"Improved Method for Producing Ceramic Superconductors" The present invention relates to the production of high temperature ceramic super conductors for a wide range of applications (see e.g. Bowen H.K. Amer. Cεra . Bull S 6_ 1191-96, 1987), and in particular the invention provides an improved production method whereby the orthorhombic phase of the ceramic material is favoured without seriously impairing superconducting properties. The discovery of oxide high temperature superconductors was made in 1986. Initial work was carried out on compounds of La, Ba, Ca and 0. However higher critical temperatures for superconductivity were achieved with the compound Y]_Ba2Cu3θ7_ _x (x 1), [123], and variants formed by substitution of the various atoms (Easterling K.E., Sorrell, C.G., Bourdillon A.J., Dou

S.X., Sloggett G.J., and MacFarlane J.C. Materials Forum, 1988, 1_1 30). These systems are technologically important because the critical temperature for superconduction is above liquid nitrogen temperature making superconducting devices and loss-free electrical transmission comparatively cheap and easy to attain.

The properties of [123] have been extensively studied. The production of superconducting bulk material involves, beside normal ceramic engineering techniques for attaining high density and good mechanical properties, consideration of phase changes that occur on cooling. Sintering normally occurs about 925°C, when a tetragonal phase is formed, depleted in oxygen (x 1). If this material is rapidly cooled the tetragonal phase is retained and it fails to superconduct. However if it is slowly cooled, oxygen diffusion into the material (x 0) accompanies a phase change to the superconducting orthorhombic phase, which occurs at 600-700°C in air. The phase transformation temperature has been found to depend on the oxygen partial pressure, and in pure oxygen

SUBSTITUTE SHEET

at normal pressure, the transformation temperature increases by about 50°C.

This phase transition has several consequences. The soaking time in oxygen is long, costly and inconvenient, particularly in dense material (for good mechanical properties and electrical conduction) where oxygen transport is inhibited by lack of pores. Moreover, the superconducting phase cannot be sintered in closed environments such as in tubes or in hot isostatic presses owing to the lack of oxygen in the local environment. Furthermore the phase transformation induces microcrystalline stresses and twinning owing to changes in lattice parameter and cell volume, and these are believed harmful to superconduction. According to a first aspect the present invention consists in an improvement in a superconducting ceramic material, the improvement comprising including in a ceramic composition prior to sintering an effective amount of an oxidizing agent, to raise the tetragonal-orthorhombic phase transition temperature of the ceramic material relative to the liquid sintering temperature.

Preferably the raise in the tetragonal-orthorhombic phase transition temperature will place this temperature above or in the near vicinity of the liquid sintering temperature.

According to a second aspect the present invention consists in an improvement in a superconducting ceramic material, the improvement comprising including in a ceramic composition prior to sintering, an effective amount of a compound selected from the group consisting of V2θ5,KM θ4 and Pt to act as a denεifying agent.

Preferably, additives will be included in the ceramics of the present invention in amount of from 0.0001 to 10 wt. percent. In the preferred embodiments of the invention the

ceramic material will be of the basic form

Y ₁ Ba ₂ Cu ₃ 0 ₇ . _x

and the addition of impurities will have the effect of producing a ceramic in either of the forms

Y ₁ Ba Cu ₃ V _y 07. _x (a)

YlBa2Cu3_.yVyO7._x (b)

where V is the dopant cation.

In one method according to the invention, impurities are selected as oxidizing agents which raise the phase change temperature of the ceramic material. Examples of suitable impurities which act as oxidizing agents are

Si2θ3, Tiθ2, ^" 2θ5, r 2 ~ ,Fe2θ3,NiO,ZnO KMnθ4, K2Cr2θ7 & KBrθ3

The principles of the invention will now be described in greater detail with reference to experimental examples and the accompanying drawings, in which: - Fig. 1 graphically illustrates superconducting transition temperatures measured from resistivity for Y_Ba2Cu3M2θ7-x for z = 0.06 and

M=Sc,Ti,V,C ,Mn,Fe,Ni,Cu,Zn. Bar represents 10-90 percent drops in resistivity along transition sigmoid. Fig. 2 graphically illustrates normalized (at 100K) resistivity as a function of temperature for [123] doped with Cr,Mn,Ti,V,Fe and Zn. (Ni omitted for clarity). Fig. 3 graphically illustrates superconducting transition temperatures, T _c (midpoint), T _Q (90 percent of sigmoid), T _c χ (10 percent) in doped [123] plotted against dopant valence.

