precooling. Cailletet used liquid SO2 for precooling oxygen before it throttles through a J-T valve. A good historical account of the liquefaction processes can be found in Cryogenic Engineering [1]. Carl von Linde [2] and Hampson [3] perfected the oxygen liquefaction technology by using more reliable ammonia cycle for precooling compressed oxygen and the counter current heat exchangers before the gas expands through a J-T valve. Linde founded Linde Eismaschinen AG in 1879 and later obtained a German Patent in 1895. Basic principle in this process is that air/oxygen is alternatively compressed, pre-cooled and expanded in a J-T valve This results each time in reducing the temperature till the gas gets liquefied. The pressure used in Linde process is rather high. For example, for air at 300 K the optimized pressure is about 40 MPa (~5,880 psi) but the actual machines use a pressure of about 20 MPa. A typical Linde-Hampson cycle is shown in Fig. 1.2.

Fig. 1.2

Linde-Hampson liquefaction cycle consisting of a compressor, a counter current heat exchanger and a J-T valve at the final stage

1.3.4 The Claude Process

Claude Process [4] is an isentropic process which is adiabatic and thermodynamically reversible and therefore more efficient than the isenthalpic process to produce cold. Another added advantage is the lower operating pressure needed for this cycle ~1.7 MPa (~250 psi). In this process the gas does an external work in an expansion engine. The engine can either be of reciprocating piston type or of a rotating type. As shown in Fig. 1.3 the gas is compressed to the required pressure and passes through the first heat exchanger. A portion of the gas (60–80 %) is then sent to an expander and the rest continues to move along the main stream path. The expanded low pressure cold gas is fed back to the returning gas just after the second heat exchange. The return gas cools down the high pressure incoming gas via the two heat exchangers. Thus the cold high pressure gas proceeds via the third heat exchanger and expands in a J-T expander and gets liquefied. The cold vapours from the liquid reservoir return to the compressor via the heat exchangers giving out cold to the incoming high pressure gas. A J-T expansion valve is still necessary because liquid formation in the cylinder of the expansion engine is not desirable. The stresses caused by the low compressibility of the liquid can damage the cylinder. Rotary turbine expander can, however, tolerate almost 15 wt% liquid without causing damage to the turbine.

Fig. 1.3

The Claude cycle consisting of a compressor, a series of heat exchangers, an expansion engine and a J-T valve in the final stage

Two great inventions which took place in the history of gas liquefaction need to be mentioned here. First in 1892 when James Dewar developed a double walled vacuum-insulated cryogenic-fluid storage vessel (popularly known as Dewar Flask). This made it possible to store, transport or pour cryogenic-fluid from one vessel to another. Experiments with cryogenic-fluids for long duration became possible. Second invention [5, 6] by Dewar was the first time liquefaction of hydrogen in 1898, lowering the temperature range for studies to 20 K and to 14 K under reduced pressure. Dewar used the Linde cycle (that is, high pressure and J-T expansion) with liquid nitrogen as the pre-cooling stage.

A masterpiece treatise on the liquefaction cycles has been written by Randall F. Barron [7]. Readers are advised to consult this book for greater details on gas liquefaction and most of the cryogenics topics.

1.3.5 Liquefaction of Helium (1908)

Heike Kamerlingh Onnes at Leiden Institute, The Netherlands had great fascination for the study of Van der Waals equation of corresponding states down to lowest ever temperatures. Nernest heat theorem and Planck’s zero point energy theory further added to his curiosity to achieve lowest possible temperature. His passion to liquefy helium became all the more stronger. He made use of the Linde technique, that is, pre-cooling compressed helium to the freezing point of hydrogen (14 K) and subjecting it to J-T expansion. He succeeded in liquefying helium on July 10, 1908. This turned out to be a turning point for the entire condensed matter physics community. This opened the flood gate for getting to lower and lower temperatures. Discoveries one after another followed in quick succession. The first one was, of course the discovery of superconductivity in 1911 by Kamerlingh Onnes himself. Kamerlingh Onnes got Noble Prize in 1913 for this work. The Noble Prize citation dated Dec. 10, 1913 states For his investigations on the properties of matter at low temperature which led inter alia, to the production of liquid helium. Figure 1.4 shows the schematic diagram of the apparatus used by him for helium liquefaction [8].

