Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE (www.oxfordhandbooks.com). © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 19 August 2019

Small Superconductors—Introduction

Abstract and Keywords

This article provides an overview of small superconductors, including some of the basic definitions, prominent characteristics, and important effects manifested by such materials. In particular, it discusses size effects, surface effects, electron-mean-free-path effects, phase slips, unusual vortex states, and proximity effects. The article first considers the two characteristic length scales of superconductors, namely the magnetic penetration depth and coherence length, before proceeding with an analysis of two size effects that account for how superconductivity responds when the bulk sample is made smaller and smaller in the nano range: the small size effects and the quantum size effects. It then examines other phenomena associated with small superconductors such as quantum fluctuations, Anderson limit, parity and shell effects, along with the behaviour of nanowires and ultra-thin fims. It also describes some of the experimental techniques commonly used in the synthesis of small superconductors.

Keywords: small superconductors, surface effects, vortex states, proximity effect, superconductivity, small size effects, quantum size effects, Anderson limit, Nanowires, ultra-thin fims

A.V. Narlikar

By small superconductors we mean superconducting objects whose one, two, or all three dimensions are shorter than some small characteristic length. As depicted in Fig. 1.1, this respectively makes a (a) three-dimensional (3D) bulk material (b) quasi-two dimensional (2D), (c) quasi-one dimensional (1D), or (d) quasi-zero dimensional (0D) in the form of a thin film, a narrow wire, and a fine particle. Although studies of small superconductors began in the 1960s, the field received a significant impetus only 15–20 years ago when technological progress made various new experimental tools of fabrication and characterization of nanosized samples more abundantly available. Studies of small superconductors are important for many reasons. One of the prime motivations is to assess if there is any miniaturization limit for superconducting nanodevices. How do various superconducting parameters such as the critical temperature Tc, critical magnetic field Hc, critical current density Jc, etc. respond to sample size reduction and is there any limiting size just below which superconductivity completely disappears or is destabilized? Further, above all, there exists a perpetual curiosity to know if some entirely new properties or phenomena emerge when the sample size is significantly reduced down to the nanoscale. Finally, in a world obsessed with miniaturization of technology, mesoscopic superconductors are acquiring even greater relevance and timeliness for revolutionary innovations in novel devices and applications.

1.1 Two characteristic length scales of superconductors

A superconductor possesses two fundamental length scales, namely the magnetic penetration depth λ and coherence length ξ. The former, according to F. and H. London (1935), arises when a superconductor is subjected to an external (p. 4) magnetic field, smaller than the critical field Hc. Due to one of its fundamental properties, namely, the Meissner effect, when a bulk material turns superconducting it expels the magnetic field from its interior. This, however, leaves a thin surface layer, called the penetration depth (or the London penetration depth) λ over which the magnetic field has exponentially decayed and there exist supercurrents that flow to keep the external magnetic field excluded from the bulk of the sample.

Small Superconductors—IntroductionClick to view larger

Fig. 1.1 (a) Bulk (3D), (b) thin film (2D), (c) fine (nano) wire (1D), (d) fine particle/nanodot (0D).

The second characteristic length ξ, after Pippard (1950) and Ginzburg and Landau (1950), represents the spatial range over which superconductivity gets affected by any local perturbation like thermal fluctuation/heating and magnetic field. For type-I, or positive surface energy superconductors, that mostly include superconducting elements, such as Al, In, Pb, Sn, Zn, etc., ξ > λ (Fig. 1.2a), while for type-II, or negative surface energy superconductors, like alloys and compounds that include Chevrel phases, A-15 superconductors, high-temperature superconducting (HTS) cuprates, and Fe-based superconductors (FBS), ξ < λ (Fig. 1.2b). In terms of microscopic theory, the coherence length measures the separation of the two charge carriers (both negative or positive), which are generally of opposite spins, forming a Cooper pair which makes the pairs function as extended entities.

Table 1.1 lists ξ and λ values along with the critical temperature for some representative superconductors. The materials with one or more dimensions smaller than ξ or λ are considered as small superconductors where the two length scales lie between a few nanometers and several hundred nanometers. Small superconductors thus clearly belong to the nanoscale, but more broadly they are also referred to as mesoscopic. The Greek word meso means “in between” and the smallness lies in between the macro, or the bulk, obeying the classical laws of physics, and the micro, or molecular size, where quantum laws dominate. On practical considerations, the nanometer is taken as the lower limit of the length (p. 5) scale for size reduction which classifies the above low-dimensional systems as taking the form (Fig. 1.1) of nanometric thin films, ultrafine nanowires, and isolated or embedded nanoparticles (or nanodots), respectively. The above length scales, in their lower limit, are comparable to the de Broglie or Fermi wavelength which causes a superconducting nanoparticle to exhibit quantum confinement effects and display unusual superconducting behavior.

Small Superconductors—IntroductionClick to view larger

Fig. 1.2 Two characteristic lengths of superconductors: λ and ξ. (a) Type-I superconductor where λ < ξ and (b) type-II superconductor with λ > ξ.

(From Narlikar, A.V. (2014), Courtesy Oxford University Press, Oxford, UK.).

In general, the sample dimensions smaller than λ give rise to unusual magnetic and electrical transport properties while those smaller than ξ manifest changes in their order parameter and Tc-related behaviors. The relevance of the former had in fact been realized as early as the mid-1930s from the London phenomenology (F. and H. London, 1935), which had showed that for superconducting thin films and fibers of thickness smaller than λ the Meissner effect is incomplete or only partial. As a result, the energy of magnetic field expulsion in them is significantly reduced in comparison to the bulk samples of the same material. Since the condensation energy density (μ0Hc2/2) for both bulk and thin samples is the same, the field expulsion from a smaller volume of thin samples would lead to a higher value of their critical (p. 6) field μ0Hc. This contention was shortly afterwards corroborated by Pontius (1937) who found the critical field of a lead wire to be markedly increased when its wire diameter was reduced. Bean et al. (1962) forced soft type-I superconducting metals like Hg, In, Pb, etc. into the interconnected 3–5-nm- sized pores of a Vycor glass which resulted in networks of very fine filaments (Fig. 1.3). The critical fields of the filamentary structures were found to be around 5 T, i.e. more than 100 times their bulk values. Further, at 4 K, the filaments were found to sustain supercurrent densities of 108A/m2, about 1000 times larger than the bulk values. A similar large increase of the critical field has more recently been reported for Pb and In nanoparticles (Li et al. 2003, 2005), (p. 7) primarily resulting from their incomplete Meissner effect. In type-II superconductors, the Meissner effect is observed only until the lower critical field Hc1 < Hc and here again the results of Blinov et al. (1993) and Fleischer et al. (1996) exhibit an order of magnitude increase in Hc1 of YBCO fine powders of diameter < λ.

Table 1.1 Critical temperature, coherence length and penetration depth of different superconducting materials. (Data compiled from Narlikar (2014) and Narlikar and Ekbote (1983).

Superconductor

Critical Temperature (K)

Coherence Length ξ(0Κ) (nm)

Penetration Depth λ(0Κ) (nm)

Al

1.2

1600

16

Cd

0.51

760

110

In

3.4

364

64

Nb

9.30

38

39

Pb

7.19

87

39

Sn

3.7

230

34

V

5.38

40

100

MgB2

39

6.0

185

Nb3Sn

18.3

3.1

100

NbN

16.1

4.5

200

PbMo6S8

13

2.5

240

BKBO

30

4.0

250

MgCNi3

7.0

4.6

225

NaxCoO2

4.5

5.7

790

KOs2O6

9.6

3.7

243

YBa2Cu3O7

91

0.3–2.0

500–150

SmFeAsO0.8F0.2

55

4.0

200

Rb3C60

29.5

2.0

440

k-(ET)2Cu(NCS)2

9.0

3.0

510

Small Superconductors—IntroductionClick to view larger

Fig. 1.3 Schematic illustration of fine filamentary networks formed by forcing soft superconducting metallic elements like Hg, In, Pb, etc. through the interconnected ultrafine pores of Vycor glass. Very fine mesh of metallic filaments of 3–5 nm diameter showed superconductivity up to magnetic fields exceeding 5 T, i.e. about more than 100 times greater than the bulk critical fields of these metals (after Bean et al. 1962).

(Figure shown along with this caption is reproduced from Narlikar (2014); Courtesy Oxford University Press, Oxford, UK).

Small Superconductors—IntroductionClick to view larger

Fig. 1.4 (a) Quasi-0D crystal structure showing part of the Chevrel phase lattice depicting small clusters of Mo-octahedrons where superconductivity resides. (b) Quasi-1D organic superconductor (TMTSF)PF6; here TMTSF molecules are linearly stacked. (c) Quasi-2D HTS cuprate where superconductivity occurs in copper–oxygen planar stacks that are mutually separated by other non-superconducting oxide planes

(For details see Narlikar 2014).

It is worth mentioning that some of the superconducting materials structurewise naturally grow as quasi-low-dimensional superlattices and interestingly a good many of them like HTS cuprates, Fe-based compounds, MgB2, and A-15 compounds are noted for their high Tc values. This has markedly emphasized the significance of low dimensionality in superconductivity. For instance, various HTS cuprates are all layered materials (Fig. 1.4c) in which metallic copper oxide layers one or a few atoms thick, where superconductivity occurs, alternate with insulating or blocking layers that make the structure quasi-2D. Similarly, the crystal structure of Chevrel phase compounds (Fig. 1.4a) possesses mutually separated localized molybdenum octahedra, important for their superconductivity, and some of the organic superconductors (Fig. 1.4b) are formed as linear stacks of molecules that superconduct. These structural features make the above two types of materials respectively quasi-0D and 1D. Interestingly, nanolevel defect structures present in bulk materials can additionally be of 0D/1D/2D (Table 1.2), which accordingly infuse in the latter the (p. 8) traits of low-dimensional systems. For example, non-interacting ultrafine particles embedded in a bulk matrix or the extra-fine grain structure of thin films can serve as a 0D system.

Table 1.2 Defect structures with different dimensionality in materials.

Defect Structure

Naturally Grown

Artificially Introduced

Zero-dimensional

  • Point defects:

  • Lattice vacancies

  • Interstitial atoms

  • Impurity atoms, and their small clusters, ultrafine grains & precipitates.

Nanoparticles, small pieces of nano-rods, nano-ribbons of additive materials like SiC, different forms of carbon, iron oxide, different carbides, and various pure metals such as silver, nickel, cobalt, etc.

One-dimensional

Crystal dislocations

Two-dimensional

  • Sample surface

  • Low and high angle grain boundaries, twin boundaries, various interphase boundaries, etc.

Three-dimensional

Precipitates of second phase, voids and inclusions, 3D networks of dislocation tangles and various boundaries

1.2 Two size effects in superconductors

How does superconductivity respond when the bulk sample is made smaller and smaller in the nano range? Although the question has been extensively pursued for the last two decades it is still a long way away from being fully understood. The issue is commonly discussed in terms of two size effects, namely (a) the small size effects, and (b) the quantum size effects. As with almost all nanomaterials, the pristine properties of bulk superconductors get markedly affected when their dimensions are reduced and made smaller than λ or ξ. Generally, the superconducting characteristics remain little affected down to 100 nm. The changes taking place on the length scale of 100 nm to about 20 nm are described by the small size effects (SSE), while in the ultra-small regime below 20 nm, until the superconductivity is quenched, the behavior is discussed in terms of the quantum size effects (QSE).

