Moulson A.J., Herbert J.M. Electroceramics: Materials, Properties, Applications

Подождите немного. Документ загружается.

2.7.1). The GL theory produces useful rationalizations concerning super-

conductors in general. One of these concerns the ‘coherence length’ x,a

characteristic of a superconductor and a measure of the distance superconductor

properties extend into a normal material. The coherence length is related to T

the smaller x the higher T

. As examples, for Pb (T

¼7.2 K) x ¼ 87 nm and for

the high temperature superconductor, Bi

CaCu

¼89 K) x ¼ 1.8 nm.

The very small x-values for the ceramic superconductors have signiﬁcant

implications as far as their processing is concerned.

The relative magnitudes of l and x determine whether a superconductor is

Type I or II. If l 4 x the superconductor is Type II.

Understanding and controlling the behaviour of the ‘vortices’ which develop

as an applied magnetic ﬂux penetrates Type II superconductors is critical to their

successful exploitation. If the superconductor is carrying a current, then the

vortices experience the expected electromagnetic Lorentz interaction force and

move. In fact, under an applied magnetic ﬁeld, vortices are continually entering

the material, moving through it and leaving. This motion of a magnetic vortex

induces an e.m.f. which leads to an eﬀective electrical resistance and energy

dissipation; in the specialist literature this resistive eﬀect is referred to as a

‘viscous drag’ opposing the vortex movement. If the process continued and the

temperature exceeded T

the material would revert to the normal state. Therefore

basic research is focused on learning more concerning how the vortex population

is arranged and how it moves. In particular, materials’ engineers seek means of

preventing its movement, that is by ‘pinning’ at irradiation-induced defects,

dislocations and microstructural defects, including grain boundaries, second

phases, etc.

Preventing vortex motion by ‘pinning’ is also critical to the development of the

HTS magnets discussed below.

4.7.3 Ceramic high- T

superconductors (HTSs)

Access to the Handbook of Superconducting Materials edited by D.A. Cardwell

and D.S. Ginley [34] and the Handbook of Superconductivity edited by C.P. Poole

Jr [35] is recommended. The review by R.J. Cava [36], focused on the structures

of the high-T

superconductors, should also be consulted.

A signiﬁcant step in the history of the HTSs was the discovery in 1966 of

superconductivity in the oxygen-deﬁcient perovskite SrTiO

3d

, containing some

barium or calcium substituted for strontium. Although the T

value was very low

(0.55 K), in retrospect it can be seen as the ﬁrst superconducting ceramic. In 1979

a T

of approximately 13 K was discovered for BaPb

0.75

0.25

, which also has

the perovskite structure.

The breakthrough came in 1986 when J.G. Bednorz and K.A. Mu

ller [37]

reported superconductivity with T

¼35 K in the semiconducting ceramic,

222 CERAMIC CONDUCTORS

TEAMFLY

Team-Fly

(La,Ba)

CuO

, which has the layer perovskite K

NiF

structure. The discovery

won them a Nobel prize.

This discovery stimulated what can only be described as frenzied research in

many laboratories throughout the world which quickly resulted in the discovery

of YBa

7d

, the so-called ‘123’ phase, or YBCO, with a transition at a

temperature as high as 93 K. There are also bismuth-based high-T

super-

conductors, which are described below.

The structure of YBCO is closely related to that of perovskite as evident in

Fig. 4.54, which shows the unit cell consisting of three perovskite-type cubes with

Cu ions at the corners and O ions at the centres of the cube edges. The top and

bottom cubes contain Ba and the centre cube Y.

The practical signiﬁcance of the discovery, apart from the fact that it suggested

the possibility of a room temperature superconductor, is that the T

is above the

boiling point of liquid nitrogen (77 K), a cheap refrigerant.

The remarkable progress stimulated by the work of Bednorz and Mu

ller is

illustrated in Fig 4.55.

