Platinum Metals Rev., 1996, 40, (2), 64
Flux Pinning by Platinum and Rhodium in High Temperature Superconductors
Platinum Additions to YBCO Thick Films Improve Critical Current Densities and Superconducting Properties
Screen printed thick film technology is recognised as an inexpensive and effective means for the production of superconductors which have potential applications in the electronics and microwave device industries. Additions of the platinum group metals have been made to superconducting thick films of YBa2Cu3O7-δ, and these have shown significant improvements, relative to melt processed films, in both their superconducting and physical properties, particularly in their critical current densities. A possible mechanism to explain the improvements is discussed.
The field of cryogenic research was begun and greatly advanced in 1908 with the discovery by Hieke Kamerlingh Onnes (1850–1926) of the liquefaction of helium. He used this discovery to investigate the electrical resistance of metallic elements and observed that the resistance of many metals fell dramatically to zero as the temperature was reduced below the critical temperature (Tc) of the material. This was to be the criteria for a new state of matter, the superconducting state. In 1933 another inherent characteristic of the superconducting state was discovered by W. Meissner and R. Ochenfeld. They found that a material in the superconducting state will act to expel magnetic flux from its bulk.
Major obstacles, such as the cost of liquid helium and the fact that the presence of any external magnetic field destroys superconductivity in a material, however, hindered research in and development of these materials. The dependence of the superconducting state upon temperature, the applied external magnetic field and the ability of the material to carry electrical current is illustrated in Figure 1.
The critical current density Jc) of a material is an extremely important parameter for electrical applications of superconductors. When a transport current is applied to a superconducting material, a magnetic field is generated in the superconductor. In Type I superconductors, if this field, coupled with external fields, exceeds the critical field of the material (Hc), then the superconductivity will be destroyed. When a magnetic field is applied to Type II superconductors, such as YBa2Cu3O7-δ (YBCO) flux quanta penetrate the material as a regular array of vortices. When transport currents are then applied they act to move these vortices, thus destroying the superconductivity. The development of microstructures made of YBCO materials has enabled engineers to increase the Jc within Type II materials by introducing “flux pinning” centres into the material. Many microstructural anomalies, such as oxygen inho-mogeneities, twins, stacking faults, cracks, dislocations and second phase particles, have been found to inhibit the possible motion of magnetic flux vortices in these materials.
Platinum Group Metal Additions
Platinum has already been widely exploited in the development of superconductors, even though it exhibits no superconducting properties itself. The principal utilisation of platinum has involved the use of its non-reactive properties for the processing and characterisation of ceramic superconductors in crucibles, substrates and buffer layers, or as electrodes and electrical contacts.
More recently, it has been found that processing YBCO in the presence of platinum can have a dramatic effect on the microstructure and can improve the superconducting properties of the material produced. It has been observed that platinum additions refine the morphology and distribution of the inherent precipitates of non-superconducting Y2BaCuO5 (211) within bulk melt processed YBCO. On refining, the precipitates become gready reduced in size and are distributed homogeneously throughout the matrix (1). This refinement of the microstructure increases the magnetic hysteresis properties due to flux pinning, and hence increases Jc to values estimated as 2 × 104 A/cm2, at 77 K and a magnetic field of 1 T. Rhodium has also been observed to give the same remarkable effect and it was suggested that the platinum group metal addition inhibits the growth of the 211 on processing.
Other work has lead to the theory that platinum plays an important part in the heterogeneous nucleation of 211, which is formed by peritectic decomposition of YBCO on melt processing (2, 3). The reactions which occur between platinum and the YBCO system have also been investigated, and it was found that platinum reacts to form many compounds, including Ba4CuPt2O9 (0412) which is considered to act as a possible heterogeneous nucleation site for 211 (4). The synthesis of the 0412 phase and its addition to YBCO have been undertaken and improvements in properties comparable to those caused by platinum were observed (5).
However, other investigators have argued that platinum does not react with the YBCO system on heating, but dissolves into the peritectic liquid phase, thus altering the diffusion rate of yttrium atoms in the system and/or changing the 211/liquid interfacial energy (6–9). The precise mechanism for the microstructural refinement which occurs on addition of platinum is, therefore, still widely debated.
This paper will present results from the systematic doping of YBCO thick films with platinum group metals, discuss the improvements in superconducting properties achieved and suggest the possible mechanism for 211 refinement in the YBCO system.