The extended size ξ of the Cooper pairs as supercurrent carriers poses an obvious constraint in their mobility through superconducting samples with dimensions smaller than ξ. Besides, there is a relevant question as to how an inherently long-range phenomenon like superconductivity can really survive on a scale that is small or ultra-small in comparison to ξ. In such a situation the Cooper pairs may get markedly squeezed and their wave functions significantly (p. 9) perturbed to adversely affect or even completely destabilize superconductivity. A third length scale, much smaller than either of λ or ξ, in the regime of QSE, is found to be necessary, which could set the limit to the superconductivity phenomenon. This third length scale is the Anderson limit.

1.3 QSE, quantum fluctuations, Anderson limit, parity and shell effects, etc.

In the ultra-small regime of 0D superconductors, with the nanoparticle size getting closer to the Fermi wavelength, quantum confinement effects occur which give rise to discretization of energy levels that forms part of the QSE. The mean spacing of the eigenspectrum, dk (the Kubo gap) ≈ (1/N0)EF; near the Fermi level EF increases with decrease of the particle diameter σ (Fig. 1.5) and superconducting properties dramatically change. The situation is akin to the electrons in a box model where a small change in the box dimensions produces large changes in the density of energy states. It is worth recalling that in superconductivity the electronic states located within a small shell of width ±hωD at EF are responsible for the phenomenon and these are markedly affected when the particle size is reduced.

Small Superconductors—IntroductionClick to view larger

Fig. 1.5 As the size of the superconductor (with energy gap of 2Δ at EF) is sequentially reduced from (a) bulk to (b) large grain, (c) small grain, and finally to (d) ultra-small grain, the inter-level spacing (Kubo gap) dk of the eigenspectrum rises. The Anderson limit for the particle size is said to be reached when the particle or grain size is so small that dk just exceeds the superconducting gap.

(p. 10) Anderson (1959) pointed out that in the ultra-small regime (< 20 nm diameter) of a 0D superconductor when the particle or grain diameter σ = σc is indeed so small that it leads to dk ≥ Δ, where 2Δ is the superconducting energy gap at the Fermi level EF, the particle will no longer be superconducting. This is known as the Anderson limit for the critical particle diameter σc, below which superconductivity is destabilized and ceases to exist. Broadly, the ratio Δ/dk relates to the number of Cooper pairs present in the system and, clearly, at σ = σc, when the above ratio becomes < 1 (i.e. Δ < dk), the σc is too small to hold any Cooper pair to sustain superconductivity. When the Anderson limit is approached from above, quantum fluctuations quickly build up to destabilize superconductivity. The series of experiments by Ralph et al. (1995, 1996, 1996a) strongly corroborated the discretization of the eigenspectrum of ultra-small superconductors and established the validity of the Anderson limit. These experiments also yielded some novel insights linked with (a) discretization of energy levels, and (b) the finite electron number present in small particles exhibiting superconductivity. Besides the suppression or destabilization of superconductivity just below the Anderson limit, (a) and (b) above give rise to a host of unexpected phenomena like quantum fluctuations, shell effects, and shape resonance effects, by modulating various features like electronic density of states (DOS), the superconducting energy gap, critical temperature Tc, electron–phonon interaction, resistivity, Hall conductivity, Hc2, etc. Most noteworthy of these are the building up of huge oscillations in Δ and Tc, first observed in Sn (Bose et al. 2010), when the particle size (or more correctly, the number of electrons therein) was reduced in the ultra-small range from 20 nm down to the Anderson limit. The behavior is described in terms of what are known as the shell effects and the shape resonance effects. It is worth pointing out that although both Δ and Tc monotonically decrease with the particle size, the amplitude of oscillations keeps on increasing as the Anderson limit gets closer. At the optimum level, in the case of nanoparticles of Sn, such oscillations were 60% higher than their corresponding bulk values (Bose et al. 2010). These oscillations are the manifestation of quantum fluctuations in the superconducting order parameter caused by large variations in the electronic DOS in a shell of width ±hωD at EF resulting from changes in the particle size. An increase in DOS implies a rise in Δ and Tc, and vice versa. Here the particle shape is also relevant as the symmetric (spherical or cubic, for example) particles contribute further degeneracies to the discrete energy spectrum and thereby enhance the fluctuations and the ensuing variation in Δ and Tc. The degenerate levels are analogous to the shells of atomic nuclei, which has prompted the name shell effects for the above phenomena. The maxima of the oscillating pattern are usually referred to as the shape resonances.

The finite electron number present in ultra-small superconductors makes the application of conventional BCS (Bardeen–Cooper–Schrieffer) theory unsuitable (p. 11) and further leads to an unusual parity effect. Accordingly, ultra-small grains with even electron numbers exhibit a larger superconducting gap than those with odd numbers, which is, however, consistent with the characteristic pair correlation of a superconducting state. Ultra-small grains with an odd number of finite electrons always possess an unpaired electron that blocks the scattering of other pairs and thereby weakens the pairing correlations and reduces Δ. Taking into consideration the parity effect, more involved calculations reveal that when the electron number present is even, superconductivity would disappear only when dk = 3.56Δ, but when the number is odd, the phenomenon will vanish when dk = 0.89Δ (Dukelsky and Sierra 2000).

Table 1.3 gives the measured values of the Anderson limit σc for some of the elemental superconductors. As can be seen, the values are much smaller than the pertinent λ or ξ values of Table 1.1. Similar studies, however, seem to be lacking for superconducting alloys and compounds. The Anderson limit for MgB2 has been found to be 2.5 nm (Li and Dou 2006) and their results for nano-grained samples are shown in Fig. 1.7. As may be seen, down to about 12 nm grain size there is essentially no change in the bulk Tc of MgB2, but below that it decreases fast and no superconductivity is observed at grain sizes of 2.5 nm and below. On theoretical considerations Ivanov et al. (2003) have pointed out that for superconductors with higher Tc and larger gap coefficients the values of σc will be much smaller than for pure elements of Table 1.3 which, experimentally, may not be easy to realize. For HTS cuprates, for example, which have Tc > 100 K and a large gap coefficient in the range 4 to 12, the σc is predicted to be < unit cell length, and therefore the Anderson limit may, in fact, be unattainable (Ivanov et al. 2003).

Interestingly, many of the mentioned effects in fact were long ago theoretically predicted (Blatt and Thompson 1963, Hwang et al. 2000, Jankó et al. 1994, Kresin and Ovchinnikov 2006, Wei and Chou 2002, Yu and Strongin 1976) and a good many of them have also since been experimentally corroborated (Bao et al. 2005, Bose et al. 2009, 2010, Czoschke et al. 2003, García-García et al. 2008, Orr et al. 1984, Ralph et al. 1995, 1996a, 1996b, Romero-Bermudez and García-García 2014, Tinkham 2000, Upton et al. 2004, von Delft and Ralph 2001).

Table 1.3 Anderson limit for nanoparticles of some elemental superconductors. The values mentioned are from Bose and Ayuub (2014) and references therein, and Yang et al. (2011).

Superconductor

Al

In

Nb

Pb

Sn

V

Anderson limit σc (nanoparticle diameter) (nm)

6.2

5.5

7.0

3.5

4.3

3

(p. 12) 1.4 Factors influencing small size effects

The superconducting behavior in the regime of small size effects, i.e. in the 100–20 nm range, is commonly linked to two main factors arising from size reduction. (a) The disruption in the periodicity at the sample surface causes the latter to possess vastly altered properties. At the nanolevel, the surface to volume ratio becomes significantly prominent to cause the surface properties to override the bulk properties (surface effects). (b) A decrease in the electron mean free path due to spatial confinement (electron mean free path effects) which, as we will see, has interesting consequences. Besides (a) and (b) there are fluctuations (thermal and quantum) in the form of phase slips that become important as the sample size diminishes. Their effect (Sec. 1.5) is primarily to broaden the superconducting transition and adversely affect superconductivity.

1.4.1 Surface effects

With samples of reduced size, their surface properties grow more relevant than bulk properties and this influences many of their important parameters like the critical temperature, the critical magnetic field, and the critical current. The atoms located on the sample surface possess a lower number of bonds than those located in the bulk interior, which makes the surface phonon modes much softer than the bulk and this thereby lowers its average phonon frequency.

The dimension-less parameter λe, which measures the electron–phonon interaction strength in conventional superconductors to determine their Tc, is related to the electron DOS N(0) at EF, ionic mass M, the electronic matrix element ⟨I2⟩, and mean square phonon frequency ⟨ω2⟩ through the relation,

(1.1)

λe=N(0)I2M*ω2

Decrease in ⟨ω2⟩ in the above situation favors an increase in λe and consequently also the critical temperature. In this way, fine pressed powders (quasi-0D category) of elemental platinum, which is otherwise non-superconducting, exhibit superconductivity at 0.02 K (Schindler et al. 2002). The softened surface phonon modes in the case of small superconductors make a positive contribution to their critical temperature. There is also a negative contribution to Tc coming from structural distortion occurring as a result of the sample size reduction. This causes a decrease in the electronic DOS N(0) at EF which suppresses Tc. As we have seen earlier, discretization of energy levels occurs in small particles due to quantum confinement effects that subsequently get broadened as the particle size is reduced. This again lowers the DOS which makes a negative contribution to Tc. The overall effect of these competing factors is found to be different in different materials. The variation of Tc with particle size results in three types of behavior (p. 13) which are schematically shown in Fig. 1.6. A pronounced peak-like behavior such as shown by curve II is manifested by weakly superconducting metal. For instance, in the case of In, which is weakly superconducting, the Tc goes through a sharp maximum at about 30 nm and thereafter, in the ultra-small regime, it shows a fast decrease until the Anderson limit is reached, below which the phenomenon disappears (Li et al. 2005). For intermediate electron–phonon coupling superconductors, the behavior is as shown by (III), i.e. the curve is initially flat, and then there is a decrease in Tc with particle size which becomes faster as the Anderson limit is approached. This kind of behavior is manifested by Nb which possesses intermediate electron–phonon coupling and its superconductivity is lost when the particle size is smaller than around 7 nm (Bose et al. 2009). In general, the negative contributions in small superconductors dominate over the aforesaid positive contribution when the sample dimensions are lowered below around 20 nm and the overall effect is therefore to suppress Tc. The behavior represented by (I) in Fig. 1.6 is for strongly coupled superconductors such as Pb. From bulk down to below 10 nm there is no change in Tc which Bose et al. (2009) attribute to the almost exact compensation of the positive phonon softening effect by the negative quantum size effect until the Anderson limit is approached. Interestingly, the behavior (Fig. 1.7) of MgB2 (Li and Dou 2006) seems very similar to that of Pb and may also be explained similarly.

Small Superconductors—IntroductionClick to view larger

Fig. 1.6 Three types of behavior observed for the size effect of critical temperature of superconducting nanoparticles.

Interestingly, in the case of weakly coupled superconductors like In, Al, and Sn, the gap coefficient 2Δ(0)/kBTc increases on size reduction; it remains invariant for Nb whose gap coefficient is of an intermediate value between weak and strongly coupled superconductors; while for Pb which is strongly coupled, its gap coefficient exhibits an increase (Bose and Ayyub 2014) on size reduction.

