High-temperature superconductors (HTS) such as RE-Ba2Cu3O7–x (REBCO, where RE represents a rare- earth element) and Bi2Sr2Ca2Cu3O10–x (Bi-2223) are available only in rectangular tape forms, making it geometrically difficult to wind them into coils, especially for large magnets where multi-kiloamp conductors are needed. So low-temperature superconductors (LTS) such as Nb-Ti and Nb3Sn, which can be formed as round wires, dominate the market for magnetic resonance imaging, nuclear magnetic resonance, and particle accelerator equipment. Now David C. Larbalestier of Florida State University (FSU), Christian Scheuerlein of the European Organization for Nuclear Research (CERN), and their colleagues have announced round, multifilamentary Bi2Sr2CaCu2O8–x (Bi-2212) HTS wires with a very high critical current density of 2500 A mm–2 at 20 T and 4.2 K. These wires exceed a vital barrier that has restricted HTS rectangular tapes to a small subset of magnet applications. The fact that high current density can be obtained even when many high-angle grain boundaries (HAGBs) are present suggests an important change in the dominant paradigm of making these conductors.
“The initial excitement in the field of superconductivity always tends to be about discovering a higher transition temperature,” said Larbalestier, the lead author of an article published in the April issue of Nature Materials (DOI: 10.1038/NMAT3887; p. 375). But applications cannot be satisfied by high T c alone. It is necessary to have a high critical current density in a useful field and temperature range.
Bi-2212 was a neglected material when Larbalestier and Eric Hellstrom, both of the National High Magnetic Field Laboratory in the university, revived it about five years ago. Although it was the first HTS material to be considered for conductor use, it was quickly superseded by Bi-2223, with a 20 K higher T c (110 K). Ultimately, the less anisotropic compound YBa2Cu3Ox (YBCO) became the favored HTS. But Bi-2212 has one intriguing property: it can develop high critical current density in round and macroscopically untextured form—but only in very short wires, not in the lengths needed for magnets.
(a) Cross sections of the round wire Bi-2212 multifilament conductor before reaction, and (b) electron backscatter diffraction images of the grain structure after reaction at a high current density state. Major misalignments shown by the red-colored grains exist without preventing high filament current densities.
Credit: Jianyi Jiang and Fumitake Kametani.
Larbalestier said, “We found that the critical problem was not the presence of many high-angle grain boundaries in the wire but rather lots of bubbles generated by residual gas,” which form during processing of the Bi-2212 powder.
Until now, the commonly held theory for low critical current densities in HTS was that the presence of many HAGBs in the superconducting phase blocked the supercurrent. Hence, all processing was devoted to producing a highly textured superconducting phase, with a minimum of HAGBs, which required the tape form. Tape-form conductors of REBCO and Bi-2223 followed. The high-performing Bi-2212 wires produced by Larbalestier and his colleagues in the present work manage to keep the supercurrent flowing despite the presence of many HAGBs.
Gas bubbles are an inherent consequence of the need for powders to slide over one another during deformation. Sliding means that the terminal mass density of the powder inside its metallic sheath is only about two-thirds, with the remainder being gas, generally air. While bubbles of gas that formed voids in the superconducting phase had been seen before, especially in short tapes, they were considered to be a minor nuisance compared to the overwhelming effects of the HAGBs. Depending on how it was viewed, a bubble could be obscured by a bridging Bi-2212 crystallite, or filled by polishing debris in cross sections. Alternatively, the gas that forms the bubbles could diffuse out of the ends of unsealed short wires, thus eliminating the bubbles. All of this led to huge variability and generally very poor properties, especially in any length suitable for magnet wires.
At CERN, researchers were able to watch the heating sequence in real time. Larbalestier said, “As the 2212 filaments enter the melt stage, Christian Scheuerlein [at CERN] was able to see the formation of little lenses of gas and observe the growth of these lenses into bubbles. The critical parameter controlling the current density was thus seen to be the formation of these gas bubbles which had to be bridged by 2212 grains that reformed during the solidification process.”
So how to eliminate the formation of gas bubbles during the heating stage? Hellstrom and Larbalestier had met a variant of this problem earlier with Bi-2223 when they had noticed some residual porosity in samples of that material. Working with American Superconductor Corporation, they had jointly patented an overpressure process that closed up the small residual pores left after rolling the wire to tape, raising the current density by about 30%. Using this process, researchers Jianyi Jiang and Maxime Matras at FSU started reacting round, 0.8-mm-diameter Ag wires embedded with 666 Bi-2212 powder filaments, each 15 µm in diameter. Inserting this wire into an overpressure furnace at pressures of 1–100 bars during the heat-treating process prevented wire diameter expansion by Ag creep, allowing for the first full densification of the Bi-2212 phase. Measurement of the whole-conductor (or engineering) current density, J E, revealed that J E increased by a factor of eight going from 1 bar to 100 bars pressure. The overpressure collapsed bubbles as they formed, leading to higher critical current density.
The research team made a test coil that reached 33.8 T. Larbalestier said, “This is not just a breakthrough result for short wire samples—it’s a breakthrough using long samples that have been tested in a very high field magnet. The wires are now being scaled up to kilometer lengths with the wire producer, Oxford Superconducting Technology.”
The research team wants to repeat this process in YBCO next which, they said, would be genuinely revolutionary for superconducting magnet technology because it would allow construction of multi-Tesla magnets in the 30–70 K regime where no other superconductor can operate.