Fig. 1.4

A schematic diagram of the apparatus used by Kamerlingh Onnes for the liquefaction of helium first time (Courtesy Peter Kes, Kamerlingh Onnes Laboratorium, Leiden University)

As seen in Fig. 1.4, compressed helium gas flows through the liquid air cooled charcoal dewar to get rid of moisture, gets cooled through a spiral immersed in pumped hydrogen (15 K) and expands in the inner most dewar through a J-T expansion valve and gets liquefied. Kamerlingh Onnes with his original liquefier is seen in Fig. 1.5. The boiling temperature of helium is 4.2 K at atmospheric pressure. For the next 20 years the Leiden Laboratory remained a most sought after place for research by the condensed matter physics community from Europe and USA and Kamerlingh Onnes enjoyed complete monopoly.

Fig. 1.5

Kamerlingh Onnes in his Low Temperature Laboratory (Courtesy Peter Kes, Kamerlingh Onnes Laboratorium, Leiden University)

John Cunningham Mclinnan built the second helium liquefier at the Toronto University in 1923. The design of this machine was borrowed from Kamerlingh Onnes but looked little more elegant. In reality low temperature research started flourishing around 1934–35. Four German scientists, namely, Franz Simon, Heinrich Kuhn, Nicolas Kurti and Kurt Mendelssohn joined Clarendon Laboratory at Oxford University at the invitation of Lindemann. Low temperature research started at Oxford when Simon built a mini He-liquefier. Around the same time Pjotr Kapitza too built a He-liquefier [9] at the Cambridge University. This machine was based upon a rotating expansion engine or the so called ‘turbine’. He made a similar liquefier at the Institute for Physical Problems, Moscow during 1935. The commercial machine [10] built by Samuel Collins (MIT) and marketed by M/S Arthur D Little of the USA was the beginning of the spread of low temperature studies using liquid helium the world over. Many countries around the globe bought this machine and the low temperature research now flourished all around. My earlier place National Physical Laboratory, Delhi too acquired Collin’s helium liquefier in 1952 and started low temperature under the leadership of David Shoenberg.

1.3.6 Collins Liquefaction Cycle

The liquefaction cycle used by Collins and shown in Fig. 1.6 is an extension of the Claude cycle (Fig. 1.3). The machine uses three stages of cooling for helium liquefaction, two expansion engines of the reciprocating piston type followed by the J-T expansion valve. Pure helium gas is compressed to about 225 psi pressure, precooled to 77 K and passes through the first heat exchanger. Thereafter, part of this cold gas expands in the first expander cooling the gas further to about 60 K. This low pressure gas goes back to the compressor via the second and first heat exchangers cooling in-turn the incoming high pressure gas. Rest of the gas continue to proceed through yet another (third) heat exchanger and again a fraction of the gas expands in the second expander bringing down the temperature of the gas to about 20 K. This temperature is well below the inversion temperature of helium gas. Low pressure gas again returns to the compressor via the series of the heat exchangers transferring its cold to the incoming high pressure gas. The cold gas now at 6 K finally throttles through the J-T valve, liquefies and gets collected in the vessel. The liquid helium can be siphoned out of the container for use. The evaporated gas from this container continues to travel to the compressor via the heat exchangers for a continuous operation.

Fig. 1.6

The flow diagram of the Collins helium liquefaction cycle

Precooling helium gas with liquid nitrogen though not essential, nevertheless increases the yield of liquid helium, by almost a factor of two. Figure 1.7 is the photograph of this first generation Collins liquefier of the ADL make. This machine used to produce 4 l/h liquid with a single compressor. In recent years reciprocating engines have been replaced by the turbo-expanders which rotate at speeds varying between 250,000 and 300,000 rpm. These machines can produce several hundred litres of liquid helium per hour. Figure 1.8 is a photograph of a modern turbo-cooled helium liquefier custom manufactured and installed by Linde Kryotechnik AG with a refrigeration capacity of 900 W (~300 l/h) at our Centre, IUAC in 2012.