Surface effects are important also in influencing the critical magnetic field of small superconductors. Saint-James and de Gennes (1963) had showed that the solution of Ginzburg–Landau equations for a type-II superconductor subjected to a magnetic field was quite different at its surface than for the bulk. Even when the magnetic field had exceeded the upper critical field Hc2 to quench the bulk superconductivity, a thin surface layer of thickness ξ, parallel to the magnetic field, continued to remain superconducting up to a much (p. 14) higher magnetic field, known as the sheath critical field Hc3 ≅ 1.695Hc2. For thin discs the numerical factor exceeds even 2.5 (Schweigert and Peeters 1999). The surface phonon modes, mentioned earlier, are believed to be responsible for the origin of sheath superconductivity. For type-II superconductors of low Ginzburg–Landau parameter κ (see Narlikar 2014), the existence of a surface sheath promotes magnetic irreversibility and a higher critical current density jc. However, for large-κ materials, which are used for winding high-field superconducting magnets, different types of pinning entities in the form of small normal (non-superconducting) particles and structural defects are essential for strong flux pinning and enhanced jc.

Small Superconductors—IntroductionClick to view larger

Fig. 1.7 Until about 12 nm grain size the critical temperature of MgB2 remains the same as of the bulk sample. Tc subsequently is fast suppressed on further lowering the grain size. No superconductivity is observed below 2.5 nm, i.e. the Anderson limit for MgB2 (based on the reported measurements of Li and Dou 2006).

Surfaces of small precipitates of a second phase, various structural defects, and of artificially introduced fine particles (Table 1.2) embedded in wires, tapes, or thin films of superconductors serve as effective pinning centers for Abrikosov’s flux vortices. Their presence vastly enhances the critical current density of the superconducting matrix in which they are embedded. In general, for optimum pinning, the number density of the pinning entities should be large, their volume fraction small, and the surface area of each entity or the pinning center should be optimum. Ideally, since the normal core of the flux vortex has a radius equal to the range of coherence ξ of the matrix, the size of the pinning centers should be in the nano range, for optimum pinning. This makes the pinning entities mesoscopic. Since the magnetization of the pinning entity, in general, differs from that of the superconducting matrix, there is always a circulating supercurrent around the pinning center at its interface with the matrix. This surface current interacts with the circulating supercurrent of the moving flux vortex and creates an irreversible surface barrier responsible for flux pinning (Bean and Livingston 1964, Campbell et al. 1968). This approach of introducing nanoparticles of different materials and metallurgical phases, such as Gd-211 in HTS cuprates (Muralidhar et al. 2004), or nano-diamond (Cheng et al. 2003, Vajpayee et al. 2007), nano-SiC (Dou et al. 2002), nano-C (Yeoh et al. 2006), nano-SiO2 (Rui et al. 2004), carbon nanotubes (Yeoh et al. 2006a), etc. as nanosized artificial pinning centers (APCs), has been effectively used to enhance the jc of superconducting MgB2. In fact, this has emerged as a successful route to increase the current carrying capacity of several high-field superconductors for various applications.

1.4.2 Electron mean free path effects

When a metallic particle or a grain is reduced in size, a wire is made narrower, or a film is made thinner, all lead to a decrease in their electron mean free path. The electron mean free path stands out as an important factor in superconductors which in different ways influences several of their important properties and parameters. These prominently include the two characteristic lengths λ and ξ, the normal stat resistivity ρn, the Ginzburg–Landau parameter κ, and the three critical fields Hc1, Hc2, and Hc3.

(p. 15) 1.4.2.1 λ, ξ, and κ

When the sample size is reduced, the electron mean free path lF decreases, which has mutually opposite effects on the coherence length ξ and the penetration depth λ. The former decreases while the latter increases, following the relations from the Ginzburg–Landau theory (1950):

(1.2)

ξ(T)=0.60(ξolF)½/(1t)½

(1.3)

λ(T)=0.62λo{ξo0/lF(1t)}½

where ξο is the coherence length for the pure material at 0 K. In the case of thin superconducting discs (thickness t) the effective penetration depth, Pearl’s penetration depth (Pearl 1964), is given by Λ = 2λ2/t. The Ginzburg–Landau parameter κ, which varies as λ/ξ, therefore increases as the sample thickness is reduced. Interestingly, this makes a type-I superconductor like Pb behave like a type-II when it is formed as a nanoparticle of around 15 nm diameter (Bose 2007). As a result very thin type-I superconductors show an interesting vortex structure, normally seen only in type-II, but, as we will see later, they also carry many other surprising features (Deo et al. 1997, Schweigert and Peeters 1998, 1999).

1.4.2.2 Critical fields Hc1 and Hc2

The lower and upper critical fields Hc1 and Hc2 are respectively related to λ and ξ,

(1.4a)

μoHc2(0)=φo/2πξ2

(1.4b)

μoHc1(0)=φo/4πλ2

where φo (= h/2e = 2×10−15Wb) is a quantum of magnetic flux.

Consequently, with size reduction the lower critical field would decrease while the upper critical field would increase. However, if the sample size is smaller than λ, due to an incomplete Meissner effect, as discussed in 1.1, the Hc1 would increase. Thus, in the case of Hc1 there is a competition occurring between the electron mean free path effect and the consequence of the incomplete Meissner effect, and the latter effect should dominate as the sample size is reduced. As mentioned earlier, the lower critical field does indeed go up with decreasing particle size (Blinov et al. 1993).

There is a similar competition also with Hc2. Decrease in the electron mean free path enhances the normal state resistivity of the small superconductor, which is related to the upper critical field through the relation (see Narlikar 2014),

(1.5)

μoHc2(0)=3.09×γρnTc(T)

(p. 16) The coefficient of the electronic specific heat in the normal state, γ, and the critical temperature Tc do not normally change significantly by electron mean free path effects and consequently an increase of ρn as per the above equation would make a positive contribution to the upper critical field. However, there is a competing negative contribution that arises from discretization of energy levels due to the quantum confinement effect which, as discussed earlier, causes a decrease in the DOS at EF and thereby suppresses Tc. The latter effect would surely dominate when the sample size is reduced down to the ultra-small regime and the upper critical field would finally vanish when the Anderson limit is reached. The reported observations and their explanation of findings by Bose et al. (2006) of such a non-monotonic behavior of the upper critical field of nano-sized grains of Nb are in accord with the above reasoning.

Nanostructured thin films of some superconductors are noted for their exceptionally high upper critical fields in comparison with their bulk values, which makes them attractive candidates for high magnetic field applications. For instance, for Chevrel phase superconductors the upper critical field for nanostructured films is reported to increase from 50 T to 100 T while for MgB2 from 16 T to about 75 T (Narlikar 2014).

1.5 Behavior of nanowires and ultra-thin films

1.5.1 Nanowires and fluctuation effects: thermally activated phase slips and quantum phase slips

Fluctuation effects, in general, become relevant for low-dimensional systems and all the more so in 1D superconducting nanowires. Quantum fluctuations were briefly mentioned in Sec. 1.3. In a strictly 1D system, superconducting long-range order and a zero-resistance state are, in fact, forbidden by the Mermin–Wagner (1966) theorem. Langer and Ambegaokar (1967) and McCumber and Halperin (1970) in their LAMH theory had pointed out that the process responsible for transient breaking down of superconductivity just below Tc was the so-called phase slips, which originated from thermal fluctuations. In addition to these thermally activated phase slips (TAPS), the experimental evidence (Giordano 1994) has indicated that for ultra-thin wires, at much lower temperatures < Tc, when thermal excitation is not relevant, quantum phase slips (QPS) can occur due to quantum fluctuations via quantum tunneling.

Superconducting order is characterized by a wave function ψ = |ψ|exp.(iϕ) where ϕ is its spatially coherent phase that is, so to speak, locked everywhere in the superconducting state. Locally, near Tc, the coherence can, however, get disturbed by thermal fluctuations or the occurrence of phase slips which results in the local transient loss of superconductivity (t ≈10−12 s) over a region of size ξ where the order parameter ψ locally fluctuates to zero and the phase ϕ becomes (p. 17) indeterminate. The minimum energy cost for this is a function of the condensation energy density ½μ0Hc2 per unit volume, and for a nanowire of cross-sectional area A becoming normal over a segment ξ, the corresponding energy needed, i.e. the barrier height, is (Langer and Ambegaokar 1967) ΔF = K0Hc2Aξ, where Hc is the thermodynamic critical field and K is a constant. Subsequently, as ψ builds back up to its original value, the phase undergoes a slip and rotates by ±2π from its value just prior to the phase slip event. This change of phase corresponds to a voltage pulse, with more slips giving rise to more resistive voltages. The process is represented in Fig. 1.8a by a spiral of constant radius where with each phase slip event the spiral loses one complete loop corresponding to a phase of 2π. As shown in Fig. 1.8b, the process (TAPS) can take place by thermal activation over the barrier of height ΔF, just below Tc or, if the wire is ultra-thin, it (QPS) can (p. 18) occur quite far below Tc, through the above barrier due to quantum fluctuations via tunneling. In both situations the system moves from one local potential minimum into the neighboring one separated by ±2π in the phase space.

Small Superconductors—IntroductionClick to view larger

Fig. 1.8 Phase slips in 1D superconductors. (a) The two axes shown represent the real and imaginary parts of the order parameter ψ, and the third axis is along the nanowire length. The spiral of constant radius represents a constant supercurrent and prior to phase slip, the spiral remains intact (the top spiral). A phase slip event (10–12s) causes the order parameter |ψ| to become locally zero and the spiral loses one loop (middle spiral). During the phase slip, the phase at some point in the nanowire becomes indeterminate. During the recovery it is changed by ±2π, resulting in a finite voltage with loss of superconductivity, and as shown in (b) the superconducting system passes from one state to another across a potential barrier. The energy to cross the barrier is provided by thermal fluctuations, leading to thermally activated phase slips (TAPS), or by quantum fluctuations via the tunneling process resulting in quantum phase slips (QPS).

Small Superconductors—IntroductionClick to view larger

Fig. 1.9 Typical Resistance–Temperature behavior of superconducting nanowires, displaying thermally activated phase slip (TAPS) and quantum phase slip (QPS).

In 2D and 3D samples the supercurrent transport is, however, little affected by the creation of the normal (non-superconducting) region that can be readily bypassed by moving Cooper pairs without interruption. But, in nanowires of radius smaller than ξ, the normal region formed blocks the entire cross-section of the nanowire, and therefore it can no longer be bypassed by electron pairs. The wire therefore ceases to be superconducting. The experimental results show that the normal state resistance RN of the nanowire has to be smaller than the quantum resistance RQ = h/(2e)2 ≈ 6.5 kOhm to follow the predictions of the LAMH theory. When RN > RQ, the nanowire fails to exhibit superconductivity (Bezryadin 2008).

Because of the two characteristic phase slips, TAPS and QPS, mentioned earlier, the resistance (R)–temperature (T) behavior of nanowires differs from that of the bulk wires (thickness/diameter of around 100 nm or more) in two ways. Instead of the very sharp transition (width < 0.01 K) of bulk wires the TAPS results in the rounding and broadening of the transition point at Tc-onset, while the QPS results in an elongated tail at T < Tc, which frequently does not lead to a R = 0 state even at 0 K. The behavior is schematically shown in Fig. 1.9. This happens when the wires are sufficiently small (≤ 30 nm) in diameter or thickness. The behavior has been studied in a number of elemental superconductors like Sn, Zn, Al, Pb, and In (Zgirski et al. 2005, Sharifi et al. 1993, Giordano 1994, Bezryadin et al. 2000, Lau et al. 2001, Tian et al. 2005, Wang et al. 2005). However, the transition from TAPS to QPS in superconducting nanowires still remains to be systematically studied.

