Better superconducting wires

Magnets are the biggest market for superconducting wire, and the favored wire for generating high fields is round and contains multiple filaments. The round geometry enables flexible conductor designs that support high magnetic stresses and can be wound in complex patterns to precisely shape the field. Although wires made from YBa₂Cu₃O_7−x (YBCO) and other high-temperature cuprate superconductors offer the allure of high current densities, that goal has been realized primarily only in broad, flat, single-filament “tape” geometries that reduce the prevalence of current-limiting misalignments between crystal grains (see the article by Alex Malozemoff, Jochen Mannhart, and Doug Scalapino, Physics Today, April 2005, page 41 , and also January 2008, page 30 ). Now David Larbalestier (National High Magnetic Field Laboratory) and colleagues have found a way to make round, multifilament wire out of the superconductor Bi₂Sr₂Ca₁Cu₂O_8−x (Bi-2212). The team’s key observation is that the usual wire fabrication process, which involves melting Bi-2212 powder packed inside a long silver tube, introduces bubbles that compromise the superconductor’s connectivity, and those voids, not grain boundaries, are what place the dominant constraint on the current density. By applying a pressure of 100 bar during the heat treatment, the researchers could suppress the bubble formation. The resulting wires are denser and better connected, and despite having numerous grain boundaries (colors in this wire image reveal the significant grain misorientation), at liquid-helium temperatures they can carry a higher current density in strong fields than cuprate tapes. The wires open up new opportunities for 20- to 30-tesla magnets for nuclear magnetic resonance spectrometry and other applications, from physics labs to particle accelerators. The incorporation of similar improvements into YBCO wires could yield a transformational technology for helium-free high-field magnets. (D. Larbalestier et al., Nat. Mater. 13, 375, 2014, doi:10.1038/nmat3887 .)