HIGH TRANSITION TEMPERATURE SUPERCONDUCTORS 223

Fig. 4.54 The structure of YBa

7d

The presence of Cu–O planes is a critically important structural feature of the

HTSs. Two such planes are evident in Fig. 4.54 separated by a plane containing

Y ions and oxygen vacancies. It is accepted that the Cu–O planes are a

requirement for high temperature superconductivity in this class of ceramic. The

structure of the prototype HTS, Bi

CaCu

(designated ‘Bi-2212’) has two

such adjacent planes and a T

of 80 K, and a similar compound, Bi

(Bi-2223) has three sets and a T

of 110 K. There are also thallium-based

analogues e.g. Tl

with a T

of 125 K, but thallium is toxic.

A compound which held out great promise has the precise and surprising

composition, Ca

0.84

0.16

CuO

. It consists entirely of Cu–O planes separated by a

layer of the alkaline earth elements, randomly arranged but in a ﬁxed ratio. In

the light of the experience outlined above, the fact that the compound consists

entirely of Cu–O planes pointed to exceptionally high T

values, if it could be

suitably compositionally modiﬁed, and attempts to achieve this were undertaken.

The task proved diﬃcult until synthesis at high pressure and high temperature

(typically 50 kbar; 1000 8C) was tried. A modiﬁed composition was prepared and,

although not of great signiﬁcance in itself, the work stimulated a strong interest

in high pressure synthesis, an interest which has continued.

The record for the highest T

is held by pressure-formed HgBa

with

134 K measured at normal pressure and 166 K under high pressure.

224 CERAMIC CONDUCTORS

Fig. 4.55 T

against time illustrating the remarkable development following the discovery in

1986 of the high temperature superconductors. (*under pressure) (Adapted from Chu, C.W.

(2000) ‘High temperature superconducting materials: present status, future challenges, and

one recent example – the superconducting ferromagnet’, Physica C, 341–348, 25–30.)

The perovskite structure is, of course, of special signiﬁcance in the

electroceramics context since the ferroelectric perovskites are dominant in the

ceramic capacitor, PTC thermistor and electromechanical transducer industries.

The structure favours the existence of soft modes (low frequency phonons) as

evidenced by its tendency to instability, for example the ferroelectric–paraelectric

transition. Instability is evident in the case of the ‘123’ compound which exhibits

a tetragonal-orthorhombic transition in the region of 700 8C (the exact

temperature depends on the oxygen content). Extensive twinning, very

reminiscent of ferroelectric domain structures, is observed.

4.7.4 The properties, processing and applications of HTSs

Reference to the texts by F.C. Moon [38] and that edited by H. Maeda and K.

Togano [39] is recommended. The Proceedings issue [40] covers the present status

of the materials, processing and applications and the article by D. Larbalestier

et al. [3, pp. 368–77] reviews HTSs for power applications.

Important characteristics

Eﬀorts directed to develop HTSs to commercial products are conﬁned to the

RE–123 compounds, where RE ¼rare earth element, Y, Nd, Sm, etc., and the

Bi-based materials. As far as the former type is concerned the discussion is

conﬁned to YBCO, the most prominent of the RE–123 family. YBCO has the

advantage of a higher J

value at 77 K than the Bi-based materials (10

2

compared to 10

2

) although the comparison is not an easy one to make

since the values are so dependent on specimen form, doping, applied ﬁeld, etc.

More importantly, the Bi-based powders have a ﬂat, clay-like morphology and

this is very signiﬁcant when it comes to fabrication. In contrast YBCO particles

tend to be equiaxed.

For the reasons summarized below, the coherence length x, and irreversibility

ﬁeld H

irr

are two parameters of special signiﬁcance in the processing of HTSs.

The coherence length (x) To attain one of the principal goals, namely to

achieve maximum values of J

and B

, the spacing of features which are not

intrinsically superconducting, such as second phases and grain boundaries

should be 44 x.

The grain boundaries constitute barriers to the supercurrent ﬂow and their

width should, ideally, be less than x. Values of x for the high-T

superconductors

lie in the range 1–100 nm (depending upon applied ﬁeld strength) and so in their

processing close attention has to be given to optimising microstructure and

texture.