Fig. 4 graphically illustrates calculated free energy changes plotted against temperature for the reactions described by equations (1), (2) and (3). See below. Fig. 5 graphically illustrates electrical resistance

SHEET

verses temperature for V doped [123] with data normalized at 100K;

Fig. δ graphically illustrates resistance versus temperature of Fe doped [123] with the data normalized at lOOK-

Fig. 7 illustrates x-ray defraction patterns for Y ₁ Ba ₂ Cu ₃ Feyθ7_ _x ;

Fig. 8 graphically illustrates calculated free energy charges plotted against temperature for the reactions described by equations (3) and (4). See below.

Fig. 9 graphically illustrates measured a) T _c (x), -D) Jc(-)' ^anc _* ^c ) Vickers hardness (+) of [123] plotted against dopant level.

Figs. 10(a) and (b) depict optical micrographs of [123] containing 0.5 wt percent Pt, showing a) exaggerated grain growth, pores (black) and excess Cu (white) , and b) twin structure.

Fig. 11 shows a scanning electron micrograph of rapidly sintered pure [123] at room temperature placed inside a furnace previously heated to 930°C for sintering.

As stated above, phase transformation problems can be overcome if the transition temperature can be raised above 700°C so that even fast-quenched material will tend to become super conducting, especially if sintering aids or hot pressing are used to lower sintering temperatures. This is to say that, by raising the transition temperature, the orthorhombic superconducting phase can be favoured to exist at higher temperatures despite physical atmospheric and processing limitations to oxygen diffusion. It has been shown by calculations and experimental results that this can be achieved by sintering with oxidizing agents, including 2O5.

The relative stabilisation of the orthorhombic phase can be achieved with small amounts of oxidizing dopant in

^SU B ^S TITUTE SHEE _T

the range 0.0001 to 10 weight percent, but preferably about 0.1 weight percent depending on which oxidizing agent is used. This quantity is small enough not to seriously adversely affect the superconductive properties of the ceramic oxides.

The raw materials for the production of [123] parent material consist of Y2O3 (99.9 percent, Molycorp), BaC03 (99 percent, E. Merck), CuO (98.5 percent, American Chemet) , to which various minor constituent impurity oxides are added as dopants (SC2O3, Tiθ2, V2O5, Cr2θ3, Mnθ2, e2©3, NiO, ZnO). The powders were mixed using [123] as the parent composition, with transition metal (M) oxides serving as dopant additions in Y^Ba2 u3M2θ7_. _x , where z = 0.06. The mixtures were ground with a mortar and pestle for about half an hour, uniaxially pressed at 215 MPa, and sintered three times at 935° in air. The samples were crushed, reground, and re-pressed after the first two firing cycles. Electrical resistivity was measured by the standard 4 probe d.c. technique with a current density of 0.05 A cm"2. Compositional studies were performed with a JEOL JSM 840 scanning electron microscope (SEM), equipped with a Link Systems energy dispersive spectrometer, and further characterisation studies with a Philips type PW 1140/00 X-ray powder diffractometer using CuK radiation.

Table 1 gives the transition temperature (midpoint, T _c ), the width of the transition (10-90 percent, T _c ), and the resistivities at 100K and 300K [ (100K), (300K)] for pure [123] and samples doped with 1 mole percent (cation basis) transition metal. It can be seen that there is a general correlation between the room temperature resistivity with that just above the transition temperature at 100K. A graphical representation of the T _c and T _c data is given in

SUBSTITUTE SHEET

Figure 1, which shows that Fe and Zn give minima in T _c and maxima in T _c .

Figure 2 shows the normalized resistivity data as a function of temperature for these samples. It can be seen that the samples doped with Ti, V Cr, Mn, and Ni show metallic behaviour, whereas those doped with Fe and Zn show semiconducting behaviour and depress T _c dramatically. It is also noted that the nonmagnetic Zn suppresses T _c more significantly than magnetic Fe. Maeno et al. Nature 328 512 and J.J. App. Phys. 2_6 L774 [1987] and Xiao et al. Phys. Rev. B 3_5 8782 [1987] examined the effect of substituting Fe for Cu and observed the same phenomenon. The latter authors proposed that the large paramagnetic moment of Fe suppresses T _c as a result of the breakage of conducting Cooper pairs by d-electron scattering at paramagnetic sites.

These authors have also studied the substitution of _*

Zn for Cu. Maeno et al. [1987] found that Zn had less effect on T _c than F _e and showed metallic behaviour; whereas Xiao et al. [1987] found that Zn had a stronger influence on T _c than Fe, and explained the observed suppression of T _c in terms of the filled 3d shell (10 electrons) of Zn, where Zn provides an extra electron for the Cu shell (9 electrons), thereby filling the antibonding d _x - y band and reducing the density of states at the Fermi level.

X-ray diffraction analysis of the samples showed that those doped with 1 mole percent Ti, V, Cr, Mn and Ni all possessed the orthorhombic structure. The samples doped with 1 mole percent Fe and Zn gave X-ray patterns indicating that the principal phase was not orthorhombic but tetragonal as evidenced by the absence of peak splitting at 47° = 20. Xiao et al. [1987] observed a new tetragonal phase in samples substituted with Fe, Co, and Ga, but they did not provide any data concerning this

SUBSTITUTE SHEET

phase .

Analytical SEM showed that this new phase possesses the same composition as the parent [123] phase. Inasmuch as very small amounts of dopant can cause the [123] phase to precipitate principally as the new tetragonal phase, it is very likely that the latter phase is a mineralized polymorph of the [123] composition. Consequently this phase is probably responsible for the semiconducting behaviou . Concerning the suppression of T _c by transition metals other than Fe and Zn, this has been attributed to the formation of a localized moment due to the effect of magnetic impurity scattering [Xiao et al., 1987], or to the valencies and ionic radii of the dopants (Maeno et al., 1987]. Figure 3 correlates T _c , T ₀ (90 percent of the transition sigmoid) and T _c ± (10 percent) values against (assumed integral) valencies for Ti^ ⁺ , V^ ⁺ , Cr^+, n^ ⁺ and i2+. it can be seen that there is a linear relationship between these values. Consequently it is suggested that the dominant effect of this range of transition metal cations is to suppress T _c by an amount depending on the oxidation potential of these cations. If it is assumed that the transition metal ions replace Cu on the orthorhombic lattice at the Cul sites [Henry et al., - Henry J.Y., Burlet P., Bourret A., Roult G., Bacher P., Jurgens M.J.G.M. and Rossat-Mignod J. (1987), Sol. St. Comm. in press.], then the associated 04 sites whose presence is critical to superconductivity [Henry et al., 1987], will tend to be filled according to the oxidation potential of the dopant cations. In this case a cation with higher oxidation state, such as V ^* 5+, should suppress T _c less than a cation with a lower oxidation state, such as Ni^+.

SUBSTITUTE SHEET

The dopant elements with higher oxidation state may oxidize Cu^ ⁺ to higher valence and provide oxygen to the lattice. It is well known that the transition metal elements with higher valence exhibit multivalence during solid state reaction depending upon the chemical potential of oxygen in the multicomponent system. Thermodynamic calculations computed with Termodata System [Turnbull and Wadeεley, Thermodata System version V, C.S.I.R.O., Port Melbourne 1987] show that the changes in standard free energy with temperature for the following reactions are as plotted in Figure 4.

V2O5 + CU2O = 2V02 + 2CuO (1)

3/2Mnθ2 + CU2O = 2CuO + l/2Mn ₃ 0 (2)

It is interesting to note that, in the case of the

SUBSTITUTE SHEET

V2O5, reaction (1) is in equilibrium at 750°C in one atmosphere of oxygen. Above 750°C, V will have valence 5+ and the reaction will not proceed, while below this temperature, reaction (1) takes place spontaneously leading oxidation of CU2O and reduction of V^÷ to V^ ⁺ . Although CU2O is not in free oxide form, the basic tendency remains the same since [123] is a labile compound. In contrast, for the corresponding reaction (2) CU2O will be oxidized over the entire temperature range. The above consideration can be applied to the orthorhombic - tetragonal transition in the [123] compound. Oxidizing agents sometimes can be used also as sintering aids to improve densi ication. The simultaneous stabilization of the orthorhombic phase is particularly important since in dense material oxygen is inhibited. Dense material has good mechanical properties and 2O5 is a particularly good sintering aid and oxidizing agent partly because of its suitable melting point and thermodynamic properties. High densification is also achieved with Pt, without serious degradation of superconducting properties.

Charge balance requires that the orthorhombic phase (x 0) unit cell contains a Cu^+ ion (symmetry and ionic radii suggest it is on the Cul site of Henry et al.), while the tetragonal phase (x 1) contains a Cu^ ⁺ ion. The C ^ ⁺ can be oxidized to Cu ²⁺ and consequently Cu ³⁺ can be stabilized since a system containing all three valencies is not thermodynamically stable. This model is also supported b thermodynamic calculations for this transition using CUO/CU2O equilibria assumed for pure [123]. The free energy changes for this reaction are plotted against temperature in Figure 4 for the reaction:

CU2O + l/20 ₂ = 2CuO (3)

in air and in argon (O2 pressure assumed 10"^ atm) . It is seen from the plots that the Cu^÷ valence state is stable in air up to 930°C and, in Ar or 2, up to 430°C, while the Cu-*- ⁴ " state is stable at higher temperatures. This explains why the tetragonal-orthorhombic phase transition can be raised from ό70°C in air to 750°C in 1 atm O2 [Uchita εt al., Proc. Int. Top. Conf on Super, June 29 - July 1, Beijung 1987} or lowered to 500°C in Ar or 2. It also shows that it is possible to stabilize the orthorhombic phase internally through the additions of high oxygen potential dopants. These include not only V2O5 but K nθ4, K2Cr2θ7, KBrθ3 and indeed by the same chemical argument all oxidizing agents.

Hereinafter given are examples illustrating the present invention.

EXAMPLE 1 2O5 additions over a range of levels; YlBa2CU3Vyθ7_. _x , 0.06 y 0.3.

Samples of material containing varied amounts of V2O5 additions in the parent [123] material were prepared as befor and the sintered material was slowly cooled in air at 60°C/hour. Measurements were made by electrical resistance, X-ray diffraction and Vickers hardness. In all cases the superconducting transition temperature was well above the temperature of liquid nitrogen (LNT), though there was some reduction in T _c with increasing concentration, y (table 2, figure 5), and also an increase in the width of the transition. X-ray diffraction patterns showed the orthorhombic structure as the main phase up to y = 0.06. SEM and energy dispersive X-ray analysis revealed that excess copper oxide phase increased with vanadium concentration, showing that V was introduced into the lattice substitutionally for Cu. Vickers hardness showed a very significant increase with

SUBSTITUTESHEET

y and this corresponds with an observed increase in density.

TABLE 2

V2O5 substitutions over a range of levels, ^γ l ^Ba 2 Cu ₃ _yV _y θ7_. _x , 0 y 0.6

Samples of material were prepared as before with varying concentration y=0, 0.045, 0.1, 0.3 and 0.6. In the present case, CuO excess was virtually eliminated by the substitutional (instead of additional, see formulae) impurity. The sintered material was cooled in four ways: i) slow cooled in air at 60°C/hour, ii) quenched in air, with specimen resting on an

- 12 -

iron plate, iii) slow cooled in oxygen at 60°/hour and iv) quenched in oxygen with specimen resting on an iron plate. The effects of these methods of specimen preparation are listed as follows. i) X-ray diffraction showed that all five specimens were orthorhombic. Electrical resistivities were similar to those described above for V2O5 additions. This shows that in both substitutional and additional doping, the V enters the lattice, ii) X-ray diffraction showed tetragonal structures for y = 0.045 and 0.1 and orthorhombic structures for y = 0.3 and 0.6. Smaller amounts of dopant failed to stabilize the orthorhombic phase up to 940°C in air. iii) Specimens slow cooled in oxygen had the same properties as those slow cooled in air. iv) X-ray diffraction showed that all specimens were orthorhombic (Figure 6a) so that this phase is more easily achieved in oxygen than in air. While dopant levels corresponding to y=0.3 were needed to stabilize the orthorhombic phase in air; in oxygen very small amounts of dopant (0.7 weight per cent) stabilized the orthorhombic phase. ^v 2°5 ^a ~s° had a densifying effect. The observed increase in hardness and density shows that V2O5, besides its oxidizing power, is also an efficient sintering aid. In its pure form, V2O5 has a melting point at 670°C so it is liquid at sintering temperature. These densifying properties are of particular value in conjunction with the oxidizing power of V2O5. We have observed a linear reduction in pellet size of 20%, without need for five particle size and large pelletizing pressure.

SUBSTITUTE SHEET

- 13 -

EXAMPLΞ 3

lBa2Cu3Feyθ7. _x ; 0.02 y 0.2

In similar experiments with Fβ2θ3 substitutions, the orthorhombic phase was destablilized and the tetragonal phase was promoted. The super conducting transition temperature fell rapidly and nearly to below LNT (figure 6, table 2) as the concentration, y , increased. Also shown in table 2 are resistivity measurements taken above the superconduction transition temperature at 300K.

The increase in these resistivities correlates with the deterioration in superconduction properties. X-ray powder diffraction patterns showed that (figure 7) for y 0.002 the structure remains essentially orthorhombic as evidenced by the splitting of the peaks at 20 = 47°, 57° and 68°; while for y 0.045 the tetragonal phase dominates as shown by the single peak at 47° and 58°. The tetragonal phase caused by the Fe substitution accounts for the large depression in T _c . Lattice parameter measurements (see table 3) show systematic changes in c and a corresponding to a slight increase with y _x in unit cell volume in the tetragonal phase.

TABLE 3 LATTICE PARAMETERS FROM X-RAY DIFFRACTION SAMPLE Crystal a b c V

YBa2Cu3. _y Feyθ7_ _x System (A) (A) (A) (A ³ ) y=

0 '0' 3.83 3.87 11.70 173.4

0.02 '0' 3.84 3.87 11.73 174.3

0.045 'T' 3.87 11.72 175.5

0.1 'T' 3.86 11.80 175.8 '0' = Orthorhombic 'T' = Tetragonal

The substitutional nature of the iron impurities was confirmed by Mossbauer spectroscopy. This showed sites of 4-fold coordination as for the Cul site (Henry et al, 1987) , Thermodynamic calculations computed with Thermodata System (Turnbull and Wadesley, (1987) Thermodata System Version V. C.3.I.R.Q., Port Melbourne, Victoria see figure 10) show that, compared to the oxidizing effect of V ³⁺ on Cu^÷ described above; Fe ³⁺ cannot oxidize Cu ⁺ to Cu^ over the range of temperatures from room temperature to sintering temperatures. In figure 8 the free energy for the reaction:

3Fe2θ3 + CU2O - 2CuO + 2Fe2θ4 (4)

is compared with the free energy of reaction (3) as functions of temperature. These explain why Fe tends to promote the tetragonal phase.

EXAMPLE 4

Pt Additions

The orthorhombic YιBa2Cu3θ7_. _x [123] superconductor is now accepted to have a superconducting transition temperature, T _c , of about 93K. Although very high values for the critical current density, J _c , have been predicted on the basis of measurements on thin films of the material (Chaudhari, Koch, Laibowitz, McGuire and Gambino, (1987), Phys. Rev. Lett. 58 2684-2686), such values have not as yet been realized in the bulk sintered material. Many attempts have been made to increase both T _c and J _c in bulk materials by "doping" or by metal substitution with suitable impurities. These can be used to promote densification through liquid phase sintering or through the formation of defects. They could also be used

-

to raise the phase transition temperature of the (superconducting) orthorhombic - (high temperature) tetragonal phase (Henry, εt al), Most of the effort has shown only minor success, for various ^' reasons including instability of materials and contact problems in determining J _c . We report here the effects produced by very small amounts (0-0.5wt percent) of finely divided Pt on the properties and structure of [123]. There is an optimum concentration of Pt { 0.2wt percent) for which T _c remains unchanged while J _c is significantly increased and the density and hardness of the material increase. Higher concentrations of Pt cause a slight drop in T _c while both the density and hardness remain high.

Kovachek, εt al (Kovachek, Vlakov, Nenkov, Lovchinov, Gospodinov, Stojanova and Dimitrov (1987), Proc. Int.

Topic Meet. High Temp. Superc, Being in Press) reported the effect of Pt addition by diffusion into [123] during _^ sintering. Samples were prepared with various dopant concentrations of 0.2, 0.3 and 0.5wt. percent Pt, and the mixes were milled for 24 hours in nylon ball mills with copper milling media and hexane. After ball milling the particle size of the parent material was reduced to lum. The milled powders were isostatically pressed at 210 MPa, and the pellets were sintered at 925°C for 6 hours followed by slow cooling of flowing oxygen at 60° per hour.

Electrical resistance measurements were made by the conventional 4 probe d.c. technique, with a current density of 0.05 A cm"2. The J _c was measured by increasing the current through the sample until the voltage drop across the specimεn ( 15 mm x 4 mm x 5 mm, necked to 1 mm x 1 mm at the centre) exceeded luV. Electrical contacts were made by sintering silver onto the ceramic superconductor and soldering with In. Microhardness values (average of 10 indentations) were

obtained with a Lεitz Durimet hardness tester using a load of lOOg and a load time of 20s.

Figure 9 shows the effects of varying the Pt concentration on T _c , on. J _c measured at 77K and on

5 Vickers hardness, H _v . It may be seen that J _c is improved nearly sevenfold from 37 to 250 Acm"-^ with a small decrease in T _c from 92.3 to 90.OK, when Pt is present at the 0.2 wt percent level. While T _c , and H _v demonstrate regular trends, J _c exhibits a maximum. 10 Consequently it may be concluded that Pt additions have a negative effect on the electrical properties and a positive effect on the mechanical properties. At the 0.2 wt percent level, improved intergranular contacts result in improved J _c ; but at higher dopant levels other

15 effects dominate.

X-ray diffractographs (Philips PW1140/00 diff actometer using CuK radiation) and microstructural analysis (JEOL JSM 840 operated at 15KV and equipped with a Link Systems energy dispersive spectrometer at 40°

20 take off angle) showed that the materials were of uniform cationic composition except for a small fraction of excess CuO which resulted from ball milling and, in the case of the 0.5 percent addition, a minority phase of Pt:Ba:Cu in the ratio 1:4:2. The fact that Pt compounds were not

25 observed in the samples doped with 0.2 and 0.3 wt percent Pt suggests solid solution or a thin grain boundary phase. The latter was not observed in the transmission electron microscope (JEOL 2000 FX) .

Optical micrography revealed several reasons for the

30 improvement in J _c and H _v . Increased density has been achieved along with exaggerated grain growth (Figure 10(a), (b)). The porosity of the specimens was measured from the pore area (assumed to be isotropic) in such images, and found to be as small as 5 and 6 percent for

35 0.2 and 0.5 wt percent Pt dopings respectively. This is

SUBSTITUTE SHEET

considerably lower than the porosity of 39 percent found in pure [123] prepared in a similar way. These effects are a likely consequence of either liquid phase sintering or of the presence of defects caused by Pt dissolution. The images, observed with crossed polarizers on unεtched specimens, show the first evidence of such dense and widespread twinning. The twinning occurs in many orientations and spacings with preferred orientations frequently occurring at angles of 90° within single grains. The contrast is attributed to variations in dielectric constant consistent with the anisotropy of the [123]. The microstructure is reminiscent of the twinned ferroelectric domains observed in the perovskite structured BaTiθ3 (Hooton and Hertz, (1955) Phys. Rev. 9_8 409), twinning indicates a history of stress, which could be a consequencε of the dεnsification, or of the tetragonal - orthorhombic phase transformation occurring at 700°C (Sawada, Iwazumi, Saito, Abe, Ikeda and Yoshizake, (1987), Jap. J. App. Phys. 2S_ 1054-1066). ^' During thε phase transformation the lattice parameter b increases while a and c decrease (Easterling, Sorrell, Bourdillon, Dou, Slogett and Macfarlanε, (1987) leading to a slight reduction ( 1 percent) in volume in spite of absorption of oxygen. The high density was shown however to have some disadvantage, εspecially in the most strongly doped specimen. The central region of the pellet was found to have properties different from those in the outer region, in spite of possessing a similar microstructure but with a lower twin density. The central region of a polished section was shinier and insulating at room temperature, while the outer region was duller and conducting. The latter effect strongly suggests that the diffusion of oxygen required for the tetragonal - orthorhombic phase transformation had been inhibited in the cεntral region.

SUBSTITUTE SHEET

The central region was also found to be unstable after three weeks exposure to the atmosphere, with surface cracks being created.

These observations (especially if twins are detrimental to superconduction) show how yet further improvements to electrical and mechanical properties might be made by increasing the stability temperature of the orthorhombic YιBa2C 3θ7_ _x up to temperatures in the sintering range. This can in principle be done by both reducing the sintering temperature with liquid phase dopants (such as Pt) and by increasing the transformation temperature with dopants (such as V, as shown by thermodynamic calculations consistent with experiment - as described above) that oxidize the [123] at high temperature, thereby stabilising the orthorhombic phase.

SUBSTITUTE SHEET