Fig. 1.7

First generation commercial Collin’s helium liquefier ADL make with 4 l/h LHe capacity

Fig. 1.8

A modern day turbine based helium liquefier with a refrigeration capacity of 900 W (~300 l/h) manufactured by Linde Kryotechnik AG installed at IUAC in 2012 (Photo courtesy IUAC Delhi)

1.4 Discovery of Superconductivity—A Fall Out of Helium Liquefaction

The liquefaction of helium gas in 1908, and just discussed above, was the culmination of a well designed programme of Kamerlingh Onnes of studying properties of materials at lower and lower temperatures achieved by the successive liquefaction of permanent gases, viz; oxygen, air, hydrogen and finally helium. Till this time, there were only conjectures that the electrical resistance of metals will drop to zero as the temperature approached absolute zero or will show a minimum and rise again and so on. Kamerlingh carried out electrical resistivity measurements on pure platinum and gold and found that the resistivity attains a temperature independent constant value below about 10 K. Purer the material smaller is the value of this residual resistivity. He then took up pure mercury for his studies as it was possible to obtain mercury in ultra-pure form through multiple distillations . What he found was quite startling and unexpected. The resistance in mercury just close to 4.2 K ‘abruptly’ dropped to zero (one thousand-millionth part of the normal temperature value) with no potential difference. He thus proclaimed that mercury just below 4.2 K has entered a new state which he named ‘suprageleider’ and when translated from Dutch to English became ‘superaconductivity’ and finally changed to superconductivity [11]. His original resistance versus temperature plot for mercury is shown in Fig. 1.9. Lead and Tin were next metals from the periodic table to have shown superconductive transition at 7.2 and 3.7 K respectively.

Fig. 1.9

First observation of superconductivity in pure Mercury by Kamerlingh Onnes (Courtesy Peter Kes, Kamerlingh Onnes Laboratorium, Leiden University)

Kamerlingh Onnes also realized soon that this zero potential difference lasts only up to a threshold current in the sample beyond which it rises sharply. Lead, for example, stood superconducting up to a threshold current density of 4.2 A/ mm² only.

He also observed that superconductivity stays in lead up to a threshold magnetic field of 600 Gauss without a sign of magneto-resistance . Beyond this field, the resistance appears and rises fast with magnetic field. Notwithstanding these limitations, Kamerlingh Onnes did realize that superconducting coils can be used to produce fields in excess of 10,000 Gauss without Joule heating. His dream did come true and it is all for us to see the great revolution these materials have brought about. How this phenomenon unfolded, understood, different classes of superconductors discovered and put to use in producing high magnetic fields, in accelerators, in fusion reactors and such other applications (NMR, MRI, SMES, levitation etc.) is discussed in the following chapters.

References

1.

K. Timmerhause, R.P. Reed (eds.), Cryogenic Engineering, Fifty Years of Progress. International Cryogenics Monograph Series, Chapter 1 (Springer, New York, 2007)

2.

C. Linde, G. Claude, Liquid Air, Oxygen, and Nitrogen, trans. by H.E.P. Cotrell (J & A Churchill, London, 1913), p. 75

3.

W. Hampson, and G. Claude, Liquid Air, Oxygen, and Nitrogen, trans. by H.E.P. Cotrell (J & A Churchill, London, 1913), p. 88

4.

G. Claude, C.R. Acad. Sci. Paris. 134, 1568 (1902)

5.

J. Dewar, Preliminary Notes on Liquefaction of Hydrogen and Helium. In: Proceedings of Chemical Society No. 158, 12 May 1898

6.

J. Dewar, Collected Papers of Sir James Dewar, ed. by L. Dewar (Cambridge University Press, Cambridge, 1927), p. 678

7.

R.F. Barron, Cryogenic Systems, Chapter 3, (Oxford University Press, 1985), pp. 60–150

8.

H. Kamerlingh Onnes, Comm. Leiden No. 1206 (1911), Noble Prize Lecture, 11 Dec 1913, http-nobleprize.​org/​noble_​prizes/​physics/​laureate/​1913/​ones-lecture-pdf

9.

P. Kapitza, Russian J. Phys. (English transl.) 1, 7, (1939)

10.

S.C. Collins, Rev. Sci. Instrum. 18, 157 (1947)

11.