HIGH TRANSITION TEMPERATURE SUPERCONDUCTORS 225

The irreversibility field (H

irr

) Figure 4.56 illustrates the events as a magnetic

field is applied to and then removed from a Type II superconductor below its

critical temperature. At applied fields up to H

the response of the material is as

expected for a Type I material. That is the field is excluded (the Meissner effect)

and the material behaves as a perfect diamagnetic. At applied fields above H

there is some flux penetration and this increases until the material becomes

normal at H

As the applied ﬁeld is removed the route the M–H curve takes depends upon

the extent to which the vortices are ‘pinned’. With no pinning the magnetization

remains negative and fully reversible, that is in the opposite sense to the applied

ﬁeld.

If the vortex population is pinned so that it is not in equilibrium with the

applied ﬁeld, then the M–H curve takes the irreversible route as illustrated. The

pinned vortices contribute an M in the same sense as the applied H since they are

the penetrated ﬁeld, and as the applied ﬁeld is reduced the þM manifests itself

just as it does for an ordinary permanent magnet. The pinned ﬁeld is aptly

described as the ‘trapped ﬁeld’.

The irreversibility ﬁeld (H

irr

), which has to be measured experimentally, is an

important parameter since it determines the size of the ﬁnal M developed. As

expected H

irr

depends upon temperature, and irreversibility ﬁeld v. temperature

curves are shown in Fig. 4.57. It is apparent that at liquid nitrogen temperature

YBCO is suited to permanent magnet applications whereas Bi-2223 is not.

Because of the energy dissipation accompanying vortex motion and referred to

earlier, vortex pinning also raises J

In summary, under the usual operating conditions the ceramic HTSs are

characterized by the presence of ﬂux vortices and materials engineers are

226 CERAMIC CONDUCTORS

Fig. 4.56 The reversible and irreversible routes taken by the magnetization of a Type II

superconductor as the applied magnetic ﬁeld is reduced from a value larger than H

learning how to ‘pin’ them. This is necessary because in current-carrying

applications their movement is accompanied by energy dissipation, and in the

production of quasi permanent magnets they constitute the ‘trapped’ magnetic

ﬂux, so determining the strength of the magnet.

Processing high-T

superconductors

Intensive eﬀorts are being made to exploit bulk and tape HTSs for a variety of

applications including magnets for magnetic levitation (‘maglev’) and for energy

storage, current leads and fault current-limiters and cables for power generation,

control and transmission.

As discussed in Chapter 3, the fabrication of bulk oxide ceramics usually starts

with a powder and this can be prepared via solid state reaction (also loosely

referred to as the ‘mixed oxide’ route), or by ‘wet chemical’ routes.

As an example, the preparation of Bi-2223 compound starts with Bi

, PbO,

SrCO

, CaCO

and CuO powders which are well mixed, in a typical case in

acetone in a mill using zirconia media. The mix is calcined at 800–900 8C and the

calcine ground and granulated. To avoid the occurrence of carbon residues in the

grain boundaries, which are said to have a deleterious eﬀect upon J

and T

certain of the processing steps are carried out in pure oxygen or in vacuo [42].

There are a variety of routes described as ‘wet chemical’. For example an acid

(HNO

) solution of the metal nitrates is prepared and the oxalates precipitated

by the addition of oxalic acid and ammonia, to control pH. Because of mixing on

HIGH TRANSITION TEMPERATURE SUPERCONDUCTORS 227

Fig. 4.57 Irreversibility curves for candidate HTS ceramics: at liquid N

temperature T/T

(YBCO) 0.83 and for Bi-2223 0.79. (After D.A. Cardwell [41].)

an atomic scale the reaction to form the oxides can be carried out at a relatively

low temperature of about 700 8C.

‘Spray-drying/calcination’ and ‘spray pyrolysis’ would typically start with a

solution of nitrates of the metals. In the former case, the solution is sprayed into

a warm (250 8C) enclosure where the granules are collected and subsequently

calcined and, in the latter, into an enclosure at a temperature suﬃciently high so

as to simultaneously eﬀect the decomposition and reaction to form the oxides.

Leads and ‘fault current-limiters’ Current leads and fault current-limiters are

the first large scale commercial applications of Bi-based HTSs. Current leads

significantly improve efficiency of liquid helium-cooled magnets.

In principle, components can be formed by any standard ceramic fabrication

route. The densiﬁcation step involves a partial melting and a controlled

recrystallization, and attention has to be given to sintering/annealing atmo-

spheres to achieve the required stoichiometry. Leads have also been formed by

melt-casting.