As with 0D superconductors, discussed earlier, 1D ultra-thin nanowires do exhibit the quantization of electron motion and discretization of energy levels, but only along the directions transverse to the length. These have been experimentally corroborated by observations of shape resonances with large oscillations in Tc of ultra-thin nanowires of Al, Sn, and Pb (Guo et al. 2004, Shanenko et al. 2006) and also theoretically explained (Shanenko and Croitoru 2006).

1.5.2 Ultra-thin films

The 2D superconducting system holds some similarity with the behavior of 1D nanowires described above. When the normal state resistance RN of a thin film exceeds the quantum resistance RQ = 6.5 kOhm, superconductivity is not observed (Haviland et al. 1989, Hebard and Paalnen 1990, Bollinger et al. 2011). Ultra-thin superconducting films with thickness < 20 nm, if they are not of a continuous form, comprise small islands and exhibit granular behavior, characteristic of 0D superconductors discussed earlier. The films are formed on substrates which can influence their intrinsic behavior through the proximity effect. Similar to what we previously discussed for 0D superconductors, fluctuations too play (p. 19) a significant role in 2D systems. The Berezinskii–Kosterlitz–Thouless (BKT) transition originating from phase fluctuations adversely affects superconductivity. Interestingly, despite this, superconductivity does occur in ultra-thin films of thicknesses much smaller than ξ and, in fact, even down to the level of a single monolayer of the superconducting elements (Zhang et al. 2010) and compounds (Lee et al. 2014, Xi et al. 2015), albeit with reduced Tc and broadened transition as compared to the bulk samples.

Small Superconductors—IntroductionClick to view larger

Fig. 1.10 Critical temperature of epitaxial ultra-thin NbN films. The filled circles present the results of Kang et al. (2011). The open circles represent the data points of Marsili et al. (2008). In both cases the films were deposited on MgO substrates by magnetron sputtering.

Fig. 1.10 depicts the results of Tc depression in ultra-thin films of NbN reported by Marsili et al. (2008) and Kang et al. (2011) when the film thickness varied from 100 nm to about 2.5 nm. Clearly, the two sets of observations are in close mutual agreement. Down to 100 nm the critical temperature remains practically invariant from the bulk value. From 100 to about 10 nm thickness the films display a small decrease of Tc, which, however, in the ultra-thin regime below 10 nm depicts a much faster drop. As also discussed previously, a decrease in the film thickness can give rise to different competing factors influencing Tc. Firstly, due to the previously discussed small size effects (SSE), the structural disorder in the film increases with decreasing thickness, causing a reduction in the DOS at EF which lowers the Tc. On the other hand, as the film gets thinner, surface phonons become more dominant to increase the electron–phonon interaction which promotes Tc. As with 0D and 1D superconductors, in the ultra-thin regime of 2D films, the electron wave function is quantized along the direction normal to the film surface, which results in discretization of energy levels as part of the quantum size effects (QSE). This contributes to a decrease in the DOS at EF and lowers Tc. The observed behavior of Tc in Fig. 1.10 is the overall impact of the above factors in which the QSE finally dominate and depress Tc through decrease in the DOS (Kang et al. 2011) at EF. Similar to 0D and 1D superconductors, discretization of energy levels in ultra-thin films gives rise to large oscillations (Bao et al. 2005, Shanenko et al. 2007) in various superconducting parameters in the form of shape resonances as we discussed earlier. (p. 20)

Small Superconductors—IntroductionClick to view larger

Fig. 1.11 Resistance–Temperature curves for ultra-thin films of superconducting (a) YBCO (after Gao et al. 1999) and (b) magnesium diboride (after Wang et al. 2009), both showing Tc decrease and broadening of the transition with decrease of film thickness.

Fig. 1.11a presents the results of RT measurements of Gao et al. (1999) on ultra-thin films of YBCO while Fig. 1.11b depicts the results on MgB2 films (Wang et al. 2009). In both cases the transition gets rounded and broadened and Tc goes down. In the case of YBCO, the 2.5 nm film does not seem to indicate any possibility of reaching the zero-resistance state even at 0 K. Interestingly, the MgB2 films of thickness ≤ 20 nm exhibit a localization effect by displaying the characteristic negative dR/dT at low temperatures, possibly due to the presence of structural disorder occurring at the ultra-small thickness. However, from the RT curves of Fig. 1.11a and Fig. 1.11b one cannot unambiguously distinguish TAPS and QPS regimes.

1.6 Vortex states of small superconductors

We have already seen above how size effects influence Tc and other related properties of superconductors. In this section we take a quick look at the influence of (p. 21) confinement on the formation of magnetic field induced vortex states in them. Normally, the vortex structure is a characteristic feature only of type-II superconductors. However, type-I superconductors, as we have seen earlier, if they are made sufficiently small, start behaving akin to type-II ones and thus may have flux vortices under an applied magnetic field. When the field exceeds the lower critical field Hc1 the macroscopic Meissner state gives way to a mixed state, comprising a triangular lattice-like arrangement of Abrikosov’s vortex lines (Fig. 1.12a) threading the bulk of the superconductor (Abrikosov 1957). The triangular vortex lattice, as imaged by Nishio using scanning Hall microscopy, is depicted in Fig. 1.12b. Each vortex line has a normal cylindrical core of radius ξ carrying a single quantum of magnetic flux ϕo (= h/2e = 2 × 10−15Wb), which is surrounded by vortices of supercurrents spread over a radius of λ (Fig. 1.12c). The smallest subdivision of magnetic flux is a consequence of the negative surface energy existing between the normal core and its superconducting surrounding, the perennial characteristic of type-II superconductivity. A single rotational turn around the vortex changes the phase of the order parameter by 2π. In general, for a bulk superconductor, it is energetically impermissible for a flux line, whose self-energy varies as ϕo2, to contain n ≥ 2 flux quanta. Such a vortex line, known as the giant vortex, will normally be unstable and split into n vortices, each with a (p. 22) single flux quantum. However, a giant vortex state (GVS), as we will see, is one of the novel and stable vortex states of small superconductors.

Small Superconductors—IntroductionClick to view larger

Fig. 1.12 (a) Schematic of triangular vortex lattice, (b) vortex lattice in a superconductor imaged using scanning Hall microscopy (courtesy Nishio), (c) structure of a vortex line (Narlikar 2014).

The triangular vortex lattice in bulk superconductors is the outcome of mutual repulsion between the macroscopic number of vortex lines. In small or confined superconductors of size <(λ, ξ), where the sample boundary casts its strong influence on the vortex configuration, the latter markedly deviates from the conventional triangular lattice. The vortex configuration is now determined not just by the repulsive inter-vortex interaction but also by the sample boundary simultaneously imposing the strong effect of its own symmetry in arranging the vortex distribution (Baelus and Peeters 2002). We have already mentioned earlier surface superconductivity and the presence of the Bean–Livingston surface barrier (1964) for flux vortices. Extensive theoretical (Schweigert and Peeters 1998, Schweigert et al. 1998, Palacios 2000, Baelus et al. 2004) calculations, based on the Ginzburg–Landau theory, the London approach, molecular dynamics, and the Bogoliubov–de Gennes formulation (Bogoliubov 1947, de Gennes 1966) have predicted the existence of two distinct vortex states, namely (a) the giant vortex state (GVS) and (b) multi-vortex state (MVS), respectively consisting of (one or more) multiply quantized and several singly quantized flux vortices. These have been corroborated by a host of experimental studies (Moshchalkov et al. 1995, Bruyndoncx et al. 1999, Geim et al. 1997, 2000, Kanda et al. 2004, Baelus et al. 2005) on mesoscopic superconducting thin (~5 nm) circular discs of lateral size of a few ξ (< 10 ξ). The total number of flux quanta present in the sample is called its vorticity or winding number denoted by L which is an integral number. Alternatively, this parameter is defined as L = Δϕ/2π, where Δϕ is the total phase change of the sample due to magnetization. The spatial arrangement of vortices in mesoscopic superconductors is controlled by the geometrical shape of the sample boundary and the vorticity L. In the case of a small circular disc, instead of Abrikosov’s triangular lattice, the vortices formed prefer to arrange in circular form or a shell along the periphery of the sample edge. Some form of triangular lattice may however appear in the central part of the disc, away from the boundary, where its geometrical effect is relatively less. With an increasing external magnetic field or by reduction of the lateral size, as more such vortices with a single flux quantum are pushed inside, some of them tend to merge to form a single giant vortex at the center of the disc, with its normal core carrying a magnetic flux of (nϕo). The geometrical boundary effect in small superconductors facilitates the formation of a GVS which, as discussed above, is otherwise not feasible in bulk samples. If the lateral size of the sample is increased, or the magnetic flux density is reduced the GVS would, however, dissociate into the multivortex state.

Unlike the bulk superconductors, in mesoscopic samples only a finite number of vortex lines are accommodated and the confinement effect increases with decrease in the sample size. For instance, if the lateral diameter of the superconducting disc D ≤ 2ξ, which represents the diameter of the normal core of a flux vortex, the disc clearly would not accept it as this would imply the destruction of superconductivity. Insertion of a single vortex in such a sample implies an immediate transition from the Meissner state to the normal state, without passing (p. 23) through the mixed state. This way, effectively, a type-II superconductor under a small magnetic field behaves like a type-I. Interestingly, the reverse situation was discussed previously in the absence of a magnetic field where a small type-I superconductor behaved like a type-II. The ultimate confinement represents the situation where the sample is large enough to accommodate just one vortex inside without the destruction of superconductivity. The above contentions have been experimentally well corroborated by studies of Nishio et al. (2008) on thin nanocrystals of Pb on Si(111) which showed that samples with lateral size smaller than (2.5–2.8)ξeffective are not able to accommodate even a single vortex.

Interestingly, much theoretical work has been done (Baelus et al. 2002, 2004) towards predicting different vortex configurations formed in nanosized superconductors or nanodots of various geometrical shapes like circular discs, equilateral triangles, squares, rectangles, etc., as a function of increasing L. Many of these have also been successfully corroborated through the Bitter decoration (Grigorieva et al. 2006) and various scanning probe techniques, like scanning tunneling, scanning SQUID, and scanning Hall microscopy techniques (Kokubo et al. 2010, 2014, 2015, Nishio et al. 2004, 2008a). In the case of circular discs, the multivortex state formed grows up in the form of discrete concentric circular shells (Xu et al. 2011). Interestingly, the process of the filling up of shells with added vortices resembles the building up of the well-known periodic table of chemical elements. Starting from the Meissner state, which corresponds to L = 0, until L = 5 all the vortices occupy the first shell. A new shell is formed at L = 6 with the added vortex occupying the center of the previous shell having 5 vortices and such a state is represented by (1,5). With the increase of vorticity up to L = 8 the state changes to (1,7), whereafter (i.e. at L = 9) with added vortices, first only the inner shell grows to the state (2,7). This pertains until L = 13, then is followed by increase of the vortices only in the outer shell, bringing about the state represented by (2,11). Subsequently, at L > 14, the inner shell grows, leading to the state (5,11) at L = 16. The third shell appears at L = 17, with the added vortex again occupying the shell center, and the state is represented by three digits, i.e. (1,5,11). The next three vortices, leading to L = 20, are all added to the outermost shell, i.e. (1,5,14) and thereafter, up to L = 32, all the shells grow intermittently to (5,11,16). The fourth shell is formed at L = 33, again in the form of a single vortex at the center of the existing three shells, and the vortex configuration is represented by four numbers, i.e. (1,5,11,16). The numbers 6, 17, and 33, at which a new shell is formed, are referred to as magic numbers. These magic numbers and the described pattern of shell formations and filling are both theoretically and experimentally in mutual agreement.