H.K. Kamerlingh Onnes, Commun. Phys. Lab. Univ. Leiden, 29, (1911)

© Springer International Publishing Switzerland 2015

R.G. SharmaSuperconductivitySpringer Series in Materials Science21410.1007/978-3-319-13713-1_2

2. The Phenomenon of Superconductivity

R. G. Sharma¹  

(1)

Cryogenics and Applied Superconductivity Laboratory, Inter-University Accelerator Centre, New Delhi, India

R. G. Sharma

Email: [email protected]

Abstract

A superconductor is not only a perfect conductor (ρ = 0) but also a perfect diamagnet (B = 0) below T c. Meissner and Ochsenfeld discovered in 1933 that the magnetic field is expelled out of the body of the superconductor. Field penetrates the material only a small distance, called London’s penetration depth, λ which is of the order of 30–60 nm in metal superconductors. The transition to superconducting phase has been found to be of the second order as confirmed by the absence of a latent heat and by the appearance of a peak in the specific heat at T c. These materials also exhibit flux quantization in so far as the field entering a superconducting ring or a cylinder has to be an integral multiple of a flux quantum Φ0 = h/2π (= 2 × 10–15 T m²). The strong evidence of the role of phonons in the occurrence of superconductivity came from the isotope effect which shows that T c is inversely proportional to the square root of the atomic mass. Pippard introduced the concept of long range coherence among the super electrons and defined a characteristic length, the coherence length ξ over which the order parameter changes in a superconductor. This parameter is of the order of 1,000 nm much larger than the parameter λ for these metal superconductors. Optical experiments strongly hinted at the existence of an energy gap in the energy spectrum of these materials. All these experimental facts led the three physicists, Bardeen, Cooper and Schrieffer, to formulate the first successful microscopic theory, the BCS theory of superconductivity. The chapter ends with a short discussion on dc and ac Josephson effect. The design of SQUID, an ultra low magnetic field/voltage measuring device, based upon the Josephson junction behavior, has also been discussed. A large number of SQUIDs are mounted on a helmet shaped cryostat and used for mapping feeble magnetic field inside the brain. This technique is called magneto-encephalography.

2.1 Electrical Resistance Behaviour at Low Temperature: Electrical Conduction in Metals

Conduction in materials is a wonderful gift of nature. All materials conduct but the conductivity can vary from one extreme to another. Pure metals like silver, copper and gold, for example, are the best conductors of electricity. The electrical conductivity in semiconductors, on the other hand, is several orders of magnitude smaller and other materials do not conduct at all and are perfect insulators. An indication to what extent the conductivity (or the resistivity) varies as one moves away from metals to semiconductors to ionic solids, glasses and finally insulators, can be seen from Fig. 2.1. It can vary by about 25 orders of magnitude. According to Drude hypothesis conduction electrons in a metal wander randomly in the background of positively charged ions rigidly fixed to their (ordered) lattice positions. These ions vibrate at quantized frequencies limited to a maximum frequency called the Debye frequency. In the absence of an electrical potential electrons do get scattered by these ions but randomly, such that they do not drift in a particular direction. So, there is no current flow. In the presence of an electric potential electrons still get scattered by the ions but now they drift in a particular direction which is opposite to that of the applied potential. A net current thus flows through the conductor. When electrons are scattered by ions they lose energy which is absorbed by the lattice in the form of heat, called dissipation. Thus in a sense, electrons face resistance in their free movement when scattered by the lattice ions vibrating at quantized frequencies , called phonons . In what follows, we will refer to this mechanism as electron-phonon interaction. Figure 2.2 shows schematically the electron motion in a lattice under these two different situations. If impurities are present in the lattice, they too will scatter electrons and will add an additional term to the resistivity. What is surprising is that the electrical resistance of most materials is governed by a simple law, the Ohms Law (V = I × R) and strangely enough this relationship is found valid over a large resistance range of the order of 10²⁴.

Fig. 2.1

Electrical resistivity in different class of materials varying as it does from very low value for metal to extremely high values for insulators, semiconductors lie some where in the middle

Fig. 2.2

Free electron motion in a lattice of a metal. a In the absence of an electric field. b In the presence of an electric field

Temperature variation of resistance in metals has been the next most exciting problem researchers aimed at after the liquefaction of helium. Since the dominant component of resistivity comes from the scattering of electrons by phonons (lattice vibrations) and the amplitude of lattice oscillations being strongly temperature dependent, such studies yield information on the nature of electron-phonon interaction. Around ambient temperature resistivity usually show a linear variation with temperature. The behaviour may become quite different at low temperature. In metals well below Debye temperature (T ≪ θ D) the resistance varies as ~T ⁵. What will happen to resistance just a few K above the absolute zero had been a curiosity of the physicists all along. In fact, this curiosity has been a motivation for getting to lower and lower temperatures. Before Kamerlingh Onnes succeeded in liquefying helium there were different conjectures as to what will happen to resistivity at absolute zero. As shown in Fig. 2.3 James Dewar had predicted the resistivity to become zero as the temperature approaches zero K because the phonon scattering too should die down. Kelvin believed that the resistivity should decrease to a minimum and rise again at still lower temperature as the electron motion will freeze. Matthiessen’s prediction that resistance will saturate at a finite value close to absolute zero turned out to be the most accurate one.

Fig. 2.3

Three different predictions about the possible resistivity-temperature behaviour in metals as one approaches zero absolute

According to Matthiessen’s Rule the resistivity at low temperature consists of two dominant components, viz; ρ 0, the residual resistivity which is temperature independent and is caused by the scattering of electrons by impurities and imperfections and another ρ i, the intrinsic resistivity caused by the scattering of electrons by phonons and is strictly temperature dependent.

(2.1)

The intrinsic resistivity ρ i always decreases with the fall of temperature. Dewar, after liquefying hydrogen, measured the resistivity of pure silver and gold down to 16 K but always found it to be saturating to a finite value (Fig. 2.4). He believed that there is always an impurity. This was precisely the reason that Kamerlingh Onnes chose to study resistivity of mercury which can be obtained in ultra high purity form by multiple-distillation process. What he observed in mercury at 4.2 K became a history and is the subject matter of the book.

Fig. 2.4

Electrical resistance behavior at low temperature (close to 0 K) of pure metals and a superconductor. Impurity in a metal raises the residual component of the resistance, ρ 0

2.2 The Phenomenon of Superconductivity

As discussed in Chap. 1 superconductivity was discovered by Kamerlingh Onnes [1] in mercury in 1911 at 4.2 K. The temperature at which superconductivity occurs is called the ‘transition temperature ’ or the ‘critical temperature ’, T c. Kamerlingh Onnes also concluded that the purity of mercury was not of consequence and superconductivity was an intrinsic property of mercury. Soon, he found superconductivity in Sn at 3.7 K and in Pb at 7.2 K. Intense research continued for discovering more and more superconductors across the periodic table. There was no rule to govern which particular element should become superconductor and which should not. Characteristic parameters like melting point and crystal structure did not show a particular trend for the occurrence of superconductivity. The position of the superconducting elements in the periodical table has been marked with their transition temperature (T c), the critical field (B c), the penetration depth (λ) and the coherence length (ξ) values in Table 2.1. The critical field parameter (B c) will be introduced and discussed in next section. As seen from the table, the T c of these elements varies from as low a value as 0.0003 K for Rh to a maximum of 9.3 K for Nb. In addition, there are elements which are superconducting only under high pressure [2]. These are shown as shaded in the Table 2.1. Pressure at which they become superconducting has been indicated in kbar unit. One striking feature of the periodical table, however, is the absence of superconductivity in the best known electrical conductors, namely copper, silver and gold.

Table 2.1

Elements in the periodic Table showing superconductivity, with their T c, B c, λ and ξ values. (Parameter values compiled from my lecture notes, large number of publications and [2]). (Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission)

Shaded elements show superconductivity at high pressure only, T c(K) and pressure values (kbar) are mentioned. Elements with* turn superconducting under high pressure or in thin film form only.

2.3 The Critical Magnetic Field

Soon after the discovery of superconductivity Kamerlingh Onnes realized the importance of these materials for winding magnets to produce high fields without dissipation but to his dismay he found in early 1913 that superconductivity gets destroyed when exposed to small magnetic field. Each superconductor was found to have a characteristic value of this field, called B c, the critical magnetic field. B c is maximum at T = 0 K and continuously decreases with the increase of temperature and becomes zero at T c. It is shown in Fig. 2.5 that the material at point P in the superconducting state can be driven to normal state by either increasing the temperature or the magnetic field and taking it outside the parabolic curve. Unfortunately, the value of B c for most of the elemental superconductors is very low, of the order of few hundred Gauss. B c of Nb metal happens to be maximum ~1980 Gauss. Values of the B c for most superconductors are shown in the periodic Table 2.1. The variation of B c with temperature is parabolic and can be expressed by the expression (2.2).