Because the J

in the a–b plane of Bi-HTSs is several orders of magnitude

larger than that in the c-direction, it is necessary to have the latter normal to the

lead length. The following tape-casting route is one which was developed to

achieve optimum texture.

The organic-based tape-casting slurry consisted of the usual solvents,

plasticizers and binders (see Section 3.6.6) and Bi-2212 powder (approximately

3 mm particle size). The slurry also contained MgO ﬁbres of diameter and length

in the range 5 mm and 100 mm respectively. These ﬁbres formed the substrate onto

which the platelets grew epitaxially during the partial melting–recrystallization

stage. The slurry was tape-cast into sheets, a process which aligned the MgO

ﬁbres in the plane of the sheet and in the casting direction. The tape-cast sheets

were cut into strips, stacked and consolidated by warm-pressing. Short lengths of

silver foil were incorporated into the ends of the stack prior to laminating to

provide the contacts. After the organics were burnt oﬀ the lead was ﬁred in

oxygen when partial melting and recrystallization occurs. The ﬁnal step was an

anneal in argon with 2 vol.% oxygen to adjust the stoichiometry. The result was

a well textured structure as shown in Fig. 4.58(a). Complete leads are shown in

Fig. 4.58(b).

An alternative manufacturing route exploits the powder-in-tube technology

described below, except that a metal with poorer thermal conductivity replaces

the silver matrix.

The function of a current lead is to improve the eﬃciency of liquid helium-

cooled superconducting electromagnets. The current is fed to the coils in three

stages. The ﬁrst comprises cooled copper leads (the resistivity of a metal falls

with falling temperature) which are joined to the HTS leads, connected in turn to

the low temperature superconductor (e.g. Nb

Sn). The heat management is such

228 CERAMIC CONDUCTORS

that the HTS is cooled to 66–77 K by the liquid helium, and the copper by cold

He vapour.

A fault current-limiter is a component which protects power transmission and

distribution systems from surges caused by, for example, a lightning strike,

fulﬁlling a function similar to that of a varistor (see Section 4.3.1). The limiter

should be capable of reducing the fault current to a fraction of its peak value in

less than a cycle. Because for this application the requirement is for low J

, fault

current limiters are already a commercial product. In the case of the lead shown

in Fig. 4.58(b) the fault current is limited to a safe value within 5 ms of the arrival

of the current ‘spike’.

HIGH TRANSITION TEMPERATURE SUPERCONDUCTORS 229

Fig. 4.58 (a) Micrograph of lead section; see text for details (b) Bi-2212 current leads; the

longer lengths are approx. 325300 mm. (Courtesy of the IRC in Superconductivity and

Department of Materials Science, University of Cambridge, UK, ABB Corporate Research,

Switzerland and Advanced Ceramics Ltd., UK.)

Cables via the ‘powder-in-tube’ (PIT) method The approach is especially

suited to processing Bi-2223 into leads and cables for power applications. A

silver or silver alloy tube, filled with the partially reacted precursor powders

formulated to yield Bi-2223, is drawn down to a wire 1–2 mm diameter. The wire

is rolled into a tape if that is the required form, usually with a width-to-thickness

ratio of approximately 10:1. The composite is then heated to 800–900 8C when

the powder partially melts. The recrystallization process is controlled and the

pure Bi-2223 phase develops with large grains oriented so that the Cu–O planes

lie parallel to the silver surface to optimize J

Cables for handling power can be made up from the tapes which are wound

in a manner similar to standard cabling practice with a surrounding corrugated

(to facilitate bending) steel duct for carrying the liquid nitrogen refrigerant.

Lengths totalling in excess of 150 km have been manufactured for electro-

magnet applications. A commercially important objective arising because of the

demand for increased power, is the replacement of existing electric power

transmission cables with high-T

superconducting cables. This is economically

attractive in large and highly developed and congested cities such as Tokyo,

London and New York, where space in the existing underground ducts is

running out and enlarging them would be unacceptably disruptive to day-to-

day city life.

Detailed cost analyses are made before such expensive ventures are

undertaken, but there seems little doubt that the next decade or so will see

HTS cables steadily replacing metallic conductors for power distribution in

locations where the economic beneﬁts are justiﬁed.