It would be interesting to ask how the vortex configuration might change in small noncircular superconducting samples possessing a discrete symmetry, such as squares and equilateral triangles having pointed edges of the size ≈ 5ξ to 10ξ? In this situation the ensuing vortex configuration grows more complex since the vorticity has to adjust with the corresponding boundary shapes respectively of C4 and C3 symmetries (Chibotaru et al. 2001, 2005, Misko et al. 2003, Teniers et al. 2003, also see Chapter 3 of this book). L = 1 and L = 4 in a square-shaped (p. 24) sample present no problem as one and four vortices, each with a single flux quantum, can respectively be accommodated at the square center and at its four corners or diagonal points without disturbing the fourfold symmetry. Similarly L = 2 can be readily realized by replacing the single quantum vortex of L = 1 by a giant vortex with L = 2 at the square center. The vorticity can be increased to L = 5 by adding a single quantum vortex at the center of the four corner points of the L = 4 configuration and further raised to L = 6 by replacing the central vortex of L = 5 by a giant vortex with L = 2. Both these configurations represent two concentric shells. Vorticities of L = 3 and L = 7 are, however, not consistent with the square symmetry of the sample. Here an additional vortex–antivortex pair is formed to geometrically adjust the vortex distribution to meet the fourfold square symmetry of the sample; here antivortex represents an oppositely directed vortex with a flux quantum –ϕo. Thus, for L = 3, we have four vortices at the square corners and one antivortex at its center (i.e. 3 vortices + 1 vortex–antivortex pair). The vortex–antivortex pair is also called a vortex–antivortex molecule. Similarly, the configuration for L = 7 is formed with four vortices on each of its two diagonals with an antivortex at the square center (i.e. 7 vortices + 1 vortex–antivortex pair/molecule) constituting a three-shell arrangement (4ϕo.+ 4ϕo.–1ϕo.). A similar situation exists for a triangle-shaped sample, for example, for L = 2 which is not compatible with the triangular shape. This can be realized by three ϕo. vortices, one each at the three apex points, and a single –ϕo. (antivortex) at the center of the triangle. Interestingly the filling up of the triangle and square centers with magnetic flux quanta, as the vorticity rises, follows distinct repetitive patterns (Table 1.4). The former is 0:+1:–1:0:+1:–1:0:+1: … while for a square it is 0:+1:+2:–1:0:+1:+2:–1:0 … This holds relevance for superconducting technology applications, which can use the triangular and square nanodots as components of supercomputers in quantum computing.

Table 1.4 Filling up of magnetic flux vortices/flux quantum at the center of mesoscopic superconducting triangles and squares with increasing vorticity.

Vorticity L

0

1

2

3

4

5

6

7

8

9

10

11

Mag.flux (ϕo) at triangle center

0

+1

  • –1

  • AV

0

+1

  • –1

  • AV

0

+1

  • –1

  • AV

0

+ 1

  • –1

  • AV

Mag.flux (ϕo) at square center

0

+1

  • +2

  • GV

  • –1

  • AV

0

+1

  • +2

  • GV

  • –1

  • AV

0

+1

  • +2

  • GV

  • –1

  • AV

GV: giant vortex state;

AV: antivortex state. In the case of a triangle the repetitive pattern is 0:+1:–1:0:+1: … …, while for a square: 0:+1:+2:–1:0:+1: … ….

The antivortex formation in square and triangle-shaped samples, described above, is however symmetry induced which to date remains to be experimentally corroborated. One of the reasons for this is that vortex–antivortex pair formation is sensitive to the presence of defects in the sample and moreover their mutual separation is too small, i.e. < ξ.

(p. 25) 1.7 Proximity effect behaviors

1.7.1 Proximity effect and Andreev reflection

The proximity effect is a mesoscopic phenomenon that takes place when a superconductor (S) is placed at close proximity and in good electrical contact with a normal metal (N). The Cooper pairs from S now, so to speak, leak into N with the result that the latter inherits superconducting characteristics over a mesoscopic range, extending from the nano- to the micrometer scale, close to their NS interface. This interesting phenomenon is called the superconducting proximity effect, and the length scale over which it occurs, the proximity range or Thouless range. The proximity behavior was first observed by Holm and Meissner (1932) which could, however, capture more serious attention only after the discovery of the Josephson effect in 1962, which subsequently led to its significant theoretical understanding (Andreev 1964, de Gennes 1966, Deutscher and de Gennes 1969). The advent of nanoscience in the late 1980s and realization of various challenges posed in its understanding gave a fresh impetus to further studies of the proximity effect for basic research and applications. The effect makes it possible to induce superconductivity in varied materials, resulting in a wide combination of diverse properties which normally would be unattainable. Proximity superconductivity has been reported in numerous novel materials such as graphene (Heersche et al. 2007, Hayashi et al. 2010), DNA molecules (Yu et al. 2009), and in a host of nanowires and nanobelts of Al, Au, Co, Pb, Zn, etc. (Singh et al. 2012). Because of the mesoscopic range of the phenomenon, the experimental studies are performed with the two materials processed in thin film form with, of course, a good electrical bonding in between. It is worth emphasizing that superconductivity induced by the proximity effect essentially represents a metastable state resulting from noninteracting charge carriers and is therefore different from the stable phenomenon originating from the regular pairing interaction of interacting electrons of the BCS theory.

At voltages and temperatures below the superconducting gap, single particle tunneling is suppressed, and for electrons with energy less than 2Δ the dominant process for transfer of electrons across the NS interface is the Andreev reflection. In this, an incoming quasi-electron from N is reflected at the NS interface as a quasi-hole. The missing charge 2e is absorbed by the superconducting condensate S as a Cooper pair. The opposite may also happen and accordingly a quasi-hole may get reflected in the Andreev way as a quasi-electron and thereby remove one Cooper pair from S. This is believed to be how the Cooper pairs in the proximity effect may leak from S to N. It is worth pointing out that the Andreev reflection, which is called a retro-reflection, is different from a normal reflection. In this there is a conservation of momentum and not of electrical charge and further it is compatible with time reversal, i.e. the reflected wave vector propagates back along the incoming path.

(p. 26) 1.7.2 Inverse proximity effect and the behavior of S–FM hybrid structures

Interestingly, just as the proximity of a superconductor makes a normal metal superconducting, there exists also a simultaneous reverse effect in which the proximity of a normal metal weakens superconductivity due to the depletion of Cooper pairs in S and diffusion of normal electrons from N to S. The Tc of S may therefore get significantly lower than the bulk value or may even get completely quenched. This is known as the inverse proximity effect which becomes particularly interesting when N is magnetic.

The proximity range (or Thouless range) may, in general, vary from less than a few nanometers to over many micrometers from the NS interface, depending on the materials used. For instance, in pure nonmagnetic metallic N the length scale at very low temperatures is in the micrometer range which is, however, markedly lowered to a few nanometers or even smaller when component N is ferromagnetic (FM). Clearly, the singlet superconductivity with Cooper pairs having antiparallel electron spins is expected to decay rapidly when leaked inside a ferromagnetic N exhibiting parallel spins. This is found true when the components involved, such as (Fe/Ni)–(In) (Chiang et al. 2007), are of macroscopic size and the proximity length is ≈1 nm. Surprisingly, however, for ultra-thin films and nanowires of ferromagnetic materials forming mesoscopic S–FM hybrid structures the superconducting proximity effect is unexpectedly enhanced to the micrometer range and also it is larger for thinner nanowires (Petrashov et al. 1999, Bergeret et al. 2005, Keizer et al. 2006, Wang et al. 2010). Researchers have argued that the large proximity range is due to a change from the conventional spin singlet superconductivity to the unconventional spin triplet superconductivity where parallel electron spins are permitted. In the case of magnetic Ni and Co nanowires proximity-connected to superconducting W electrodes, Wang et al. (2010) find the proximity length is around 500 nm and, interestingly, RT measurements depict an unusual peak near the Tc-onset which is absent with nonmagnetic nanowires. According to the authors, the peak is possibly related to the coexistence of superconductivity and ferromagnetic order.

1.7.3 Giant proximity effect

An anomalously large proximity effect, extending over a length scale of 10–100 micrometers, is observed when the component S is an optimally doped high-Tc cuprate and N is also of the same family, but in the underdoped normal state (Bozovic et al. 1994, 2004, Kasai et al. 1992, Marchand et al. 2008). It is called the anomalous or giant proximity effect. The unconventional normal state is considered one of the distinguishing marks of high-Tc cuprates (see Narlikar 2014) and in respect of the proximity effect too, their observed behavior is found to be equally anomalous. The normal state of N in the case described is believed to be the so-called pseudogap state, which is quite unusual in that (p. 27) even above Tc it carries preformed electron pairs and flux vortices which normally occur only in the superconducting state. High-Tc material with a giant proximity effect has the advantage that a more easily synthesized and thicker normal material can be used for applications. Excitingly, high-temperature proximity superconductivity with Tc = 80 K has been realized (Zareapour et al. 2012) in the topological insulators Bi2Se3 and Bi2Te3 by keeping them in proximity to HTS Bi2212. These results hold considerable promise for future device applications.

1.7.4 Anti-proximity effect and the proximity behavior of nanowires; minigap state

Yet another category of proximity effect, called the anti-proximity effect (APE), was first reported (Tian et al. 2005a, 2006) with Zn nanowires of 40 nm thickness and 6 μm length sandwiched in between two bulk superconducting (BS) electrodes (Sn or In). Surprisingly, the nanowires did not exhibit superconductivity when the BS electrodes were superconducting and they showed superconductivity only after the superconductivity of the electrodes was quenched by an external magnetic field. The effect was, however, not observed for nanowires of thickness t ≥ 70 nm. The observed behavior, being quite contrary to the expected proximity effect, is termed the anti-proximity effect and has since been observed by several researchers (Chen et al. 2009, 2011, Singh et al. 2011) and also for Al single crystalline nanowires (Singh et al. 2010). In the case of the latter, the authors (Singh et al. 2011) find that the critical current of an individual nanowire, contacted by superconducting electrodes, in the absence of magnetic field was significantly smaller than that for normal electrodes, suggesting that the origin of the APE is not linked with the application of the magnetic field.

To date, there seems to be no convincing explanation of the APE, however. But, the above studies carried out (de Horne et al. 2007, Wang et al. 2009, Liu et al. 2012, Stanescu and Sarma 2013) with a host of normal metallic, semiconducting, and superconducting nanowires in contact with superconducting electrodes have yielded some unexpected novel insights into mesoscopic proximity superconductivity. The studies of Wang et al. (2009) on 70 nm thick and 1–2 μm length Au nanowires, with proximity superconductivity induced through contact with a superconducting tungsten (W) electrode, have yielded many new results. While the proximity superconductivity of short wires of length L ≤ 1.0 μm displayed the characteristic zero resistance drop in RT curves, the longer samples with L = 1.9 μm revealed some residual resistance down to the lowest temperature reached. This shows that small lengths of nanowires prepared from normally non-superconducting metals can serve as suitable proximity superconductors for various nanodevice applications. Interesting RT behavior is observed for the samples of intermediate lengths of 1.2 μm, which depicted a resistive drop in two steps (p. 28) corresponding to two Tc values. This was further corroborated through RH curves manifesting two critical fields. The higher values, closer to the contact point with the W electrode, matched well with the reported critical parameters of the W electrode used for the creation of the proximity superconductivity. The lower values are attributed to the formation of the proximity-induced minigap state close to the center of the Au nanowire which constitutes a novel feature of the proximity superconductivity. The gap in the Au wire at the electrode is the same as the bulk gap of W. Low temperature differential magneto resistance behavior of the proximity-connected gold nanowires described shows (He and Wang 2011) uniform oscillatory behavior with a period proportional to the quantum of magnetic flux ϕo. The previously discussed phase slip process occurring in such nanowires is considered responsible for the observed resistance jumps in the oscillatory behavior. This holds relevance to the possible application of proximity superconductivity in quantum computing.