Fig. 2.5

The parabolic B–T curve represents the boundary between the superconducting state and the normal state

(2.2)

B c is maximum (= B 0) at T = 0 and drops to zero at the T c. The transition to normal state in magnetic field can be very sharp depending upon the purity and perfection of the material. Transition also depends strongly upon the direction of the applied magnetic field. Transition is sharp if the field is parallel to the axis of the cylindrical sample. Transition starts at B c/2 if the field is perpendicular to the axis. Figure 2.6 shows the plots of B c versus temperature for a number of metal superconductors.

Fig. 2.6

The B c versus temperature plots of some of the metallic superconductors. (Data from ‘Superconductivity’ by Shoenberg, 1952, p. 224. (Reproduced with the permission of Cambridge Uni. Press))

Francis Silsbee [3] proposed in 1916 that a superconductor has a critical value of current too which will produce a self field equivalent to B c and destroy superconductivity. We thus have three critical parameters characterizing a superconductor, namely, critical temperature (T c), the critical field (B c) and the critical current (I c). All the three parameters are inter dependent. Thus a superconductor remains superconducting within the confine of these three critical parameters and turns normal, the moment any of critical parameter is exceeded as shown in Fig. 2.7.

Fig. 2.7

Critical surface of a superconductor. Notice that the three critical parameters are inter-dependent

2.4 The Meissner Effect (Field Expulsion)

Low temperature research spread beyond the confines of the Low Temperature Laboratory of Leiden around the world after 1930. Race to discover superconductivity in variety of materials and especially in alloys continued unabated. Some thing very extraordinary happed in 1933 when Walter Meissner and Robert Ochsenfeld [4] at Berlin found out that a magnetic field is not frozen within the body of a superconductor when cooled down to below its critical temperature T c. Instead, the field is expelled from the interior of the superconductor. This observation was quite startling and unexpected. It was expected that a magnetic field will freeze inside a superconductor (being a perfect conductor) until the superconductor is warmed up above T c. We know a perfect conductor cannot sustain an electric voltage, that is E = 0, it therefore follows from Maxwell Equations

(2.3)

(2.4)

This implies that $$ \partial B = \, 0 $$ or the magnetic flux B inside a superconductor should be constant. This means that if a perfect conductor is placed in a magnetic field and then cooled down to T c the magnetic flux remains trapped inside even when the field is removed. Meissner and Ochsenfeld however observed that it does not happen in a superconductor. They found that the flux is expelled from the body of the superconductor the moment it is cooled down to below T c. Irrespective of the fact whether the superconductor is kept in a magnetic field and cooled below T c or it is cooled below T c first and then a field is applied, the magnetic flux does not enter a superconductor. The lines of force now pass around the sample as shown in Fig. 2.8. A superconductor thus behaves like a perfect diamagnet.

Fig. 2.8

Magnetic flux is expelled from the body of a superconductor the moment it enters the superconducting state when cooled to below T c

We can therefore state that a superconductor is a perfect conductor as well as a perfect diamagnet below its transition temperature T c. Field expulsion can be explained in terms of screening current running across the surface so as to produce a magnetic field equal and opposite to the applied field. The consequence of this induced field is that a magnet will levitate over a superconductor provided, of course, the weight of the magnet is less than the force of levitation. Figure 2.9 is a typical picture of a superconductor floating over a permanent magnet. One can also levitate a magnet over a superconductor.

Fig. 2.9

Levitation of a superconductor over a magnet

2.4.1 Perfect Diamagnetism

One immediate consequence of Meissner effect is that superconductivity is a thermodynamical phase in so far as the final state of magnetization does not depend upon the sequence of cooling below T c and applying magnetic field. This behaviour is totally different to that observed in a normal metal. The fact, that the magnetic flux inside a superconductor remains zero (B = 0) when an external field B a is applied, implies that a magnetization is induced in the superconductor which exactly cancels out this magnetic flux. Thus,

(2.5)

(2.6)

A superconductor thus has a magnetic susceptibility of −1 and is a perfect diamagnet and quite distinct from the known diamagnetic materials.

As stated earlier, the screening currents flow along the surface of a superconductor to prevent the entry of the flux. Since the resistance of a superconductor is zero, these currents never decay (supercurrents ) and flow persistently without Joule heating.

2.4.2 The Penetration Depth

The just discussed Meissner effect or the so called perfect diamagnetic property of a superconductor implies that screening currents flow along the external surface of the superconductor. If these currents were to flow only at the surface the current density will be infinite which will be an impossible proposition. Current sheet should therefore extend into the material to a depth on an atomic dimension. In fact, the magnetic field penetrates a superconductor a very small distance, falling exponentially to zero with a characteristic depth, called the ‘penetration depth’ λ. This is shown in Fig. 2.10. This penetration depth in most pure metals turns out to be of the order of 10–100 nm. The flux at a distance of x inside the material is given by the expression:

Fig. 2.10

Field penetration in a superconductor. The magnetic flux drops exponentially inside the material. Penetration depth λ is defined as the depth at which the flux density drops to its eth value

(2.7)

It is rather difficult to measure the penetration depth in bulk material, flux penetration being so small. In specimens of the dimension of λ, the field penetration could be across the material and can therefore be estimated to some accuracy.

The penetration depth is not constant but varies widely with temperature. Well below T c and until 0 K there is hardly any variation but close to T c (>0.8T c) λ rises exponentially to infinity at T c. Figure 2.11 is a typical λ–T curve for a superconductor. Equation (2.8) and (2.9) represent the observed λ–T behaviour.

Fig. 2.11

Penetration depth is nearly constant until about 0.8T c and rises sharply thereafter, reaching infinite at T c

(2.8)

Very close to T c

(2.9)

Table 2.2 gives typical values of λ at 0 K for Pb, In and Al. The exponential variation of λ near T c makes measurements close to T c most difficult. The technique employed to measure field penetration is to have cylindrical superconductor of possibly high purity snugly fitted into a solenoid magnet. The inductance of this system will depend upon the extent of field penetration. If this inductance is connected to a LCR circuit, change in inductance can be evaluated in terms of frequency which can be measured with high accuracy.

Table 2.2

Penetration depth of some metals at 0 K

2.4.3 Magnetization in Superconductors

We have already discussed in Sect. 2.4.1 that superconductivity is an equilibrium thermodynamic state and a magnetization is induced in the superconductor when an external field B a is applied. This behavior too is different from that of the normal metal. From Fig. 2.5 we find that at temperature higher than T c and in field higher than B c, a normal state is more stable. The magnetic flux inside a superconductor remains zero but the flux enters the material rather sharply as soon as the field exceeds B c. The material turns normal. This is shown in Fig. 2.12a. The magnetization versus field behavior is shown in Fig. 2.12b. In the superconducting state magnetization rises with the increase of magnetic field to oppose the field penetration and drops to zero as soon as B c is approached. Interestingly both these processes turn out to be reversible confirming the thermodynamical nature of superconductivity.

Fig. 2.12

Magnetic flux (a) and magnetization (b) behaviour of a superconductor in an applied magnetic field

2.4.4 The Intermediate State

The value of critical field B c usually gets modified because of the concentration of field lines at the surface. This, so called, the demagnetization effect, depends upon the shape of the material and the field orientation. For example a long thin cylindrical superconductor has no demagnetization effect if the magnetic field is parallel to its axis. In perpendicular field, however, the field penetration starts at ½B c, that is, the demagnetization factor is 1/2. For a sphere the demagnetization factor is 1/3. The uniform field B i inside the superconductor in an applied field B a is given by the expression

(2.10)

where n is the demagnetization factor. Since n  = 1/3 for a superconducting sphere, at B a =B c, B i = 2/3B c and the sphere has to turn normal. If, however, it turns normal M ≈ 0 and B i = 2/3B c which is less than the critical field B c and the sphere has to stay superconducting. This simply means that between B c and 2/3B c the sphere can neither be fully normal nor fully superconducting. Instead, the sphere undergoes an intermediate state. Figure 2.13 schematically shows the splitting of the material into laminae of alternate normal and superconducting material parallel to the field in a long thin cylinder when the field exceeds ½B c. As field increases superconducting fraction diminishes and the normal regions grow at the expense of the superconducting regions. At B c, the entire material turns normal. Consequently, there is a continuous increase of resistance. These laminae are parallel to field. Since magnetic induction in the normal regions is parallel to field and goes to zero in superconducting regions the boundary has to be parallel to the field direction only. That the laminae extend right through the sample is confirmed by the observation of a finite resistance in the intermediate state. Finite resistance obviously comes from the current flow through the normal