The strong inverse dependence of critical current density on magnetic ﬁeld at

77 K is a limitation associated with Bi-2223 tape. This necessitates operation at

lower temperatures for certain applications (e.g. electromagnets) when the cable

is subjected to strong magnetic ﬁelds. This limitation does not arise with the

(RE)BCO system and intensive eﬀorts are being made to manufacture suitable

tapes.

Engineering the required texture in a (RE)BCO tape is a much more complex

matter than in the case of the Bi-2223 tape. One approach is to prepare a nickel

tape with a carefully controlled crystalline surface texture. This forms the

substrate for the development of a correspondingly textured thermally grown

oxide layer which, in turn, forms the substrate for the epitaxial growth of MgO

by electron beam evaporation. The textured MgO serves as the substrate for the

epitaxial growth of the 1 mm thick (RE)BCO via pulsed laser deposition.

High ﬁeld gradient ‘magnetic separation’ is an old technology being

increasingly exploited for the puriﬁcation of liquids. Applications include the

removal of heavy metal ions from laboratory waste water, puriﬁcation of

china clay for the paper industry, recycling waste water used in the steel

industry and recycling of grinding sludge in glass polishing industries, for

example in the manufacture of cathode ray tubes. Bi-2223-based electromagnets

230 CERAMIC CONDUCTORS

are being introduced leading to signiﬁcant increases in eﬃciency. It is anticipated

that the Bi-2223 cables will eventually be replaced with those incorporating

(RE)BCO.

Tapes by dip-coating The process is essentially as described in Section 3.6.6.

A silver or silver alloy tape is passed through a slurry of the partially reacted Bi-

2223 powder dispersed in an organic vehicle. The coated tape is heated at

800 8C in air to remove organics and then at nearly 900 8C when partial melting

and controlled recrystallization takes place.

Bulk magnets Because of the many potential applications, including magnetic

levitation for transport, magnetic separation, bearings and energy storage, the

development of ‘quasi-permanent magnets’ is an important branch of HTS

technology. The practical maximum field developed by the best rare earth

permanent magnets is 1.0 T. From Fig. 4.57 the H

irr

field for YBCO at 77 K is

approximately 5 T, and trapped fields in excess of 3 T have been reported at this

temperature.

The various common processing routes to producing bulk (RE)BCO can be

exploited. Although there seems to be no obstacle to obtaining high T

values

with sintered material, there is with regard to obtaining high critical current

densities J

. The basic reason for this is the presence of ‘defects’ of various kinds,

particularly grain-boundaries, which interfere with current ﬂow. This has led to a

focus on the growth of single crystal forms of (RE)BCO by ‘melt-processing’. By

this approach, crystals possessing values of H

irr

and J

attractive for permanent

magnets are being grown. The technology is reviewed by D.A. Cardwell [41] (see

Fig. 4.59) and by M. Murakami [43].

Melt-processed crystals contain many defects, including dislocations, twin

planes and non-superconducting particles, which act as the essential pinning

centres. The deliberate engineering of a ﬁne dispersion of precipitates of the non-

superconducting ‘211’ phase, Y

BaCuO

, has been a major breakthrough

permitting trapping of the strongest ﬁelds, for example 10 T at 42 K. A penalty is

the occurrence of microcracking around the inclusions due to diﬀerential thermal

expansivities between inclusion and matrix.

During the magnetizing process, and in the ﬁnal fully magnetized state,

complex mechanical stress ﬁelds which can be large enough to cause fracture are

developed. The stresses originate from the Lorentz forces (see Section 9.1.2) on

the ﬂuxoids which, through ﬂux-pinning, are transferred to the material. In the

ﬁnal fully magnetized state, when the magnetizing ﬁeld is removed, the screening

currents associated with the non-uniform ﬂuxoid distribution result in the critical

current density J

ﬂowing throughout the volume. The stresses are the result,

then, of the interaction between J

and the penetrated induction B

. They are

tensile throughout the bulk and any defect of ‘critical’ dimensions (or larger) will

lead to immediate fracture or, in the case of smaller defects, possible eventual

HIGH TRANSITION TEMPERATURE SUPERCONDUCTORS 231