1.8 Synthesis of small superconductors

Several physical, chemical, and metallurgical techniques are currently being used to fabricate nanoparticles, nanowires, and ultra-thin films of a host of different materials. In this section we take a brief look only at some of the processing routes that are commonly used for synthesizing small superconductors.

Historically, Shalnikov (1938) was the first to study the vapor condensation of superconducting Sn and Pb on liquid helium cooled substrate and report their respectively enhanced Tc of 4.7 K and 7.5 K. However, this approach gained significance only in the mid-1950s when Buckel and Hilsch (1954, 1956) found that the structure of quench-condensed films of various superconducting elements like Pb, Bi, In, Sn, etc. could be described either as liquid like amorphous or as lattice like amorphous, both showing a wide deviation from their bulk superconductivity. The latter represents really nothing but the state of nano-grained films while the former manifests the situation of the sub-nano structure, having a short range order, that of a liquid. A modified form of this technique—inert gas condensation—is schematically displayed in Fig. 1.13. The method involves vaporizing the molten metal from heated boats placed in a chamber (evaporation rate of about 0.005 nm/s), under an inert gas pressure. The metal vapor is allowed to condense and quench at the cold finger from where the condensed particles formed can be easily scraped off and studied ex situ without substrate. The particle size, from 100 nm down to a few nm, as determined by XRD (X-ray diffraction) or TEM (transmission electron microscopy), is primarily controlled by the inert gas pressure of the deposition chamber. The method has been widely used for preparing ultra-small particles of low melting point superconducting metals (Brun et al. 2009, Buhrman and Halperin 1973, (p. 29) Bernardi et al. 2006, Li et al. 2003, 2005, 2008) to understand their nanostate behavior. Interestingly, nanoparticles of superconductors can also be easily prepared down to the size of a few nanometers without any substrate by using the popular metallurgical technique of ball milling. Performing the ball milling in tungsten carbide bowls purged with argon, Li and Dou (2006) produced nanoparticles of MgB2 in the range of 64.1 to 2.5 nm. The heat accumulation during the milling process was kept under control by following a continuous cycle of 5 mins milling and 3 mins of rest time.

Small Superconductors—IntroductionClick to view larger

Fig. 1.13 Schematic setup for synthesizing nanoparticles by inert gas condensation.

Among the established PVD (pulsed vapor deposition) routes, both high pressure dc magnetron sputtering (Ayyub et al. 2001, Banerjee et al. 2003, Bose et al. 2006) and the pulse-laser ablation technique (Yang et al. 2011) have been successfully used to synthesize nanostructured films of Nb and V, with average grain diameters respectively down to 5 and 2.5 nm deposited on Si substrates. In the case of the former, the base pressure in the deposition chamber was 10−8 torr and the average grain size was controlled from 60 nm to 5 nm by varying the process parameters, namely the Ar gas pressure from 3 to 100 mtorr and the dc power, from 25 to 200W. The structure essentially comprises weakly connected nano-sized grains, akin to independent nanoparticles.

Novel approaches have been followed to realize nanowires of various superconducting metals, alloys, and compounds. The two that have emerged as the most popular are the (a) suspended molecular template technique (Bezryadin et al. 2000, Hopkins et al. 2005, Zhang and Dai 2000) and (b) porous template technique (Martin 1994, Michotte et al. 2002). The former can be used down to (p. 30) 10 nm diameter/width, while the latter is suitable for growing single crystalline wires of diameter in the range 40–100 nm. The molecular template technique involves sputter depositing the superconducting material on the surface of a carbon nanotube (CNT) (or a nanotube string/bundle), serving as a template, mounted over a trench-like configuration (Fig. 1.14) formed on a tri-layered Si/SiO2/SiN substrate. Since the template CNT is very thin, with diameter of < 3 nm, the diameter of the coated nanowire, after deposition of the desired superconductor, can be kept below 10 nm. The cross-section of the nanowire formed is determined by the width of the CNT and the amount of material deposited and this can be conveniently measured by TEM. The morphology of the deposited wire is controlled by sputtering parameters like gas pressure and dc power. Instead of CNT, the DNA molecule can alternatively be used as a molecular template for the synthesis of nanowires (Bezryadin et al. 2004).

Small Superconductors—IntroductionClick to view larger

Fig. 1.14 Schematic mounting arrangement of molecular template for synthesizing superconducting nanowire.

The porous template technique uses membranes with long aligned arrays of nano-sized pores which are filled with the desired superconductor by electro-deposition to form an array of nanowires. Two types of polymer membranes have primarily been used for this application. These are (a) track etched poly-carbonate (PC), and (b) anodic aluminum oxide (AAO) membrane, which are chosen according to the requirements. The pore density of the latter is in between 1010 and 1011 pores/cm2, while the PC possesses a lesser pore density of about 108/cm2. Both types possess long (a few tens of micrometers in length) parallel arrays of channels free from interconnections which are aligned almost perpendicular to the membrane surface. The pore diameter tends to vary from 40 to 300 nm. Synthesis of metallic nanowires involves growing them inside the aforementioned channels by electro-deposition, using the conventional three-electrode cell and a suitable electrolyte. A gold layer is coated on one surface of the membrane to serve as the cathode with a graphite rod as the anode. The deposition is stopped when the pores are filled. The nanowires formed are single crystalline and their morphology and microstructure can be controlled by varying the deposition parameters, e.g. deposition potential, pH value, and temperature. As the nanowires are placed inside the membrane they are protected from oxidation and the transport measurements can be made in situ by attaching leads on either sides.

Lithography provides yet another fabrication route for synthesizing nanowires on a substrate. The method is more versatile, albeit more expensive than those (p. 31) described above, and also complicated and is generally used for synthesizing complex nanostructures. Both photo (using UV) and electron beam lithography (EBL) are commonly used although the latter has a higher resolution, but is more costly and time consuming. The process for fabricating an array of nanowires is summarized stepwise in Fig. 1.15. Basically, the process involves five steps. A Si chip is coated with a resist (Fig. 1.15a) suitable for photo-lithography or EBL to be used and, in turn, it is covered (Fig. 1.15b) with a mask for the synthesis of nanowires. The structure is exposed to UV radiation or an electron beam (p. 32) (in SEM—scanning electron microscope). The exposed regions are chemically altered to get dissolved in a developer (Fig. 1.15c). Next, the superconducting material whose nanowires are to be made is sputter deposited on the structure (Fig. 1.15d). The substrate structure is then subjected to the lift-off process to remove the left-over resist, which leaves an array of nanowires on the substrate (Fig. 1.15e). Low energy (≈ 1 keV) Ar ion irradiation of the nanowires provides a smooth polishing effect. It is also used for progressively reducing the nanowire cross-section to allow measurements on the same sample with reduced thickness/diameter.

Small Superconductors—IntroductionClick to view larger

Fig. 1.15 Sequential five processing steps (a–e) for fabricating nanowires on Si-substrate by lithography, (based on Singh et al. 2012).

Among the chemical routes the sol-gel method has been successfully pursued (Nath and Parkinson 2006) to synthesize 20 μm lengths of MgB2 nanowires of 50–100 nm diameter. In this, as the first step, a gel was prepared by mixing magnesium bromide and sodium borohydride reagents in the presence of cetyltrimethylammonium (CTAB) and leaving this solution open to the atmosphere for many hours. The pyrolysis of this gel was the next step, carried out for about 5 min at 800oC in an atmosphere of diborane and nitrogen and thereafter it was allowed to cool very slowly to room temperature. The XRD and SEM studies of the black powdery product thus formed confirmed it to be mainly MgB2 in the form of fine nanowires. Hundreds of micrometers long MgB2 nanowires of 50–400 nm diameter have been reported as synthesized by converting long boron nanowires under Mg vapor at high temperature (Wu et al. 2001).

For synthesizing 2D samples of superconducting elements, alloys, and compounds, various PVD techniques like sputtering, laser ablation, thermal evaporation, etc. are commonly used for depositing films of a few nanometers to several tens of nanometers thickness on suitable substrates; the methods are described elsewhere (Narlikar 2014). For example, epitaxial NbN films of thickness 2.5–100 nm have been prepared on MgO substrates using magnetron sputtering (Kang et al. 2011, Marsili et al. 2008). Similarly, high-quality ultra-thin MgB2 films of thickness down to 7.5 nm have been reported on SiC substrates via the hybrid physical–chemical vapor deposition (HPCVD) technique (Wang et al. 2009). Various thin film deposition techniques used for high-Tc cuprates, Fe-based superconductors, etc. are described elsewhere (Narlikar 2014). However, as mentioned in Sec. 1.5.2, in the case of small 2D superconductors, one is particularly interested in the behavior of ultra-thin films down to the lowest 2D limit of monolayer thickness. To prepare such films, the technique most popularly used is molecular beam epitaxy (MBE), which is a sophisticated form of thermal evaporation that allows a single layer of atoms to be deposited at a time. The high quality of the performance is assured by the combined MBE system having various analytical tools like STM (scanning tunneling microscopy), ARPES (angle resolved photoemission spectroscopy), and RHEED (reflection high energy electron diffraction), all functioning at the base pressure of 10−11 torr, for in situ characterization and monitoring of the deposited films.

(p. 33) 1.9 Summary and outlook

In this introductory chapter we saw that as the size of a superconductor was significantly reduced, it exhibited novel properties, displaying a marked departure from the behavior of bulk superconductors. At this level of smallness, which may vary from a fraction of a micrometer to a few nanometers or even shorter, both quantum confinement effects and sample surface effects become significant controlling factors for their physical properties. The bizarre change in the properties and the new phenomena displayed by small superconductors are manifested by all categories of superconductors that may include type-I, type-II, low Tc, high Tc, metallic, ceramic, organic, hybrid, etc. In this chapter we briefly explored the general mesoscopic superconducting features and phenomena that include phase slips, quantum fluctuations, different types of proximity effects, Andreev reflection, strange vortex matter, etc. The fabrication techniques commonly used for the synthesis of small samples and structures have been briefly presented. The upcoming chapters of this handbook will be taking stock in depth of the recent advances in the field of small superconductors, in the form of basic research, materials-specific developments, and the exciting progress in their rapidly evolving device technology. Various new challenges posed and the way they are being effectively met have been assessed and discussed in the reviews and extended articles comprising the present book, with each chapter promising an exciting outlook for the future of small superconductors.