Developments in Superconducting YBCO Thick Films
The application of bulk ceramic superconductors in the microelectronics and microwave industries is expected to be somewhat limited. However, these industries are already beginning to benefit from the introduction of different forms of these materials which have highly desirable intrinsic properties, such as extremely low resistive loss characteristics. Many research groups and industrial sources have fabricated high quality thin film materials, which have very high current densities and Tc values that are equivalent to those of the bulk material. However, the processing techniques used are expensive and would be very difficult to incorporate into the production of integrated circuits, etc. Thick film technology may offer important advantages in the production of large area coatings on, for example, magnetic shields and/or complex shapes, such as microwave components. These can be produced economically and with relative ease, but until recently their properties (Jc and Tc) were inferior to those of their polycrystalline bulk and thin film equivalents.
While material developments have been made in bulk YBCO by improvements in processing and the implementation of many doping additions, similar improvements in YBCO thick films have not been so dramatic. Previous studies of the superconducting properties of YBCO thick films deposited on a variety of substrates have shown that the film characteristics are highly dependant upon the substrate material, the YBCO precursor material and the processing conditions of the film. YBCO films processed on alumina substrates have yielded poor properties with Tc values around 90 K and Jc values in the region of 100 A/cm2 (10).
The properties of YBCO thick films can be significantly enhanced, however, by processing them on yttria-stabilised zirconia (YSZ) substrates, and values of Jc = 1.5 × 103 A/cm2 and Tc = 92 K are easily achieved (11). Also, certain additives to YBCO films have led to increased flux pinning characteristics; for example BaSnO3 additions have increased Jc values up to 2 × 103 A/cm2, and controlled amounts of Ag2O have increased Jc values up to 3 × 103 A/cm2. However, these mechanisms are different to that of platinum group metal additions.
The work reported here was undertaken to gain an increased understanding of the influence of microstructure on the properties of YBCO thick films and the mechanisms by which platinum group metal additions change micro-structure, and to improve the power handling capabilities of YBCO thick films by improving the flux pinning characteristics of the material.
Results of Experimental Work
The first series of experiments which were performed involved the addition of platinum sponge (2–250 μm) to YBCO powder prepared by solid state calcination at 900°C for 24 hours. Previous studies had shown this YBCO precursor material was good for the production of thick films on YSZ substrates (12) and the addition of large particulate platinum sponge was intended as a feasibility study.
The addition of platinum to the YBCO material was found to have a dramatic effect upon the microstructure and the superconducting properties of the processed films. As the concentration of platinum in the system was increased from 0 to 5 weight per cent, the morphology, size and distribution of the inherent non-superconducting 211 precipitates were observed to become refined. Increased volumes of CuO were also seen within the films at 123 YBCO grain boundaries, see Figures 2 and Figures 3.
Quantitative analysis of phases by energy dispersive X-ray analysis (EDX) in the electron microscope has shown that the larger 211 precipitates were composed of yttrium, barium, copper, platinum and oxygen in proportions consistent with the reported supposition that platinum reacts to form a 0412 compound, which then acts as a heterogeneous nucleation site for 211 precipitates (4). Electron microprobe analysis of the same sample section revealed the possible coring of platinum within the largest particles. Coring is the increase in concentration of a particular element towards the centre of a particle.
Additions of rhodium sponge were also observed to increase the superconducting properties by a similar refinement of 211 precipitates, although not to the same extent as with platinum. Maximum improvements in superconducting properties occurred for additions of 0.5 weight per cent: Jc increased from 1.2 × 103 A/cm2 to 2.2 × 103 A/cm2 for platinum, and from 1.2 × 103 A/cm2 to 2.0 × 103 A/cm2 for rhodium, both held at 77 K and zero applied field.
Following these initial developments, additions of finer particulate platinum (0.8–1.2 μ m) were made to the same YBCO precursor material, in order to refine further the morphology of the 211 precipitates and to increase the super-conducting properties. As the concentration of platinum in the system was increased, the texture of the films was observed to diminish considerably and the concentration of non-superconducting second phases increased dramatically. It is thought that the increased surface area of platinum, as platinum is added, and the subsequent increased reaction to form barium-rich Ba4CuPt2O9 is responsible for the larger quantity of copper-rich phases seen in the films. Transport Jc values measured within these films were not observed to exceed 500 A/cm2. The poor superconducting properties are primarily due to the presence of high volumes of these second phases, which inhibit the percolation of supercurrents through the films.
In order to try and prevent the generation of the detrimentally high volumes of copper-rich second phases during processing, a higher purity commercial YBCO precursor material was utilised (Rhône Poulenc ‘Superamic Y123’). The superconducting properties of control films made of ‘Superamic Y123’ were found to be comparable to those of films processed from the YBCO precursor described earlier. Platinum additions of size (0.8–2.5 μm) made to this material were seen to refine dramatically the microstructure of the films – with only limited formation of copper-rich phases – and to improve markedly the Jc characteristics of the material. The variation in transport Jc with platinum concentration is shown in Figure 4.
The optimised doping level of 0.15 weight per cent coincides with a maximised reduction in 211 precipitate size, while maintaining relatively large 123 grains and good c -axis texture.
The addition of platinum having smaller particle sizes (0.5–1.2 μm) was observed to increase the Jc characteristics even further. Transport Jc values up to 6 × 103 A/cm2 were readily achieved with 0.1 weight per cent platinum. Examination of thin film specimens by transmission electron microscopy showed that the presence of increased dislocation densities was associated with refined 211 precipitates of greater surface curvature, see Figures 5 and Figures 6. This increase in dislocation density is thought to be the main factor contributing to increased flux pinning.
As previous studies have suggested, the refinement of 211 appears to be due to the formation of 0412 nucleation sites (4); the 0412 phase was synthesised by a solid state calcination route and as an addition to YBCO. Films subsequently processed from the doped material were found to contain highly refined and anisotropic 211 precipitates, as do platinum-doped specimens. Transport critical current densities greater than 7.0 × 103 A/cm2 were readily achieved with the addition of approximately 0.3 weight per cent of the synthesised 0412 nucleation phase.
The addition of controlled concentrations of platinum and/or Ba4Cu1+xPt2-x.09-z particulates to precursor powders of YBCO, subsequently used for the deposition of thick film superconductors, can effectively improve the critical cur-rent characteristics of the processed films. Microstructural observations have shown that the size, morphology and distribution of Y2BaCuO5 precipitates within the films become highly refined on doping with platinum and Ba4CuPt2O9 (0412). The smaller precipitates in the doped films are observed to be relatively spherical in form, while the larger ones are acicular. An increased number of defects is observed at the 123/211 interface, associated with increased surface curvature of these precipitates. It is proposed that platinum based additions act as nucleation sites for the deposition of refined 211 particles, which in turn act as nucleation sites for dislocation networks and aid flux pinning in the films.
- 1 N. Ogawa, I. Hirabayashi and S. Tanaka, Physica C, 1991, 177, 101
- 2 M. Yoshida, N. Ogawa, I. Hirabayashi and S. Tanaka, Physica C, 1991, 185 – 189, 2409
- 3 M. Morita, M. Tanaka, S. Takebayashi, K. Kimura, K. Miyamoto and K. Sawano, Jpn. J. Appl Phys., 1991, 30, (5A), 813
- 4 Y. Saito, T. Shishido, N. Toyota, K. Ukei and T. Sakai, J. Cryst. Growth, 1991, 109, 426
- 5 H. J. Park, H. W. Kim and J. T. Song, J. Mater. Sci.: Mater. Electron., 1993, 4, 77
- 6 C. Varanasi and P. McGinn, Pkysica C, 1993, 207, 79
- 7 C. Varanasi, P. McGinn, V. Pavate and E. P. Kvam, Physica C, 1994, 221, 46
- 8 C J. Kim, K. B. Kim, H. W. Park and G. W. Hong, Physica C, 1994, 250, 153
- 9 T. Izumi, Y. Nakahara, T. H. Sung and Y. Shiohara, J. Mater. Res., 1992, 7, 801
- 10 T. C Shields and J. S. Abell, Supereond. Sei. Technol., 1992, 5, 627
- 11 Y. J. Bi, F. Wellhofer, M. J. Day and J. S. Abell, Mater. Sci. Eng., 1993, B21, 19
- 12 M. J. Day, S. D. Sutton and J. S. Abell, Cryogenics, 1993, 33, 113