References (Chapter-1)

Abrikosov, A.A. (1957) Sov. Phys. JETP 5, 1174.Find this resource:

Anderson, P.W. (1959) J. Phys. Chem. Solids 11, 26.Find this resource:

Andreev, A.F. (1964) Sov. Phys. JETP 19, 1228.Find this resource:

Ayyub, P., Chandra, R., Taneja, P., Sharma, A.K. and Pinto, R. (2001) Appl. Phys. A 73, 67.Find this resource:

Baelus, B.J. and Peeters, F.M. (2002) Phys. Rev. B 65, 104515.Find this resource:

Baelus, B.J., Cabral, L.R.E. and Peeters, F.M. (2004) Phys. Rev. B 69, 064506.Find this resource:

Baelus, B.J., Kanda, A., Peeters, F.M., Ootuka, Y. and Kadowaki, K. (2005) Phys. Rev. B 71, 140502(R).Find this resource:

Banerjee, R., Vasa, P., Thompson, G.B., Fraser, H.L. and Ayuub, P. (2003) Sol. State. Commun. 127, 349.Find this resource:

Bao, X.Y., Zhang, Y.F., Wang, Y., Jia, J.F., Xue, Q.K., Xie, X.C. and Zhao, Z.X. (2005) Phys. Rev. Lett. 95, 247005.Find this resource:

Bean, C.P., Doyle, M.V. and Pincus, A.G. (1962) Phys. Rev. Lett. 9, 93.Find this resource:

Bean, C.P. and Livingston, J.D. (1964) Phys. Rev. Lett. 12, 14.Find this resource:

(p. 34) Bergeret, F.S., Volkov, A.F. and Efetov, K.B. (2005) Rev. Mod. Phys. 77, 1321.Find this resource:

Bernardi, E., Lascialfari, A., Rigamonti, A., Romanò, L., Iannoti, V., Ausanio, G. and Luponio, C. (2006) Phys. Rev. B 74, 134509.Find this resource:

Bezryadin, A., Lau, C.N. and Tinkham, M. (2000) Nature 404, 971.Find this resource:

Bezryadin, A., Bollinger, A., Hopkins, D., Murphey, M., Remeika, M. and Rogachev, A. (2004) Dekker Encyclopedia of Nanoscience and Nanotechnology DOI:10.1081/E-ENN120013540 p. 3761.Find this resource:

Bezryadin, A. (2008) J. Phys. Cond. Matter 20, 043202.Find this resource:

Blatt, J.M. and Thompson, C.J. (1963) Phys. Rev. Lett. 10, 332.Find this resource:

Blinov, E., Vlasenko, L.S., Kufaev, J.A., Sonin, E.B., Stepanov, Y.P., Takanzev, A.K. and Fleischer, V.G. (1993) J. Exp. Theor. Phys. 76, 308.Find this resource:

Bogoliubov, N.N. (1947) J. Phys. (Moscow) 11, 23.Find this resource:

Bollinger, A.T., Dubuis, G., Yoon, J., Pavuna, D., Misewich, J. and Božoić, I. (2011) Nature 472, 458.Find this resource:

Bose, S., Raychauhuri, P., Banerjee, R. and Ayyub, P. (2006) Phys. Rev. B 74, 224502.Find this resource:

Bose, S. (2007) Size effects in nanostructured superconductors, PhD thesis, TIFR, Mumbai.Find this resource:

Bose, S., García-García, A.M., Ugeda, M.M., Urbina, J.D., Michaelis, C.H., Brihuega, I. and Kern, K. (2010), Nat. Mater. 9, 550.Find this resource:

Bose, S., Galande, C., Chockalingam, S.P., Banerjee, R., Raychaudhuri, P. and Ayyub P. (2009) J. Phys. Cond. Matter 21, 205702.Find this resource:

Bose, S. and Ayyub, P. (2014) Rep. Prog. Phys. 77, 116503.Find this resource:

Bozovic, I., Eckstein, J.N., Virshup, G.F., Chaiken, A., Wall, M., Howell, R. and Fluss, M. (1994) J. Supercond. 7, 187.Find this resource:

Bozovic, I., Logvenov, G., Verhoeven, M.A.J., Caputo, P., Goldobin, E. and Beasley, M.R. (2004) Phys. Rev. Lett. 93, 157002.Find this resource:

Brun, C., Hong, I.P., Patthey, F., Sklyadneva, I.Y., Heid, R., Echenique, P.M., Bohnen, K.P., Chulkov, E.V., and Schneider, W.D. (2009) Phys. Rev. Lett. 102, 207002.Find this resource:

Bruyndoncx, V., Van Look, L., Verschuere, M. and Moshchalkov, V.V. (1999) Phys. Rev. B 60, 10468.Find this resource:

Buhrman, R.A. and Helperin, W.P. (1973) Phys. Rev. Lett. 30, 692.Find this resource:

Buckel, W. and Hilsch, R. (1954) Z. Physik 138, 109Find this resource:

Buckel, W. and Hilsch, R. (1956) Z. Physik 146, 27.Find this resource:

Campbell, A.M., Evetts, J.E. and Dew-Hughes, D. (1968) Phil. Mag. 18, 313.Find this resource:

Chen, Y., Snyder, S.D. and Goldman, A.M. (2009) Phys. Rev. Lett. 103, 127002.Find this resource:

Chen, Y., Lin, Y.H., Snyder, S.D. and Goldman, A.M. (2011) Phys. Rev. B 83, 054505.Find this resource:

Cheng, C.H., Zhang, H., Zhao, Y., Feng, Y., Rui, X.F., Munroe, P., Zheng, H.M., Koshizuka, N. and Murakami, M. (2003) Supercond. Sci. Technol. 16, 1182.Find this resource:

Chiang, Y.N., Shevchenko, O.G. and Kolenov, R.N. (2007) Low Temp. Phys. 33, 314.Find this resource:

(p. 35) Chibotaru, L.F., Ceulemans, A., Bruyndoncx, V. and Moshchalkov, V.V. (2001) Nature 408, 833.Find this resource:

Chibotaru, L.F., Ceulemans, A., Morelle, M., Teniers, G., Carballeira, C. and Moshchalkov, V.V. (2005) J. Mathematical Phys. 46, 095108.Find this resource:

Czoschke, P., Hong, H., Basile, L. and Chiang, T.C. (2003) Phys. Rev. Lett. 91, 226801.Find this resource:

de Gennes, P.G. (1966) Superconductivity of metals and alloys, Benjamin, NY.Find this resource:

de Horne, F.D., Piraux, L. and Michotte, S. (2007) Physica C 460, 1441.Find this resource:

Deo, P.S., Schweigert, V.A. and Peeters, F.M. (1997) Phys. Rev. Lett. 79, 4563.Find this resource:

Deutscher, G. and de Gennes, P.G. (1969) Superconductivity, edited by R.D. Parks, Marcel Dekker, NY, p. 1005.Find this resource:

Dou, S.X., Soltanian, S., Horvat, J., Wang, X.L., Munroe, P., Zhou, S.H., Ionescu, M., Liu, H.K. and Tomsic, M. (2002) Appl. Phys. Lett. 81, 3419.Find this resource:

Dukelsky, J. and Sierra, G. (2000) Phys. Rev. B 61, 12302.Find this resource:

Fleischer, V.G., Laiho, R., Lahderanta, E., Stepanov, Y.P. and Traito, K.B. (1996) Physica C 264, 295.Find this resource:

Gao, J., Chui, T.C., and Tang, W.H. (1999) IEEE Trans. Appl. Superconductivity 9, 1661.Find this resource:

García-García, A.M., Urbina, J.D., Yuzbashyan, E.A., Richter, K. and Altshuler, B.L. (2008) Phys. Rev. Lett. 100, 187001.Find this resource:

Geim, A.K., Grigorieva, I.V., Dubonos, S.V., Lok, J.G.S., Maan, J.C., Filippov, A.E. and Peeters, F.M. (1997) Nature (London) 390, 256.Find this resource:

Geim, A.K., Dubonos, S.V., Grigorieva, I.V., Novoselov, K.S., Peeters, F.M. and Schweigert, V.A. (2000) Nature (London) 407, 55.Find this resource:

Ginzburg, V.L. and Landau, L.D. (1950) Sov. Phys. JETP 20, 1064.Find this resource:

Giordano, N. (1994) Physica B 203, 460.Find this resource:

Grigorieva, I.V., Escoffier, W., Richardson, J., Vinnikov, L.Y., Dubonos, S. and Oboznov, V. (2006) Phys. Rev. Lett. 96, 077005.Find this resource:

Guo, Y., Zhang, Y.F., Bao, X.Y., Han, T.Z., Tang, Z., Zhang, L.X., Zu, W.G., Wang, E.G., Niu, Q., Qiu, Z.Q., Jia, J.F., Zhao, Z.X. and Xue, Q.K. (2004) Science 306, 1915.Find this resource:

Haviland, D.B., Liu, Y. and Goldman, A.M. (1989) Phys. Rev. Lett. 62, 2180.Find this resource:

Hayashi, M., Yoshioka, H. and Kanda, A. (2010) J. Phys: Conf. Ser. 248, 012002.Find this resource:

He, L. and Wang, J. (2011) Nanotechnology 22, 445704.Find this resource:

Hebard, A.F. and Paalnen, M.A. (1990) Phys. Rev. Lett. 65, 927.Find this resource:

Heersche, H.B., Jarillo-Herrero, P., Oostinga, J.B., Vandersypen, L.M.K. and Morpurgo, A.F. (2007) Nature 446, 56.Find this resource:

Holm, R. and Meissner, W. (1932) Z. Phys. 74, 715.Find this resource:

Hopkins, D.S., Pekker, D., Goldbart, P.M. and Bezryadin, A. (2005) Science 308, 1762.Find this resource:

Hwang, E.H., Das Sarma, S. and Stroscio, M.A. (2000) Phys. Rev. B 61, 8659.Find this resource:

Ivanov, V.A., Misko, V.R., Fomin, M. and Devreese J.T. (2003) Sol. State Commun. 125, 439.Find this resource:

Jankó, B., Smith, A. and Ambegaokar, V. (1994) Phys. Rev. B 50, 1152.Find this resource:

(p. 36) Kanda, A., Baelus, B.J., Peeters, F.M., Kadowaki, K. and Ootuka, Y. (2004) Phys. Rev. Lett. 93, 257002.Find this resource:

Kang, L., Jin, B.B., Liu, X.Y., Jia, X.Q., Chen, J., Ji, Z.M., Xu, W.W., Wu, P.H., Mi, S.B., Pimenov, A., Wu, Y.J. and Wang, B.G. (2011) J. Appl. Phys. 109, 033908.Find this resource:

Kasai, M., Kanke, Y., Ohno, T. and Kozono, Y. (1992) J. Appl. Phys. 72, 5344.Find this resource:

Keizer, R.S., Goennenwein, S.T.B., Klapwijk, T.M., Miao, G., Xiao, G. and Gupta, A. (2006) Nature 439, 825.Find this resource:

Kokubo, N., Okayasu, S., Kanda, A. and Shinozaki, B. (2010) Phys. Rev. B 82, 014501.Find this resource:

Kokubo, N., Okayasu, S., Nojima, T., Tomochi, H. and Shinozaki, B. (2014) J. Phys. Soc. Jpn. 83, 083704.Find this resource:

Kokubo, N., Miyahara, H., Okayasu, S. and Nojima, T. (2015) J. Phys. Soc. Jpn. 84, 043704.Find this resource:

Kresin, V.Z. and Ovchinnikov, Y.N. (2006) Phys. Rev. B 74, 024514.Find this resource:

Langer, J.S. and Ambegaoker, V. (1967) Phys. Rev. 164, 498.Find this resource:

Lau, C.N., Markovic, N., Bockrath, M., Bezryadin, A. and Tinkham, M. (2001) Phys. Rev. Lett. 87, 2170031.Find this resource:

Lee, J.J., Schmitt, F.T., Moore, R.G., Johnston, S., Cui, Y.T., Li, W., Yi, M., Liu, Z.K., Hashimoto, M., Zhang, Y., Lu, D.H., Devereaux, T.P., Lee, D.H. and Shen, Z.X. (2014) Nature 515, 245.Find this resource:

Li, S. and Dou, S.X. (2006) Studies of high temperature superconductors Vol. 60, edited by A.V. Narlikar, Nova Science Publishers, NY, 145.Find this resource:

Li, W.H., Yang, C.C., Tsao, F.C. and Lee, K.C. (2003) Phys. Rev. B 68, 184507.Find this resource:

Li, W.H., Yang, C.C., Tsao, F.C., Wu, S.Y., Huang, P.J., Chung, M.K. and Yao, Y.D. (2005) Phys. Rev. B 72, 214516.Find this resource:

Li, W.H., Wang, C.W., Li, C.Y., Hsu, C.K., Yang, C.C. and Wu, C.M. (2008) Phys. Rev. B 77, 094508.Find this resource:

Liu, H., Ye, Z., Luo, Z., Rathnayaka, K.D.D. and Wu, W. (2012) J. Phys. Conf. Ser. 400, 022136.Find this resource:

London, F. and London, H. (1935) Proc. Roy. Soc. London A149, 74.Find this resource:

Marchand, D., Covaci, L., Berciu, M. and Franz, M. (2008) Phys. Rev. Lett. 101, 097004.Find this resource:

Marsili, F., Bitauld, D., Fiore, A., Gaggero, A., Mattioli, F., Leoni, R., Benkahoul, M. and Lévy, F. (2008) Optics Express 16, 3191.Find this resource:

Martin, C.R. (1994) Science 266, 1961.Find this resource:

McCumber, D.E. and Halperin, B.I. (1970) Phys. Rev. B 1, 1054.Find this resource:

Mermin, N.D. and Wagner, H. (1966) Phys. Rev. Lett. 17, 1133.Find this resource:

Michotte, S., Piraux, L., Dubois, S., Pailloux, F., Stenuit, G. and Govaerts, J. (2002) Physica C 377, 267.Find this resource:

Misko, V.R., Fomin, V.M., Devreese, J.T. and Moshchalkov, V.V. (2003) Phys. Rev. Lett. 90, 147003.Find this resource:

Moshchalkov, V.V., Gielen, L., Strunk, C., Jonckheere, R., Qiu, X., Van Haesendonck, C. and Bruynseraede, Y. (1995) Nature (London) 373, 319.Find this resource:

(p. 37) Muralidhar, M., Sakai, N., Jirsa, M., Murakamai, M. and Koshizuka, N. (2004) Supercond. Sci. Technol. 17, S66.Find this resource:

Narlikar, A.V. and Ekbote, S.N. (1983) Superconductivity and superconducting materials, South Asian Publishers, New Delhi, India.Find this resource:

Narlikar, A.V. (2014) Superconductors, Oxford University Press, Oxford, UK.Find this resource:

Nath, M. and Parkinson, B.A. (2006) Adv. Mater. 18, 1865.Find this resource:

Nishio, T., Okayasu, S., Suzuki, J. and Kadowaki, K. (2004) Physica C 412–414, 379.Find this resource:

Nishio, T., An, T., Nomura, A., Miyachi, K., Eguchi, T., Sakata, H., Lin, S., Hayashi, N., Nakai, N., Machida, M. and Hasegawa, Y. (2008) Phys. Rev. Lett. 101, 167001.Find this resource:

Nishio, T., Okayasu, S., Suzuki, J., Kokubo, N. and Kadowaki, K. (2008a) Phys. Rev. B 77, 052053.Find this resource:

Orr, B.G., Jaeger, H.M. and Goldman, A.M. (1984) Phys. Rev. Lett. 53, 2046.Find this resource:

Palacios, J.J. (2000) Phys. Rev. Lett. 84, 1796.Find this resource:

Pearl, J. (1964) Appl. Phys. Lett. 5, 65.Find this resource:

Petrashov, V.T., Sosnin, I.A., Cox, I., Parsons, A. and Troadec, C. (1999) Phys. Rev. Lett. 83, 3281.Find this resource:

Pippard, A.B. (1950) Proc. Roy. Soc. London, A203, 210.Find this resource:

Pontius, R.B. (1937) Nature 139, 1065.Find this resource:

Ralph, D.C., Black, C.T. and Tinkham, M. (1995) Phys. Rev. Lett. 74, 3241.Find this resource:

Ralph, D.C., Black, C.T. and Tinkham, M. (1996) Phys. Rev. Lett. 76, 688.Find this resource:

Ralph, D.C., Black, C.T. and Tinkham, M. (1996a) Physica B 218, 258.Find this resource:

Romero-Bermúdez, A. and García-García, A.M. (2014) Phys. Rev. B 89, 024510.Find this resource:

Rui, X.F., Zhao, Y., Xu, Y.Y., Zhang, I., Sun, X.F., Wang, Y.Z. and Zhang, H. (2004) Supercond. Sci. Technol. 17, 689.Find this resource:

Saint-James, D. and de Gennes, P.J. (1963) Phys. Lett. 7, 306.Find this resource:

Schindler, A., König, R., Herrmannsdörfer, T., Braun, H.F., Eska, G., Günther, D., Meissner, M., Mertig, M., Wahl, R. and Pompe, W. (2002) Europhys. Lett. 58, 885.Find this resource:

Schweigert, V.A. and Peeters, H.M. (1998) Phys. Rev. B 57, 13817.Find this resource:

Schweigert, V.A., Peeters, H.M. and Deo, P.S. (1998) Phys. Rev. Lett. 81, 2783.Find this resource:

Schweigert, V.A. and Peeters, H.M. (1999) Phys. Rev. Lett. 83, 2409.Find this resource:

Shalnikov, P. (1938) Nature 142, 74.Find this resource:

Shanenko, A.A. and Croitoru, M.D. (2006) Phys. Rev. B 73, 012510.Find this resource:

Shanenko, A.A., Croitoru, M.D., Zgirski, M., Peeters, F.M. and Arutyunov, K. (2006) Phys. Rev. B 74, 052502.Find this resource:

Shanenko, A.A., Croitoru, M.D. and Peeters, F.M. (2007) Phys. Rev. B 75, 014519.Find this resource:

Sharifi, F., Herzog, A.V. and Dynes, R.C. (1993) Phys. Rev. Lett. 71, 421.Find this resource:

Singh, M., Sun, Y. and Wang, J. (2012) Superconductors-properties, technology and applications, edited by Y. Grigorashvili, InTech Publications, Croatia & China.Find this resource:

(p. 38) Singh, M., Wang, J., Tian, M., Mallouk, T. and Chan, M. (2010) APS March Meeting.Find this resource:

Singh, M., Wang, J., Tian, M.L., Mallouk, T.E. and Chan, M.H.W. (2011) Phys. Rev. B 83, 220506.Find this resource:

Stanescu, T.D. and Sarma, S.D. (2013) Phys. Rev. B 87, 180504(R).Find this resource:

Teniers, G., Moshchalkov, V.V., Chibotru, L.F. and Ceulemans, A. (2003) Physica B 329–333, 1340.Find this resource:

Tian, M.-L., Wang, J.-G., Kurtz, J.S., Liu, Y., Mayer, T.S., Mallouk, T.E. and Chan, M.H.W. (2005) Phys. Rev. B 71, 104521.Find this resource:

Tian, M.-L., Kumar, N., Xu, S., Wang, J.-G., Kurtz, J.S., and Chan, M.H.W. (2005a) Phys. Rev. Lett. 95, 076802.Find this resource:

Tian, M.-L., Wang, J.G., Kumar, N., Tianheng, H., Kobayashi, Y., Liu, Y., Mallouk, T.E. and Chan, M.H.W. (2006) Nano Lett. 6, 2773.Find this resource:

Tinkham, M. (2000) J. Superconductivity 13, 801.Find this resource:

Upton, M.H., Wei, C.M., Chou, M.Y., Miller, T. and Chiang, T.C. (2004) Phys. Rev. Lett. 93, 026802.Find this resource:

Vajpayee, A., Huhtinen, H., Awana, V.P.S., Gupta, A., Rawat, R., Lalla, N.P., Kishan, H., Lahio, R., Felner, I. and Narlikar, A.V. (2007) Supercond. Sci. Technol. 20, S155.Find this resource:

Von Delft, J. and Ralph, D.C. (2001) Phys. Rep. 345, 61.Find this resource:

Wang, J., Shi, C., Tian, M.L., Qi, Z., Kumar, N., Jain, J.K., Mallouk, T.E. and Chan, M.H.W. (2009) Phys. Rev. Lett. 102, 247003.Find this resource:

Wang, J., Singh, M., Tian, M.L., Kumar, N., Liu, B., Shi, C., Jain, J.K., Samarth, N., Mallouk, T.E. and Chan, M.H.W. (2010) Nature Physics 6, 389.Find this resource:

Wang, J.-G., Tian, M.-L., Kumar, N. and Mallouk, T.S. (2005) Nano Lett. 5, 1247.Find this resource:

Wang, Y., Zhuang, C., Sun, X., Huang, X., Fu, Q., Liao, Z., Yu, D. and Feng, Q. (2009) Supercond. Sci. Technol. 22, 125015.Find this resource:

Wei, C.M. and Chou, M.Y. (2002) Phys. Rev. B 66, 233408.Find this resource:

Wu, Y., Messer, B. and Yang, P. (2001) Adv. Mater. 13, 1487.Find this resource:

Xi, X.X., Zhao, L., Wang, Z., Berger, H., Forró, L., Shan, J. and Mak, K.F. (2015) Nature Nanotechnology 10, 765.Find this resource:

Xu, B., Milošević, M.V., Lin, S.H., Peeters, F.M. and Jankó, B. (2011) Phys. Rev. Lett. 107, 057002.Find this resource:

Yang, C.C., Huang, W.L., Lin, Y.H., Weng, C.Y., Mo, Z.Y. and Chen, Y.Y. (2011) IEEE Trans. Magnetics 47, 3535.Find this resource:

Yeoh, W.K., Kim, J.H., Horvat, J., Dou, S.X. and Manroe, P. (2006) Supercond. Sci. Technol. 19, L5.Find this resource:

Yeoh, W.K., Kim, J.H., Horvat, J., Xu, X., Qin, M.J., Dou, S.X., Jiang, C.H., Nakane, T., Kumakaru, H. and Manroe, P. (2006a) Supercond. Sci. Technol. 19, 596.Find this resource:

Yu, M. and Strongin, M. (1976) Phys. Rev. B 14, 996.Find this resource:

Yu, Y., Aczel, A.A., Williams, T.J., Budko, S.L., Ni, N., Canfield, P.C. and Luke, G.M. (2009) Phys. Rev. B 79, 020511.Find this resource:

(p. 39) Zareapour, P., Hayat, A., Zhao, S.Y.F., Kreshchuk, M., Jain, A., Kwok, D.C., Lee, N., Cheong, S.W., Xu, Z., Yang, A., Gu, G.D., Jia, S., Cava, R.J. and Burch, K.S. (2012) Nature Commun. 3, 1056.Find this resource:

Zgirski, M., Riikonen, K.-P., Touboltsev, V. and Arutyunov, K. (2005) Nano Lett. 5, 1029.Find this resource:

Zhang, T., Cheng, P., Li, W.J., Sun, Y.J., Wang, G., Zu, X.G., He, K., Wands, L., Ma, X., Chen, X., Wang, Y., Liu, Y., Lin, H.Q., Jia, J.F. and Xue, Q.K. (2010) Nature Physics 6, 104.Find this resource:

Zhang, Y. and Dai, H. (2000) Appl. Phys. Lett. 77, 3015.Find this resource: